Tumors of the Hemolymphatic System

7
Tumors of the Hemolymphatic System


Victor E. Valli,1 Dorothee Bienzle,2 and Donald J. Meuten3


1University of Illinois, USA


2University of Guelph, Ontario Veterinary College, Canada


3North Carolina State University, USA


INTRODUCTION


Classification and nomenclature schemes for tumors of the hemolymphatic system are fairly numerous and have undergone multiple revisions through the decades.1–6 As we understand more about normal and neoplastic cells in this system, including their biology, molecular, and genetic features, as well as how the different tumors respond to therapies, our classifications will need to be modified. The two general groups for tumors of the hemolymphatic system are lymphoid (lymphoma and leukemia) and myeloid leukemias. The tumors in each can be characterized by their anatomic distribution in the body, histologic distribution within hemolymphatic organs, morphology of cells, immunophenotype, cytopenias, and, when known, their biologic behavior and molecular characteristics. The latter is just being explored in veterinary medicine but is relied on for classifying human hematopoietic tumors. Biologic behavior is a key component in classifying human tumors and determining whether classifications have clinical utility. However, accurate, consistent follow‐up data on large numbers of each type of lymphoma in animals are sadly lacking.


Lymphomas arise in lymphoid tissues outside of the bone marrow and leukemias arise in bone marrow or spleen. However, this is biology and these dividing lines merge. A diagnosis of lymphoma or leukemia is made from cytology or histopathology and in many cases that part of the diagnosis is straightforward. A reality in veterinary medicine is that the diagnosis from cytology may provide sufficient information for clinicians and owners. In most situations there is no third party payment available to explore all diagnostic avenues. Nevertheless, immunophenotyping to determine if the neoplasm is of B‐ or T‐cell lineage or an atypical phenotype is now a central component of classifying lymphomas; it is requested by approximately 75% of veterinary oncologists and the results of phenotyping may influence treatment recommendations.1–9 It is reasonable to assume most labs can phenotype into B‐cell or T‐cell types, but to subtype and fully characterize lymphomas and leukemias (or other tumors) requires a broad battery of antibodies and expertise (see Tables 3.3, 7.2, and 8.2). Furthermore, the methods used to recognize B cell versus T cell (immunohistochemistry (IHC), flow cytometry, and PCR for antigen receptor rearrangements (PARR)) can produce different results and there are contrasting results in the literature.7,10,11 PARR is a polymerase chain reaction (PCR)‐based assay for antigen receptor rearrangement that determines clonality and has also been used to determine B‐cell versus T‐cell phenotype. Determination of phenotype by PARR agreed with IHC in approximately 70% of cases and with flow cytometry in 63%; IHC and flow cytometry agreed in 94%.7 Phenotype is determined well with IHC, which is considered the gold standard but requires histopathology, or with flow cytometry, which can utilize a wide variety of antibodies and allows assessment of additional parameters, such as enumeration and size, and may identify atypical phenotypes but requires live cells. Flow cytometry as used in most laboratories detects only cell membrane antigens, whereas PARR can be used on a wide variety of samples, including cytology or histopathology preparations. This sometimes could avoid the collection of a new sample. We need standardized definitions of the tumors and the techniques used to help classify them. Until this is established there will be conflicting reports in the literature.


A large battery of antibodies can be used in flow cytometry and these may be needed to distinguish B‐cell versus T‐cell acute lymphoblastic leukemia (ALL) and to distinguish both from acute myeloid leukemia (AML). Flow cytometry is a highly developed research technique that is nowadays used to detect antigens, nuclear properties, organisms, and RNA anywhere inside and on the outside of cells. However, flow cytometry is also a very artifact‐prone technique that requires comprehensive controls and should not be practiced by the casual user. Detecting surface antigens is fairly straightforward with either directly or indirectly labeled antibodies. But detecting intracellular antigens requires permeabilization of cell membranes with a low concentration of paraformaldehyde plus detergent, which renders cells very “sticky,” meaning antibodies are more likely to bind nonspecifically. Only antibodies directly conjugated with fluorochromes should be used for intracellular staining, and additional controls are required to assure specificity of signal. For example, antibodies to CD79a, such as clone HM57, tend to bind nonspecifically to nuclei and some types of collagen when the actual target is an intracellular epitope in B cells. If such antibodies are used in intracellular flow cytometry in unconjugated format they are very likely to yield false‐positive signals. Hence, when setting up flow cytometry it is important to incorporate a comparison of detection of signal to microscopic localization of signal.12


Flow cytometry and IHC identify phenotypes but do not differentiate neoplastic cells from reactive cells. However, the number of cells and homogeneity of patterns in tissue sections or in flow cytometry help to identify clonal versus heterogeneous cellular proliferations. Flow cytometry has also been used to provide prognoses for dogs with lymphocytosis.13


In addition to phenotype and cytologic morphology lymphomas are also subtyped based on cell size (small, intermediate, large), high, intermediate, or low grades, and by histologic patterns in lymphoid organs. They are also categorized by anatomic location in the body: enteric, mediastinal, nasal, etc. Molecular characterizations of canine lymphomas are being investigated and results indicate that phenotypic variability is reflected in molecular heterogeneity.14–16 Therefore, the classifications of lymphomas are difficult to grasp and explain and conflicting reports in the literature make summaries almost impossible. Depending on the classification system(s) used there may be 50–100 different lymphomas and subtypes. In veterinary pathology we can identify about 20 B‐cell lymphomas and almost as many T‐cell lymphomas. However, approximately 85% of canine lymphomas fall within one of these five diagnoses, in descending frequency (percentages in parentheses are approximations only): diffuse large B‐cell lymphoma (50%); peripheral T‐cell lymphoma not otherwise specified (15%), T‐zone lymphoma (5–10%), T‐lymphoblastic lymphoma (4%), and marginal zone lymphoma (5–10%).4–6,17


Other types of lymphoma are uncommon in dogs. The uncommon lymphomas and non‐neoplastic lesions such as lymphoid hyperplasia make up the remaining approximate 15% of diagnoses in lymph nodes of dogs.4,5 In principle, the diagnosis of lymphoma is still easy but subtyping can be complicated depending on how specific one wants to be and requires histopathology and immunophenotyping. Cytology is an excellent means to diagnose lymphoma but it cannot classify each type. Flow cytometry can diagnose and determine cell type but it cannot classify lymphomas via the Revised European–American Lymphoma (REAL)/World Health Organization (WHO) system, which requires histology. Neither flow cytometry or cytology requires general anesthesia and they are both practical means to diagnose and offer prognostic information.


Approximately 60–70% of canine lymphomas are B‐cell, 30–40% T‐cell, and less than 1% null, non‐B non‐T‐cell.4,6 If a lymphoma is a large cell type and of B‐cell origin then diffuse large B‐cell lymphoma (DLBCL) is the most likely diagnosis by a wide margin, especially for dogs but most species as well. Approximately 90% of large cell B‐cell lymphomas in dogs will be DLBCL or marginal zone lymphoma (MZL). The majority of DLBCLs will be multicentric.4,6,17 DLBCL can be subdivided into immunoblastic or centroblastic (more common), and each of these may be further divided by their mitotic count into low, intermediate, and high grade.4–6,17 The end result is a final subtype but the total number of cases per group in retro‐ or prospective studies is invariably small and statistical comparisons biased. As molecular tools are used to study canine lymphomas it is likely that groups of lymphomas such as DLBCL will be declared as heterogeneous, as they have been in human lymphomas. We already know that B‐cell lymphomas can be subdivided by parameters such as flow cytometry, survivin, class II MHC, or size of cells into prognostically important subgroups.13,18 Molecular characterization is likely to further enhance our ability to segregate and provide clinically relevant prognoses.14,16,19,20


If a canine B‐cell lymphoma is of intermediate size than MZL is likely, especially if there is a histologic pattern of neoplastic lymphoid cells around fading germinal centers. However, the diagnosis of MZL requires histology to see the organizational pattern of the tumor in a lymph node or spleen. MZL is the most common lymphoma in the spleen of dogs and humans but it can also arise in or spread to lymph nodes. It can be difficult to distinguish MZL and DLBCL morphologically, and molecular studies suggest the two diseases could be a continuum.14,19


T‐cell lymphomas in dogs are suggested by cytology, histology, and confirmed by immunophenotyping. The most common T‐cell lymphoma is peripheral T‐cell lymphoma not otherwise specified (PTCL‐NOS), which is an umbrella term for T‐cell lymphomas not yet fully characterized. It is followed by T‐zone lymphoma (TZL) (also referred to as small clear cell type), lymphoblastic T‐cell lymphoma, and numerous reported cases of cutaneous T‐cell and large granular lymphocytic (LGL) lymphomas.5,6 PTCL‐NOS will be subdivided into disease entities as new information is discovered from immunophenotyping, molecular signatures, and biologic behavior. It is hoped that these will be correlated with treatment outcomes. TZL is an important diagnosis to recognize as it has an indolent behavior that multiple studies have agreed on. Furthermore it has a molecular signature that complements its unique histologic, cytologic, immunophenotypic, and biologic behavior.14


Of the lymphoid leukemias, chronic lymphocytic leukemia (CLL) and ALL are the most common in dogs and their distinction has profound clinical importance.9,10,21 Morphologically the tumor cells can look similar but correct diagnosis is usually straightforward if all the data is looked at in a case (e.g., cytopenia, CD34, clinical course). CLL and small cell lymphoma (SLL) are important diagnoses to identify correctly as they are indolent neoplasms with survival times of up to 3–4 years with or without treatment. However, dogs with atypical phenotypes of CLL have survival times of only days to weeks.21


Lymphoblastic tumors (ALL, lymphoblastic lymphoma (LBL)) are some of the most aggressive neoplasms in veterinary medicine with survival times measured in weeks to months. Lymphoblastic neoplasms can be T‐ or B‐cell and the literature on prevalence is conflicting.8,9,11,17,22 The term “lymphoblastic” has been used imprecisely in veterinary pathology. Morphologically it should be reserved for an intermediate‐size cell, one with dispersed nuclear chromatin and inconspicuous nucleoli. Large cells with large nuclei and prominent nucleoli are large cell lymphomas, B‐ or T‐cell. Use of the term lymphoblastic is discussed in the sections on these neoplasms.


Approximately 75% of canine CLLs are T‐cell, 25% B‐cell and a few may be dual negative or have atypical patterns. Of the T‐cell CLLs the majority are LGL, and within this subtype >90% are CD8+ cytotoxic T lymphocytes. Spleen is their likely tissue of origin. The lymphoma counterpart is SLL, and most of these are T‐cell. CLLs are less common in cats and they are also T‐cell neoplasms (90%) but most are CD4+ T‐helper lymphocytes. In humans, the majority of CLLs are B‐cell neoplasms.


Many other lymphomas and leukemias are described in dogs but these are the most common, at least with our present classifications, antibodies, and differentiating features. We tend to focus on exceptions and rarities but for most of our diagnostic work we should be able to recognize the common lymphomas and leukemias. As more molecular and genetic markers are applied there will be revisions to our present classifications. The most practical application of these classifications is to provide prognoses and possibly help guide treatment selections. For the latter we need accurate clinical outcome assessments complete with autopsies and large enough numbers of animals in each type of lymphoma and treatment group (versus no treatment) for conclusions to be meaningful.


Lymphomas account for approximately 80% of all hematopoietic tumors. They develop from lymphocytes and the most common site is in lymph nodes. Less commonly they arise from lymphocytes located in lymphoid centers outside lymph nodes, including spleen, tonsil, gastrointestinal tract, thymus, skin, nasal, and other tissues. If the predominant clinical signs are referable to a specific system, that anatomic location is often appended to the diagnosis; hence terms such as enteric, mediastinal, and cutaneous lymphoma. Extranodal lymphoma is a term used for lymphomas that are believed to be confined to locations other than lymph nodes, mediastinum, gastrointestinal, or multicentric; examples are nasal, respiratory, cutaneous, nervous, ocular, etc. Rarely is it proven that the lymphoma arises in one of these systems and remains confined to that system. This terminology is generic and the tumor should still be identified by cellular features, phenotype, and how generalized the lymphoma is. Sometimes the lymphoma is reported or implied to be confined to that system but in many cases a thorough clinical search or autopsy has not been performed to determine accurately how widespread or confined the neoplasm is.


Enteric lymphoma accounts for approximately 50% of all lymphomas in cats and presently is the most common anatomic site of lymphoma in cats.17,23–27 It is an example of a lymphoma that arises in extranodal mucosal lymphoid centers and is unique to an anatomic location; in fact, in cats the type of lymphoma differs between stomach, jejunum, and the ileal–cecal region.23 The majority of enteric lymphomas in the jejunum of cats are of T‐cell phenotype, however, enteric lymphomas in the stomach or ileal–cecal region may be of T‐ or B‐cell phenotype, and there are small cell, large cell, and granulated types.23,26 Some of the enteric lymphomas are indolent, usually small cell type and mucosal, and others are aggressive, usually large cell type, transmural, and often granular lymphocytes. Cats with this form of transmural lymphoma live less than 1 month post diagnosis.23 However, it is not reported if the tumors are also in lymph nodes, spleen, liver, or bone marrow and if this influences survival. Data like this are often not reported because retrospective studies have incomplete clinical and/or pathology reports. The biology of the normal enteric lymphoid cells, including trafficking signals and their neoplastic counterparts is intriguing.23


Immunophenotyping is an essential part of the classification of lymphomas. Even in topographic locations where a B‐ or T‐cell neoplasm is more likely, sometimes the opposite phenotype is found and without special techniques cell lineage cannot be identified from routine cytology or histopathology. However, there are some lymphomas in which the phenotype can be predicted with high accuracy because of historical characterization. Immunophenotyping is needed to be certain. For example, follicular lymphoma in humans is B‐cell, jejunal lymphoma of cats is T‐cell, marginal zone lymphoma and mantle cell lymphoma in dogs are B‐cell, large granular lymphoma is T‐cell, mediastinal lymphoma in hypercalcemic dogs and cats is T‐cell, CLL in cats is T‐cell predominant, and nasal nasopharyngeal lymphoma in cats is B‐cell (exceptions are noted). For some anatomic forms of lymphoma, in some species the determination of B‐ versus T‐cell type, subtypes and/or presence of lymphocytosis have prognostic significance.4,6,8,13,21 As lymphoma morphology and immunophenotype are more thoroughly correlated with biological behavior and response to therapy, knowledge of cell type and subtypes will be increasingly relevant.


Occasionally a lymphoma will also have neoplastic cells in the peripheral blood; this is secondary leukemia. For leukemia to be present the lymphoma must also be in the bone marrow or spleen. There are no accurate data for each type of lymphoma but probably less than 20% of all lymphomas in animals are leukemic, or at least are recognized to be leukemic with our present means of detection. The methods used to determine if neoplastic cells are in circulation influences the percentage of animals that have lymphoma and concurrent leukemia. Many T‐cell lymphomas have circulating neoplastic cells, and this is less common for B‐cell lymphomas. Many small cell T‐cell lymphomas are of splenic red pulp origin (CD11d positive), which manifests as CLL and splenomegaly. It is rare for splenic lymphomas other than CLL to cause leukemia; again, how thoroughly bone marrow involvement is searched for will influence final conclusions.


When neoplastic lymphoid cells are identified in circulation it may be difficult to determine if they are secondary to lymphoma (stage V) or if the neoplasm is a primary lymphoid leukemia. Prognoses vary with the degree of lymphocytosis and the type and size of the neoplastic cells.13,21 Lymphopenia is commonly seen with lymphomas in all species and may be more common than lymphocytosis. It is attributed to endogenous steroids from the stress of the cancer, which causes lymphocytolysis and redistribution of lymphocytes.


Leukemias are neoplasms arising in bone marrow. Three general groups are acute myeloid leukemia, myeloproliferative neoplasms, and myelodysplastic syndromes. Neoplastic cells are usually myeloid or myelomonocytic, but erythroid, lymphoid, megakaryocytic, and rarely mast cell leukemias also occur in animals. The neoplastic cells may or may not be observed in routinely prepared blood films. Some forms have marked increases of neoplastic cells in the peripheral blood and in others the neoplastic cells are only found in low numbers and when searched for diligently: feathered edge of blood films, or buffy coat preparations or flow cytometry. Although few or no leukemic cells may be apparent on the blood smears of animals with AML, the greater volume of blood assessed in automated analyzers and flow cytometers generally identifies abnormal leukocytes. When few or no leukemic cells are seen in peripheral blood, terms such as “aleukemic leukemia” are sometimes used, but these are cumbersome and not necessary for naming the type of neoplasm. The most important means to classify leukemias is to assess their morphology and to analyze hematology results. Flow cytometry and immunophenotyping is helpful to rule out lymphoid leukemia and may identify lineage of blast cells.


This chapter is about tumors of the hemolymphatic system and it is divided into lymphomas, followed by descriptions of myeloid tumors, and nonlymphoid tumors in the spleen.


LYMPHOID TUMORS


Biological implications of tumor classification


The classification of lymphomas in this chapter follows the updated World Health Organization system (REAL/WHO) outlined in Box 7.1. Classification of lymphomas has progressed through criteria largely related to cytological characteristics (Kiel) to one based on distinct diseases encompassing the total information related to that neoplasm.1–5 This current interpretation of the lymphomas has been through a gradual progression of classification systems that led to the universal understanding that there is no gold standard of cell classification that applies to all lymphoma subtypes in all species. In many respects the classification adopted by the WHO for human lymphomas and leukemias was driven by immunophenotyping that showed that cells with similar morphology can have differing phenotypes and markedly different biology in both normal and neoplastic situations. The present system uses morphology (cytology and histology), immunophenotype (flow cytometry, IHC), molecular characteristics (if known) and the biologic behavior (aggressive, indolent, response to treatments) to arrive at a diagnosis (disease). The apparent deficiency of a system that does not group the subtypes of lymphoma by grade is made up for by each subtype of lymphoma representing a well‐characterized disease.


For most lymphomas and leukemias in humans, these diseases and therefore the defining characteristics are based on hundreds to thousands of examples. In animals our numbers in each diagnosis are much smaller and these numbers are further reduced by subtypes, different treatment groups, and limited follow‐up data for each classification. Our database to characterize the disease therefore has inherent constraints.


There is also likely interpathologist variation in our diagnoses, especially since many diagnostic veterinary pathologists are not hematopathologists. In human pathology biopsies and/or cytologic samples are examined by clinical pathologists and/or hematopathology specialists. In veterinary pathology we have cost constraints that limit the number of antibodies and the techniques used to characterize the neoplastic cells. We also have the reality that owners may stop the diagnostic work‐up at the point they know their animal has “cancer.” An advantage, however, is that not all of our patients will be treated and therefore we can compare survival times between treatment groups versus no treatment. A confounder in these data is that owners will elect when their pet is euthanized and these decisions are not always based on the extent of the tumor.


Although we work with animals, our ability to determine final outcomes, which is critical to know if these classification systems have practical utility, is mediocre to terrible. Too many cases are lost to follow‐up or an autopsy is not performed to determine recurrence, metastasis, which hemolymphatic organs are involved, concurrent diseases and/or tumors, etc.


Despite these differences and limitations the REAL/WHO system should be the one used by veterinary clinical and anatomic pathologists to classify lymphomas fully. Having said that, there will be many cases evaluated without histopathology, or phenotyping, or without a thorough search for anatomic distribution. It is a practical consideration that the oncologist and owner will decide what parameters they choose to be evaluated. Many cases will be diagnosed accurately as lymphoma from cytology, without histopathology or flow cytometry or IHC and will be classified from the data that was gathered and treated accordingly. It becomes difficult to justify the collection of additional tissues unless there are clear outcome assessments that will affect prognosis or treatment selections. Furthermore, histologic classifications of nodal lymphomas are made from an entire lymph node; trying to determine morphologic patterns in a core sample is difficult. Flow cytometry can be performed on samples obtained from fine‐needle aspiration (FNA) but it requires live cells which can be problematic. These factors and the costs involved will influence what data are available to classify a lymphoma and to correlate that diagnosis with follow‐up data.


The WHO/REAL classification has been tested by a group of MD pathologists who retrospectively evaluated 1403 cases of human lymphoma at eight sites worldwide. The specific goals of that review were as follows: First, to evaluate the ability of the hematopathologists to apply the International Lymphoma Study Group (ILSG) classification system to a large group of lymphomas collected at different sites. Second, to determine the role of immunophenotyping and clinical data in diagnostic use. Third, to determine the intra‐observer and inter‐observer reproducibility.1–3 This model was followed to test the value and applicability of the WHO classification when applied to lymphomas in dogs with 17 veterinary pathologists5 and another with 7 veterinary pathologists6 forming the reviewing groups. The conclusions of these smaller studies were that experienced veterinary pathologists who were not specialists in hematopathology could accurately apply the WHO classification after review of a CD describing the application of the WHO system.5 There was good agreement between pathologists and if only the six most common tumors were reviewed the agreement was almost 90%. The corollary to this conclusion is that the WHO system of classification is applicable to animal lymphomas. The review by the MD pathologists concluded that IHC was essential for all lymphomas other than follicular lymphoma because the immunophenotype of follicular lymphoma could be predicted from the characteristic follicle formation. Since follicular lymphoma is very uncommon in animals, phenotypic identification is required to classify all lymphomas in animals.


Presently the diagnosis of lymphoma in animals is made from a cytologic or histologic specimen. Classification is accomplished by immunophenotyping (T‐cell, B‐cell) and histology or cytology that assesses cellular and nuclear features, histologic organization, as well as growth patterns of the lymphomas, all of which are, or should be, correlated with outcome assessment. Of these characteristics used to classify lymphomas, outcome assessment is our weakest link. All of these parameters may not be done on suspected cases of lymphoma. If the diagnostic investigation stops at FNA cytology then the tumor cannot be fully classified. However, enough information may be available for an oncologist and owner to make decisions without additional tests. Detailed diagnoses that use all the parameters available need to be linked to accurate outcome assessments, including data from autopsies to correlate diagnoses with survival times, disease‐free intervals, treatment responses, relapses, etc. to know if these classifications have clinical value. In veterinary medicine we also need comparable studies that use cytology and information that can be collected without histopathology. Immunocytochemistry and flow cytometry are valuable in this regard and are a part of standard diagnoses at referral centers. Molecular cytogenetic characteristics are used in human lymphomas to help predict prognoses and direct therapies. These tools are being used to evaluate lymphomas in dogs, primarily to determine the usefulness of canine models for human lymphomas, but we will benefit greatly.14–16,19,20,28–37 It is easier to correlate molecular characteristics with classifications but it will be years before they are correlated with outcome assessment.


Types of lymphoma by species


The incidence of the various types of lymphomas in domestic animals can be expected to follow the guidelines in Box 7.1. Throughout this chapter the reader should consider percentages as approximations. Studies will not replicate exact percentages or incidences. Indeed, we should feel fortunate if the results between studies are close, especially given the small numbers of animals per group that are often reported. Additionally there will be different prevalences between species and there may be tumors in humans that do not exist in animals and vice versa. For example, in humans follicular lymphoma is very common and Burkitt’s lymphoma is common but both are rare in animals; T‐zone lymphoma is common in dogs but is not classified separately in humans; and for some types of lymphoma described in dogs no counterpart apparently exists in humans.6,17


Approximately 60–70% of canine lymphomas are B‐cell, 30–40% T‐cell, <1% null‐cell (non‐B, non‐T), and 75% are high grade and 25% low grade.4,6,17 Some tumors may dual mark and there are uncommon aberrant phenotypes.9,13,21 The term null cell is used generically for a lymphocyte that was not immunoreactive for B or T cell. It may be a poorly differentiated cell of unknown lineage, there could be a methodology issue, or it may be a natural killer (NK) cell. We do not have antibodies or clear definitions to recognize NK‐cell lymphomas in animals as we do in humans. Regardless, >95% of lymphomas in animals will be B‐ or T‐cell phenotype. In dogs the two most common lymphomas are diffuse large B‐cell at 50% and peripheral T‐cell lymphoma not otherwise specified at 15%.4,6,17 Reports differ on percentages for the different types but the next most common lymphomas in dogs are T‐zone at 3–15% and marginal zone at 5–10%; in one study T‐cell cutaneous lymphomas made up approximately 12% of the lymphomas.6 Lymphoblastic lymphomas of T‐cell account for 3–5%, Burkitt’s‐like lymphomas 2%, and mantle cell lymphoma just under 2%.4–6,17 The T‐cell‐rich large B‐cell lymphoma is 1% and all other subtypes of lymphomas are under 2%. Differences between studies will always appear and can be attributed to the subjectivity of classifications, definitions of tumors, antibodies used, and caseloads. The latter may explain the high prevalence of cutaneous T‐cell lymphomas if a large portion of cases are derived from surgical skin biopsies.6


Approximately 15–20% of canine lymph nodes biopsied because they are enlarged and because lymphoma is a differential diagnosis are not neoplastic and have a type of lymphoid hyperplasia. However, the survival time in this group was approximately one year, suggesting misdiagnosis or that lymphoid hyperplasia indicates an underlying serious disease.5 Unfortunately, follow‐up data in this study was too limited to know all the final outcomes.


There are many other types of lymphoma in dogs but they are uncommon to rare. Unique types of lymphoma in dogs include plasmacytoid T‐cell6 and multicentric small mature T‐cell.17 By anatomic location the most common site in dogs is multicentric (75%), followed by extranodal, skin, and other sites.5,6,17 Sometimes epidemiologic, clinical, and clinical pathology data are reported for specific types of lymphomas but most of the reported data is for lymphomas in general. From a clinical perspective the most important diagnoses to make are the indolent versus aggressive lymphomas (see Appendix for summary). Common indolent lymphomas include TZL, small mature lymphocytic (CLL, SLL), and possibly MZL, but current data suggest MZL may be more aggressive than previously thought.14,19 The common aggressive lymphomas include DLBCL, PTCL‐NOS, and lymphoblastic (ALL, LBL). There are other indolent and aggressive lymphomas but these are the more common tumors. The majority of this chapter is on hematopoietic tumors of dogs. This reflects the literature and the prevalence of hematopoietic tumors by species.


Since the introduction of feline leukemia virus (FeLV) vaccination, the most common type of lymphoma in cats is enteric T‐cell lymphoma, which makes up approximately 50% of all feline lymphomas.17,23–27 Feline enteric lymphoma has been characterized and illustrated in detail23 and is described in the section on T‐cell lymphomas. Testing and vaccinating cats for FeLV has shifted the prevalence of lymphomas from younger cats with mediastinal lymphoma to older cats with extranodal types of lymphomas.24–27 FeLV‐associated lymphomas are usually multicentric or mediastinal and in young cats. Feline immunovirus (FIV)‐associated tumors in cats are usually high‐grade B‐cell lymphomas. By anatomic location enteric, mediastinal, and multicentric are the most common locations of lymphoma in cats.17,23–27


Recent characterization of upper respiratory tract (URT) lymphomas in cats indicates the majority, almost 90%, are B‐cell; they have an aggressive course and are the most common neoplasm in the URT of cats.25 They can be subdivided by anatomical location into nasal (67%), nasopharyngeal (16%), or both locations (18%). Approximately 25% had an epitheliotropic pattern and this was associated with longer survival but the number of cases was small.25 Approximately 56% are FeLV positive and 60% have submucosal inflammation. Survival times ranged from 0 to 300 days with a mean of 53 days.25


Using the REAL/WHO system, DLBCL is one of the most common types of lymphoma in cats.17,24,26,27 Locations may be mediastinal, in URT, in segments of the bowel, or multicentric. Cats also have large cell lymphomas, T‐cell‐rich B‐cell lymphomas (TCRLBCL), that are “Hodgkin’s‐like.”17,24,26 Cytologic characterizations are usually reported as histopathology is often not performed.26,27 Prognostic information for cats is embedded in descriptive manuscripts of the tumors.23–27 Enteric LGL lymphomas that are transmural have short survival times, and if concurrent lymphocytosis is present survival may be only days post diagnosis.


In horses, TCRLBCL is the most common lymphoma at about 40–45% and this is consistent between studies.38–41 After TCRLBCL, studies report different prevalences, but the more common lymphomas in horses are (approximations are in parentheses): peripheral T‐cell lymphoma (PTCL) (20%), diffuse large B‐cell (10%), enteric T‐cell type (5%), and cutaneous T‐cell.38,39 Anaplastic B‐cell is reported, as are most of the other types of lymphoma but much less frequently than those just listed. By anatomic location multicentric is most common, but cutaneous and enteric are also common.38,39,41 How carefully it is determined that a lymphoma arose or remained confined to a system will influence the data between studies. Multicentric lymphomas in horses can be TCRlBCL (n = 28), PTCL (n = 26), DLBCL (n = 11), or other types.38 Cutaneous lymphoma in horses is most frequently TCRLBCL by a wide margin (80–85%) followed by T‐cell and DLBCL.


TCRLBCL is classified as a type of large B‐cell lymphoma. It has the unusual presentation in horses of a few to hundreds of subcutaneous nodules. The clinical presentation is approximately 50:50 solitary nodule versus multiple. If only a few nodules are present and they are excised fully this has proven an effective treatment in some cases.39 Those cases in which the tumor does not recur post excision have long survival times of up to 10 years and those that recur have shortened survival.39 In a study of 203 lymphomas multicentric was the most common location (n = 83), then cutaneous (n = 38) and enteric (n = 24).38 Fourteen subtypes could be classified and TCRLBCL was most common (n = 87), followed by PTCL (n = 45) and DLBCL (n = 26).38 Equine lymphomas are heterogeneous and can be difficult to classify by morphology.41 A study of 37 equine cases reported 26 as T‐cell, 7 as B‐cell, and 4 could not be classified.41 The age range of horses with lymphoma is wide, from months old to 30 years with a mean around 11 years.38,41


In cattle, in cases submitted for slaughter 65% were of large cell type and are likely to be large B‐cell lymphoma, but immunophenotyping was not used to characterize lymphomas in the early studies in cattle. The lymphomas assumed to be of sporadic type (non‐bovine leukemia virus (BLV) associated) based on the anatomic location of the tumor and age of the animal, were 10% of total cases and are likely T‐cell type.


Lymphoma was the most common tumor identified in 100 goats and the most common distribution was multicentric.42 Seventeen cases were reported, and ages ranged from 1 to 9 years with a median of 3 years. Nine goats had thymomas and two were predominantly lymphocytes.


Standardization for diagnoses of lymphomas is needed, not only for morphologic definitions but also for flow cytometry, IHC, PARR, antibodies, primers, and the techniques and parameters used to establish diagnoses. These standards need to be agreed upon for different species worldwide. We need to determine where lymphomas arise and how confined or disseminated the tumors are. Identification of molecular and genetic profiles needs to be expanded and correlated with morphologic and clinical data. A pragmatic feature of veterinary medicine is that we need diagnostic and prognostic tools that can be used with the least invasive and expensive procedures. FNA is a practical tool with distinct advantages, but for nodular lymphomas and tumors that are not diffuse in a node or organ, sampling of the representative cells is a concern. Follow‐up data that predicts outcome is critical and it needs to be collected as accurately and carefully as the laboratory data. Expertise, antibodies used, and how tumors are defined greatly influence final conclusions. Central or regional laboratories that standardize these and other factors as well as gather large numbers of cases and interpret the results are needed if we are to develop accurate and uniform data based on hundreds to thousands of cases. These labs can maintain expensive and current infrastructures. In clinical pathology we have done this for years with some of the less common laboratory tests (e.g., those used in endocrinology). It is not necessary for each laboratory to maintain all the tests that are now available as well as interpret results, some of which are subjective. Without standardization for diagnoses and techniques there will continue to be contradictory results and relatively small case numbers which will hinder clinical applications.


Molecular considerations


Lymphoma is a heterogeneous group of tumors phenotypically and genetically. The Kiel and WHO/REAL classifications use multiple parameters to subdivide lymphomas into many diagnoses. Molecular techniques complement these classifications and additionally further subdivide lymphomas and leukemias, demonstrating how heterogeneous phenotypically similar tumors are.14,15,19,20,28–34 However lymphomas or other cancers are subdivided, the critical discovery is to find classifications that predict prognoses and responses to treatments. Molecular classifications of human DLBCL is predictive of biologic behavior and treatment outcomes and this will likely be true for animals. CLL in humans can be subdivided by molecular signatures into multiple groups that have different prognoses but we do not know this for animals.36 In animals we use immunophenotype, cell numbers, and size of cells to subdivide CLL into groups with different outcomes.13,21,32,33 There are many similarities between canine and human lymphomas and comparative studies enhance our understanding of lymphomas in both species, however differences are equally numerous. The same is true for canine and feline lymphomas and lymphomas in other animals. Put quite simply, dogs are not cats or humans but studying their differences may prove insightful.


Across species, DLBCL is the most common or one of the most common lymphomas. Some notable differences between species are as follows: CLL in humans is a B‐cell predominant tumor and in dogs and cats it is T‐cell predominant; T‐zone is rare and not listed as a separate classification in humans yet it is common in dogs; one of the most common lymphomas in humans is follicular and this is rare in animals, although perhaps molecular characterization in animals will demonstrate that we have mislabeled follicular lymphoma and some DLBCL; there are other differences but there are many similarities.


The principle of shared genomic information that may be similar between phenotypically divergent species is intriguing and rewarding. Finding the shared molecular pathways or comparable chromosomal regions between species that predispose to certain cancers will lead to molecular signatures that are predictive of outcomes independent of other parameters for each species. Molecular investigations on canine lymphomas are designed to determine the comparative usefulness of dogs as a model for human non‐Hodgkin’s lymphomas. Although that may be the primary goal, these investigations will help us understand the pathogenesis of lymphomas in animals and, importantly, improve the patient care we provide to pets and hopefully patient care to humans.


These collaborative investigations are numerous.14–16,19,20,28–37 For example, dogs and humans with a subtype of DLBCL have a mutation in TRAF3 which results in a loss of TRAF3 and therefore dysregulation of nuclear factor (NF)‐κB.30 Approximately 44% of 84 canine B‐cell lymphomas had a mutation of somatic and/or germ‐line variants. These results open up new possibilities for treatment with the proteasome inhibitor bortezomib, used to treat multiple myeloma in humans with a comparable mutation.30 TRAF3 is considered a tumor suppressor gene associated with multiple myeloma, Hodgkin’s lymphoma, splenic MZL, and other hematopoietic cancers in humans.


Canine and human DLBCLs share the signaling pathways NF‐κB, PI3/AKT, Notch, and JAK/STAT, which makes a spontaneous canine model useful, especially for treatment options.20 Gross genetic abnormalities in copy number aberrations (increases or decreases) in regions of dog chromosomes 13 and 31 may contribute to B‐ and T‐cell lymphomas in dogs.34 Gain in Canis familiaris autosome (CFA) 13 was the most frequent aberration in a study of 150 canine lymphomas and increases have been associated with other canine tumors, such as osteosarcoma, urothelial carcinoma, and prostate tumors.34 Copy number increase in the MYC oncogene (CFA 13q13) in canine lymphoma is of comparative interest because of similar increases in humans with B‐cell lymphoma. However the gain in MYC occurred in both B‐ and T‐cell canine lymphomas.34


Pet dogs have become a relevant model to study human lymphomas and different types of lymphoma are associated with certain breeds of dogs.31 Canine lymphomas have different molecular characteristics and there may be genetic susceptibilities in these breeds that allow specific types of lymphoma to develop. It is estimated that almost 70% of deaths in golden retrievers are due to cancer. Approximately 13% of golden retrievers develop lymphoma and they are split 50:50 B‐cell/T‐cell lymphomas, in contrast with the general dog population in which most lymphomas are B‐cell (70%). Approximately 20% of the risk for golden retrievers to develop B’cell lymphoma (and hemangiosarcoma) is due two loci located on chromosome 5 associated with differential expression of immune‐related genes, including BIRC3.28 The mutation and the mechanisms that lead to these cancers still need to be unraveled. Golden retrievers tend to develop more indolent TZLs than aggressive lymphomas, LBL, or PTCL but approximately 90% of lymphomas in boxer dogs are T‐cell, many of which are aggressive.14,29,33 Their lymphoma typically involves the PTEN‐mTOR pathway, which is important in several tumors in humans and animals.29 Mutations in the PTEN pathway were noted in approximately half of the samples biopsied and 81% of all boxers sequenced had a germ‐line variant in the proto‐oncogene FOS. Part of FOS transcription factor controls some of the downstream targets in the PTEN pathway, which may explain why the boxer breed is genetically more susceptible to mutations in the PTEN pathway.29


These researchers investigated somatic mutations by exome sequencing of tumor in three breeds of dogs that develop lymphoma: boxer with aggressive T‐cell, cocker spaniel with B‐cell, and golden retriever with B‐ and T‐cell.29 Golden retrievers and cocker spaniels had recurrent mutations in TRAF3‐MAP3K14 (28% of cases), FBXW7 (25%), and POT1 (17%). The T‐cell lymphomas in golden retrievers, which tend to be less aggressive, had mutations in genes involved in cellular metabolism, which suggests the immune system of the host is an important component in the tumorigenesis of lymphoma in this breed. Tumor‐initiating cells are important in the pathogenesis of hematopoietic and solid tissue cancers. A possible lymphoma‐initiating cell population has been studied in canine B‐cell lymphomas that expressed hematopoietic progenitor antigens.16


Chronic infection with gammaherpesvirus Epstein–Barr in humans is usually asymptomatic but a portion of infected individuals co‐infected with malaria will develop lymphomas, most notably Burkitt’s lymphoma. Serologic evidence indicates privately owned dogs with and without lymphoma are exposed to a gammaherpesvirus and further studies implicated its presence in B‐cell lymphomas of dogs and suggested it may be part lymphomagenesis in a subset of canine lymphomas.37 Retroviral particles have been seen and reported in canine lymphomas but are not considered causative.


Molecular diagnostic tests have been used in human oncology to identify tumors and provide clinically relevant prognostic features. Molecular signatures help to classify tumors and are part of the diagnostic arsenal in oncology. In humans DLBCLs that look morphologically identical can be subdivided by molecular means to identify groups that have different biological behaviors and responses to treatments. Gene expression profiles applied to six common canine lymphomas revealed that three groups could be identified and prognoses provided: TZL, high‐grade T‐cell (LBL, PTCL), and B‐cell (MZL, DLBCL, Burkitt’s lymphoma).14 Samples were obtained via biopsy and tissues disaggregated into single‐cell suspensions, however similar techniques can be used on cells derived with FNA. A diagnostic test was developed based on the expression of four genes. Gene expression profiles could distinguish the three groups indicated but could not distinguish MZL and DLBCL, even though they did find genes that were differentially expressed. The molecular similarities in MZL and DLBCL were such that they may represent a continuum rather than separate lymphomas.14 Although MZL is considered morphologically a nodular lymphoma and DLBCL is diffuse, both have stages in which their morphologies overlap and they are difficult to distinguish. Additional studies that correlate molecular, clinical, and light microscopic characteristics are needed to determine the relationship between MZL and DLBCL.


Burkitt’s lymphoma was defined by a translocation t(8;13) at the IGH locus of canine chromososome 8 (CFA 8) and the MYC locus in CFA 13.14,35 Burkitt’s or Burkitt’s‐like lymphoma has morphologic features that can be identified in dogs but sometimes the distinction from DLBCL is subtle. Having a molecular means to separate these tumors (and other tumors) would be beneficial, especially if they have clearly different behaviors or treatments. The methods used identified indolent (TZL) and aggressive T‐cell lymphomas (LBL and PTCL).14 The results also complemented prior observations that CD21 may be useful to separate indolent and aggressive T‐cell tumors.32,33 The present investigation identified higher expression of the CR2 gene,14 which codes for CD21 in TZL as compared to LBL and PTCL. Although CD21 is used as a B‐cell marker it is also expressed by T cells, and unique flow cytometry profiles of CD4, CD21, CD45, and class II MHC expression appear to separate aggressive and indolent T‐cell tumors.32,33 Subgroups of human DLBCL are believed to arise from the germinal center B cell (GCB) or from cells exiting the germinal center and referred to as activated B‐cell lymphoma (ABC). Gene expression profiles that can distinguish these two subtypes of human DLBCL could not distinguish canine DLBCL from MZL in this or a prior study.14,19 Furthermore, profiles of IHC that could identify different groups of human DLBCL could not do the same in canine B‐cell lymphoma.


A canine set of genes similar to the signature genes that identify GCB and ABC did separate the canine tumors into two groups and although the biologic behavior of the two groups had statistically significant differences, the distinctions in survival data were not sharp. DLBCL in humans can be separated into subtypes ABC or GCB but the data for this separation in dogs is not uniform.14,19,35 Although genes were not conserved between human and canine DLBCL it appeared that pathways were shared across species (NF‐κB signaling and B‐cell receptor signaling). The number of dogs in the MZL groups of both studies was small.14,19 Although morphologic differences between MZL and DLBCL can be defined, the two tumors are not genetically distinct and perhaps they represent a continuum.14,19,35 The majority of the samples studied were from resected lymph nodes but FNA was used to obtain cells in some of the cases.


Morphology is one of the main parameters used to diagnose lymphomas with the REAL/WHO classification. Histopathology is excellent but if diagnostic samples require general anesthesia to obtain them applications will be limited in veterinary medicine. It is a practical reality that diagnoses and prognoses that can be obtained from FNA will gain wider use by practicing veterinarians. A diagnosis of lymphoma from cytology is reliable and if sampling at the same time can yield flow cytometry profiles, immunophenotyping, and molecular tests that are cost acceptable, predict prognoses, and/or help guide treatment options they will be applied more widely than histopathology. A concern of FNA will always be whether the sample is representative of the tumor, especially in nodular lymphomas.


References



  1. 1. Harris, N.L, Jaffe, E.S., Diebold, J., et al. (2000) The World Health Organization classification of neoplastic diseases of the haematopoietic and lymphoid tissues: report of the Clinical Advisory Committee Meeting, Airlie House, Virginia, November 1997. Histopathology 66:69–89.
  2. 2. Armitage, J.O. (1997) NHL Classification project. A clinical evaluation of the International Lymphoma Study Group classification of non‐Hodgkin’s lymphoma. By the Non‐Hodgkin’s Lymphoma Classification Project. Blood 98:3909–3918.
  3. 3. Swerdlow, S.H., Campo, E., Harris, N.L., et al. (2008) WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues . International Agency for Research on Cancer (IARC), Lyon, France.
  4. 4. Valli, V.E., Kass, P., San Myint, M., and Scott, F. (2013) Canine lymphoma: The effect of age, stage of disease, subtype of tumor, mitotic rate and treatment protocol on overall survival. Vet Pathol 50:738–748.
  5. 5. Valli, V.E., San Myint, M., Barthel, A., et al. (2011) Classification of canine malignant lymphomas according to the World Health Organization criteria. Vet Pathol 48:198–211.
  6. 6. Ponce, F., Marchal, T., Magnol, J.P., et al. (2010) A morphological study of 608 cases of canine malignant lymphoma in France with a focus on comparative similarities between canine and human lymphoma morphology. Vet Pathol 47:414–433.
  7. 7. Thalheim, L., Williams, L.E., Borst, L.B., et al. (2013) Lymphoma immunophenotype of dogs determined by immunohistochemistry, flow cytometry, and polymerase chain reaction for antigen receptor rearrangements. J Vet Intern Med 27:1509–1516.
  8. 8. Regan, R., Kaplan, M., and Bailey, D. (2012) Diagnostic evaluation and treatment recommendations for dogs with substage – a high grade multicentric lymphoma: Results of a survey of veterinarians. Vet Comp Oncol DOI: 10.1111/j.1476‐5829.00318.x
  9. 9. Vernau, W. and Moore, P.F. (1999) An immunophenotypic study of canine leukemias and preliminary assessment of clonality by polymerase chain reaction. Vet Immunol Immunopathol 69:145–164.
  10. 10. Tasca, S., Carlil, E., Caldin, M., et al. (2009) Hematologic abnormalities and flow cytometric immunophenotyping results in dogs with hematopoietic neoplasia: 210 cases (2002–2006). Vet Clin Pathol 38:2–12.
  11. 11. Stokol, T., Schaefer, D.M., Shuman, M., et al. (2015) Alkaline phosphatase is a useful cytochemical marker for the diagnosis of acute myelomonocytic and monocytic leukemia in the dog. Vet Clin Pathol 44:79–93.
  12. 12. Reggeti, F. and Bienzle, D. (2011) Flow cytometry in veterinary oncology. Vet Pathol 48:223–235.
  13. 13. Williams, M.J., Avery, A.C., Lana, S.E., et al. (2008) Canine lymphoproliferative disease characterized by lymphocytosis: Immunophenotypic markers of prognosis. J Vet Intern Med 22:596–601.
  14. 14. Franz, A.M., Sarver, A.L., Ito, D., et al. (2013) Molecular profiling reveals prognostically significant subtypes of canine lymphoma. Vet Pathol 50:693–703.
  15. 15. Zhang, J., Grubor, V., Love, C.L., et al. (2013) Genetic heterogeneity of diffuse large B‐cell lymphoma. Proc Natl Acad Sci USA 110:1398–1403.
  16. 16. Ito, D., Frantz, A.M., and Modiano, J. (2014) Canine lymphoma as a comparative model for human non‐Hodgkins lymphoma: recent progress and application. Vet Immunol Immunopathol 159:192–201.
  17. 17. Vezzali, E., Parodi, A.L., Marcato, P.S., and Bettini, G. (2009) Histopathologic classification of 171 cases of canine and feline non‐Hodgkin lymphoma according to the WHO. Vet Comp Oncol 8:38–49.
  18. 18. Rao, S., Lana, S., Eickhoff, J., et al. (2011) Class II major histocompatibility complex expression and cell size independently predict survival in canine B‐cell lymphoma. J Vet Intern Med 25:1097–1105.
  19. 19. Richards, K.L., Motsinger‐Reif, A.A., Chen, H.W., et al. (2013) Gene profiling of canine B‐cell lymphoma reveals germinal center and post germinal center subtypes with different survival times, modeling human DLBCL. Cancer Res 73:5029–5039.
  20. 20. Mudaliar, M.A., Haggart, R.D., Miele, G., et al. (2013) Comparative gene expression profiling identifies common molecular signatures of NF‐kB activation in canine and human diffuse large B cell lymphoma (DLBCL). PLOS ONE 8:e72591.
  21. 21. Comazzi, S., Gelain, M.E., Martini, V., et al. (2011) Immunophenotype predicts survival time in dogs with chronic lymphocytic leukemia. J Vet Intern Med 25:100–106.
  22. 22. Novacco, M., Comazzi, S., Marconato, L., et al. (2015) Prognostic factors in canine acute leukaemias: a retrospective study. Vet Comp Oncol DOI: 10.1111/vc0.12136
  23. 23. Moore, P.F., Rodriguez‐Bertos, A., and Kass, P.H. (2012) Feline gastrointestinal lymphoma: mucosal architecture, immunophenotype, and molecular clonality. Vet Pathol 49: 658–668.
  24. 24. Valli, V.E., Jacobs, R.M., Norris, A., et al. (2000) The histologic classification of 602 cases of feline lymphoproliferative disease using the National Cancer Institute working formulation. J Vet Diagn Invest 12:295–306.
  25. 25. Santagostinol, S.F., Mortellarol, C.M., Boracchi, P., et al. (2015) Feline upper respiratory tract lymphoma: site, cyto‐histology, phenotype, FeLV expression, and prognosis. Vet Pathol 52:250–259.
  26. 26. Sato, H., Fujino, Y., Chino, J., et al. (2014) Prognostic analyses on anatomical and morphological classification of feline lymphoma. J Vet Med Sci 76: 807–811.
  27. 27. Chino, J., Fujino, Y., Kobayashi, T., et al. (2013) Cytomorphological and immunological classification of feline lymphomas: clinicopathological features of 76 cases. J Vet Med Sci 75:701–706.
  28. 28. Tonomura, N., Elvers, I., Thomas, R., et al. (2015) Genome‐wide association study identifies shared risk loci common to two malignancies in golden retriever. PLOS Genet DOI: 10.1371journal.pgen.1004922
  29. 29. Elvers, I., Turner‐Maier, J., Swofford, R., et al. (2015) Exome sequencing of lymphomas from three dog breeds reveals somatic mutation patterns reflecting genetic background. Genome Res 25:1634–1645.
  30. 30. Bushell, K.R., Kim, Y., Chan, F.C., et al. (2015) Genetic inactivation of TRAF3 in canine and human B‐cell lymphoma. Blood 125:999–1005.
  31. 31. Modiano, J.F., Breen, M., Burnett, R.C., et al. (2005) Distinct B‐cell and T‐cell lymphoproliferative disease prevalence among dog breeds indicates heritable risk. Cancer Res 65:5654–5661.
  32. 32. Seelig, D.M., Avery, P.R., Webb, T., et al. (2014) Canine T‐zone lymphoma: unique immunophenotypic features, outcome, and population characteristics. J Vet Intern Med 28:878–886.
  33. 33. Avery, P.R., Burton, J., Bromberek, J.L., et al. (2014) Flow cytometric characterization and clinical outcome of CD4+ T‐cell lymphoma in dogs: 67 cases. J Vet Intern Med DOI: 10.1111/vc0.12043
  34. 34. Thomas, R., Seiser, E.L., Motsinger‐Reif, A., et al. (2011) Refining tumor‐associated aneuploidy through ‘genomic recoding’ of recurrent DNA copy number aberrations in 150 canine non‐Hodgkin lymphomas. Leuk Lymphoma 52:1321–1335.
  35. 35. Breen, M. and Modiano, J. (2008) Evolutionarily conserved cytogenetic changes in hematological malignancies of dogs and humans – man and his best friend share more than companionship. Chromosome Res 16:145–152.
  36. 36. Döhner, H., Stilgenbauer, S., Benner, A., et al. (2000) Genomic aberrations and survival in chronic lymphocytic leukemia. N Engl J Med 343:1910–1916.
  37. 37. Huang, S.H., Kozak, P.J., Kim, J., et al. (2012) Evidence of an oncogenic gammaherpesvirus in domestic dogs. Virology 427:107–117.
  38. 38. Durham, A.C., Pillitteri, C.A., San Myint, M., and Valli, V.E. (2012) Two hundred three cases of equine lymphoma classified according to the World Health Organization (WHO) classification criteria. Vet Pathol 50:86–93.
  39. 39. Miller, C.A., Durham,A.C., Schaffer, P.A., et al. (2015) Classification and clinical features in 88 cases of equine cutaneous lymphoma. J Vet Diagn Invest 27:86–91.
  40. 40. Kelley, L.C. and Mahaffey, E.A. (1998) Equine malignant lymphoma: morphologic and immunohistochemical classification. Vet Pathol 35:241–252.
  41. 41. Meyer, J., DeLay, J., and Bienzle, D. (2006) Clinical, laboratory, and histopathologic features of equine lymphoma. Vet Pathol 43:914–924.
  42. 42. Lohr, C.V. (2013) One hundred two tumors in 100 goats. Vet Pathol 50:668–675.

Precursor B‐cell neoplasms


B‐cell lymphoblastic leukemia/lymphoma


Defining the neoplasm


These are neoplasms of lymphoblasts that produce leukemia or lymphoma (Figure 7.1A,B).1–4 Immunophenotyping is required to determine if they are of B‐ or T‐cell type as the cells look identical via cytology.5–8 B‐cell acute lymphoblastic leukemia (B‐ALL) is a malignant proliferation of lymphoblasts that arise in the bone marrow, there are numerous neoplastic cells in circulation, cytopenias are expected and it has a rapidly progressive course with a fatal outcome. B‐ALL is reported to be more common than T‐ALL in dogs but the data are not uniform.9–12 ALL and CLL are the most common lymphoid leukemias in dogs and they have very different prognoses. Their distinction is important. ALL is rare in cats and CLL is uncommon.13 B‐cell lymphoblastic lymphoma (B‐LBL) is also a tumor of lymphoblasts. It is uncommon to rare: it was not recognized in 600 dogs with lymphoma14 and was less than 2% in a report of 1000 dogs.15 T‐cell lymphoblastic tumors are more common than B‐cell ones in dogs.13–15

Micrograph of blood film.
Micrograph of lymph node aspirate, with arrows pointing to the burst nuclei and an arrowhead pointing to the mature lymphocyte with dense nuclear chromatin.
Micrograph of the bone marrow core.
Micrograph of the marrow core with CD79a IHC stain. Inset: Marrow core with CD3 IHC  displaying the area of lymphoid proliferation (pointed by arrow).

Figure 7.1 B‐cell acute lymphoblastic leukemia (ALL) in a 4‐year‐old dog. (A) Blood film. The nuclei are of intermediate size with finely dispersed chromatin that obscures nucleoli; cytoplasm is minimal but apparent. The lymphocyte count was >50,000/μL, the majority of which were of similar appearance. (B) Lymph node aspirate, nuclei are 1.5 RBC in diameter (intermediate size), uniform. There is one oval nucleus at lower center typical of lymphoblastic lymphoma. The finely dispersed chromatin irregularly obscures the nucleoli. Large nuclei with prominent nucleoli are features of large cell lymphoma not lymphoblastic lymphoma. Cytoplasm is minimal and lightly basophilic. Mature lymphocyte with dense nuclear chromatin is at arrowhead. Burst nuclei (arrow) simply indicate the fragility of the tumor cells and do not help with identification. (C) Bone marrow core. The area at the top of the image has fewer adipocytes and is densely cellular with partial phthisis of normal marrow; a few megakaryocytes remain. (D) Marrow core with CD79a IHC stain. The area of lymphoma (upper right) is uniformly and strongly marked for B cells while the lower area of marrow is not. Note loss of fat cells in the area of lymphoid colonization. There is some nonspecific staining of megakaryocytes subjacent to the infiltrate. Inset: CD3 IHC: The area of lymphoid proliferation is completely negative and lacks fat cells (arrow).


LBL arises in a peripheral lymph node and does not involve the bone marrow or blood, at least not initially. These are defining lines between the tumors, but as each disease progresses their morphologic features start to overlap. ALL (B‐ or T‐cell) will enter lymph nodes and other organs, and LBL (B‐ or T‐cell) will infiltrate organs and bone marrow and leukemia can develop. The latter is referred to as secondary leukemia, leukemic phase of lymphoma, or stage V lymphoma, if it was certain the origin was in solid lymphoid tissues such as lymph nodes.


Some classifications rely on where the greatest tumor burden is suspected to be – bone marrow, nodal, or extranodal. In some cases the distinction is difficult or impossible. Regardless of definitions, at these stages when the tumors are advanced and neoplastic cells can be found in nodes, bone marrow, blood, and organs, the prognosis is poor and survival is measured in days to weeks. B‐ALL and B‐LBL are tumors of lymphoblasts and they cannot be differentiated in advanced disease states.


The lines also become blurred between B‐cell and T‐cell ALL and therefore throughout this section many times only the abbreviation ALL is used. One study reported that the majority of ALL in dogs were B‐cell in origin (>90%),9 another study that T‐cell types were more common (>90%),10 and two other studies reported approximately 60–70% B‐ALL and 30–40% T‐ALL, but a few had atypical phenotypes.11,12 Differences this wide make summarizing data problematic. Level of expertise, antibodies used, and how tumors are defined influence final conclusions. Until we have standardized techniques, antibodies, and definitions there will be conflicting data which makes how to use that data, especially for non‐hematopathologists, difficult in diagnostic settings. Regardless, ALL of B‐cell or T‐cell origin is a lethal disease with short survival times.11,15,16


Also, confounding diagnoses is the use of the term “lymphoblast.” Lymphoblasts can be defined by morphologic, immunologic, or functional parameters. The term lymphoblastic is used in human and veterinary histology to describe a relatively small cell fully capable of cell division. In contrast, in veterinary pathology large dividing lymphoid cells, especially if nucleoli are seen, are loosely called “blasts,” which renders the cytologic recognition of this cell type less specific. True stem cell lymphomas are rare or not recognized in veterinary pathology. Lymphoblastic is a term we have used imprecisely in veterinary pathology, probably most frequently by non‐hematology specialists, for a large immature cell with large nuclei and nucleoli. Many of these were likely large cell lymphomas and not lymphoblastic tumors. The designation lymphoblastic based on morphology should be reserved for cells that are intermediate in size, larger than a mature lymphocyte. They have minimal cytoplasm, a high nucleus‐to‐cytoplasmic ratio, and nuclei that are approximately 1.5–2 red blood cells in diameter and that have uniformly fine dispersed chromatin with no or inconspicuous nucleoli. If evaluated they will have a high proliferative rate. They could be confused with CLL or AML. Lymphoblasts are not large lymphoid cells with open chromatin and prominent nucleoli, these are large cell lymphomas. B‐LBL is uncommon but DLBCL is the most common lymphoma in dogs and common in cats. B‐LBL and DLBCL are different tumors but there are no antibodies or clear morphologic criteria to distinguish these two diagnoses in veterinary medicine. A neoplasm that is positive for CD34 favors ALL over lymphoma but exceptions will exist. CLL are considered CD34 negative. Until markers are developed that recognize B‐LBL it is likely that intermediate to large B‐cell lymphomas will be classified as DLBCL which is the most common lymphoma of dogs.


Reports of “high‐grade lymphoma” of B‐cell type also do not clearly define or separate B‐LBL from DLBCL.9 Therefore to distinguish B‐LBL from DLBCL we rely on subtle morphologic differences in nuclear size and chromatin. Canine B‐cell (<2%) and T‐cell (3%) lymphoblastic lymphomas are uncommon lymphomas.7,9,14,15 They are also uncommon in cats and horses, less than 1%. Large B‐cell lymphomas are common in dogs, cats and horses but these are not lymphoblastic.


Epidemiology, occurrence, and clinical presentation


Lymphoblastic leukemias and lymphomas occur in all species4,17 and are most often seen in cattle17 and dogs.3,9,10,17 ALL can be identified as T‐ or B‐cell, but cases reported in the older literature were not separated into T‐ or B‐cell phenotypes. The typing of acute leukemias in animals needs continued investigations as it is likely that many undifferentiated leukemic cells have been assumed to be lymphoid and actually are of myeloid origin. Broad batteries of antibodies have not been available until recently to verify this distinction (see Myeloid section of this chapter). The morphology of poorly differentiated neoplasms can look similar and should not be relied on to identify cell of origin.


Lymphoblastic leukemia and lymphoma can occur in calves less than 6 months of age.17 Rarely, calves may develop the tumor in utero and it causes dystocia due to marked hepatosplenomegaly. Cases of ALL in the calf are not associated with BLV. These calves are in poor body condition and infarcts in bones are present. Dairy cows with ALL, in contrast to lymphoma, present in good body condition and are first noticed because of a rapid reduction in milk production.17 Dairy bulls also maintain good body condition and the diagnosis is sometimes made from blood taken during an annual examination. The laboratory result of a leukemia is a paradox compared to the apparent excellent health of the bull. In contrast to most other animals, calves that present with ALL are in poor body condition with a rough dull hair coat and enlarged peripheral lymph nodes, up to 10 cm in diameter.17 Calves are often recumbent, likely due to bone pain from multiple areas of marrow infarction (see Figure 10.44).


Horses are rarely identified with ALL. The few cases reported were horses in racing or other disciplines and were examined because of a rapid decline in performance.17


Dogs with LBL are also usually in good condition and anorexia or physical exam abnormalities are recent developments as the disease has an acute onset and progresses rapidly. Only one or two lymph nodes may be enlarged at presentation but multiple nodes quickly become enlarged. Dogs with ALL present from 4 to 12 years of age.3,9,12,17 In dogs the disease seems to have a bimodal age distribution, appearing most often in dogs less than 5 years of age and in old dogs.17 In a study of 210 dogs with hematopoietic neoplasia 51 (24%) were ALL and 61 (29%) were CLL, 92% of ALL were B‐cell and 89% of CLL were T‐cell.9 The average age of dogs with ALL was 7.4 years, there was no sex or specific breed predominance but larger breeds were overrepresented.9 Younger dogs that develop B‐cell ALL or CLL seem to have a worse prognosis than older dogs.18 In one study the golden retriever was overrepresented.12


The incidence of ALL in cats has decreased with increased vaccination for FeLV. ALL occurs in cats but is a rare neoplasm compared to other leukemias and lymphomas.13


Pathology


Blood

The majority of animals with ALL have a marked cytosis of immature lymphoid cells; there may be over 100,000 neoplastic cells/μL (Figure 7.1A).8,9,11,12,16 One study reported a mean of approximately 64,000/μL for ALL and 53,000/μL for 47 dogs with B‐ALL and 109,000/μL for 4 dogs with T‐ALL.9 Some cases may exceed 500,000/μL. The increase in neoplastic cells in the blood is often greater with ALL than in AML. Recognizing leukemia is usually easy as there are so many neoplastic cells visible in a routine complete blood cell count (CBC). However, the neoplastic lymphoid cells of ALL may appear identical to AML, acute undifferentiated leukemia, or even CLL and immunophenotyping or histochemistry will be needed to differentiate them and distinguish B‐ versus T‐cell origin. A bone marrow core or aspirate from ALL patients often reveals marrow packed with neoplastic cells and phthisis of benign cell lines. Tumor cells are >20% of the nucleated cell count.12


Anemia, neutropenia, and thrombocytopenia are expected in the majority of dogs with ALL and AML (bone marrow is involved) and in some cases the pancytopenia is severe. Cytopenia is not predominant in CLL or dogs with lymphoma (greater tumor burden in nodes). Anemia was present in 50/51 (98%) dogs with ALL, the anemia was mild in 6/50 (12%), moderate in 27/50 (54%), severe in 7/50 (14%), and very severe in 10/50 (20%).9 Neutropenia was present in 40/51 (78%), neutrophilia in 5/51 (10%).9 Thrombocytopenia was present in 46/51 (90%). In this study 47 of the 51 (92%) cases of ALL were classified as B‐ALL and 4/51 (8%) were T‐ALL.9 Cytopenia is a characteristic of ALL and is due to myelophthisis and/or cytokines produced by the neoplastic cells. If neutropenia is present the dogs are also anemic and thrombocytopenic.


In the of 210 dogs with hematopoietic tumors 65 were classified as high‐grade lymphoma but with concurrent leukemia.9 Forty‐one of the 65 (63%) were B‐cell type and 24 (37%) were T‐cell type.9 Anemia was seen in 50 of the 65 dogs (77%) and the anemia was mild in 33/50 (66%), moderate in 12 (24%), marked in 4 (8%), and very severe in one dog (2%).9 Neutrophilia was seen in 30/65 (46%), neutropenia in 7 (11%), lymphocytosis in 56 (86%), thrombocytopenia in 26 (40%), and thrombocytosis in 2 (3%). It is likely that these 65 dogs included cases of DLBCL and LBL as these two diseases look similar. Anemia was seen in 32 of the 41 dogs (78%) with B‐cell high‐grade lymphoma, and the anemia was mild in 22 (69%), moderate in 7 (22%), and marked in 3 dogs. Anemia was seen in 18 of the 24 dogs (75%) with T‐cell high‐grade lymphoma, and the anemia was mild in 11 (61%), moderate in 5 (28%), and marked in 2 dogs.9 Neutrophilia was seen in 20 of 41 (49%) dogs with B‐cell lymphoma and neutropenia in 2 (5%). Neutrophilia was seen in 10 of 24 (42%) dogs with T‐cell lymphoma and neutropenia in 5 (21%). Thrombocytopenia was seen in 9 of 41 (22%) dogs with B‐cell lymphoma and in 17 of 24 (71%) dogs with T‐cell high‐grade lymphomas.9 Anemia and thrombocytopenia are likely in dogs with ALL or high‐grade lymphoma and leukemia but neutropenia is more likely with ALL.


Bone marrow and lymph nodes

In ALL the bone marrow is always hypercellular with loss of most of the fat cells and with frequent hemorrhage, perhaps due to thrombocytopenia (Figure 7.1C,D). By definition the marrow contains >20% lymphoblasts. In affected bones the marrow may be 90% cellular and less than 10% adipose. There is filling of the marrow sinuses with the neoplastic cells and eventually there is marked phthisis of all normal marrow cell lines. There is characteristic focal infarction of the marrow that is most apparent in calves with ALL. The necrotic areas usually appear pale and slightly yellow and are usually surrounded by a thin pink halo of hemorrhage (see Figure 10.44). They will only be found if bones are cut longitudinally and will be present in many bones including vertebrae.


In LBL the bone marrow is not neoplastic, at least not in the early stages as the tumor starts in lymph nodes. Usually the submandibular, prescapular, or popliteal lymph nodes are enlarged. If tumor is present in the node biopsied the entire node and its architecture will be effaced by the neoplastic population. Tumor extends beyond the capsule and may be present in perinodal adipose. Most lymph nodes have or eventually will have tumor, and aspirational cytology is diagnostic for lymphoma (Figure 7.1B). The mitotic count is high, usually 10 or more mitoses/400× field. However the mitotic figures are not as obvious in LBL as in other lymphomas because the metaphase figures have hazy or poorly defined features. Nuclear chromatin is dispersed and deeply stained, which obscures nucleoli. Chromatin does not aggregate and form cleared regions as in some other lymphomas.


There are several characteristic changes that assist recognition of LBL. One is the high number of mitotic figures for a relatively small to intermediate cell type. There are usually few or no tingible body macrophages, which is in contrast to high‐grade B‐cell lymphomas and Burkitt’s‐like lymphomas, which have numerous tingible body macrophages. The node appears more deeply stained than usual because of the prominent and dispersed nuclear chromatin. It is important to examine the node near the capsule or outer edge of the tissue section since this characteristic appearance of dispersed chromatin, rather than aggregated chromatin with open nuclei, is lost in deeper areas of the node due to delayed fixation. In this respect a Tru‐Cut needle biopsy is ideal because of immediate and uniform fixation. Cell and nuclear details are seen well in cytologic preparations. Some nodes in B‐ALL can be atrophic or irregularly enlarged, with taut and thinned capsules and compression of the peripheral sinus. Atrophy of the thymus may be present with disproportionate loss of cortex.


Spleen

When the spleen is involved with either LBL or ALL it is diffusely and symmetrically enlarged, with filling of the sinus areas by neoplastic cells. There is atrophy of the periarteriolar lymphoid sheaths and an absence of normal germinal centers. In aggressive leukemias such as ALL the tumor cells may colonize subendothelial areas of large muscular veins, which indicates the neoplastic cells are intravascular. When there is colonization of splenic endothelium there is usually tumor in the liver. The majority of the tumor is periportal or perivascular with LBL, but if there are neoplastic cells in circulation with either LBL or ALL then hepatic sinusoids will also contain tumor cells. In ALL hepatomegaly is often diffuse and an FNA of liver from any location will establish the diagnosis of lymphoma.


Cellular morphology


Cytologically, lymphoblastic tumors are intermediate in size with nuclei about 1.5 red cells in diameter. However, some nuclei are long and oval‐shaped and could be called large cells in any other subtype of lymphoma. The chromatin does not have coarse aggregations; it is dispersed and this pattern makes visualization of the multiple small nucleoli that are present difficult (Figure 7.1A). In histologic preparations this feature should be examined at the periphery where fixation is best. Cytoplasm is minimal and lightly basophilic. Nuclei reside close together and there is a high mitotic count with 10 or more mitoses/400× field. The mitotic figures are unique and are not the characteristic metaphase type that is easy to recognize. There will be numerous apoptotic cells in each field and these should not be counted as mitotic figures. Having provided these features, it is a reality that many cases of ALL will look identical to AML or even CLL. Morphology can be used but should not be the sole diagnostic criterion. All the case data need to be interpreted, including onset of disease (rapid or indolent), phenotype, broad batteries of antibodies, and which cytopenias are present if any.


Cytochemistry and immunohistochemistry


Immunohistochemistry or flow cytometry are needed to identify B‐ versus T‐cell immunophenotype. B‐ALL in tissues is routinely positive with the B‐cell markers CD79a, CD20, and CD21 after antigen retrieval is carried out on FFPE samples.2 The cells in blood can also be effectively phenotyped by flow cytometry or fixing unstained blood films for up to a minute in formalin. This adheres the cells to the glass slide and permits a wide range of stains to be used after heat‐activated antigen retrieval, including CD79a, CD3, and CD45.7 If only stained films are available it is possible to decolorize Wright–Giemsa‐stained slides with acetone before or after formalin fixation.1,7 A large battery of antibodies can be used in flow cytometry and these may be needed to distinguish B‐ versus T‐cell ALL and to distinguish both from AML and CLL. Detection of cytoplasmic antigens by flow cytometry requires permeabilization of cell membranes, which makes the technique prone to artifact. Therefore, flow cytometry data that use antibodies directed at cytoplasmic antigens should be interpreted cautiously. Flow cytometry and IHC identify phenotypes but do not differentiate neoplastic cells from reactive cells. However, the homogeneity of patterns in tissue sections or in flow cytometry help to identify clonal versus heterogeneous cellular proliferations.


A 2015 report used alkaline phosphatase (ALP) to distinguish lymphoma and CLL (all were negative) from myeloid and myelomonocytic leukemias (all positive).10 ALP did not completely distinguish ALL from AML as it was weakly positive in about one‐third of the cases of ALL.10 AML is usually CD34 positive, ALL is variable but CLL is CD34 negative. Several genetic changes have been identified in humans but in animals the diagnosis rests on morphologic recognition of lymphoblastic cells in cytology and histology combined with immunophenotyping and histochemistry.


See T‐cell lymphoblastic leukemia/lymphoma in this chapter for details on IHC. If the neoplasm is positive to CD79a in histology or cytology it is considered B‐cell. However, some AMLs may be positive to CD79a and some B‐ALL may be CD79a negative. Therefore consider using more than one B‐cell antibody and technique. A dog with tumor cells in circulation that are immunoreactive to CD79a, CD20, CD21, and CD34 and negative to CD3 via flow cytometry is considered B‐ALL. B‐CLL is possible but it is uncommon (80% are T‐CLL), cells are CD34 negative and the onset and progression of CLL should be indolent and neutropenia absent. Another potential problem is that CD79a may bind with the nucleus and this cannot easily be differentiated from cytoplasmic binding and therefore could cause a false‐positive. If this is a concern use IHC or immunocytochemistry (ICC) to determine the staining pattern of CD79a (membrane staining is a positive reaction) and/or consider other antibodies.


Differential diagnosis


The diagnosis of lymphoma is usually easy with cytology or histopathology as neoplastic cells are numerous in the blood or have effaced or at least filled lymph nodes and/or bone marrow. Categorizing as lymphoblastic versus large cell is subjective and is based on size of cell, nuclei, and prominence of nucleoli. The larger the cells and nuclei and the more prominent the nucleoli the more likely the diagnosis is a large cell lymphoma and not lymphoblastic. Most large cell lymphomas in dogs and cats are DLBCL. The high‐grade B‐cell lymphomas and Burkitt’s‐like lymphoma will have numerous tingible body macrophages, which are low or absent in LBL. The node appears more deeply stained than usual because of the prominent and dispersed nuclear chromatin. ALL has an immature lymphocytosis and CLL has mature lymphocytosis, ALL is usually CD34 positive and CLL is CD34 negative. Neutropenia and other cytopenias are expected with ALL and not CLL. The onset of disease and the duration of survival are entirely different. B‐cell ALL is differentiated from T‐cell ALL by immunophenotyping techniques, but techniques and antibodies can influence results. Both B‐ and T‐cell ALL can look similar to myeloid leukemia especially if the cells in AML have few or no granules. If an AML does not demonstrate maturation, that further confounds the distinction. See the Myeloid neoplasms section of this chapter for complete antibody profiles that could be used.


If leukemic cells react with T‐ or B‐cell markers it is considered lymphoid. Both B‐ and T‐cell ALL should be negative for mature myeloid markers. If a tumor was CD34 positive and CD45 positive, but negative for B‐cell and T‐cell markers then the diagnosis from immunophenotyping is AML. AML should also be positive for myeloid markers CD11b, CD11c, and/or CD14 on flow cytometry, and/or histochemistry positive with myeloperoxidase, chloroacetate esterase, or Sudan black B. Histochemistry for ALP will be negative for lymphoma and CLL and positive for monocytic myeloid leukemias and about one‐third of ALL.10 If a leukemia was CD34 positive, ALP positive, and had morphologic features of myeloid lines then AML is favored over ALL. See T‐ALL in this chapter for further use of histochemistry and IHC to differentiate these neoplasms.


Prognosis


B‐ALL and T‐ALL are malignant tumors and some of the most aggressive cancers in veterinary medicine. Dogs with ALL and AML have short survival times measured in days to weeks.11,15,16,18 Although the numbers of dogs in the different groups were small, those with B‐ALL had median survival time (MST) of 8 days, range 5–46 days; dogs with T‐ALL had MST of 10 days, range 4–120; dogs with AML had MST of 10 days, range 3–73 days.11 The clinical features of advanced disease include lymphocytosis >30,000/μL, cytopenias, neoplastic cells 50–100% in the marrow, systemic involvement of liver, spleen and other organs and signs of hepatic or renal failure. CD34 is expressed by early precursors, dogs with CD34‐positive leukemic cells have short survival times (MST of 16 days).11,16 Dogs with a neutrophil count in the reference interval have longer survival times and perhaps the absence of anemia is a more favorable indicator.11 Regardless, ALL is a lethal disease.


Dogs with B‐cell lymphocytosis (CD21 positive) were separated by the size of the neoplastic cells.16 Dogs with large cell B‐lymphocytosis had an MST of 130 days and the MST was not reached for B‐lymphocytosis of small cells. Cell size was determined by flow cytometry on the peripheral blood. This study did not try to distinguish leukemias from stage V lymphoma or to classify the neoplasms with the REAL/WHO system. Therefore there were likely dogs with different diseases in each group (e.g. B‐ALL, B‐LBL with leukemia, B‐CLL, DLBCL, etc.). This is not a criticism; in fact it makes identifying groups of dogs easier as there are fewer parameters evaluated and histology is not needed for prognostic purposes. See Precursor T‐cell lymphoblastic leukemia/lymphoma in this chapter for additional information.


With advancing ALL an FNA of any area of the liver will provide tumor cells similar to those in the blood. With LBL, more favorable signs at diagnosis are probable if the tumor is limited to a single node, and there is an absence of systemic disease and retention of appetite and activity. With advancing disease multiple nodes are enlarged, liver and spleen is involved, leukemia can be present, the animal is lethargic, and there are signs of systemic disease.


References



  1. 1. Caniatti, M., Roccabianca, P., Scanziani, E., et al. (1996) Canine lymphoma: immunocytochemical analysis of fine‐needle aspiration biopsy. Vet Pathol 33:204–212.
  2. 2. Jubala, C.M., Wojcieszyn, J.W., Valli, V.E., et al. (2005) CD20 Expression in normal canine B cells and in canine non‐Hodgkin lymphoma. Vet Pathol 42:468–476.
  3. 3. Matus, R.E., Leifer, C.E., and MacEwen, G. (1983) Acute lymphoblastic leukemia in the dog: A review of 30 cases. J Am Vet Med Assoc 183:859–862.
  4. 4. Pui, C.H., Relling, M.V., and Downing, J.R. (2004) Acute lymphoblastic leukemia. N Engl J Med 350:1535–1548.
  5. 5. Reggeti, F. and Bienzle, D. (2011) Flow cytometry in veterinary oncology. Vet Pathol 48:223–235.
  6. 6. Reichard, K.K., Kang. H., and Robinett, S. (2011) Pediatric B‐lymphoblastic leukemia with RUNX1 amplification: clinicopathologic study of eight cases. Mod Pathol 24:1606–1611.
  7. 7. Valli, V., Peters, E., Williams, C., et al. (2009) Optimizing methods in immunocytochemistry: one laboratory’s experience. Vet Clin Pathol 38:261–269.
  8. 8. Vernau, W. and Moore, P.F. (1999) An immunophenotypic study of canine leukemias and preliminary assessment of clonality by polymerase chain reaction. Vet Immunol Immunopathol 69:145–164.
  9. 9. Tasca, S., Carlil, E., Caldin, M., et al. (2009) Hematologic abnormalities and flow cytometric immunophenotyping results in dogs with hematopoietic neoplasia: 210 cases (2002–2006). Vet Clin Pathol 38:2–12.
  10. 10. Stokol, T., Schaefer, D.M., Shuman, M., et al. (2015) Alkaline phosphatase is a useful cytochemical marker for the diagnosis of acute myelomonocytic and monocytic leukemia in the dog. Vet Clin Pathol 44:79–93.
  11. 11. Novacco, M., Comazzi, S., Marconato, L., et al. (2015) Prognostic factors in canine acute leukaemias: a retrospective study. Vet Comp Oncol DOI: 10.1111/vc0.12136
  12. 12. Adam, F., Villiers, E., Watson, S., et al. (2009) Clinical pathological and epidemiological assessment of morphologically and immunologically confirmed canine leukaemia. Vet Comp Oncol 7:181–195.
  13. 13. Vezzali, E., Parodi, A.L., Marcato, P.S., and Bettini, G. (2009) Histopathologic classification of 171 cases of canine and feline non‐Hodgkin lymphoma according to the WHO. Vet Comp Oncol 8:38–49.
  14. 14. Ponce, F., Marchal, T., Magnol, J.P., et al. (2010) A morphological study of 608 cases of canine malignant lymphoma in France with a focus on comparative similarities between canine and human lymphoma morphology. Vet Pathol 47:414–433.
  15. 15. Valli, V.E., Kass, P.H., Myint, M.S., and Scott, F. (2013) Canine lymphomas: association of classification type, disease stage, tumor subtype, mitotic rate, and treatment with survival. Vet Pathol 50:738–748.
  16. 16. Williams, M.J., Avery, A.C., Lana, S.E., et al. (2008) Canine lymphoproliferative disease characterized by lymphocytosis: Immunophenotypic markers of prognosis. J Vet Intern Med 22:596–601.
  17. 17. Valli, V.E. (2007) Precursor B‐cell lymphoblastic lymphoma and lymphoblastic leukemia. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 124–137.
  18. 18. Comazzi, S., Gelain, M.E., Martini, V., et al. (2011) Immunophenotype predicts survival time in dogs with chronic lymphocytic leukemia. J Vet Intern Med 25:100–106.

Mature peripheral B‐cell neoplasms


B‐cell chronic lymphocytic leukemia/small cell lymphocytic lymphoma/prolymphocytic leukemia


Defining the neoplasm


These are neoplasms of small lymphocytes, either B‐ or T‐cell types that result in leukemia or lymphoma (Figure 7.2). Immunophenotyping is needed to distinguish B versus T, but if cytoplasmic granules are present then the tumor is of T‐cell type. In humans, CLL is a B‐cell tumor (>95%),1,2 but in dogs only 7 of 61 cases (11%),3 26% of 73 cases,4 or 17 of 435 were of B‐cell type and 1 of 18 cats was B‐cell.6 Approximately 80% (75–90%) of canine CLL cases are T‐cell,3,4 95% (17/18) of cats are T‐cell, and 94% of these were T‐helper CD3+/CD4+/CD8− phenotype.6 Aberrant phenotypes have been reported for dogs.4,5 Horses have been inadequately studied. Cows with CLL are usually of B‐cell type. If tumors are not phenotyped then the diagnosis is CLL.

Micrograph of chronic lymphocytic leukemia (CLL) in feline (top) and light scatter plot (bottom).
Micrograph of the marrow core of a dog with chronic lymphocytic leukemia.
Magnified image of the marrow core with megakaryocyte found on the upper right.

Figure 7.2 Chronic lymphocytic leukemia (CLL), feline. (A) Nine‐year‐old cat with a lymphocyte count of approximately 11,000/μL. The lymphocytes are small to intermediate size with irregular shaped or indented nuclei. A few cytoplasmic granules are faintly apparent (see Figures 7.39 and 7.40). Flow cytometric analysis indicates that lymphocytes (light scatter plot, arrow) express CD4 and CD3 (T cells) and do not express CD8 and CD21. In healthy cats there would be a mixture of CD4‐, CD8‐, and CD21‐expressing lymphocytes in blood. Neutrophils (light scatter and control plots, arrowhead) do not express CD3, CD4, CD8, or CD21. Most CLL in cats is of T‐cell type, as is this case, and cells more commonly express CD4 than CD8. This cat tested negative for feline leukemia virus. Platelets are adequate in this cat, anemia is expected, but neutropenia is not usually seen with CLL. (B) B‐CLL, dog. Marrow core, the marrow is completely infiltrated, few fat cells remain, bone is of normal density and volume. Typically dogs with B‐CLL will have a marked lymphocytosis and paraproteinemia. (C) Marrow core, higher magnification. The marrow is heavily infiltrated with mature lymphoid cells of small to intermediate size, nuclei are about 1.5 RBC in diameter and nucleoli are not obvious. There is very little stromal proliferation. IHC is needed to identify phenotype, most CLL are T‐cell type. A young megakaryocyte is in the upper right.


B‐cell CLL is a neoplasm of the bone marrow and is characterized by marked lymphocytosis (leukemia) of small mature B lymphocytes. Animals often have few or no clinical signs and the disease has a slow but persistent progression. Spleen and liver are usually enlarged due to tumor cells but lymph nodes may be normal size. B‐cell small lymphocytic lymphoma (SLL) is a tumor of small mature B lymphocytes involving lymph nodes, spleen, and solid organs but leukemia is not present or neoplastic cells in blood are few. SLL is an uncommon lymphoma in dogs and is much less common than CLL. It has a slow rate of progression.


B‐cell prolymphocytic leukemia (PLL) is related but is an earlier stage of maturation than mature lymphocytes and is more aggressive. In terms of organ distribution it is a mirror image of CLL. Morphologic identification is accomplished better with cytology in which the nuclei appear more “immature,” are larger, the chromatin is in aggregates joined by fine chromatin bands, and the nucleolus is prominent (Figure 7.3). Since the prognosis and therapy differ from those in CLL the distinction is clinically important.

Micrograph of prolymphocytic leukemia (PLL) in a cow.

Figure 7.3 Prolymphocytic leukemia (PLL) in a cow. The cells of PLL have a characteristic chromatin pattern that is composed of multiple aggregates of chromatin that are joined by fine chromatin bands. Nucleoli are small and inconspicuous, and there is abundant lightly stained cytoplasm. B‐ versus T‐cell phenotype cannot be determined from cytology.


Epidemiology and occurrence


Lymphoma and leukemia of small mature lymphocytes are indolent, B‐cell and T‐cell. They are slowly progressive neoplasms that may involve bone marrow or peripheral tissues or both. They are considered neoplasms of accumulation rather than proliferation. Aggressive variants have been identified, usually associated with atypical phenotypes.4,5 B‐cell CLL is uncommon in domestic animals, although reported in dogs, cats, and other species.4,7,8–12 In the American College of Veterinary Pathologists (ACVP) study of 1000 cases of canine lymphoma there were 8 cases of B‐cell SLL. In humans there is a strong male predominance but there is no sex or no breed predominance in animals. In animals, as in humans, the tumor is found most often on routine examination of blood. Cells of PLL type are seen in humans,13 mature cows, and dogs, but are relatively rare.9


Clinical presentation


Most cases are asymptomatic and the disease is discovered from a CBC taken at an annual exam that reveals lymphocytosis. Weight loss is reported in approximately 50% of the cases. Cats and dogs with CLL and SLL are usually over 5 years of age, the mean age for 134 dogs with CLL (B and T) was 10 years, the range was 1.5–19 years3,4 and 17 cats with T‐cell CLL the mean age was 12.5 years.6 Only one cat in this series had B‐cell CLL. Animals in an accelerated phase of the disease occasionally present with epistaxis, likely due to thrombocytopenia; some will have diarrhea and vomiting. In dairy cattle the most common clinical problem is a drop in milk production and feed consumption.


Pathology


Blood, bone marrow, lymph nodes, and spleen

The peripheral blood count and cytologic review of a film is diagnostic, as the nucleated cell count will be 100,000–400,000/μL and >90% will be lymphocytes (Figure 7.2). Reported mean lymphocyte counts are approximately 100,000/μL but the ranges are wide, from near reference interval to >1,000,000/μL.3,4 One study reported a mean of 137,000/μL for B‐CLL and 60,000/μL for T‐CLL LGL type.3 In cats the total median lymphocyte count reported was 34,000/μL (>90% of cases were T‐cell),6 and counts may be as high as 400,000/μL. The higher the lymphocytosis the more likely the diagnosis is CLL; however, there are non‐neoplastic causes of lymphocytosis in all species that are more common than CLL.


Anemia to some degree is present in approximately 60–75% of dogs and platelets are decreased in 15–25%.3,4 The anemia is non‐regenerative and can be mild, moderate, or severe; the latter is least common. Neutropenia is not present in dogs with CLL and this is a distinguishing feature from ALL and AML in which neutropenia is expected. B‐CLL originates in the bone marrow and numerous bones will be infiltrated with tumor cells in dogs, less so in cats (Figure 7.2B,C). Neoplastic foci are identified by an absence of fat cells and in these regions there are small uniform lymphocytes. Some cases will have bone marrow nearly filled by neoplastic cells and others may be as low as 20% lymphocytes.


Core samples are more reliable to identify CLL than aspirates.6 T‐CLL is likely of splenic origin. A consistent finding in CLL is hepatosplenomegaly. Animals with SLL are not leukemic or the neoplastic cells in circulation are at a low level because the bone marrow is not involved or only focally so. In SLL all organs and nodes may eventually be involved as there is slow progression and gradual dissemination of the neoplastic cells.


The lymph nodes in cats and dogs with CLL can be atrophic or enlarged. With SLL the nodes are neoplastic and the peripheral sinus and medullary cord regions are compressed by the diffuse cortical expansion of neoplastic cells. The tumor forms solid sheets that appear densely stained because the nuclei are small and close together. Neoplastic cells have little cytoplasm. Lymphatic vessels within and close to nodes may be dilated and packed with neoplastic cells.


The spleen in B‐cell CLL has atrophy of the thymic‐dependent periarteriolar lymphoid sheaths. There is scattered infiltration of the sinus areas and in later stages of the disease these may become confluent. There can be fading germinal centers with hyalinized centers. Hematopoiesis is restricted to adjacent non‐affected red pulp with megakaryocytes appearing to be the most persistent of the hematopoietic cells. With progression there is colonization of subendothelial areas of the large muscular veins and colonization of hepatic sinusoids. With SLL there is an absence of germinal centers and the infiltrations of the spleen are larger and more solid and coalescing.


Animals with B‐cell CLL may have a macroglobulinemia like the gammopathy of the Waldenstrom type in humans, but it is uncommon. Routine chemistry may reflect this change by an unusually low albumin‐to‐globulin ratio. Serum electrophoresis will verify the presence of a monoclonal gammopathy, usually IgM. When present, macroglobulinemia suggests that the tumor is of B’cell origin and is more likely of a prolymphocytic type. Although immunophenotyping is needed to definitively separate B‐ and T’cell CLL, the presence of a monoclonal gammopathy suggests B’cell and granular lymphocytes favor T‐cell.


Neoplastic cells

B‐CLL and SLL cells look like mature lymphocytes and are identified as B‐cell with immunophenotyping. Small lymphocytes have nuclei 1–1.5 RBC in diameter with a narrow rim of cytoplasm and dense nuclear chromatin. It is the marked lymphocytosis that suggests neoplasia (Figure 7.2). Dogs usually have cells of this type or, like humans, can have larger cells, especially if they enter blast crisis with nuclei 2 RBC in diameter and relatively abundant lightly basophilic cytoplasm. Dogs also have a CLL in which the cells are large granular T lymphocytes. The cytologic diagnosis of SLL can be problematic as the cells are mature. Absence of plasma cells, uniform small lymphocytes from multiple enlarged lymph nodes, and hepatosplenomegaly all support the diagnosis. Clonality and immunophenotyping may be needed.


The cells of PLL are larger than those of CLL or SLL and the chromatin pattern is distinctive (Figure 7.3). PLL cells have round and moderately irregular nuclei that are approximately 2.0 RBC in diameter. The chromatin distribution is characteristic, with large, densely stained chromocenters that are about one‐third the diameter of a red cell apart but joined by narrow chromatin bands. A distinctive feature is the parachromatin clear areas that tend to surround large chromocenters. There is mild anisokaryosis; the larger cells have small but relatively prominent central nucleoli that may be multiple. The cytoplasm is relatively abundant and lightly stained (Figure 7.3). These cells have a low but consistent mitotic count in tissues with 0–2 mitoses/400× field.


Other organs

Both CLL and SLL have areas of infiltration in and around hepatic portal tracts. Lymphocytes in the hepatic sinusoids are usually present with CLL. Dilated veins containing neoplastic cells may be present in lung, kidney, central nervous system and eyes in CLL.


Cytochemistry and immunohistochemistry


B‐cell CLL and SLL are differentiated from the more common T’cell phenotype by IHC or flow cytometry (Figure 7.2). The B‐cell CLL and SLL cells are positive in tissues with CD79 alpha, CD20, or CD21 and are negative with CD3. B‐cell CLL cells are negative with CD5 (unlike humans) and should express CD1c (95%) and CD1a (78%).4 Cells from animals and humans with CLL are negative for CD34, which is expected to be positive in ALL and AML. CD34 is a surface glycoprotein expressed on hematopoietic stem cells (and other cells) and is useful to distinguish acute and chronic leukemia. It should be negative in CLL and most lymphomas and positive in ALL and AML.


All of these neoplastic cells can have stages or types in which the morphology of the cells are identical or so similar that phenotyping, especially with flow cytometry and multiple antibodies, may be necessary to correctly identify the neoplasm. Neutropenia is not present in CLL and is a feature of acute leukemias. Acute leukemias are symptomatic and have a rapid course with a fatal outcome, while CLL is often asymptomatic and has a clinical course of months to years. Differential diagnoses and how to distinguish them is discussed in the section on T‐cell chronic lymphocytic leukemia.


Tumor cell transformation and prognosis


These are indolent lymphomas. Animals may live for years without treatments. Since these tumors are slowly progressive and clinically occult they are usually not diagnosed until the tumor has been present for an extended period, perhaps 1–2 years. Despite the slow progression a diagnosis of leukemia is considered to be more serious than a lymphoma. Humans with SLL have a 2‐year survival of 60–70% and a 5‐year survival of 50–70%.


A monoclonal antibody has been used for treatment in humans.14,15 Dogs are treated with a variety of chemotherapies that may include prednisolone. Some are treated with only prednisolone and others are not treated.


In animals, CLL and SLL can undergo an accelerated phase similar to these diseases in humans. When this occurs in the tumor’s life is not known, but the cells gain in mitotic rate and the cells enlarge to prolymphocytic type (Figure 7.3). The next stage is similar to Richter’s syndrome, in which the small cells evolve into large cell lymphoma.11,13,16 Further progression in humans is classified as blastic transformation, plasmacytoid transformation, and, finally, immunoblastic transformation.9–11,15 The diagnostic features taken to predict the progression of human CLL are the clinical stage, lymphocyte count and doubling time, pattern of marrow involvement, cytogenetic abnormalities, and atypical immunophenotype.1,16 These methods are not available on a diagnostic basis in veterinary medicine.


The distinction of blast transformation in CLL or SLL from a large cell lymphoma is not clear unless there was documentation of a prior small cell CLL or SLL. Most canine large cell lymphomas are B‐cell, multicentric, and some may have concurrent leukemia so the distinction may be difficult when either disease is seen at one point in time. Dogs with CD21‐positive lymphocytosis have different survival times that can be separated by size of the neoplastic cells as determined from flow cytometry.17 Dogs with large cell B‐lymphocytosis had MSTs of 130 days versus an MST that was not reached for small cell lymphocytosis (>1–2 years) as determined by flow cytometry.17 See T‐cell chronic lymphocytic leukemia section in this chapter and Appendix of this book for more information on survival and prognosis.


References



  1. 1. Admirand, J.H., Knoblock, R.J., Coombes, K.R., et al. (2010) Immunohistochemical detection of ZAP70 in chronic lymphocytic leukemia predicts immunoglobulin heavy chain gene mutation status and time to progression. Mod Pathol 23:1518–1523.
  2. 2. Chiorazzi, N., Rai, K.R., and Ferrarini, M. (2005) Chronic lymphocytic leukemia. N Engl J Med 352:804–815.
  3. 3. Tasca, S., Carlil, E., Caldin, M., et al. (2009) Hematologic abnormalities and flow cytometric immunophenotyping results in dogs with hematopoietic neoplasia: 210 cases (2002–2006). Vet Clin Pathol 38:2–12.
  4. 4. Vernau, W. and Moore, P.F. (1999) An immunophenotypic study of canine leukemias and preliminary assessment of clonality by polymerase chain reaction. Vet Immunol Immunopathol 69:145–164.
  5. 5. Comazzi, S., Gelain, M.E., Martini, V., et al. (2011) Immunophenotype predicts survival time in dogs with chronic lymphocytic leukemia. J Vet Intern Med 25:100–106.
  6. 6. Campbell, M.W., Hess, P.R., and Williams, L.E. (2012) Chronic lymphocytic leukaemia in the cat: 18 cases (2000–2010). Vet Comp Oncol 11:254–264.
  7. 7. Gerou‐Ferriani, M., McBrearty, A.R., Burchmore, R.J., et al. (2011) Agarose gel serum protein electrophoresis in cats with and without lymphoma and preliminary results of tandem mass fingerprinting analysis. Vet Clin Pathol 40:159–173.
  8. 8. Harvey, J.W., Terrell, T.G., Hyde, D.M., and Jackson, R.I. (1996) Well‐differentiated lymphocytic leukemia in a dog: Long‐term survival without therapy. Vet Pathol 18:37–47.
  9. 9. Valli, V.E. (2007) Mature (peripheral) B‐cell neoplasms. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 142–161.
  10. 10. Valentine, B.A. and McDonough, S.P. (2003) B‐cell leukemia in a sheep. Vet Pathol 40:117–119.
  11. 11. Su, Y.C., Wu, W.M., Wu, M.F., and Chiang, B.L. (2001) A model of chronic lymphocytic leukemia with Ritcher’s transformation in severe combined immunodeficiency mice. Exp Hematol 29:1218–1225.
  12. 12. Vezzali, E., Parodi, A.L., Marcato, P.S., and Bettini, G. (2009) Histopathologic classification of 171 cases of canine and feline non‐Hodgkin lymphoma according to the WHO. Vet Comp Oncol 8:38–49.
  13. 13. Reininger, L., Bodor, C., Bognar, A., et al. (2006) Richter’s and prolymphocytic transformation of chronic lymphocytic leukemia are associated with high mRNA expression of activation‐induced cytidine deaminase and aberrant somatic hypermutation. Leukemia 20:1089–1095.
  14. 14. Mavromatis, B. and Cheson, B.D. (2003) Monoclonal antibody therapy of chronic lymphocytic leukemia. J Clin Oncol 21:1874–1881.
  15. 15. Zhou, Y., Tang, G., Medeiros, L.J., et al. (2012) Therapy‐related myeloid neoplasms following fludarabine, cyclophosphamide, and rituximab (FCR) treatment in patients with chronic lymphocytic leukemia/small lymphocytic lymphoma. Mod Pathol 25:237–245.
  16. 16. Leuenberger, M., Frigerio, S., Wild, P.J., et al. (2010) AID protein expression in chronic lymphocytic leukemia/small lymphocytic lymphoma is associated with poor prognosis and complex genetic alterations. Mod Pathol 23:177–186.
  17. 17. Williams, M.J., Avery, A.C., Lana, S.E., et al. (2008) Canine lymphoproliferative disease characterized by lymphocytosis: Immunophenotypic markers of prognosis. J Vet Intern Med 22:596–601.

Plasmacytoma, plasmablastic lymphoma, lymphoplasmacytic lymphoma, and myeloma‐related disorders


Plasmacytoma, plasmablastic lymphoma, and lymphoplasmacytic lymphoma (LPL) are neoplasms of mature, differentiated B lymphocytes, and the neoplastic cells look like or are plasma cells. Plasmacytomas occur outside of lymph nodes, and plasmablastic lymphoma and LPL originate in nodes and may be found in organs. Very rarely does a plasmacytoma arise in a node and equally rare is a primary osseous plasmacytoma. There are other B‐cell neoplasms reported in humans, including lymphoplasmacytoid lymphoma and immunocytoma, however, comparable tumors of similar morphology, immunophenotype, and biologic behavior have not been characterized in animals.1,2


In the ACVP study of 1000 cases of canine lymphoma there were 15 cases of plasmacytoma, 7 cases of plasmablastic lymphoma and one case classified as a B‐cell plasmacytoid lymphoma.3 It is not clear if the plasmacytomas were extramedullary.3 In the 600 dog study there were 6 plasmacytoid B‐cell, 9 lymphoplasmacytic B‐cell, and 19 plasmacytoid T‐cell lymphomas.4


There also is a disease category called myeloma‐related diseases (MRD) that is used for similar tumors and disease entities in cats and humans.5,6 MRD is a group of tumors in cats that are derived from plasma cells or immunoglobulin‐secreting B lymphocytes.5 How tumors are defined influences prevalence and the numbers of tumors in subjective categories.


Plasmacytoma is the most common tumor in this group in dogs by a wide margin. Plasmacytomas are a proliferation of differentiated B lymphocytes that originate primarily in soft tissues, oral and subcutaneous locations and rarely in nodes, organs, or bone. Plasmacytomas are common tumors in dogs that were misdiagnosed in the older literature and are reported under a wide variety of names, including reticulum cell sarcoma.7


Plasmablastic lymphoma and LPL are uncommon tumors which develop from B lymphocytes in lymph nodes and/or organs. MRD is an umbrella term for a group of about six diseases that arise from B lymphocytes. It has been used to define tumors in cats. Within this group are plasmacytomas, cutaneous and extracutaneous and B‐lymphocyte tumors that secrete immunoglobulins. Some of these latter tumors may be classified as plasmablastic lymphoma or LPL by others, especially if organs were involved and/or leukemia was present. As a generalization the plasma cell–like neoplasms in cats that are not well differentiated are aggressive.


Clinical presentation and occurrence


Plasmacytomas are common in dogs, are seen in cats, and are unusual in horses and cattle.1,2,5,6,8–11 They are described in more detail in Chapter 5. Plasmacytomas are usually solitary, benign tumors in the skin, subcutis, or gastrointestinal tract and most do not recur following excision. Solitary tumors have been identified in the liver of dogs and cats and rarely have been associated with paraneoplastic syndromes of hypoglycemia, hypercalcemia, or paraproteinemia. Paraproteinemia is a feature of myeloma, MRB, and LPL. Plasmacytomas are also found in spleen, kidney, and intestines, but in these locations consider LPL, plasmablastic lymphoma, and MRD. Plasmacytomas are well described in cats. They are divided into cutaneous and extracutaneous and compared to MRD.5,6,8 In cats the data suggest they originate in soft tissues but later may be present in bones, which is a different pathogenesis from that proposed for human MRD.5 Cats with MRD are FeLV and FIV negative. Some of the classifications and categories in MRD are similar to other diagnoses that use different nomenclature.


Regardless of names, several important conclusions from these investigations in cats are that cutaneous plasmacytomas respond well to excision, well‐differentiated tumors are associated with survival times of approximately 250 versus 14 days for poorly differentiated tumors, there is good concordance between cytology and histopathology, many tumors will not label with CD79a, mitotic figures are not frequent, and giant cells and giant nuclei may be seen in more aggressive tumors, as they are in dogs and humans.5


Most plasmacytomas are benign but infrequently cutaneous plasmacytomas behave aggressively and shorten the life of the dog or cat. When plasmacytomas are present in the aerodigestive system of dogs they are almost always solitary and benign. However, those that occur in the stomach or rectum may be clonal and aggressive and may cause vomiting and weight loss. When multiple tumors with plasma cell features are present in the intestines, and especially if they are also found in liver, spleen, or lymph nodes, the diagnosis in dogs is more likely plasmablastic lymphoma or LPL. Plasmacytomas in the rectal area usually present with some history of difficulty in defecation and are accompanied by recurrent bloody stool.10 The potential for plasmacytomas in the gastric and rectal areas to be aggressive necessitates wide surgical excision of tumors in these areas.9,10,12 Plasmacytomas in dogs are rarely associated with monoclonal gammopathy. If a monoclonal gammopathy and hypercalcemia are present the dog is more likely to have multiple myeloma.


Cats with MRD are reported to have paraproteinemia and a few have had hypercalcemia.5,6 Degree of differentiation is useful to predict survival times in cats. Well‐differentiated tumors had MSTs of 254 days versus 14 days for poorly differentiated tumors.5 Well‐differentiated tumors are more common.


LPLs are rare tumors in animals and humans. In humans this lymphoma as well as other lymphoproliferative diseases are associated with Waldenstrom‐type macroglobulinemia (IgM) that in about 20% of cases may have cryoglobulinema.1,2 Clinically this protein causes hyperviscosity that results in cold extremities and may require plasma exchange to reverse the syndrome. In animals this syndrome is exceedingly rare but has been associated with multiple myelomas that produced IgM. A category of MRD includes this syndrome.


Plasmablastic lymphomas are uncommon tumors seen in mature dogs and cats. The animals usually present with severe systemic disease and have profound weight loss and a poor hair coat. On clinical examination multiple nodes are enlarged. Leukemic manifestations are rare but focal lesions in the spleen and perivascular cuffs in the liver occur. The tumor tends to be aggressive. Others may conceivably classify these as plasmablastic MRD or plasmablastic large cell lymphoma.


Light microscopic features


The features common to plasmacytoma, LPL, plasmablastic lymphoma, and MRD are that they are composed of tumors with plasma cell features and they are derived from B lymphocytes. Histologic and cytologic features of plasmacytomas range from well‐differentiated tumors that are easily recognized to poorly differentiated neoplasms that require IHC to confirm the diagnosis (Figure 7.4).5,13 The majority have characteristic plasma cell morphology: uniform round cells, moderate amount of basophilic cytoplasm, perinuclear semi‐clear Golgi zone, single or binucleate eccentric uniform nuclei, chromatin that forms aggregates in the centers of nuclei or along the nuclear membrane, and a low mitotic count. The diagnosis on cytology or histopathology of well‐differentiated types is straightforward, IHC is not needed and these types of plasmacytomas are benign. Some tumor cells may contain intracytoplasmic inclusions that range from semi‐clear vacuoles to distinct eosinophilic globules or rhomboid‐shaped packets of immunoglobulin (Figure 7.4B,C). These are best appreciated in cytologic preparations and are referred to as Mott cells. The tumors are well delineated and surgical excision is usually curative. Unfortunately, other types have a wider range of cellular and nuclear pleomorphism such that IHC may be required to identify them correctly. These types of plamacytomas will have more binucleated cells, multinucleated cells, marked anisokaryosis, no Golgi, no cytoplasmic inclusions and may resemble amelanotic melanoma and histiocytic tumors. Plasmacytomas with these features are aggressive and considered malignant.

Micrograph of plasmacytoma in dog's skin.
Micrograph displaying a round cell tumor with the cytoplasm being packed with numerous cytoplasmic globules.
Micrograph displaying the Mott cells with large eosinophilic inclusions.

Figure 7.4 Plasmacytoma, dog skin. These three preparations demonstrate a range of cytologic patterns. (A) Plasmacytoma can be diagnosed with confidence from cytologic preparations. This example has individual round‐shaped cells with eccentric, round to oval nuclei and variable amounts of basophilic cytoplasm. Binucleation is a feature of this case and is a characteristic of plasma cell tumors. When binucleation and multinucleation are prominent then a differential is histiocytic sarcoma. The cytoplasm of some of these cells has indistinct regions of Golgi apparatus, but cytoplasmic inclusions are not present (see other examples). (B) Round cell tumor in which the cytoplasm is packed with numerous cytoplasmic globules characteristic of Mott cells. In this example the cytoplasm is abundant, markedly basophilic and intracytoplasmic inclusions of globulin are numerous. In many cases this diagnostic feature will not be this obvious. The cells and nuclei in this tumor are more uniform than in the previous example. (C) In this example the Mott cells have large eosinophilic inclusions that could be confused with erythrophagocytosis. Rarely plasma cell tumors can be phagocytic. Cellular and nuclear variability are features present in more aggressive tumors. In less well‐differentiated tumors MUM1 can be used to establish B‐cell origin and differentiate from histiocytic tumors.


Differentiation of cells was a good means to identify cats with longer survival times.5 Giant multinucleated plasma cells have been described in MRD of cats. Mitotic activity is low in cats with MRD. Only 5 of 26 cases had mitoses seen and when they were found they were <10/400× field.5 Plasmacytomas in animals are infrequently (10%) associated with deposition of amyloid, but when present in the tumor amyloid is a helpful diagnostic feature.5,9


LPL and plasmablastic lymphomas are much less common than plasmacytoma. They look similar to plasmacytoma but they arise in a lymph node rather than skin or oral cavity. The node may be focally involved or diffusely. Both may cause generalized lymphadenomegaly, whereas plasmacytomas do not. They may also be in spleen or liver or gastrointestinal tract, whereas plasmacytomas do not have this widespread organ distribution unless malignant. LPLs have a plasmacytic morphology. They have well‐delineated cell borders, the cytoplasm is usually intensely pink, and nuclei are eccentric and may have a distinct shallow nuclear indentation (Figures 7.5 and 7.6). Nuclei are approximately 1.5 RBC in diameter. The mitotic count is low, 0–2 mitoses/400× field. They are low‐grade, indolent, mature B‐cell lymphomas. Plasmablastic lymphomas also look plasmacytoid but have larger nuclei, 2.0–2.5 RBC in diameter, and prominent multiple central nucleoli and a coarsely aggregated chromatin that is hyperchromatic (Figure 7.7). Cytoplasm is usually abundant and deeply stained. Neoplastic cells have a high mitotic count with several in each field at 400× magnification. In lymph nodes they can form clusters or sheets that may be mistaken for benign plasmactyic hyperplasia or plasmacytoma. They will expand from these regions and can be diffuse in nodes or spleen. They are high‐grade mature B‐cell lymphomas and some classify these as a plasmablastic variant of DLBCL.

Micrograph displaying the nodal architecture effaced by an expanding tumor with indistinct irregular nodularity.
Micrograph displaying the tumor cells having definite plasma cell morphology with nuclei in 1.5 RBC diameter and densely stained. Cell boundaries are defined.
Micrograph displaying diffused and strong immunostaining with CD79a.
Micrograph displaying the tumor cells stained positive for CD79a, negative for CD3, and with cytologic features of plasma cells.

Figure 7.5 Lymphoplasmacytic lymphoma (LPL), dog, Tru‐Cut biopsy from mandibular lymph node. (A) Nodal architecture is effaced by an expanding tumor with indistinct irregular nodularity; the tumor has extended beyond the capsule of the node and invaded perinodal fat. (B) Tumor cells have a definite plasma cell morphology. The nuclei are 1.5 RBC in diameter and very densely stained. Nuclei are eccentric and associated with a shallow nuclear indentation. Cell boundaries are sharply defined. Cytoplasm stains intensely pink. (C) There is diffuse, strong immunostaining with CD79a. Plasmacytomas do not react this well with CD79a or CD20. There is a type of “capping” effect with the eccentric area of cytoplasm staining much more densely. (D) CD3: Only a few scattered positive cells are present, with the majority of cells unmarked. Tumor cells stained positive for CD79a, negative for CD3, and had cytologic features of plasma cells, all confirming B‐cell differentiation for this lymphoma.

Micrograph of lymphoplasmacytic lymphoma in dog node, with nuclei in intermediate size having anisokaryosis.
Micrograph displaying strong and uniform marking of the cytoplasm. Inset: Micrograph of tumor cells negative for CD3.

Figure 7.6 Lymphoplasmacytic lymphoma (LPL), dog node. (A) The nuclei are of intermediate size with little anisokaryosis and small or not apparent nucleoli. The cytoplasm is abundant and eccentric to the nucleus. Nuclei are often indented adjacent to the cytoplasmic extension. This example does not look as “plasmacytic” as the previous case. LPLs are in lymph nodes and may be diffuse, whereas plasmacytomas do not exhibit this growth pattern. (B) CD79a: There is very strong and uniform marking of the abundant cytoplasm. The tumor cells were negative for CD3 (inset).

Micrograph of plasmablastic lymphoma in dog, with abundant cytoplasm in neoplastic cells and a round material pointed by black arrow representing Russell body. Presence of mitotic figure is depicted by white arrow.
Micrograph of plasmablastic lymphomas with CD79a.
Micrograph displaying the MUM1 with nuclei which are most likely to be of lymphoid type.
Micrograph displaying positive staining with two markers for B-cell differentiation combined with negative staining with CD3.

Figure 7.7 Plasmablastic lymphoma, lymph node, dog. (A) Neoplastic cells have abundant cytoplasm that is eccentric to the nucleus and some have a distinct pale area interpreted as a Golgi zone. The round pink material at left (black arrow) is likely a Russell body (packet of immunoglobulin). The nuclei have moderate anisokaryosis and one to several nucleoli. A mitotic figure is present (white arrow); binucleation is characteristic of this tumor. (B) CD79a: About half of the cells are labeled. Plasmablastic lymphomas are better marked with CD20. (C) MUM1: The majority of the nuclei are strongly labeled, particularly the larger nuclei that are likely to be of lymphoid type. (D) CD4: The positive staining with two markers for B‐cell differentiation combined with negative staining with CD3 is interpreted as a lymphoma with plasmacytic differentiation. MUM1 is a regulatory protein involved in differentiation of B lymphocytes to plasma cells and is useful to identify plasma cell tumors and B‐cell lymphomas. Plasmablastic lymphomas are slightly larger cell lymphomas than LPL and are likely to be more aggressive.


Cytochemistry and immunohistochemistry


Canine plasmacytomas stain more reliably with MUM1 than with CD79a or CD20. In one study 101 of 109 plasmacytomas (93.5%) were positive for MUM1, 59/105 (56%) were positive for CD79, and only 21/108 (19%) were positive for CD20.13 In this same study the authors applied MUM1 on 139 other tumors, including tumors that can resemble plasmacytomas: 13 melanomas, 7 mast cell tumors, 17 histiocytomas, 2 cutaneous and 3 systemic histiocytoses, 7 histiocytic sarcomas, 17 epitheliotropic lymphomas, and 17 T‐cell lymphomas, and the only other positive tumors were 10 B‐cell lymphomas and one anaplastic lymphoma.13


MUM1 can be used on FFPE tissue sections. It primarily stains nuclei and there may be slight staining of the cytoplasm. MUM1 is in the family of interferon regulatory factors. It is required for immunoglobulin light‐chain rearrangement and is expressed in B cells, plasma cells, activated T cells and a subset of macrophages and dendritic cells in humans. Not all nuclei of a plasmacytoma mark with the MUM1, therefore interpretation is needed when only a proportion of tumor cells are labeled. In general if 20% or more of the nuclei are labeled the immunoreactivity is considered positive and if less than 10% are labeled it is considered negative. Tumors in cats with MRD do not stain reliably with CD79a: 12 cases were negative for CD79a and CD3, one case was positive for both and 6 cases stained well with CD79a.5


Plasmablastic lymphomas mark poorly with CD79a (Figure 7.7B) and strongly with CD20. LPLs generally stain strongly with CD79a and reliably with CD20. Both tumors are negative with CD3 (see Figures 7.2325). If the tumor is in a lymph node, it looks like a plasma cell tumor and stains strongly with CD79a or CD20 it is likely a plasmacytoid lymphoma or LPL. Twelve of 19 MRD cases did not label with CD79a and 18 of 19 had Ig immunolabeling.5


In animals (exception is swine) the “normal” ratio of immunoglobulin light chains is approximately 90:10 lambda to kappa.11 Therefore light‐chain restriction is a part of normal plasma cells and should not be relied on to identify a clonal expansion. If the predominant light chain was kappa, a rare occurrence, then it should represent clonality. See the section on Myeloma in this chapter. Use of lambda light chain to identify neoplastic proliferations of plasmacytic tumors in animals should be done with other diagnostic tests and should not be relied on alone.


Differential diagnosis


Tumors with plasma cell morphology include plasmacytoma, LPL, plasmablastic lymphoma, myeloma, MRD, and even some osteosarcomas that have numerous osteoblasts. The more differentiated the tumor and the more fully developed the associated syndromes, the easier the distinction. Generalizations follow and exceptions are numerous. Myeloma is in bone marrow, multiple osteolytic lesions and paraproteinemia are present. Plasmacytoma is a singular soft tissue tumor without paraproteinemia. LPL and plasmablastic lymphoma are in nodes and/or organs, LPL is more differentiated and paraproteinemia may be present.


Osteosarcomas produce osteoid. They have a characteristic clinical appearance and are the only type listed that are not of B‐cell origin. Osteosarcomas can be stained reliably with ALP and are negative to lymphoid markers.


Well‐differentiated plasmacytomas are not a diagnostic challenge but less‐differentiated tumors may be confused with other round cell tumors, melanoma, histiocytic sarcoma, or poorly differentiated mast cell tumors. The presence of binucleated and occasionally tri‐ or multinucleated cell types as well as scattered large nuclei are suggestive of a plasmacytoma or histiocytic tumor.14 The more bizarre the nuclear pleomorphism the more likely the tumor is of histiocytic origin. Cytophagia favors histiocytic, but plasma cell tumors can be added to the growing list of tumors that are capable of cytophagia.15 Differentiation of these two tumors in FFPE sections can be accomplished by IHC using MUM1, which recognizes plasma cell tumors, and CD18, which favors histiocytic tumors. Many of the antibodies used for histiocytic lineage work better on frozen sections rather than FFPE (see Chapter 3 for IHC details and Chapter 8 or a review article for details about histiocytic neoplasms and how best to recognize them).14


If the tumor is in nodes and organs it is more likely a histiocytic tumor or one of the other lymphomas, plasmablastic lymphoma, or LPL. Plasmablastic lymphoma and LPL will immunostain well with CD20, whereas plasmacytomas stain weakly or are negative with CD20 but stain very well with MUM1. If the tumor is in the lungs it is much more likely to be histiocytic than lymphoid.


Plasmablastic lymphoma and LPL look like plasmacytoma but are in lymph nodes, focally or diffusely, and may also be in intestine, liver, spleen, or bone marrow. The more widespread the neoplasm the more likely it is plasmablastic lymphoma or LPL and not plasmacytoma. LPL is more indolent and mature. Plasmablastic lymphoma is less differentiated and the prominent nucleoli and high mitotic activity differentiate it from LPL. LPL is associated with macroglobulinemia. These same generalizations can be used for cats but manuscripts describing MRD should be read for the details provided.5,6 The authors also discuss the different pathogenesis between cats and humans with MRD.


References



  1. 1. Harris, N.L. and Bhan, A.K. (1985) B‐cell neoplasms of the lymphocytic, lymphoplasmacytoid, and plasma cell types: Immunohistologic analysis and clinical correlation. Hum Pathol 16:829–837.
  2. 2. Patsouris, E., Noel, H., and Lennert, K. (1990) Lymphoplasmacytic/lymphoplasmacytoid immunocytoma with a high content of epithelioid cells. Am J Surg Pathol 14:660–670.
  3. 3. Valli, V.E., Kass, P., San Myint, M., and Scott, F. (2013) Canine lymphoma: The effect of age, stage of disease, subtype of tumor, mitotic rate and treatment protocol on overall survival. Vet Pathol 50:728–738.
  4. 4. Ponce, F., Marchal, T., Magnol, J.P., et al. (2010) A morphological study of 608 cases of canine malignant lymphoma in France with a focus on comparative similarities between canine and human lymphoma morphology. Vet Pathol 47:414–433.
  5. 5. Mellor, P.J., Haugland, S., Smith, K.C., et al. (2008) Histopathologic, immunohistochemical, and cytologic analysis of feline myeloma‐related disorders: further evidence for primary extramedullary development in the cat. Vet Pathol 45:159–173.
  6. 6. Mellor, P.J., Haugland, S., Murphy, S., et al. (2006) Myeloma‐related disorders in cats commonly present as extramedullary neoplasms in contrast to myeloma in human patients: 24 cases with clinical follow up. J Vet Intern Med 20:1376–1383.
  7. 7. Rakich, P.M., Latimer, K.S., Weiss, R., et al. (1989) Mucocutaneous plasmacytomas in dogs: 75 cases (1980–1987). J Am Vet Med Assoc 194:803–810.
  8. 8. Mazoub, M., Breuer, W., Platz, S.J., et al. (2003) Histopathologic and Immunophenotypic characterization of extramedullary plasmacytomas in nine cats. Vet Pathol 40:249–253.
  9. 9. Ramos‐Vara, J.A., Miller, M.A., Pace, L.W., et al. (1998) Intestinal multinodular Aλ‐Amyloid deposition associated with extramedullary plasmacytoma in three dogs: Clinicopathological and immunohistochemical studies. J Comp Pathol 119:239–249.
  10. 10. Rannou, B., Helie, P., and Bedard, C. (2009) Rectal plasmacytoma with intracellular hemosiderin in a dog. Vet Pathol 46:1181–1184.
  11. 11. Arun, S.S., Breuer, A.W., and Hermanns, W. (1996) Immunohistochemical examination of light‐chain expression in canine, feline, equine, bovine and porcine plasma cells. Zentralbl Veterinarmed A 43:573–576.
  12. 12. Valli, V.E. (2007) Plasmacytoma. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 189–207.
  13. 13. Ramos‐Vara, J.A., Miller, M.A., and Valli, V.E.O. (2007) Immunohistochemical detection of multiple myeloma 1/interferon regulatory factor 4 (MUM1/IRF‐4) in canine plasmacytoma: comparison with CD79a and CD20. Vet Pathol 44:875–884.
  14. 14. Moore, P.F. (2014) A review of histiocytic diseases of dogs and cat. Vet Pathol 51:167–184.
  15. 15. Yearley, J.H., Stanton, C., Olivry, T., et al. (2007) Phagocytic plasmacytoma in a dog. Vet Clin Pathol 36:293–296.

Myeloma


Defining the neoplasm


Myeloma is a clonal proliferation of malignant plasma cells that originates in bone marrow and is associated with monoclonal gammopathy and multiple osteolytic bone lesions. It is considered a primary tumor of bone and is also described in Chapter 10. Bone pain and radiographically visible bone lesions are present in approximately 50–70% of dogs and humans with myeloma. The diagnosis is established by visualization of tumor cells in an aspirational cytology or a bone marrow core biopsy from an osteolytic lesion (Figure 7.8; see also Figure 10.43A–D). In most cases the diagnosis is straightforward as there are multiple, discrete osteolytic foci (see Figure 10.43) seen on radiographs and the tumor yields numerous cells with FNA. If needed, an electrophorectic examination of serum demonstrating a monoclonal gammopathy offers further confirmation.

Micrograph of myeloma in dog's marrow core, displaying focal neoplastic infiltration pointed by arrows.
Micrograph of myeloma displaying distinct cell borders, abundant cytoplasm, and the cells are well differentiated. Nuclei are uniform and eccentric with few binucleated cells.
Micrograph of a lytic bone lesions displaying plasma cell characterized by round nuclei, abundant basophilic cytoplasm with indistinct clear vacuoles, eccentric dense nuclei, and pale perinuclear Golgi zones.
Micrograph of myeloma with nuclear characteristic and cellular features of plasma cell.

Figure 7.8 Myeloma, dog, marrow core. (A) There is focal neoplastic infiltration (arrows); normal marrow can be identified by the presence of fat cells which are absent in the areas of tumor. (B) In this region almost all the cells are myeloma, cell borders are distinct, cytoplasm is fairly abundant and the cells are well differentiated. Nuclei are uniform and eccentric; a few binucleated cells are present. Mature trabecular bone is at the top of image and an osteoclast is immediately subjacent. The multiple bone lesions characteristic of this disease are due to cytokines produced by the myeloma cells that indirectly stimulate osteoclastic osteolysis. (C) Mature male labrador had lameness and multiple lytic foci in the axial skeleton. Cytological aspirate from one of the lytic bone lesions is highly cellular and approximately 60% of the cells have plasma cell features characterized by round nuclei, abundant basophilic cytoplasm with indistinct clear vacuoles, eccentric dense nuclei, and pale perinuclear Golgi zones. Their plasma cell differentiation is not as obvious as the case shown in (D). Hematopoietic cells are dispersed throughout. (D) All the cells have characteristic nuclear and cellular features of plasma cells. Neoplastic cells in myeloma can range from well‐differentiated, as in this example, to less‐differentiated tumors that require IHC to establish the diagnosis. IHC is rarely needed to confirm this diagnosis. Monoclonal gammopathies and bone lesions aid the diagnosis of multiple myeloma.


The classification of human monoclonal gammopathies according to the International Working Group defines three categories of gammopathies based on the concentration of serum M‐protein and percent of clonal plasma cells in the bone marrow.1–3 Plasmacytoma may progress to myeloma. In humans, altered expression of the c‐myc oncogene is associated with multiple myeloma and the cell cycle protein cyclin D is associated with myeloma in humans and dogs.4,5


Occurrence and clinical presentation


Myeloma is seen primarily in dogs, and less frequently in cats and has been reported in the cow, horse, and pig.3,6–9 In cats it accounts for approximately 1% of all feline malignancies; it is not associated with FeLV or FIV and osteolytic lesions are rare.6–8 In cats the myeloma‐related diseases (MRD) are a group of about seven disorders, one of which is myeloma. Most cases of MRD are extramedullary and are described in detail in several publications.8 The majority of the cats with MRD present with extramedullary tumors (see section on Plasmacytoma, plasmablastic lymphoma, lymphoplasmacytic lymphoma, and myeloma‐related disorders) and most will have organ involvement and paraproteinemia. About half will have bone marrow involvement, but bone lesions are not a prominent feature of MRD.8


In the dog, multiple myeloma accounts for approximately 0.5% of all malignant tumors, with no apparent sex or breed predisposition. No cases were found in the ACVP study of 1000 cases of canine lymphoma, likely because the disease was diagnosed from characteristic clinical presentations including a high serum globulin and monoclonal gammopathy. There is no known cause, but chronic immune stimulation is suspected to be involved in humans, where dendritic cells infected with Herpesvirus 8 secrete high levels of IL‐6 that drives plasma cell differentiation and possible malignant transformation.


A common finding in dogs is lameness and radiographs demonstrate multiple lytic lesions (multiple myeloma) in bones.8 The lesions can be found in any bone of the axial or appendicular skeleton but are seen more commonly in areas of active hematopoiesis. Animals in later stages of the disease may present with a history of weight loss, decreased vigor, and signs of renal malfunction. A characteristic feature of myeloma is a monoclonal or less commonly a biclonal gammopathy (paraprotein). Serum total protein values over 9.0 g/dL are a clinical aid to suspect myeloma and suggest electrophoresis to characterize the serum proteins or search for bone lesions. Theoretically, any of the immunoglobulins could be produced but IgG is the most common, followed by IgA and IgM. Cats also have paraproteinemia, usually IgG. A small percent of myelomas will be nonsecretory with no gammopathy present. IgM‐secreting tumors can cause a hyperviscosity syndrome. Most myelomas with gammopathy are accompanied by overproduction of light‐chain immunoglobulins small enough to pass the kidney and be detected in the urine as Bence‐Jones protein. SSA (sulfosalicylic acid) is a more sensitive technique to recognize these proteins than are colorimetric dipsticks used for routine urinalyses. These light‐chain immunoglobulins may also be deposited in the basement membranes of glomeruli, resulting in glomerulonephritis. The light‐chain immunoglobulins may also form amyloid, which is deposited in glomeruli, resulting in the nephrotic syndrome: proteinuria, hypoalbuminemia, ascites, and increased serum cholesterol. Patients with the nephrotic syndrome are not necessarily azotemic. Amyloid may also be deposited subendothelially in other organs, most notably spleen and liver.


Monoclonal gammopathy is highly suggestive of myeloma but can also be seen in canine ehrlichiosis, leishmaniasis, pyoderma, B‐CLL, B‐cell lymphoma, plasmacytoma, and feline infectious peritonitis.10 Hypercalcemia occurs in approximately 10% of canine myelomas as a paraneoplastic syndrome and is less common in cats.8 The mechanism of hypercalcemia is attributed to the paracrine effects of cytokines produced by the tumor cells, which induce local osteoclastic osteolysis (see Figure 10.43D). In addition there may be humoral substances (parathyroid hormone‐related protein (PTHrP)) that act on renal tubules to reabsorb calcium from the glomerular filtrate and block the absorption of phosphorus, resulting in hypercalcemia and hypophosphatemia. Concurrent hypercalcemia in dogs with myeloma suggests a worse prognosis.3


Although the tumor originates in bone marrow, leukemia is rarely detected. If present the prognosis is worse. At the time of diagnosis there is usually a mild or moderate non‐regenerative normocytic normochrocmic anemia, most likely due to cytokine production associated with the anemia of chronic inflammatory disease. Petechiae and bleeding problems are fairly common and are associated with thrombocytopenia and/or adherence of paraproteins to platelet membranes that alter platelet function.10


Pathology


Bone marrow

If the marrow aspirations and/or core biopsy are from osteolytic lesions, the diagnosis is usually straightforward. Myeloma forms focal to extensive colonies of solid tumor cells devoid of normal marrow hematopoietic and adipose cells (Figure 7.8). Cytology usually yields numerous fairly well‐differentiated tumor cells that often have a greater volume of cytoplasm than normal plasma cells (Figure 7.8). Despite their well‐differentiated morphology they still will behave in a malignant fashion. The tumor cells are tightly packed but the cell margins are readily apparent. In cytological preparations the tumor cells with Wright–Giemsa staining have uniformly dense amphophilic cytoplasm that usually is not vacuolated (Figure 7.8). Occasionally, the outer rim of cytoplasm is densely purple or eosinophilic and these are referred to as flame cells. Visualization of a Golgi apparatus aids the diagnosis. Nuclei are round with densely stained and aggregated chromatin with one or more prominent nucleoli. Less‐differentiated tumors have marked anisokaryosis, with bi‐ and multinucleated cells, but these cells are less common in myeloma than in aggressive types of plasmacytomas. The mitotic count of the tumor cells varies with their differentiation; it is usually low but may be as high as 5/400× field.


An approximation of >20% plasma cells is a criterion to diagnose myeloma; however, this relative percentage can also be seen with infectious diseases such as ehrlichiosis or leishmaniasis. Each of these usually has lymphocytosis of the marrow and peripheral blood, sometimes marked. Polyclonal gammopathy is common with both but monoclonal gammopathies have rarely been observed, further confounding the correct diagnosis. Persistent high titers to Ehrlichia and other tick‐borne diseases combined with PCR are useful to rule out these differentials. In leishmaniasis there are numerous intracellular organisms in macrophages, so the infection can be ruled out by cytology.


Lymph nodes and spleen

The damage to organs other than the bone marrow appears to be due to infiltration with amyloid or M‐protein accumulations rather than actual metastatic tumor in dogs. Metastasis to the spleen does occur. The lymph nodes and spleen are much more likely to be involved with benign reactions of plasma cells, as seen in canine leishmaniasis. In this disease there may be generalized plasmacytosis of bone marrow, lymph nodes, and spleen. Leishmania organisms are very plentiful and easy to see in cytologic or histologic preparations. However, cats will have tumor in abdominal organs, spleen, liver, and lymph nodes and much more frequently than occurs in dogs. The combination of plasma cell proliferation in the marrow, monoclonal gammopathy, and multiple lytic lesions of bones is diagnostic for multiple myeloma in all species.


Cytochemistry and immunohistochemistry


Immune staining is usually not needed for a diagnosis. If performed, the results may vary but most canine cases stain positively with CD79a in both cytological and histological preparations. CD20 is not reliable in myeloma but the nuclei will stain very well with MUM1.11 CD38 and CD138 are used in humans and myelomas are negative with CD3.


Homogeneity of light‐chain restriction can be used in human plasma cell tumors to differentiate neoplasia versus inflammation. This is based on the principle that there is a balanced distribution of lambda and kappa light chains.12 However, in animals (exception is swine) the “normal” ratio is approximately 90:10 lambda to kappa.12 Therefore light‐chain restriction is a part of normal plasma cells and should not be relied on to identify a clonal expansion. If the predominant light chain was kappa, a rare occurrence, then it should represent clonality.


Staging and survival


Treatment of multiple myeloma in animals, as in humans, has not been curative, with most animals succumbing to the tumor. Chemotherapy is used but most cases survive less than 2 years. Factors associated with a poor prognosis are: hypercalcemia, leukemia, azotemia, cytopenias, and Bence‐Jones proteinuria. One parameter that may be useful to predict prognosis is the relation of the rate of increase in serum M‐globulin to the number of surviving neoplastic cells secreting protein. New treatments in humans have greatly improved the survival of younger patients. A staging system has been developed for humans based on the levels of β2‐microglobulin and albumin.


References



  1. 1. Swerdlow, S.H., Campo, E., Harris, N.L., et al. (2008) WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues . International Agency for Research on Cancer (IARC), Lyon, France.
  2. 2. Valli. V.E. (2007) Myeloma. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 189–206.
  3. 3. Matus, R.E., Leifer, C.E., MacEwen, G., et al. (1986) Prognostic factors for multiple myeloma in the dog. J Am Vet Med Assoc 188:1288–1292.
  4. 4. Tonon, G. (2007) Molecular pathogenesis of multiple myeloma. Hematol Oncol Clin North Am 21:985–1006.
  5. 5. Bergsagel, P.L., Kuehl, W.M., Zhan, F., et al. (2005) Cyclin D dysregulation: an early and unifying pathogenic event in multiple myeloma. Blood 106:296–303.
  6. 6. Patel, R.T., Caceres, A., French, A.F., and McManus, P.M. (2005) Multiple myeloma in 16 cats: a retrospective study. Vet Clin Pathol 34:341–352.
  7. 7. Bienzle, D., Silverstein, D.C., and Chaffin, K. (2000) Multiple myeloma in cats: variable presentation with different immunoglobulin isotypes in two cats. Vet Pathol 37:364–369.
  8. 8. Mellor, P.J., Haugland, S., Murphy, S., et al. (2006) Myeloma‐related disorders in cats commonly present as extramedullary neoplasms in contrast to myeloma in human patients: 24 cases with clinical follow up. J Vet Intern Med 20:1376–1383.
  9. 9. Edwards, D.F., Parker, J.W., Wilkinson, J.E., and Helman, R.G. (1993) Plasma cell myeloma in the horse. J Vet Intern Med 7:169–176.
  10. 10. Thrall, M.A. (2012) Lymphoproliferative disorders and myeloid neoplasms. In Veterinary Hematology and Clinical Chemistry , 2nd edn. (eds. M.A. Thrall, G. Weiser, R.W. Allison, and T.W. Campbell). Wiley‐Blackwell, Ames, Iowa, pp. 179–181.
  11. 11. Ramos‐Vara, J.A., Miller, M.A., and Valli, V.E. (2007) Immunohistochemical detection of multiple myeloma 1/interferon regulatory factor 4 (MUM1/IRF‐4) in canine plasmacytoma: comparison with CD79a and CD20. Vet Pathol 44:875–884.
  12. 12. Arun, S.S., Breuer, A.W., and Hermanns, W. (1996) Immunohistochemical examination of light‐chain expression in canine, feline, equine, bovine and porcine plasma cells. Zentralbl Veterinarmed A 43:573–576.

Marginal zone lymphoma


Defining the neoplasm


Marginal zone lymphoma (MZL) originates in lymph nodes or spleen and it is the most common type of lymphoma in the spleen of dogs and humans.1–14 Although it is common in the spleen of humans, it makes up <1% of lymphomas. It is one of the more common lymphomas in dogs, at 5–15%.5,7,9 MZL is considered a low‐grade B‐cell lymphoma that has an indolent course but more cases need to be followed and morphology and molecular studies integrated.10–14 It is one of the lymphomas that forms discrete nodules which become diffuse in only late stages. Histologic sections are required to make this diagnosis as the architectural arrangement in a node or spleen is critical and this cannot be appreciated in cytological preparations. MZL has a characteristic arrangement of neoplastic lymphocytes surrounding fading germinal centers which results in a nodular pattern (Figure 7.9). IHC reveals neoplastic lymphocytes are B‐cell type. There is no specific marker for MZL in humans or animals. The diagnosis is dependent on their histologic organization combined with anatomical location in body, cytologic features, and immunophenotyping. MZL was one of the more common lymphomas in the studies that total approximately 1700 canine lymphomas.4,7,8

Micrograph of marginal zone lymphoma (MZL) in dog, featuring fading germinal centers surrounded by a corona of proliferating neoplastic B lymphocytes with the presence of tingible body macrophages.
Micrograph of MZL with CD79a, displaying intense and uniform staining of the MZL corona and a few unlabeled antigen presenting or T cells in the germinal centers.
Micrograph of MZL with CD3, displaying scattered positive cells in the area of the B-cell corona and clusters are present in the fading germinal centers.
Micrograph of MZL displaying uniform population of immature lymphoid cells with rare mature, non-neoplastic lymphocytes (pointed by arrows), abundant cytoplasm, and with amphophilic strands.
image

Figure 7.9 Marginal zone lymphoma (MZL), dog, lymph node. (A) The histologic signature of MZL is fading germinal centers surrounded by a corona of proliferating neoplastic B lymphocytes. Tingible body macrophages are visible at this low magnification; they increase as the tumor expands but are absent in MZL in the spleen. (B) CD79a: There is intense and uniform staining of the MZL corona and a few unlabeled antigen‐presenting or T cells in the germinal centers. (C) CD3: There are scattered positive cells in the area of the B‐cell corona and clusters are present in the fading germinal centers. (D) In a cytological preparation there is a uniform population of immature lymphoid cells with rare mature, non‐neoplastic lymphocytes for size comparison (arrows). Cytoplasm is fairly abundant and is more visible than in the histologic preparation. Nuclei are 1.5–2 RBC un diameter and there are no mitotic figures. The amphophilic strands coursing across the image are artifacts created by nuclear chromatin from lyzed cells. From the cytological preparation a diagnosis of lymphoma is possible but histologic organization of the tumor and immunophenotyping of the cells are required to diagnose B‐cell MZL. Lymphomas such as MZL have probably been underdiagnosed.


The marginal cell zone of lymph nodes is only seen in nodes that are hyperplastic (reactive) or neoplastic (Figures 7.10 and 7.11). The nodular architecture in the cortex of nodes is due to a central germinal center, B‐cell rich, surrounded by a distinct mantle cell zone that is composed of approximately 90% B lymphocytes and 10% T lymphocytes. When a node is antigenically stimulated, the marginal zone layer proliferates and becomes visible, forming a polar pattern in which the marginal zone is thinner toward the paracortex and thicker toward the subcapsular sinus (Figure 7.10). The lighter pole has the proliferating cells (centroblasts).

Micrograph of marginal zone hyperplasia (MZH) in dog’s spleen. The surrounding corona of cells that are lighter staining with more cytoplasm are B lymphocytes of marginal zone type (depicted by arrows).
Micrograph of MZL in dog’s spleen, displaying large mass composed of multiple neoplastic nodules that coalesce and compress normal spleen.
Micrograph (in higher magnification) of MZH displaying medium-sized neoplastic lymphocytes with cytoplasm, prominent nuclei and nucleoli, but a low mitotic count.
Micrograph of MZH with CD79a, displaying strongly marked areas of neoplastic proliferation. Inset: MZH with CD3, displaying neoplastic cells that are negative with few positively stained remaining T cells.

Figure 7.10 (A) Marginal zone hyperplasia (MZH), dog, spleen. The terminal arteriole at the right has a focus of small lymphocytes near the artery. These lymphocytes have very little cytoplasm and are of T‐cell periarteriolar lymphoid sheath type. The surrounding corona of cells that are lighter staining with more cytoplasm are B lymphocytes of marginal zone type (arrows). Nodules of MZH do not coalesce and nodules of marginal zone lymphoma (MZL) will form masses of variable sizes as seen in next images. (B) MZL, dog, spleen. This is the most common lymphoma in the spleen of dogs. These tumors typically form a large mass composed of multiple neoplastic nodules that coalesce and compress normal spleen (right). The spleen is not diffusely involved in dogs as it is in humans. (C) Higher magnification reveals medium‐sized neoplastic lymphocytes with visible cytoplasm, prominent nuclei and nucleoli, but a low mitotic count. Nuclei are “open,” chromatin is peripheralized to give the appearance of a thick nuclear membrane, and nucleoli are singular, prominent, and centrally placed. The diagnosis is based on location in spleen and histologic organization more than on cytologic features. Tingible body macrophages are seen in late stages when the tumor has a high mitotic count. (D) CD79a: The areas of neoplastic proliferation are strongly marked. Inset: CD3: The neoplastic cells are negative; there are a few positively stained remaining T cells.

Micrograph of MZL in dog’s spleen displaying the neoplastic cells forming a nodule around mature lymphocytes, with cluster of benign lymphocytes at center surrounding a pink focus (pointed by arrow).
Micrograph of MZL featuring single central nucleolus with peripheralized chromatin and apoptotic cells (pointed by arrows) with densely stained nuclei.
Micrograph of MZL with CD20, displaying the neoplastic B cells. Inset: MZL with CD3, displaying smaller positive T cells and negative B cells.

Figure 7.11 Marginal zone lymphoma, dog, spleen. (A) The neoplastic cells have formed a distinct nodule around residual mature lymphocytes. There is a cluster of darker stained benign lymphocytes in the center, surrounding a pink focus that is likely follicular hyalinosis (arrow). This is an early case confirmed by histology and clonality. MZL is not diffuse in the spleen and it will penetrate into the surrounding splenic sinus. (B) Note the very obvious single central nucleolus with peripheralized chromatin which adds prominence to nucleoli and nuclear membranes. This is a characteristic feature of MZL. Nuclei and nucleoli look like an aggressive lymphoma but MZL is reported to have survival times of almost 2 years. The nuclear size is intermediate at 1.5 RBC in diameter. The other cells present with densely stained nuclei are apoptotic cells (arrows). (C) CD20: The neoplastic B cells of MZL type are strongly labeled. Inset: CD3: Smaller T cells are positive and B cells are negative.


Marginal zone hyperplasia (MZH) is an important differential for MZL, as both originate in the marginal zone of lymphoid follicles. MZH is a common finding in the spleen of aged dogs and can be observed concurrently with faded germinal centers. Hyperplastic marginal zones can have a quite monomorphic appearance with intermediate‐sized lymphocytes that often have a single prominent nucleolus. In MZH this change is generally widespread within the white pulp of the spleen; however, coalescence of proliferating marginal zones into nodules of varying sizes is a feature of MZL not MZH.


Epidemiology and clinical presentation


MZL is seen in mature dogs and cats and occurs in horses and likely in all domestic animals. In the ACVP study of 1000 cases of canine lymphoma 62 cases of MZL included 13 in the spleen, 2 cases in both node and spleen, and 1 case of mucosal‐associated lymphoid tissue (MALT) type in a salivary gland.4 The gender was about even and the mean age of the cases of MZL in nodes was 9.7 years and those with MZL in the spleen had a mean age of 11.4 years.4 In the review of 600 canine cases there were 66 MZL; most were in lymph nodes and only a few were extranodal.7 In some reviews MZL represents approximately 15% of selected canine lymphomas.7 MZL may be found in other organs, including the central nervous system in humans6 and in the spleen of p53‐deficient mice.15


MZL occurs in mature dogs, typically large breeds, with a single enlarged node that remains fully mobile. Dogs with MZL usually have no other clinical signs and leukemia is not present; however, leukemia is seen in humans if MZL is in the spleen. A series of 75 dogs with indolent lymphomas detailed the clinical characteristics and survival data;9 two other studies reported on approximately 45 dogs.13,14 MZL in lymph nodes becomes slowly generalized and even with multiple large nodes the dogs generally feel well, perhaps due to the low mitotic count and slow growth of the tumor. The tumor may also start in the spleen and abdominal ultrasound of these dogs reveals focal or multifocal splenic enlargement. Spread from the spleen is slow and likely first to the splenic hilar nodes and then to mesenteric nodes. In human patients MZL cells may be found in the peripheral blood and in darkfield examination with a wet mount these cells have villous cytoplasm. This aspect has not been reported in dogs. In humans MZL is rarely associated with a macroglobulinemia like that of the Waldenstrom type.


Blood, bone marrow, and lymph nodes

The histologic hallmark of MZL is a nodular pattern composed of fading germinal centers surrounded by a lighter staining corona of neoplastic B lymphocytes (Figures 7.9 and 7.11). The initial involvement is in the outer cortex of nodes with slow spread to the paracortical areas. The lymph node is only diffusely involved in late stages of an indolent process. Tingible body macrophages are not seen in early cases. There is usually hyperplasia of the medullary fibrovascular supporting elements which suggests a preceding period of benign, likely follicular hyperplasia. Medullary sinuses and cords are relatively uninvolved. The tumor does not invade into perinodal adipose. Cytologically, the proliferating cells are intermediate in size, nuclei are about 1.5 RBC in diameter (Figure 7.9). The chromatin is peripheralized to the nuclear membrane, which adds prominence to the nuclear membrane and a distinctive large single central nucleolus (Figure 7.10). There is parachromatin clearing around the nucleolus that gives the nuclei a more aggressive look than their size would suggest (Figure 7.11). Cytoplasm is fairly abundant and lightly stained, which contributes to the histologic corona seen at lower magnifications (Figure 7.9). The mitotic count in early cases is always low, usually with none in a 400× field. In advanced cases mitoses will increase to 2–4/400× field, accompanied by a few tingible body macrophages.


Spleen

In dogs, MZL forms one or more nodules of proliferating tumor cells that may coalesce and compress the unaffected adjacent spleen (Figure 7.11). A surgical biopsy or an FNA from a dog spleen must involve one of these nodules to ensure a diagnosis. This is in contrast to the human splenic MZL, which is always diffuse and therefore a biopsy from any site of an enlarged spleen is diagnostic in humans. Some nodules appear to arise on terminal arterioles and they may coalesce to form larger nodules, a feature that does not happen with MZH. However, if the nodules are not large it is difficult to distinguish MZL and MZH. There is generalized atrophy of the thymic‐dependent regions around arterioles. There are few or no functional germinal centers in the unaffected areas of the spleen. Arterioles have a fading cuff of mantle cells and a small surrounding area of MZH (Figure 7.10). The center of some nodules may have a darker appearing area surrounded by a lighter staining peripheral corona of typical MZL cells as previously described (Figure 7.11). Tumor cells have lightly staining cytoplasm that separates the nuclei and is the cause of the lighter staining in the area of the tumor. The mitotic count is low, with none found in most MZL at 400×. Affected areas are never diffuse, always nodular.


In MZH the foci of lymphoid proliferations are smaller and remain confined around a blood vessel and they do not coalesce to form large nodules (Figure 7.10). In addition, in hyperplasia, of both the spleen and the node, the marginal zone cells will be mixed with a few benign mantle cells, whereas in MZL all the cells in a nodule are neoplastic. In problematic cases consider all the data and possibly clonality PCR (PARR) for B cells or IHC for homogeneity of cells. Splenic MZL and MZH are indolent, but MZL in nodes may not be as indolent as previously considered.10,11


A more aggressive form of this lymphoma occurs in the spleen of humans and is known as “blastoid” type of splenic lymphoma. In humans this variant is most often of the mantle cell (MCL) type. The human MCL is identified by the genetic alteration t(11;14)(q13;q32) that gives MCL the overexpression of cyclin D1 protein identified by IHC. This protein has not been identified in the dog and in dogs the blastoid type of change in the spleen has large nucleoli and appears often like an aggressive tumor of MZL type. The blastoid type of splenic lymphoma is included in the section on Mantle cell lymphoma.


Other organs

The only other organs reported to be involved with MZL of dogs are the salivary glands, tonsil, and nictating membrane.4,7 MZL invades the salivary tissue, producing a “lymphoepithelial‐like lesion” characteristic of MALT lymphoma (Figure 7.12). Occasionally the same lesion may be present in lymphomas of the intestine in dogs and, much less commonly, in cats. Concurrent involvement of a node is likely but the tumor may only be found in the salivary gland if a thorough autopsy is performed. Rarely a MALT lesion may be found in the thyroid gland, and in humans the lung and breast may also be involved. The key to diagnose this subtype of lymphoma is invasion of the salivary epithelium by neoplastic lymphocytes, producing a lymphoepithelial lesion.

Micrograph of the mucosal-associated lymphoid tissue (MALT) in dog’s salivary gland infiltrated with  neoplastic lymphoid cells of mantle zone lymphoma type.
Micrograph of MALT in salivary gland, with salivary acini and ducts being pushed aside by lymphoid cells (depicted by arrows). Inset: Micrograph of the cytoplasm of neoplastic cells that is CD79a positive.

Figure 7.12 Mucosal‐associated lymphoid tissue (MALT), dog, salivary gland. (A) The gland is enlarged and infiltrated with neoplastic lymphoid cells of mantle zone lymphoma type. This is a rare lymphoma. (B) Salivary acini and ducts are pushed aside by numerous lymphoid cells. Lymphocytes were seen invading the salivary duct and adjacent acini (arrows) fulfilling the “lymphoepithelial” requirement for this diagnosis. Inset: Cytoplasm of neoplastic cells is strongly CD79a positive. Lymphoid cells are uniform morphologically and phenotypically; inflammatory cells and plasma cells are not present as in sialoadenitis.


Immunohistochemistry and clonality


MZL in all sites, in all animals, is strongly positive with CD79a and CD20 antibodies and negative with CD3 (Figures 7.97.12). There is no specific marker for MZL and the diagnosis is dependent on anatomical location, histologic organization, cytologic features and B‐cell immunophenotype.


Clonality was detected in 28 of 35 cases (80%) of canine MZL.5 In a separate study, 24/33 dogs had clonal rearrangement of Ig loci, 6 polyclonal and 3 pseudoclonal rearrangements.14 Absence of detectable clonality in 20% of cases is consistent with results from PARR that may have false‐negative results up to 30%. False negatives can be attributed to multiple possibilities, including gene rearrangements which the primers used do not recognize, mediocre quality of sample, or partial rearrangements of IGH. It is also noteworthy that 6 of 28 cases with a clonal IGH gene rearrangement had a concurrent TCRG gene rearrangement.5 Multiple possibilities exist, including reactive but clonal T cells within the lymphoma, which can happen with other neoplasms. Material for clonality was from FFPE material; fresh material might have revealed different results.5 Regardless, clonality must be interpreted as only one part of the case data and, depending on the results, it is important to consult with the laboratory that ran the tests.


Differential diagnosis


Neoplastic expansion of marginal zone lymphocytes surrounding germinal centers is the hallmark of this lymphoma but a nodular pattern can be seen with MZH, follicular lymphoma, TZL, MCL, and a faintly nodular form of DLBCL. Histology and phenotype provides identification and differentiating features for these lesions and therefore cytology is not useful to identify MZL. Differentiation of MZL from hyperplasia is guided by the size of the lesions and how uniform the cells are within the lesion. Hyperplasia should be confined to a region around the terminal blood vessel (Figure 7.10) and should not form coalescing nodules. MZH has a polarity to some of the nodules. Hyperplasia does not compress the adjacent spleen (Figure 7.10). However, the distinction is subjective. MZL has a uniform population of immature cells and hyperplasia has a mixture of smaller mature lymphocytes that have less cytoplasm. If the distinction is not apparent from H&E consider clonality (PARR) testing of histologic sections as well as all the data in the case. Clonality testing may not be definitive.


The late stage of MZL can be diffuse and the nodular pattern not apparent, hence these cases will resemble DLBCL. However, the distinction between these two diagnoses may not be clinically important at late stages of MZL where biological aggressive behavior more closely resembles DLBCL. Recent molecular investigations would also suggest these two diseases are not as distinct as we believed.10,11 In early stages use the nodular pattern to help differentiate but if this is absent the distinction is difficult and the two diseases may behave similarly. Both DLBCL (immunoblastic) and MZL will have prominent central nucleoli. Late‐stage MZL retains the same cellular characteristics as earlier stages, but the nodules coalesce, the mitotic count increases, and tingible body macrophages become apparent, which makes DLBCL a differential. Splenic location favors MZL, especially if peripheral nodes are not involved or mildly so. The separation of MZL and DLBCL is interpretive.


MCL resembles MZL at the architectural level as both types are nodular. Each is found surrounding central fading germinal centers, which may be pale (depleted of cells) or dark, non‐neoplastic small lymphocytes. MCL is the “inverse” of MZL in that the dark center of MZL is residual lymphocytes from the collapsed and compressed mantle cell layer. In MCL the “center” is a residual germinal center and the mantle cells proliferate around it; the marginal zone is not seen. MCL in the spleen is oriented on or near a terminal arteriole. The cells of MCL have little cytoplasm, nuclei are smaller than MZL nuclei, and they lack the large single nucleolus so characteristic of MZL nuclei.


T‐zone lymphoma (TZL) also has a nodular architectural pattern similar to MCL and MZL, but TZL is T‐cell phenotype and that is the easiest distinction. TZL cells have even smaller nuclei, the chromatin is densely stained, and nucleoli are not obvious. TZL has pale‐staining cytoplasm (clear cells) that has a “hand‐mirror” appearance cytologically (see Figures 7.47 and 7.48) and a low mitotic count. Immunophenotyping readily identifies T cells. A major but subtle histologic difference between indolent TZL and MZL or MCL is the corona of neoplastic cells that surround fading germinal centers in a concentric pattern in MZL and MCL. In TZL the proliferating T cells push the B cells away from the center of the follicle rather than surrounding them (eccentric pattern). Therefore the neoplastic T cells are adjacent to and not surrounded by a collar of B cells.


Follicular lymphoma is rare but it resembles MZL in that they both have nodular patterns and are B‐cell lymphomas. MZLs have a central area of dark non‐neoplastic lymphocytes that are from the mantle region and which have collapsed into the center of the follicle. Surrounding this is the corona of MZL. In contrast to MZL, follicular lymphoma is characterized by expansion of the germinal center with a mixed proliferation of immature lymphocytes. Follicular lymphoma expands uniformly and compresses the marginal zone and mantle cell cuff. The nodular pattern becomes less obvious as the neoplasm fills the node. Follicular lymphoma is rare and MZL common, at least in dogs.


Evaluation of treatment


MZL is considered an indolent type of lymphoma due to its low mitotic count, slow growth of tumors, and slow tendency to metastasize or develop in other lymphoid sites. Splenectomy in early MZL could be effective since the tumor may be confined to the spleen.4,9,12–14 MZL and MCL in the spleen had an indolent behavior when treated with splenectomy alone: 20 dogs had a 1‐year survival rate of 75%, median follow‐up time of 428 days, and MST was not reached.13 In another study, 29 dogs had an MST of 383 days and 14 of 29 that were asymptomatic had an MST of 1153 days.14 Spread to hilar or mesenteric lymph nodes is slow and can be evaluated grossly at surgery. If these nodes are determined to be of normal size during splenectomy it is highly likely that no spread has occurred. When internal nodes are enlarged, a biopsy of an enlarged peripheral node will often yield the correct diagnosis as well. If the MZL is not recognized until the tumor is advanced, an FNA may be misinterpreted as a high‐grade lymphoma since the large nucleoli look threatening and the low mitotic count is not as easily appreciated on cytology.


Typical survival time


Owner decisions, treatment protocols, and how tumors are defined can greatly affect the survival data that are reported for different lymphomas. As we further define MZL, some of the existing survival data will likely be modified. MZL of the spleen is associated with long survival times when treated with splenectomy alone, especially in asymptomatic dogs.13,14


MZL in lymph nodes has been considered one of the indolent lymphomas and as such survival times will approach 2–3 years.9 A study of indolent lymphoma in 75 dogs reported MST of approximately 21 months for MZL and 33 months for dogs with TZL.9 There was no statistical difference in survival between dogs with MZL and TZL. Different systemic treatments did not alter survival. The authors indicated a prospective study to evaluate treatment strategies is indicated and that a “watchful waiting” approach for indolent lymphomas has logic.9


However, the distinction of MZL from DLBCL in lymph nodes is problematic using morphology or molecular techniques.10–12 Recent work suggests that both neoplasms can be aggressive and they may represent a continuum rather than separate neoplasms.10,11 Gene expression profiles could distinguish groups of canine lymphoma but could not distinguish MZL and DLBCL. The molecular similarities in MZL and DLBCL were such that they may represent a continuum rather than separate lymphomas.10,11 Furthermore MZL and DLBCL have stages in which their morphologies overlap. More cases of MZL need to be studied morphologically and molecularly and then followed to confirm (or refute) its indolent nature.


References



  1. 1. Nathwani, B.N., Anderson, J.R., Armitage, J.O., et al. (1999) Marginal zone B‐cell lymphoma: A clinical comparison of nodal and mucosa‐associated lymphoid tissue types. J Clin Oncol 17:2486–2492.
  2. 2. Oscier, D., Own, R., and Johnson, S. (2005) Splenic marginal zone lymphoma. Blood Rev 19:39–51.
  3. 3. Takino, H., Li, C., Hu, S., et al. (2008) Primary cutaneous marginal zone B‐cell lymphoma: a molecular and clinicopathological study of cases from Asia, Germany and the United States. Mod Pathol 21:1517–1526.
  4. 4. Valli, V.E., Kass, P., San Myint, M., and Scott, F. (2013) Canine lymphoma: The effect of age, stage of disease, subtype of tumor, mitotic rate and treatment protocol on overall survival. Vet Pathol 50:738–748.
  5. 5. Valli, V.E., Vernau, W., de Lorimier, L.P., et al. (2006) Canine indolent nodular lymphoma. Vet Pathol 43:241–256.
  6. 6. Venkataraman, G., Rizzo, K.A., Chavez, J.J., et al. (2011) Marginal zone lymphomas involving meningeal dura: possible link to IgG4‐related disease. Mod Pathol 24:355–366.
  7. 7. Ponce, F., Marchal, T., Magnol, J.P., et al. (2010) A morphological study of 608 cases of canine malignant lymphoma in France with a focus on comparative similarities between canine and human lymphoma morphology. Vet Pathol 47:414–433.
  8. 8. Thalheim, L., Williams, L.E., Borst, L.B., et al. (2013) Lymphoma immunophenotype of dogs determined by immunohistochemistry, flow cytometry, and polymerase chain reaction for antigen receptor rearrangements. J Vet Intern Med 27:1509–1516.
  9. 9. Flood‐Knapik, K.E., Durham, A.C., and Gregor, T.P. (2013) Clinical, histopathological and immunohistochemical characterization of canine indolent lymphoma. Vet Comp Oncol 11:272–286.
  10. 10. Franz, A.M., Sarver, A.L., Ito, D., et al. (2013) Molecular profiling reveals prognostically significant subtypes of canine lymphoma. Vet Pathol 50:693–703.
  11. 11. Richards, K.L., Motsinger‐Reif, A.A., Chen, H.W., et al. (2013) Gene profiling of canine B‐cell lymphoma reveals germinal center and post germinal center subtypes with different survival times, modeling human DLBCL. Cancer Res 73:5029–5039.
  12. 12. Ito, D., Frantz, A.M., and Modiano, J. (2014) Canine lymphoma as a comparative model for human non‐Hodgkins lymphoma: recent progress and application. Vet Immunol Immunopathol 159:192–201.
  13. 13. van Stee, L.L., Boston, S.F., Singh, A., et al. (2015) Outcome and prognostic factors for canine splenic lymphoma treated by splenectomy (1995–2011). Vet Surg 44:976–982.
  14. 14. O’Brien, D., Moore, P.F., Vernau, W., et al. (2013) Clinical characteristics and outcome in dogs with splenic marginal zone lymphoma. J Vet Intern Med 27:949–954.
  15. 15. Ward, J.M., Taddesse‐Heath, L., Perkins, S.N., et al. (1999) Splenic marginal zone B‐cell and thymic T‐cell lymphomas in p53‐deficient mice. Lab Invest 79:3–14.

Mantle cell lymphoma


Defining the neoplasm


Mantle cell lymphoma (MCL) is an uncommon indolent lymphoma of B lymphocytes seen in animals and humans. It starts focally in the spleen in mantle cell regions. Most cases in dogs occur as a primary splenic lymphoma and lymph nodes are not involved. Two types of MCL are seen in the spleen, a focal mass, which is the more common type, and a generalized or “blastoid” type. MCL was not recognized as an entity in prior classifications because it was included with the small B‐cell non‐Hodgkin’s lymphomas.


In humans, MCL has a slow progressive course that is unresponsive to treatment and therefore the prognosis is not good. In humans MCL has a unique genetic signature arising from a translocation t(11;14) and is evident by the overexpression of cyclin D1. Immunoreactivity for this protein has not been demonstrated in the dog and it is a difficult stain to perform accurately on human tissues.1


Epidemiology, occurrence, and clinical presentation


In one study, 16 MCLs were seen in 1000 cases of canine lymphoma, suggesting that approximately 1–2% of canine lymphomas are of MCL type.2 Fifteen of the 16 cases were in the spleen. Dogs with blastoid type of MCL may have a low‐level gammopathy and recurrent fever that resolve with splenectomy. The latter changes could be related to the multiple foci of necrosis in hemorrhagic neoplastic nodules. Nine dogs were female, 5 were male, and in 2 cases the sex was not indicated. The ages ranged from 6 to 12 years and the mean age was 9.5 years.2 Only four cases were classified as MCL in the 600‐dog study. MCL occurs rarely in the cat and pig. MCL likely occurs in most mammals. In humans it constitutes about 5% of lymphomas, males predominate 3:1, and the disease is seen in elderly patients with a median age of 65–75 years.3–7


Blood and bone marrow

Neoplastic cells are not expected in the bone marrow or blood of dogs with MCL. A mild non‐regenerative anemia may occur. In humans there is involvement of the marrow in about half of cases, particularly in late stages, and neoplastic cells may be found in the blood.7


Spleen

Most cases in dogs are in the spleen, where it is a primary neoplasm that forms a locally expansile mass. Localized masses are obvious because of compression of surrounding tissue. The splenic tumor may have other concurrent lesions, including myelolipoma, plasmacytoma, hemorrhagic infarction, nodular hyperplasia, and the fibrotic reaction previously termed fibrohistiocytic nodules (Figures 7.137.15). Adjacent lymph nodes are usually not affected early in the development of the tumor. In early cases the proliferating neoplastic B lymphocytes will be in separate colonies with focal areas of coalescence. In advanced splenic MCL the infiltrations become contiguous and may be diffuse, resulting in wide separation and thinning of the splenic smooth muscle trabeculae. In the blastoid type of MCL the entire spleen is involved. The lymphoid nodules are about twice the size of a normal germinal center. These nodules enlarge until the pressure of the proliferation causes the central area to undergo ischemic necrosis that is followed by hemorrhage (Figure 7.15A). This type is described in humans3–6 and in dogs.1,2,8 In both types of splenic MCL other areas of the spleen have numerous but non‐neoplastic lymphocytes and plasma cells.

Microraph of mantle cell lymphoma (MCL) in dog’s spleen displaying a mass composed of multiple discrete nodules.
Micrograph of MCL displaying large cells consist of cluster of dendritic cells with fading germinal center.
Micrograph of MCL with the nuclei having coarsely aggregated chromatin. The  nucleoli are small and only apparent in the larger nuclei.
Micrograph of the MCL cells stained with CD79 displaying a fading germinal center. Inset: Micrograph of MCL with CD3 displaying the neoplastic cells which are negative.

Figure 7.13 Mantle cell lymphoma (MCL), spleen, dog. (A) A mass in the spleen of a 6‐year‐old pit bull is composed of multiple discrete nodules that were clonal for B cells. The tumor consists of central lymphoid proliferations surrounded by cuffs of lighter staining fibrous tissue. (B) The focus of large cells is a cluster of dendritic cells in a fading germinal center. The surrounding cells are mature‐appearing but neoplastic lymphoid cells. The diagnosis is dependent on histologic organization of a nodular pattern composed of remnants of germinal centers surrounded by small to intermediate‐size lymphocytes. (C) The nuclei have coarsely aggregated chromatin; nucleoli are small and only apparent in the larger nuclei. The largest nuclei are about 1.5 RBC in diameter. The cytoplasm is minimal. MZL is a differential but compare these nuclei with those from a MZL in Figure 7.11 and note how clear the differences are. (D) The MCL cells are strongly immunostained with CD79a; fading germinal center is at the top. Inset: CD3: The neoplastic cells are negative.

Micrograph of MCL in dog’s lymph node displaying cells in uniform size. Cells are present in multiple lymph nodes in nodular pattern, in the spleen, and in the marrow. Mitotic figures are present (depicted by arrows).
Micrograph of mantle cell lymphoma (MCL) with CD79a.
Micrograph of MCL with CD3 displaying scatteredlarger T cells stained positively through the node, similar to the pattern in the spleen (Figure 7.13). The MCL cells are negative.
Micrograph of MCL through FNA, displaying neoplastic lymphoid cells with rim of deeply basophilic cytoplasm and large nuclei with prominent nucleoli. Anuclear cytoplasm (inset) are scattered in the background.

Figure 7.14 Mantle cell lymphoma (MCL), lymph node, dog. (A) The cells are small and of uniform size. They were present in multiple lymph nodes in a nodular pattern, in the spleen, and in the marrow of this dog. The density of the nuclei indicates cells are small to intermediate with a moderate amount of cytoplasm. Mitotic figures are present in this example but are usually not common (arrows). (B) CD79a: The MCL cells are strongly and uniformly marked. (C) CD3: There is a scattering of larger T cells stained positively through the node, similar to the pattern in the spleen (Figure 7.13). The MCL cells are negative. (D) Cytology of FNA reveals neoplastic lymphoid cells with a rim of deeply basophilic cytoplasm and large nuclei with prominent nucleoli (inset). The larger, pale‐staining nuclei are tumor cells that have lyzed (burst nuclei) and are adhered to one another. This is a relatively common event in cytologic preparations from lymphoma, as are the pieces of anuclear cytoplasm (inset) scattered in the background that look like platelets (“lymphoglandular bodies”). Diagnosis of lymphoma is obvious but classification as MCL requires histology and phenotyping.

Micrograph of mantle cell lymphoma (blastoid type) in dog’s spleen displaying enlarged lymphoid nodules around periarteriolar lymphoid sheaths. Two of the nodules have central areas of hemorrhage and necrosis.
Micrograph of MCL (blastoid type) displaying nuclei having open chromatin pattern with prominent nucleoli indicating a “blastoid” transformation
Micrograph of MCL (balstoid type) stained with CD79a for B cells. The cells are strongly and uniformly marked.
Micrograph of MCL (blastoid type) stained with CD3 for T cells, displaying cells in a fading germinal center. The proliferating neoplastic cells are negative.

Figure 7.15 Mantle cell lymphoma blastoid type, spleen, dog. (A) There is marked enlargement of lymphoid nodules around periarteriolar lymphoid sheaths. Two of the nodules have central areas of hemorrhage and necrosis. The formation of these blood‐filled cysts appears to result from ischemia due to lymphoma proliferation that compresses thin‐walled vessels in the original germinal centers, causing necrosis and bleeding. There is decreased lymphocyte density and marked congestion in the sinuses. (B) Nuclei have open chromatin pattern with prominent nucleoli, indicating this MCL is undergoing a “blastoid” transformation. The amorphous protein at right is follicular hyalinosis in the remaining germinal center. (C) CD79a staining for B cells: The proliferating cells are strongly and uniformly marked. (D) CD3 staining for T cells: The labeled cells at right are in a fading germinal center. The proliferating neoplastic cells are negative.


MCL and MZL occur as primary splenic lymphomas and both are characterized by fading germinal centers surrounded by an enlarging envelope of neoplastic B lymphocytes that create a distinctive nodular pattern (Figure 7.13A,B). In the focal form of MCL there is usually a locally extensive, multifocal proliferation of lymphocytes around end arterioles (Figure 7.13) with several foci starting to coalesce. Nuclei are situated close together as the cells have relatively little cytoplasm. The nuclei have coarsely aggregated chromatin with small nucleoli that are not readily apparent. Nuclei are small, the largest nuclei are about 1.5 RBC in diameter (Figure 7.13). The mitotic count is low. Cytology of FNA reveals the neoplastic lymphoid cells have more cytoplasm and larger nuclei than appreciated in histologic sections (Figure 7.14D). Although a cytologic diagnosis of lymphoma is readily apparent, the classification of MCL requires knowledge of location (spleen), histologic organization (fading germinal centers surrounded by neoplastic lymphocytes), and B‐cell immunophenotype. Neoplastic cells of the blastoid type are larger, cytoplasm is fairly abundant and deeply stained. Nuclei are nearly 2.0 RBC in diameter. Some nuclei are irregularly shaped and have sharp shallow nuclear indentations and prominent nucleoli (Figure 7.15B). There can be 4–6 mitoses/400× field. In the blastoid type the nuclei are larger, more irregular, nucleoli are more prominent and mitoses are more numerous than in the focal type of MCL. In the blastoid type MCL neoplastic cells tend to colonize the endothelial lining of the large muscular veins.


Lymph nodes

MCL is more commonly located in the spleen. When in a node it is characterized by a nodular pattern created by fading germinal centers surrounded by neoplastic B lymphocytes and this is best appreciated in the paracortex. This pattern may not be readily apparent in H&E sections, however it becomes obvious in IHC sections stained with CD79a. In the IHC sections the neoplastic cells are strongly positive and the depleted and unstained germinal centers are more apparent. This pattern can also be seen easily in H&E when tumor foci surround germinal centers that are hypocellular or contain central hyalinosis. The cytoplasm is minimal and the nuclei are closely apposed, which gives lymphoid foci their dark appearance in H&E‐stained histologic sections (Figure 7.13B,C). Nuclei are round and regular and have coarsely aggregated chromatin with minimal parachromatin clear areas. Nucleoli are small and central and unapparent in about half of the cells. Mitoses are not present in most fields at 400 × .


Other organs

The clinical course is indolent; however, in late stages all of the hematopoietic organs may be involved to some degree. In the liver there will be irregular‐sized clusters of neoplastic lymphocytes in the portal and central vein areas.


Cytochemistry and immunohistochemistry


The cells of MCL are uniformly positive with CD79a and CD20 and are negative with CD3.2,8 The IHC staining will highlight the fading germinal centers and nodular pattern better than H&E. In humans, the cells of MCL are strongly positive with cyclin D1.1


Differential diagnosis


Because of its histologic association with the end arterioles in the spleen and fading germinal centers the most likely differential is MZL, which is also an indolent B‐cell lymphoma with a nodular organization. The major difference is that the nucleoli in MZL are large and very prominent and in the focal type of MCL they are small and inconspicuous. In the blastoid type of MCL the nucleoli are larger and centrally located (Figure 7.15). Tumors in lymph nodes favor MZL. MCL should have a light center in the nodules. This is from the remaining germinal center, which is surrounded by the tumor. MZL should have darker staining mantle cells left as remnants and they are surrounded by neoplastic B cells. TZL can look similar via histology and cytology but is readily distinguished with IHC for CD3. Hyperplasia is also a differential. The larger the mass, the more coalescence of nodules that is seen, and the more uniform the cells the more likely the diagnosis is lymphoma. Hyperplasia or a “reactive process” has a mixture of lymphocytes and other cells in the proliferation and this can be appreciated in H&E‐ or IHC‐stained sections. Clonality testing is an option to help distinguish hyperplasia.


Survival time


Too few cases have been followed with accurate clinical outcomes to predict therapies and reliable survival times. A recent study reported 9 cases of MCL treated with splenectomy alone had a 1‐year survival rate of 89%.9 In the 16 cases seen in the ACVP study the range of survival was 126–155 days and the mean survival was 140 days.2 Despite these observed survival times this is considered an indolent lymphoma. All dogs were assumed to have been treated. Survival times for all tumors are affected to some degree by the owner of the dog or cat and their willingness to continue treating versus electing euthanasia. In humans the median survival of patients with MCL is 3 years and the most prognostic factors are patient age, leukemia, and splenomegaly.4–6 Genetic studies have been done in humans to determine changes associated with survival.5,6


References



  1. 1. Valli, V.E., Vernau, W., de Lorimier, L.P., et al. (2006) Canine indolent nodular lymphoma. Vet Pathol 43:241–256.
  2. 2. Valli, V.E., Kass, P., San Myint, M., and Scott, F. (2013) Canine lymphoma: The effect of age, stage of disease, subtype of tumor, mitotic rate and treatment protocol on overall survival. Vet Pathol 50:738–748.
  3. 3. Belaud‐Rotureau, M.A, Parrens, M., Dubud, P., et al. (2002) A comparative analysis of FISH, RT‐PCR, PCR, and immunohistochemistry for the diagnosis of mantle cell lymphomas. Mod Pathol 15:517–525.
  4. 4. Hao, S., Sanger, W., Onciu, M., et al. (2002) Mantle cell lymphoma with 8q24 chromosomal abnormalities: a report of 5 cases with blastoid features. Mod Pathol 15:1266–1272.
  5. 5. Hui, D., Reiman, T., Hanson, J., et al. (2005) Immunohistochemical detection of cdc2 is useful in predicting survival in patients with mantle cell lymphoma. Mod Pathol 18:1223–1231.
  6. 6. Lai, R., Lefresna, S.V., Franko, B., et al. (2006) Immunoglobulin VH somatic hypermutation in mantle cell lymphoma: mutated genotype correlates with better clinical outcome. Mod Pathol 19:1498–1505.
  7. 7. Singleton, T.P., Anderson, M.M., Ross, C.W., and Schnitzer, B. (1999) Leukemic phase of mantle cell lymphoma, blastoid variant. Hematopathology 111:495–500.
  8. 8. Valli, V.E. (2007) Mantle cell lymphoma. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 227–238.
  9. 9. van Stee, L.L., Boston, S.F., Singh, A., et al. (2015) Outcome and prognostic factors for canine splenic lymphoma treated by splenectomy (1995–2011). Vet Surg 44:976–982.

Follicular lymphoma


Defining the neoplasm


Follicular lymphoma is a tumor of B lymphocytes that forms characteristic follicles in affected lymph nodes and spleen. It is a very uncommon lymphoma in animals but is one of the most common lymphomas in humans; more than 30% of all lymphomas in humans are follicular lymphoma and all are B‐cell phenotype.1–4 Follicular lymphoma is one of the nodular indolent lymphomas; others are MZL and MCL. MZL appears to arise in regions with lymphoid hyperplasia.


Follicular lymphoma needs to be distinguished from follicular hyperplasia, which is a common lesion in animals, but interestingly also has an unfavorable outcome, at least in dogs.5 These two lesions will be described and compared in this section as they have similar morphology. Follicular lymphoma, MZL, and MCL all have such a characteristic nodular histologic organization that their B‐cell phenotype can be predicted in H&E‐stained sections. In the review of the REAL/WHO classification of hematopoietic neoplasms the MD pathologists proposed that phenotyping was essential for all human lymphomas except for those with a follicular architecture, as these will be follicular lymphoma and they are all B‐cell.


Epidemiology, occurrence, and clinical presentation


Follicular lymphoma is recognized in most domestic animals but is very uncommon.6,7 In a review of lymphoma in cats there were 6 cases in a group of 6028 and in 1198 cases of bovine lymphoma there were 4 cases of follicular lymphoma.9 One case was found in over 200 cases of equine lymphoma.10 In 600 canine lymphomas there were 4 follicular lymphoma and in 1000 canine cases 0.3% were diagnosed as follicular lymphoma and approximately 4% follicular hyperplasia.5 In 300 canine cases there were only 3 cases of follicular lymphoma and 11 (3.7%) diagnoses of lymphoid hyperplasia, 2 of which were benign follicular hyperplasia (BFH) and 9 benign atypical follicular hyperplasia (BAFH).11 However, dogs with benign hyperplasia had an MST of approximately only 350 days.5,11 Perhaps there were misdiagnoses or follicular hyperplasia is associated with occult but potentially lethal disease processes.5


Animals with follicular lymphoma present with one or more enlarged lymph nodes but they have a normal appetite and level of activity.12 Dogs with follicular lymphoma, MCL, and MZL are often not ill, or only mildly so. Any peripheral node(s) may be involved as well as thoracic nodes, but oddly, follicular lymphoma has not been described in mesenteric nodes. The nodes are mobile and not fixed. Like many other types of lymphoma those in the head and neck region are most likely to be involved.


Bone marrow and blood

The marrow is involved in late stages of follicular lymphoma in all species and is characterized by clusters of follicles, often in subendosteal areas. These follicles have the same histologic characteristics as those in lymph nodes with follicular lymphoma. Neoplastic cells in peripheral blood have not been recognized in animals but they do occur in humans with follicular lymphoma.1–4,12


Lymph node

In both follicular lymphoma and follicular hyperplasia the diagnostic feature is the presence of distinct follicles (Figures 7.167.18). Both can have infiltration of perinodal fat by lymphocytes and both may have areas where the follicles are coalescing and are separated by narrow bands of connective tissue. The more infiltrate that is extranodal, especially if the cells form extranodal follicles, and the more coalescence of nodules, the more follicular lymphoma is favored. The distinction between follicular lymphoma and follicular hyperplasia is interpretative and the following features are used. Try to evaluate the cortical region of the node as this region has germinal centers and within these evaluate for polarity (follicular hyperplasia) or its absence (follicular lymphoma).

Micrograph of atypical follicular hyperplasia (AFH), dog’s lymph node displaying numerous large distinct follicles.
Micrograph of atypical follicular hyperplasia (AFH) with CD79a for B cells, featuring light staining of germinal centers and the surrounding mantle cell cuffs are deeply stained.
Micrograph of AFH with CD3 for T cells displaying regions peripheral to the CD79a-stained cells that are positive and have colored polar cap to the outer region of each follicle. Germinal centers are unstained.

Figure 7.16 Atypical follicular hyperplasia (AFH), lymph node, dog. (A) There are numerous large distinct follicles. Some have coalesced to make follicular structures which are so large that lymphoma is a consideration. Most follicles still retain a rim of darker staining cells that is the mantle cell cuff and the interfollicular region is well defined. Dogs with AFH can have survival times comparable to dogs with lymphoma. (B) CD79a for B cells: There is light staining of germinal centers and the surrounding mantle cell cuffs are deeply stained. (C) CD3 for T cells: The regions peripheral to the CD79a‐stained cells are strongly positive, imparting a brown‐colored polar cap to the outer region of each follicle. T lymphocytes in the interfollicular region are also strongly labeled. Germinal centers are unstained and there is light staining in the mantle cell cuffs. Compare these images with Figure 7.9 and note the homogeneity of lymphocytes populating the nodules in the MZL in Figure 7.9 versus the heterogeneity of lymphocytes in the nodules of hyperplasia here.

Micrograph of follicular lymphoma in dog’s lymph node, displaying marked and uniform proliferation of neoplastic lymphocytes within follicles which imparts a faint nodular pattern.
Micrograph of folicular lymphoma with CD79a. Inset:  Micrograph displaying small blue interfollicular regions in the CD79a-stained section stained positively for T cells with CD3.

Figure 7.17 Follicular lymphoma, lymph node, dog. (A) There is a marked and uniform proliferation of neoplastic lymphocytes within follicles which imparts a faint nodular pattern. The nodules in follicular lymphoma are not as obvious as in MZL or MCL. The uniform proliferation across the entire follicle differentiates follicular lymphoma from MZL and MCL. Mantle cell cuffs cannot be seen. Tingible body macrophages are not prominent. The tumor has infiltrated medullary cords and perinodal fat (top). Although proliferation of lymphocytes beyond the capsule can occur in follicular hyperplasia it is more characteristic of lymphoma. This node was markedly enlarged and is nearly effaced by the tumor. (B) CD79a: The follicles, perinodal region and almost the entire section are uniformly and strongly stained. The small blue interfollicular regions in the CD79a‐stained section are now stained positively for T cells with CD3 (inset).

Micrograph of follicular lymphoma, B-cell, in dog’s spleen, displaying follicles of varying sizes and are surrounded by a narrow cuff of more densely stained cells that are likely T cells.
Micrograph of follicular lymphoma, B cell, with CD79a, displaying round, sharply defined, and deeply and uniformly marked follicles. Thin outer mantle of cells marks for B cells. Inset: Follicle lymphoma with CD3.

Figure 7.18 Follicular lymphoma, B‐cell, spleen, dog. (A) Follicular lymphoma is uncommon in animals, but it is one of the lymphomas that produce a nodular pattern. It is of B‐cell type. The follicles are of varying sizes and are surrounded by a narrow cuff of more densely stained cells that are likely T cells. (B) CD79a: The follicles are round, sharply defined and deeply and uniformly marked. There is a thin outer mantle of cells that also marks for B cells. Inset: CD3: The peripheral cuff of T cells is sharply demarcated; a few lymphocytes in the follicles and scattered cells in the surrounding sinus areas are also positively stained. Lymphoid hyperplasia is a differential, especially since the nodules are not large and there is still histologic organization of B and T cells. Follicular hyperplasia and follicular lymphoma are distinguished by size of the lesions, confluence of nodules, polarity in germinal centers, and if similar lesions are in lymph nodes. Clonality is an option for ambiguous cases.


The following features all favor follicular hyperplasia: retention of nodal architecture, antigen polarity in the follicles with larger lighter stained cells to one side of the follicle usually facing the capsule, the mantle cell cuffs remain and form a dark rim around the germinal center, any perinodal lymphocytes are scattered and do not form follicles, there is a mixture of cell types in the perinodal and interfollicular regions, and there should be numerous plasma cells in medullary cords and interfollicular (Figure 7.16). Tumor and follicles may form in the perinodal fat of follicular lymphoma. Depending on how advanced the follicular lymphoma is the overall cellular proliferation will be greater and the mantle cell cuffs and polarity in the follicle are lost (Figure 7.17). The mantle cell cuffs may be obliterated by tumor cells and their absence (Figure 7.17B) or presence (Figure 7.16B) is highlighted in IHC‐stained sections. Intrafollicular tingible body macrophages should be more prominent in follicular hyperplasia and there are few or none with follicular lymphoma, which seems a contradiction but there is inactivation of the apoptotic gene by the neoplastic cells.


At highest magnifications another feature to look for is the proportion of small cells (centrocytes) and large cells capable of division (centroblasts).12 In follicular lymphoma there is the same proportion of centrocytes and centroblasts across individual follicles and between follicles, imparting a uniform color pattern (Figure 7.17A). In follicular hyperplasia there should be no centrocytes between follicles, only within the germinal centers.12 IHC should identify heterogenity of lymphocytes in follicular hyperplasia and homogeneity in follicular lymphoma. Numerous plasma cells is a feature of follicular hyperplasia.


Most nodes with follicular lymphoma are not diffusely involved, but if nodes are sampled late in the disease then neoplastic follicles occupy the entire node (Figure 7.17). The capsule is thinned and the peripheral sinus is compressed and may be focally invaded. Intrafollicular tingible macrophages, common in many lymphomas, are not prominent in follicular lymphoma, even in nodes heavily laden with tumor, and may be more prominent in germinal centers of follicular hyperplasia. The likely explanation is inactivation of the apoptotic gene by the neoplastic cells. Inactivating the apoptotic caspase enzyme system results in few or no cells dying and therefore the need for tingible body macrophages is reduced or absent.1,2,4


Spleen and liver

Follicular lymphoma is rarely diagnosed in the spleen of animals. The diagnostic features of follicular lymphoma in the spleen are the same as in nodes. There are distinct follicles that have pale germinal centers which are CD79a positive surrounded by a region of darker stained non‐neoplastic T lymphocytes (Figure 7.18A–C). The spleen is not diffusely involved. The splenic sinus between follicles is unaffected and congested.


In late stages follicular lymphoma may involve the liver. Tumor cells are located perivascular in the triads and around central veins.


Cytochemistry and immunohistochemistry


The neoplastic cells in follicles are strongly positive with CD79 alpha and CD20 but are negative with CD3 (Figures 7.17 and 7.18). There may be a few CD3‐positive T cells scattered in neoplastic follicles and in interfollicular regions. IHC will reveal more intrafollicular T lymphocytes in follicular hyperplasia than in follicular lymphoma.


CD79a will label cells in germinal centers of follicular hyperplasia, but in contrast to follicular lymphoma the mantle cell cuffs are present in follicular hyperplasia and are well stained (Figure 7.16B). CD3 for T cells leaves the germinal centers relatively unstained but the lymphocytes in the outer region of each follicle and in the interfollicular region are strongly positive in follicular hyperplasia. This staining pattern may impart a distinct polar cap to each follicle (Figure 7.16C).


In humans there are follicular dendritic cells in the follicles of follicular lymphoma that consistently stain positive with CD21 and CD23 and these are presumably present in follicular lymphoma of animals. Since the neoplastic follicles have produced the BCL2 protein and inactivated their apoptotic system, the cells of follicular lymphoma will have staining for the BCL2 protein not seen in benign follicles.


Differential diagnosis


The light microscopic criteria to differentiate follicular hyperplasia and follicular lymphoma are described in the section under lymph nodes and under IHC. If the differentiation is not clear then consider PARR to determine B‐cell clonality and all the other data in the case. In gray zone cases do not rely on only one test. In addition, benign follicles do not express the BCL2 protein while those that are malignant will be positive to some extent.


Staging


Human follicular lymphoma is graded on the proportions of centroblasts to centrocytes present in each follicle. The larger cells (centroblasts) capable of division increase in numbers as the disease progresses. The WHO grading system for human follicular lymphoma may apply to animals. If there are 0–5 centroblasts/400× field the tumor is graded 1. If there are 6–15 centroblasts/400× field, the tumor is graded 2. If there are over 15 centroblasts/400× field the tumor is graded as 3. These criteria have been altered slightly as the original description was made on oculars with field number (FN) of 20 mm diameter and most oculars now have an FN of 22–25 mm diameter.12 As the FN increases so does the total field‐of‐view area.


Typical survival time


Too few canine follicular lymphomas have been followed to predict survival times with or without treatment, but if they behave in dogs as they do in people then survival times are likely less than 1 year.5 More cases of follicular hyperplasia need to be followed but it appears that dogs with follicular hyperplasia also have survival times of approximately 1 year, suggesting that follicular hyperplasia is an indicator of an underlying and potentially lethal disease process or that it leads to lymphoma or there may be misdiagnoses.12


References



  1. 1. Cook, J., Craig, F., and Swerdlow, S.H. (2003) bcl‐2 expression by multicolor flow cytometric analysis assists in the diagnosis of follicular lymphoma in lymph node and bone marrow. Am J Clin Pathol 119:145–151.
  2. 2. Janikova, A., Tichy, B., Supikova, J., et al. (2011) Gene expression profiling in follicular lymphoma and its implication for clinical practice. Leuk Lymphoma 52:59–68.
  3. 3. Federico, M., Vitolo, U., Zinzani, P.L., et al. (2000) Prognosis of follicular lymphoma: a predictive model based on a retrospective analysis of 987 cases. Blood 95:783–789.
  4. 4. Renard, N., Valladeau, J., Barthelemy, C., et al. (1999) Characterization of germinal center dendritic cells in follicular lymphoma. Exp Hematol 27:1768–1775.
  5. 5. Valli, V.E., Kass, P., San Myint, M., and Scott, F. (2013) Canine lymphoma: The effect of age, stage of disease, subtype of tumor, mitotic rate and treatment protocol on overall survival. Vet Pathol 50:738–748.
  6. 6. Ashley, P.F. and Bowman, L.A. (1999) Symmetric cutaneous necrosis of the hind feet and multicentric follicular lymphoma in a cat. J Am Vet Med Assoc 214:211–214.
  7. 7. Valli, V.E., Vernau, W., de Lorimier, L.P., et al. (2006) Canine indolent nodular lymphoma. Vet Pathol 43:241–256.
  8. 8. Valli, V.E., Jacobs, R.M., Norris, A., et al. (2000) The histologic classification of 602 cases of feline lymphoproliferative disease using the National Cancer Institute working formulation. J Vet Diagn Invest 12:295–306.
  9. 9. Vernau, W., Valli, V.E.O., Dukes, T.W., et al. (1992) Classification of 1,198 cases of bovine lymphoma. Vet Pathol 29:183–195.
  10. 10. Durham, A.C., Pillitteri, C.A., San Myint, M., and Valli, V.E. (2012) Two hundred three cases of equine lymphoma classified according to the World Health Organization (WHO) classification criteria. Vet Pathol 50:86–93.
  11. 11. Valli, V.E., San Myint, M., Barthel, A., et al. (2011) Classification of canine malignant lymphomas according to the World Health Organization criteria. Vet Pathol 48:198–211.
  12. 12. Valli, V.E. (2007) Follicular lymphoma. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 215–227.

Burkitt’s type lymphoma


Defining the neoplasm


Burkitt’s lymphoma is a B‐cell lymphoma which is a well‐characterized disease in humans but not animals.1–3 In veterinary pathology there is no uniform agreement on this diagnosis. The distinction from DLBCL in dogs is accomplished by subjective features of cell size and nuclear and nucleolar features. In humans it is associated with the Epstein–Barr virus (EBV) and dysregulation of the c‐myc gene by one of several chromosomal translocations located at 8q24.3,4 Chronic infection with EBV in humans is usually asymptomatic but a portion of infected individuals co‐infected with malaria will develop lymphomas, most notably Burkitt’s lymphoma.


Serologic evidence indicates that privately owned dogs with and without lymphoma are exposed to a gammaherpesvirus, and further studies implicated its presence in B‐cell lymphomas of dogs and suggested it may be associated with lymphomagenesis in a subset of canine lymphomas.5


Burkitt’s lymphoma is a unique tumor with interesting epidemiologic associations that brought some historical firsts to human hematopathology.2 It was the first tumor curable by chemotherapy, it was the first human tumor to be associated with a virus, and it was one of the first tumors in which epidemiology helped define the tumor and its pathogenesis.


Several types of Burkitt’s lymphoma are recognized, one is the type seen in regions of the world where EBV and malaria are endemic. It is thought that chronic malaria reduces the resistance to EBV, allowing the virus to proliferate and the incidence of lymphoma to increase. An interesting aspect of Burkitt’s lymphoma is the tendency for the tumor to arise in an area of benign lymphoid cell proliferation, as occurs in the mouths of young children and in the breast tissue of girls reaching menarche. This accounts for the tumor arising in the jaws of young Africans and later in the breasts of teenage girls.1 In endemic areas of EBV and malaria Burkitt’s lymphoma characteristically involves the mouth, teeth, jaw bones, and other facial bones, resulting in large tumors and disfigurement.


The descriptions for Burkitt’s lymphoma use some of the classical descriptors to characterize lymphoma: sheets of monotonous neoplastic lymphoid cells with a high mitotic count and a “starry sky” appearance due to numerous tingible body macrophages. In veterinary pathology similar descriptors are used for DLBCL. The classical form of Burkitt’s lymphoma is composed of cells of uniform, intermediate size, described as like “peas in a pod.” In histological preparations the mild separation of the cytoplasm of each cell tends to form a geometric pattern of cell interfaces known as a “squaring off” of cytoplasmic surfaces. This form of the human Burkitt’s lymphoma is associated with prominent cytoplasmic vacuoles. “Burkitt’s‐like” lymphoma is a provisional entity in the WHO classification and is the type recognized in animals. This is a high’grade B‐cell lymphoma that has numerous tingible body macrophages and looks similar to DLBCL.1


Occurrence and clinical presentation


In the ACVP study of 1000 cases of lymphoma in dogs there were 21 cases classified as Burkitt’s‐like lymphoma. The dogs had a mean age at diagnosis of 8.8 years. This lymphoma is seen most often in dogs and cats and likely occurs in the pig, cow, horse, and mouse.6 In humans there is often leukemia and marrow involvement but this is not seen or not recognized in animals. Of all the domestic animals it is described most frequently in cats.7 In the series of 600 feline lymphomas 79 were considered to be Burkitt’s lymphoma or Burkitt’s‐like.7 A neurotrophic form was described in detail in a cat with clinical signs and lesions of peripheral neuropathy.8 The tumor was only found in peripheral nerves even after an autopsy. The tumor and cat were negative for FeLV and multiple human (EBV) and animal herpesviruses. The tumor was of B‐cell phenotype, had intermediate to large cells, and was similar to human diffuse neurolymphomatosis.8


Pathology


Lymph nodes

Nodes are filled with tumor that consists of sheets of monotonous medium‐sized lymphocytes with a high mitotic count (Figure 7.19). A “starry sky” appearance is expected at low magnification and is due to numerous tingible body macrophages (phagocytic cells containing dead apoptotic tumor cells). Macrophages will be fewer where the mitotic count is lower. An old descriptive term of “small non‐cleaved cell” is misleading. The tumor cells are intermediate in size. Tumor cells possess small amount of basophilic cytoplasm. Nuclei have mild anisokaryosis and are up to 1.5 RBC in diameter. Chromatin is aggregated such that it is open and nucleoli are prominent and numerous; often 3–5 nucleoli/nucleus. Cytologic preparations of DLBCL look similar. The mitotic count is high, at up to 10/400× field.

Image described by caption.
Micrograph of Burkitt’s-like lymphoma displaying round nuclei with mild anisokaryosis and are up to 1.5 RBC in diameter and with three mitotic cells in center. Nucleoli are prominent and there are 1–5/nucleus.
Micrograph of Burkitt’s-like lymphoma with CD79a. Inset: Micrograph of the Burkitt’s-like lympoma with CD3.
Image described by caption.

Figure 7.19 Burkitt’s‐like lymphoma, lymph node, dog. (A) Biopsy of popliteal node from a 7‐year‐old German shepherd with lymphadenopathy. A diffuse proliferation of lymphoid cells has effaced normal architecture. This pattern is characteristic of Burkitt’s lymphoma, DLBCL, and high‐grade lymphomas, T‐ or B‐cell. The numerous clear foci are tingible body macrophages, they impart a “moth‐eaten” appearance, and are most prominent in tumors with a high mitotic count. A differential diagnosis is DLBCL. (B) The round nuclei have mild anisokaryosis and are up to 1.5 RBC in diameter. There are three mitotic cells in center. Nucleoli are prominent and there are 1–5/nucleus; there is a moderate volume of cytoplasm. (C) CD79a: The tumor population is strongly and uniformly marked for B cells. Inset, CD3: The neoplastic cells are uniformly negative and only an occasional cell is identified as T‐cell type. (D) Cytologic preparation: There are three cells in mitosis in this small field; nucleoli are present, central, and often multiple (arrows); chromatin is coarsely aggregated. Cytologic preparations of DLBCL look similar.


The capsule of nodes is usually thickened and at least focally invaded by the lymphoid cells. The nodal tissue is friable and ischemic necrosis is not a feature.


Spleen

The spleen is not a target organ in Burkitt’s‐like lymphoma. There is patchy involvement of the sinus areas and irregular foci of neoplastic cells around terminal arterioles.


Other organs

All tissues may be involved, including liver, kidneys, reproductive organs, adrenals, pancreas, marrow, and nodes in both body cavities and peripheral nodes. Facial nerve involvement and blindness are reported anecdotally.


Cytochemistry and immunohistochemistry


The neoplastic cells are uniformly labeled with CD79a and CD20 and are negative with CD3. In humans, the cells are positive with CD10 and BCL6 and are negative with BCL2. The Burkitt’s lymphoma cells have light‐chain restriction and in humans they may have cytoplasmic IgM.


Differential diagnosis


Burkitt’s lymphoma and DLBCL, high‐grade lymphoma of B‐ or T‐cell types, as well as ALL and LBL look similar. DLBCL and Burkitt’s lymphoma are B‐cell lymphomas, the tumors are in multiple nodes and organs, the cells are immature, both have numerous mitotic figures, tumor fills or effaces node, and tingible macrophages are common in each. The differentiation is primarily based on cell size. If there are large cells in every field then the diagnosis is DLBCL. Burkitt’s lymphoma has intermediate‐sized cells. Since there is mild anisokaryosis in Burkitt’s lymphoma it is easy to confuse the predominant intermediate‐sized cells of Burkitt’s lymphoma with a large cell lymphoma. The nucleoli in Burkitt’s lymphoma are numerous, small, and always central. In DLBCL the nucleoli frequently impinge on the nuclear membrane, which is not seen in Burkitt’s lymphoma. The chromatin distribution in Burkitt’s lymphoma is coarsely aggregated and in DLBCL it is dispersed. This difference in nuclei and nucleoli helps distinguish them. Burkitt’s lymphoma has been defined by a translocation t(8;13) at the IGH locus of canine chromososome 8 (CFA 8) and the MYC locus in CFA 13 in the few dogs studied.9,10 If molecular techniques can separate canine tumors in the way they do human lymphomas it would obviously be invaluable to differentiate tumors that are morphologically similar. ALL and LBL B‐cell have smaller cells, intermediate in size, and the nuclear chromatin is dispersed such that nucleoli are not readily apparent.


Evaluation of treatment


Initially at least, the cells of Burkitt’s‐like lymphoma respond to chemotherapy and the nodes are described as “melting away.” The remissions are usually short and the tumor returns with increasing frequency. It is difficult to obtain lasting remission in dogs. Burkitt’s lymphoma is the tumor in human oncology that is most likely to result in tumor lysis syndrome.


Typical survival time


There are few dogs with accurate follow‐up data to predict survival. Twenty‐one dogs with Burkitt’s lymphoma had a range of survival of 14–398 days with an MST of 150 days.11 In the Kaplan–Meier survival curves the data from the animals with Burkitt’s‐like lymphoma were pooled with those of the B‐ and T‐cell lymphoblastic type lymphomas (LBL) with a joint MST of about 225 days.11 Although frequently cited, the survival data in this reference needs to be substantiated with larger numbers of dogs, greater clinical input, and uniform treatments. Another study reported very short survival times of less than one month for dogs with Burkitt’s lymphoma.12 There were only four dogs and all had alimentary involvement.12 Burkitt’s lymphoma, DLBCL, ALL, and LBL are all aggressive in dogs.


References



  1. 1. Diebold, J., Jaffe, E.S., Raphael, M., and Warnke, R.A. (2001) Burkitt lymphoma. In World Health Organization Classification of Tumours, Pathology and Genetics. Tumours of the Haematopoietic and Lymphoid Tissues . IARC Press, Lyon, France, pp. 181–184.
  2. 2. Gascoyne, R.D., Magrath, I.T., and Sehn, L. (2010) Chapter 22. In Non‐Hodgkin Lymphomas , 2nd edn. Wolters Kluwer/Lippincott Williams & Wilkins, Philadelphia, pp. 334–357.
  3. 3. Kanungo, A., Medeiros, J.L., Abruzzo, L.V., and Lin, P. (2006) Lymphoid neoplasms associated with concurrent t(14;18) and 8q24/c‐MYC translocation have a poor prognosis. Mod Pathol 19:25–33.
  4. 4. Nakamura, N., Nakamine, H., Tamaru, J., et al. (2002) The distinction between Burkitt lymphoma and diffuse large B‐cell lymphoma with c‐myc rearrangement. Mod Pathol 15:771–776.
  5. 5. Huang, S.H., Kozak, P.J., Kim, J., et al. (2012) Evidence of an oncogenic gammaherpesvirus in domestic dogs. Virology 427:107–117.
  6. 6. Kovalchuk, A.L., Qi, C.F., Torrey, T.A., et al. (2000) Burkitt lymphoma in the mouse. J Exp Med 192:1183–1190.
  7. 7. Valli, V.E., Jacobs, R.M., Norris, A., et al. (2000) The histologic classification of 602 cases of feline lymphoproliferative disease using the National Cancer Institute working formulation. J Vet Diagn Invest 12:295–306.
  8. 8. Higgins, M.A., Rossmeisl, J.H. Jr., Saunders, G.K., et al. (2008) B‐cell lymphoma in the peripheral nerves of a cat. Vet Pathol 45:54–57.
  9. 9. Frantz, A.M., Sarver, A.L., Ito, D., et al. (2011) Molecular profiling reveals prognostically significant subtypes of canine lymphoma. Vet Pathol 50:693–703.
  10. 10. Breen, M. and Modiano, J. (2008) Evolutionarily conserved cytogenetic changes in hematological malignancies of dogs and humans – man and his best friend share more than companionship. Chromosome Res 16:145–152.
  11. 11. Valli, V.E., Kass, P., San Myint, M., and Scott, F. (2013) Canine lymphoma: The effect of age, stage of disease, subtype of tumor, mitotic rate and treatment protocol on overall survival. Vet Pathol 50:738–748.
  12. 12. Ponce, F., Magnol, J.P., Ledieu, D., et al. (2004) Prognostic significance of morphological subtypes in canine malignant lymphomas during chemotherapy. Vet J 167:158–166.

Diffuse large B‐cell lymphoma


Defining the neoplasm


Diffuse large B‐cell lymphoma (DLBCL) is the most common type of lymphoma in dogs. Almost half of nearly 1700 lymphomas in dogs were of this subtype.1–3 It is also one of the most common lymphomas in all species and is the “classical” form of lymphoma seen frequently in cytologic or histologic specimens, appearing as enlarged lymph nodes filled or effaced by a homogeneous population of large immature lymphoid cells, large nuclei, small to moderate amount of basophilic cytoplasm, a high mitotic count, high‐grade lymphoma, and numerous tingible body macrophages. The diagnosis of lymphoma is made easily, but other immature large cell lymphomas look almost identical in routine specimens and the distinction between the different types of large cell lymphomas requires immunophenotyping, cytology, and histologic organization of the tumor within lymph nodes.


DLBCL can be divided into centroblastic or immunoblastic (single, central nucleolus) based on examination of nuclei and nucleoli. The majority are high‐grade lymphomas. T‐cell‐rich B’cell lymphoma is a subtype of DLBCL, it is common in cats and horses and rare in dogs.4–8 Two other rare variants of DLBCL are mediastinal and intravascular. However, most mediastinal and intravascular lymphomas in dogs are of T‐cell phenotype.9 DLBCL is associated with BLV and FeLV.10 As more parameters, and careful follow‐up information on biologic behavior are collected and correlated with molecular techniques it is likely that DLBCL will be subdivided in animals as it is in humans. Human DLBCL can be subdivided into two subtypes: tumors that are believed to arise from the germinal center B cell are called GCB lymphomas and those arising from cells exiting the germinal center, referred to as activated B cells, are called ABC lymphomas. Expression of different NF‐kB pathway–related genes is part of the molecular classification of human DLBCL: GCB subtype originates from centroblasts and expresses BCL6 and ABC subtype originates from plasmablasts and expresses XBP1, PRDM1, and IRF4.11 This may be the case for canine B‐cell lymphomas, but the data for this separation in canine studies is not uniform.12,13 At least the shared signaling pathways between the two species make the dog an attractive model.11


Epidemiology and occurrence


DLBCL is the most common lymphoma in most domestic species, birds, and many wild animals.14 It is a common type of lymphoma in cats using the REAL/WHO system. Locations may be mediastinal, in URT, in segments of the bowel, or multicentric. By anatomic location, enteric lymphoma has emerged as the most common lymphoma in cats, and most are T‐cell.4–6,15 In dogs, DLBCL accounts for almost 50% of all lymphomas.1,2,16,17 TCRLBCL, a variant of DLBCL, is the most common lymphoma in horses (40–50%).7,8 DLBCL makes up another 8–10% of the equine cases.7


In cattle the large cell lymphomas comprised almost 70% of 1200 cases, but immunophenotyping was not performed.18 Large cell lymphoma, particularly those with cleft nuclei, are usually related to BLV.14 The less common types that occur in calves as leukemia and in 2‐year‐olds in the thymus or skin (mycosis fungoides) are more likely to be T‐cell lymphomas and are not BLV related. In swine, approximately 60% of 136 cases of lymphoma were of large cell type.14 In the cat, DLBCL has been associated with FeLV but there also are DLBCLs that are not virally induced.4,5,14 One study in cats reported that all cats infected with FeLV or FIV had high‐grade lymphomas.5 Six of 7 cats with FIV had B‐cell lymphomas and the phenoptype was split fairly evenly in the 9 cats infected with FeLV.5 In dogs no association has been found between lymphoma of any type and viral infection. In birds there is a constant association of large T‐cell lymphoma affecting peripheral nerves with herpes virus (Marek’s disease).14 HIV lymphomas in humans are also T‐cell lymphomas.


Types of diffuse large cell lymphoma


Mediastinal large B‐cell lymphoma

A variant of large B‐cell lymphoma occurs in humans in the mediastinum and this tumor is the major reason cited for the WHO classification to include topography as part of the disease description. Lymphoma in the mediastinum is relatively common in dogs but it is usually of T‐cell type. Mediastinal B‐cell lymphoma is rarely identified in the dog but may be more frequently found if we immunophenotyped all lymphomas. Approximately one‐third of cats with mediastinal lymphoma were of B‐cell type.4 In the study of 600 cats large B‐cell lymphoma was common and many were in the mediastinum.6 Mediastinal lymphoma is felt to arise in the thymus. Prior to FeLV testing and vaccinating, mediastinal lymphoma in young cats was very common. In dogs the neoplastic cells may be lobated and even multinucleated with much more anisokaryosis than is usually seen in large B‐cell lymphomas. In humans this subtype is very aggressive and requires extensive therapy.


Intravascular large B‐cell lymphoma

Intravascular lymphoma is a rare subtype of lymphoma in which the neoplastic cells are in blood vessels but the tumor does not form neoplastic masses in solid tissues such as nodes or spleen. In humans most are of B‐cell type; in dogs the majority are T‐cell lymphomas but rarely B‐cell intravascular lymphomas are documented.9 Intravascular cell lymphomas seem to have a predilection for the central nervous system.9 The neoplastic cells range from few to numerous and may fill and dilate the veins in the meninges of the brain and spinal cord. The neoplastic cells do not appear to be attached to the vessel wall and they will be intravascular in other organs as well but they are not detected in circulation, at least not with our present methods. Some of the lymphoma cells may be multinucleated with marked anisokaryosis. Tumor cells are accompanied by fibrin thrombi with marked irregular dilation of the veins. Immunophenotyping is required to determine whether they are T‐cell, B‐cell or NK lymphomas. See section on Intravascular large T‐cell lymphoma in this chapter.


Clinical presentation


Most dogs with DLBCL have multicentric lymphoma with a generalized lymphadenopathy, but approximately 10% are diagnosed when only a single peripheral node is enlarged.1,2,14 Up to 20% will involve abdominal nodes or viscera and 10% will be in the skin.1 Large cell lymphomas are common in cats; they may be T‐cell or B‐cell and are found in different anatomic locations.4–6,10,19–21 Nasal lymphoma in cats are almost all large B‐cell lymphomas.4,5,10,19 Large B‐cell lymphomas are common in the stomach and ileal–cecal region of cats.15 The number of cases of mediastinal lymphomas in cats has decreased since the implementation of testing and vaccination for FeLV but it still represents a frequent anatomic site.


Horses may present with multiple subcutaneous tumors (TCRLBCL) or with enlarged abdominal nodes that cause colic.7,8 Reports indicate that multicentric lymphoma with numerous enlarged lymph nodes is the most common lymphoma in horses and the great majority are DLBCL,7,8 but cutaneous lymphoma is also common in horses and unique. The most common problems were weight loss (25 of 37 horses) and ventral edema (21 horses, 19 of which had hypoalbuminemia).7 Immunophenotyping revealed 26 T‐cell and 7 B‐cell lymphomas; 4 could not be classified. Morphology of tumors was heterogeneous (n = 17) and it was difficult to assign definite classifications. Eight tumors had a marked histiocytic component with giant cells.


Multiple tumors were considered TCRLBCL, which is a common lymphoma in horses. Horses can have hemolytic anemia and a huge splenic lymphoma that may weigh several hundred pounds, but these cases are rare.14 In the era before ultrasound, splenic lymphoma was difficult to diagnose because the spleen was not palpable on rectal examination and an important differential for the hemolytic anemia was equine infectious anemia virus.


Pathology


Blood and bone marrow

DLBCL is common but most cases do not have neoplastic cells in circulation on routine microscopic review. A few dogs with DLBCL have tumor cells in circulation, however, the cells in circulation may not have the same morphology as those seen in the solid tumors.15 Some cases will have cytopenia, most commonly thrombocytopenia. Dogs with abnormal cells in peripheral blood or with cytopenia have diffuse or multifocal lymphoma in bone marrow. Twenty to 40% of affected dogs without circulating neoplastic cells or cytopenia will also have neoplastic lymphocytes in bone marrow, but this does not alter the prognosis. Involvement of the marrow is identified by solid areas of neoplastic cell proliferation with displacement of fat cells. Markers to identify DLBCL subtypes that have a propensity to involve bone marrow are not available for dogs as they are in humans.


Of 210 dogs with hematopoietic tumors, 65 were classified as high‐grade lymphoma but with concurrent leukemia.22 Forty‐one of the 65 (63%) were B‐cell type and 24 (37%) were T‐cell type and it is likely that many of the dogs within this group had DLBCL. Anemia was present in 50 of the 65 dogs (77%), neutrophilia was seen in 46%, neutropenia in 11%, lymphocytosis in 86%, thrombocytopenia in 40%, and thrombocytosis in 3%.22


Lymph nodes

One or more peripheral lymph nodes are enlarged, some markedly so. Often the entire node is effaced by a diffuse infiltration. This is the prototype lymphoma that completely fills and effaces lymph nodes with uniform large immature lymphocytes. If seen early, the tumor seems to start in the cortex. The subcapsular sinus is compressed or filled with tumor, the capsule is thin and tumor extends through the capsule into the perinodal tissues. The interior architecture is diffusely involved and medullary cords are replaced by neoplastic cells. Nodes that are not completely effaced will have remnants of normal structures remaining: small darkly stained regions of germinal centers in the outer cortex, outlines of the medullary cords may be evident, subcapsular sinus focally compressed and fibrovascular supporting structures present. The rate at which the tumor progressed can be deduced from the absence or prominence of these normal histologic structures. In rapidly advancing lymphomas there is little medullary structure present, while in slower developing cases the medullary supporting structures will be hyperplastic and more apparent. DLBCL neoplasms are not sclerotic and ischemic necrosis is not present. On gross inspection the nodes are white and soft and cut surfaces bulge. Touch imprints from cut surfaces yield many large, high‐grade type immature lymphocytes (Figure 7.20).

Micrograph of diffuse large B-cell lymphoma, (high-grade, centroblastic) in dog’s lymph node. The nuclei have multiple nucleoli and some of which are located close to or impinge on the nuclear membrane.
Image described by caption.
Micrograph of diffuse large B-cell lymphoma (high-grade, centroblastic) CD79a, displaying elongated negative cells that are endothelial and supporting stromal cells.
Micrograph of diffuse large B-cell lymphoma with CD3, displaying the scattered nonneoplastic cells marked for T cells. Multiple nucleoli in the tumor cells are apparent with the hematoxylin counterstain.

Figure 7.20 Diffuse large B‐cell lymphoma, high‐grade, centroblastic, lymph node, dog. (A) The nuclei are 1.5–2.5× the diameter of a red blood cell and most have multiple nucleoli, some of which are located close to or impinge on the nuclear membrane. The cytoplasm is relatively abundant and of moderate staining density. Two cells are in mitosis, center. (B) Cytological preparation: Cytoplasm is densely basophilic, nuclei have dense, uniformly stained chromatin and there are multiple nucleoli per nucleus. The two cells with more spaces in nuclei and a ropey appearance to the chromatin are undergoing lysis (arrow). Nucleoli are prominent in any lytic cell, even non‐neoplastic cells, but this is an artifact. (C) CD79a: The tumor population is strongly and uniformly marked. The elongated negative cells are endothelial and supporting stromal cells. (D) CD3: Only scattered non‐neoplastic cells marked for T cells are present. Multiple nucleoli in the tumor cells are apparent with the hematoxylin counterstain.


Enlarged nodes are diffusely filled with neoplastic cells in monotonous sheets (Figures 7.207.22). In histologic sections the cells are best interpreted near the tissue margins where fixation has been rapid. DLBCL is composed of large cells, most of the nuclei are 2.0 or more red blood cells in diameter. Nuclei are round to oval and almost never elongated as seen in lymphoblastic lymphoma of B‐ or T‐cell types. Nuclei are rarely cleaved or indented. Nuclei tend to fill the entire cell such that only a small amount of basophilic cytoplasm is seen. Mitotic counts vary considerably from less than 1/400× field to 20 or more. As the mitotic count increases so do macrophages. Tumors with high mitotic counts will have numerous tingible body macrophages (Figure 7.19), seen easily at low magnification as a moth‐eaten appearance. Tumors with a lower mitotic count will have a paucity of these phagocytic macrophages.

Micrograph of diffuse large B-cell lymphoma (high-grade, immunoblastic) in dog’s lymph node with the nuclei in 1.5–2.5 RBC in diameter. The nucleoli are large, very prominent, single and centrally located.
Image described by caption.

Figure 7.21 Diffuse large B‐cell lymphoma, high‐grade, immunoblastic, lymph node, dog. (A) The nuclei are 1.5–2.5 RBC in diameter, generally round, with some nuclei having shallow nuclear indentations. The nucleoli are large, very prominent, single and centrally located. Compare to Figure 7.20A,B. (B) Cytological preparation: Nuclei are round with densely stained chromatin and a single prominent central nucleolus. The cytoplasm is densely basophilic and forms a thin rim around nuclei. There are numerous anuclear cytoplasmic fragments in the background (lymphoglandular bodies). These can be seen in preparations from neoplastic and hyperplastic lymph nodes. Burst nuclei are present at bottom of the image. They do not help with any diagnosis and are simply an artifact of preparation because of the fragility of the tumor cells.

Micrograph of diffuse large B-cell lymphoma with CD79a, displaying the tumor population that are heavily mark. Pale areas are vessels.
Micrograph of diffuse large B-cell lymphoma with  CD3, displaying the tumor population which is unmarked and only a few cells stain positive for T cells.

Figure 7.22 Diffuse large B‐cell lymphoma, same case as Figure 7.21. (A) CD79a: The tumor population is uniformly heavily marked. The pale areas are vessels. (B) CD3: The tumor population is unmarked and only a few cells stain positive for T cells.


DLBCL can be divided into centroblastic or immunoblastic based on examination of nuclei and nucleoli. Neoplastic cells with a single prominent central nucleolus are termed immunoblastic (Figure 7.21A,B). Centroblastic types have multiple nucleoli. These are not as obvious as in immunoblastic types and they are often situated near the nuclear margin (Figure 7.20A,B). Approximately 25% of canine DLBCLs are immunoblastic and 75% centroblastic. Many tumors will have a mixture of each type, and the designation immunoblastic was only used when 90% or more of the nuclei were of this type.1


T‐cell‐rich large B‐cell lymphoma (TCRLBCL) is a type of large B‐cell lymphoma composed predominantly of small, non‐neoplastic T cells with variable numbers of large neoplastic B cells scattered through the tumor. They usually have abundant fine stroma.


Spleen and liver

In most species, the spleen is involved and it may be diffusely enlarged and meaty or contain a single large mass. If the lymphoma is diffuse then cut surfaces will have prominent white to pink, slightly raised foci diffusely spread throughout. All lymphomas should be white on cut surface, however some may be pink or have red regions. This is simply due to congestion, hyperemia, and/or hemorrhage. The neoplastic lymphocytes are identical to DLBCL cells in lymph nodes. Architecturally the smooth muscle trabeculae of the spleen are displaced by the proliferation and the tumor is surrounded by a condensation of smooth muscle and connective tissues. Poorly defined areas of fading germinal centers will have only the dendritic cells remaining. If the spleen is neoplastic then the liver is as well. Tumor cells tend to be periportal and perivascular in most lymphomas rather than sinusoidal, as seen in myeloproliferative diseases or the unique hepatosplenic γδ T‐cell lymphomas. Hepatosplenomegaly can be marked and FNA of the liver or spleen will provide the diagnosis of large cell lymphoma.


Other organs


In late‐stage or advanced cases in dogs almost every lymph node and the internal organs are involved. This is seen in 15–20% of cases. These animals usually have severe cachexia. Skin involvement is seen in about 10% of cases and approximately 5% are considered primary in the spleen or mediastinum. DLBCL can be found in skeletal or cardiac muscle, where the neoplastic cells can be seen dissecting between bands of muscle, pushing the muscle fibers apart without causing necrosis of muscle (see Figure 11.24C,D). A similar pattern of infiltration without causing necrosis or inflammation may be present in fat and other tissues throughout the body. Cases with hepatosplenomegaly, enlarged mesenteric and peripheral lymph nodes, and visceral involvement make a gross differential diagnosis of lymphoma easy and touch imprints will confirm lymphoma.


Cytochemistry and immunohistochemistry


The cells of DLBCL label strongly with CD79 alpha and CD20 and are negative with CD3. In horses, CD20 is a preferable B‐cell marker. Neoplastic cells can readily be stained in cytological preparations after fixation in formalin and with heat‐activated antigen retrieval. Phenotyping can also be performed on slides stained with Wright–Giemsa. These slides need to be decolorized in the antigen retrieval process and then stained via IHC. In the dog, most of the tumors of DLBCL type have surface or cytoplasmic Ig and are positive for light chains.1,2,14 It is likely that in dogs, as in people, there are biologic and morphologic variations in DLBCL that are not presently detectable by our limited immunophenotypic analyses.


Differential diagnosis


The diagnosis of lymphoma is straightforward at gross examination in an autopsy; multiple nodes are enlarged and there may be hepatospenomegaly. The diagnosis of lymphoma from cytology or histopathology is easy and the neoplasm is identified as a large cell lymphoma, likely DLBCL. Differentials for DLBCL include peripheral T‐cell lymphoma of large cell type, T‐cell lymphoma not otherwise specified, LBL, MZL, and Burkitt’s‐type lymphoma; even benign lymphoid proliferations can be considered in some cases, especially those in the gastrointestinal tract. DLBCL and large T’cell lymphoma look identical and are distinguished by immunophenotype. LBL has smaller nuclei. They are intermediate in size with dispersed chromatin that may obscure nuclear detail, while the nuclei of DLBCL are larger, with aggregated chromatin, and prominent nucleoli. MZL is of B‐cell type and the characteristic arrangement of neoplastic cells around fading follicles combined with smaller nuclei, a low mitotic count, and more abundant cytoplasm help to identify it. However, stages of both MZL and DLBCL are morphologically similar and molecular studies indicate the two diseases may represent a continuum rather than separate tumors.12,13 Molecular separation of the two tumors was not distinct.12 Follicular lymphoma is B‐cell but the tumor should form distinct follicles as opposed to diffuse sheets of neoplastic cells as seen in DLBCL. Follicular lymphoma is rare in animals. DLBCL and Burkitt’s lymphoma can look similar morphologically and both are B‐cell lymphomas. Burkitt’s lymphoma has intermediate‐sized cells and DLBCL has large cells. Nucleoli in Burkitt’s lymphoma are numerous, small, and always central. However, in DLBCL the nucleoli are single, and larger in immunoblastic types. In centroblastic DLBCL the multiple nucleoli should impinge on the nuclear membrane, which is not seen in Burkitt’s lymphoma.


DLBCL is a common type of lymphoma in animals and Burkitt’s lymphoma is rare. In the few dogs studied, Burkitt’s lymphoma had a translocation t(8;13) at the IGH locus of canine chromososome 8 (CFA 8) and the MYC locus in CFA 13.12 If confirmed on larger series, having molecular signatures to complement morphology will obviously improve our ability to distinguish between tumors that look similar.


Tumor cell transformation and progression


No cytologic changes are associated with progression of DLBCL. In human large B‐cell lymphoma the hypermutation of the Ig gene is detectable by PCR and the subtype of lymphoma can be related to gene expression profiling.


Survival time


See sections on B‐ and T‐cell CLL and the appendix on lymphoma (p. 961) for additional prognostic information. The majority of DLBCLs are high‐grade lymphomas that eventually cause the death of dogs, cats and horses, usually in less than one year. As a generalization, B‐cell lymphomas have a better prognosis than T‐cell lymphomas, but tumors within each of these two groups are heterogeneous and subdividing them has prognostic implications. Size of tumor cells has been used to subdivide B‐cell tumors. Small cell tumors (SLL, CLL) had MST that was not reached and dogs with large cell B‐lymphocytosis had MST of 130 days.23,24 Medium to large B‐cell lymphomas can be subdivided to help predict outcome via histopathology, survivin expression, or MHC expression. Using histopathology, the MST of B‐cell lymphoma ranged from 15 days for Burkitt’s‐like lymphoma to 510 days for other subtypes.25 Using survivin as a marker, the time to first remission was 171 days for high levels and 321 days for low levels of survivin expression.26 Approximately 160 dogs with B‐cell lymphoma could be divided with flow cytometry by class II MHC expression or cell size to help predict outcome when treated with a multi‐agent chemotherapy protocol. Dogs with low level of expression of class II MHC had decreased adjusted survival times of 120 days versus 314 days for dogs with high expression of class II MHC.23 The subset of dogs with low expression of class II MHC was small. Most dogs had high expression and therefore shorter survival.


The poorer prognosis identified is similar to that observed in humans. The explanation may be decreased presentation of tumor antigen to CD4 T cells, resulting in decreased host immune response. The data from both species suggests that the biology and the immune system of the host, not just the type of neoplasm present, is an important or more important component of the prognosis. Dogs with large B‐cell lymphoma (via flow cytometry) had adjusted MST of 120 days versus 275 days for medium‐sized tumor cells. Surprisingly, a CD34 (or CD5 or CD21)‐positive phenotype was not predictive of prognosis for dogs.23 Previous studies indicated that dogs with large B‐cell leukemia (some of which were likely stage V lymphoma) and CD34‐positive cells had a very poor prognosis, with an MST of 16 days.24 Only about 10% of dogs with B‐cell lymphoma express CD34. These authors concluded that in dogs CD34‐positive lymphoma and CD34‐positive leukemia are different types of lymphoproliferative disease.23 This same study indicated that young age was a negative prognostic factor.


An older and large study found an overall survival time of 330 days in dogs with B‐cell lymphoma treated with multi‐agent chemotherapy.27 A study of 63 dogs with lymphoma treated with a CHOP‐based chemotherapy protocol reported on the stage of lymphoma, histologic types, morphologic groups, and clinical parameters.28 The majority of the dogs had DLBCL and the results indicated that aggressive B‐cell lymphomas had an MST of 256 days. The study correlated remissions with different histologic types as well as indolent and aggressive B‐ and T‐cell lymphomas.28 In the ACVP study the DLBCL cases were divided into centroblastic and immunoblastic types.2 Each of these were subdivided into three groups based on mitotic count as low (0–5 mitoses/400× field), intermediate (6–10), and high (>10) grade. Most dogs received chemotherapy, however some cases were terminated shortly after the diagnosis. Treatments varied from practice to practice and the stage of the lymphoma was not provided. From the available information the MST of dogs treated for DLBCL was less than 1 year. A few cases survived longer and none survived 2 years or longer.


Therefore within the classification of B‐cell lymphoma, most of which will be DLBCL, there are subsets with different prognoses and there are multiple methods that can be used to identify these subsets and predict approximate survival times. In humans, morphologically similar DLBCL can be subdivided by gene expression profiles into GCB and ABC patients that have different survival times.12,13,29 GCB patients have a 5‐year survival of 75% and ABC patients 15% at 5 years. It appears that canine DLBCL or at least large B‐cell lymphomas have germinal center and post‐germinal center subtypes with modestly different survivals; however, the distinction is not as definitive as in humans and needs clarification.12,13 Molecular classifications of human DLBCL are predictive of biologic behavior and treatment outcomes but are based on hundreds to thousands of analyses. Present data from dogs are based on a low number of cases in which the data are promising.


One 2015 study on high‐grade multicentric lymphomas in dogs did not use phenotyping but used cytology to provide prognostic information.3 Presumably these were a mixture of different lymphomas but certainly DLBCL were in the study group. Subjective cytologic features were evaluated in FNA preparations from 20 dogs with high‐grade multicentric lymphoma. In addition, a numerical scoring system was created and morphometry was performed. At diagnosis, multinucleation in the neoplastic cells was associated with decreased survival time and binucleation was associated with shorter remission. None of the other cytologic features evaluated at diagnosis or the scoring system or morphometry were predictive of survival. At relapse, the number of mitoses in lymph node aspirates and the total cytologic scores assigned were greater than those obtained at the time of diagnosis.3 However, the increased total score was due to the increased mitoses and not increased morphologic abnormalities. Mitotic counts as determined in this study did not correlate with survival time or other time points assessed. The overall MST of the dogs followed was 236 days.3 Despite relatively low numbers of cases, no immunophenotyping data, and no staging data, the results suggest that cytologic evaluation for bi‐ and multinucleation at the time of diagnosis has predictive value for survival. Cytology of good‐quality preparations can diagnose 80–90% of lymphomas. Incorporating assessment of bi‐ and multinucleation may provide additional information to help oncologists and owners make decisions. The original paper should be read for the many details and comparisons that were studied.3


Lymphomas can be classified by pathologists as outlined in Box 7.1 and DLBCL in animals can be divided into centroblastic and immunoblastic types. Molecular techniques are being used to subdivide these and other lymphomas to determine their utility as models for similar diseases in humans.12,13,29,30 For future studies we need standardized classification of lymphomas between pathologists and for clinicians to standardize treatment protocols. These data will then need to be correlated with accurate clinical follow‐up, including autopsies on a large number of cases, if we are to determine reliable survival times, prognoses, and effective therapeutic protocols. It is unlikely there will be a sufficient number of cases in which no treatment is received and the dogs are allowed to live out their lives to compare survival times between treated and nontreated dogs. A confounder in all types of clinical studies will be the level of commitment to continue therapy and how different clinicians and owners will use the option of euthanasia.


References



  1. 1. Valli, V.E., San Myint, M., Barthel, A., et al. (2011) Classification of canine malignant lymphomas according to the World Health Organization criteria. Vet Pathol 48:198–211.
  2. 2. Valli, V.E., Kass, P., San Myint, M., and Scott, F. (2013) Canine lymphoma: The effect of age, stage of disease, subtype of tumor, mitotic rate and treatment protocol on overall survival. Vet Pathol 50:738–748.
  3. 3. Munasinghe, L.I., Kidney, B.A., MacDonald‐Dickinson, V., et al. (2015) Evaluation of lymph node aspirates at diagnosis and relapse in dogs with high‐grade multicentric lymphoma and comparison with survival time. Vet Clin Pathol 144:310–319.
  4. 4. Sato, H., Fujino,Y., Chino, J., et al. (2014) Prognostic analyses on anatomical and morphological classification of feline lymphoma. J Vet Med Sci 76:807–811.
  5. 5. Chino, J., Fujino, Y., Kobayashi, T., et al. (2013) Cytomorphological and immunological classification of feline lymphomas: clinicopathological features of 76 cases. J Vet Med Sci 75:701–706.
  6. 6. Valli, V.E., Jacobs, R.M., Norris, A., et al. (2000) The histologic classification of 602 cases of feline lymphoproliferative disease using the National Cancer Institute working formulation. J Vet Diagn Invest 12:295–306.
  7. 7. Meyer, J., DeLay, J., and Bienzle, D. (2006) Clinical, laboratory, and histopathologic features of equine lymphoma. Vet Pathol 43:914–924.
  8. 8. Durham, A.C., Pillitteri, C.A., San Myint, M., and Valli, V.E. (2012) Two hundred three cases of equine lymphoma classified according to the World Health Organization (WHO) classification criteria. Vet Pathol 50:86–93.
  9. 9. McDonough, S.P., Van Winkle, T.J., Valentine, B.A., et al. (2002) Clinicopathological and immunophenotypical features of canine intravascular lymphoma (malignant angioendotheliomatosis). J Comp Pathol 126:277–288.
  10. 10. Santagostinol, S.F, Mortellarol, C.M., Boracchi, P., et al. (2015) Feline upper respiratory tract lymphoma: site, cyto‐histology, phenotype, FeLV expression, and prognosis. Vet Pathol 52:250–259.
  11. 11. Mudaliar, M.A., Haggart, R.D., Miele, G., et al. (2013) Comparative gene expression profiling identifies common molecular signatures of NF‐kB activation in canine and human diffuse large B cell lymphoma (DLBCL). PLOS ONE 8:e72591.
  12. 12. Frantz, A.M., Sarver, A.L., Ito, D., et al. (2011) Molecular profiling reveals prognostically significant subtypes of canine lymphoma Vet Pathol 50:693–703.
  13. 13. Richards, K.L., Motsinger‐Reif, A.A., Chen, H.W., et al. (2013) Gene profiling of canine B‐cell lymphoma reveals germinal center and post germinal center subtypes with different survival times, modeling human DLBCL. Cancer Res 73:5029–5039.
  14. 14. Valli, V.E. (2007) Diffuse large B‐cell lymphoma. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 238–258.
  15. 15. Moore, P.F., Moore, P.F., Rodriguez‐Bertos, A., and Kass, P.H. (2012) Feline gastrointestinal lymphoma: mucosal architecture, immunopheotype, and molecular clonality. Vet Pathol 49:658–668.
  16. 16. Ponce, F., Marchal, T., Magnol, J.P., et al. (2010) A morphological study of 608 cases of canine malignant lymphoma in France with a focus on comparative similarities between canine and human lymphoma morphology. Vet Pathol 47:414–433.
  17. 17. Grindem, C.B., Page, R.L., Ammerman, B.E., et al. (1998) Immunophenotypic comparison of blood and lymph node from dogs with lymphoma. Vet Clin Pathol 27:16–20.
  18. 18. Vernau, W., Valli, V.E.O., Dukes, T.W., et al. (1992) Classification of 1,198 cases of bovine lymphoma Vet Pathol 29:183–195.
  19. 19. Little, L., Patel, R., and Goldschmidt, M. (2007) Nasal and nasopharyngeal lymphoma in cats: 50 cases (1989–2005). Vet Pathol 44:885–892.
  20. 20. Flatland, B., Fry, M.M., Newman, S.J., et al. (2008) Large anaplastic spinal B‐cell lymphoma in a cat. Vet Clin Pathol 37:389–396.
  21. 21. Higgins, M.A., Rossmeisl, J.H. Jr., Saunders, G.K., et al. (2008) B‐cell lymphoma in the peripheral nerves of a cat. Vet Pathol 45:54–57.
  22. 22. Tasca, S., Carlil, E., Caldin, M., et al. (2009) Hematologic abnormalities and flow cytometric immunophenotyping results in dogs with hematopoietic neoplasia: 210 cases (2002–2006). Vet Clin Pathol 38:2–12.
  23. 23. Rao, S., Lana, S., Eickhoff, J., et al. (2011) Class II major histocompatibility complex expression and cell size independently predict survival in canine B‐cell lymphoma. J Vet Intern Med 25:1097–1105.
  24. 24. Williams, M.J., Avery, A.C., Lana, S.E., et al. (2008) Canine lymphoproliferative disease characterized by lymphocytosis: immunophenotypic markers of prognosis. J Vet Intern Med 22:596–601.
  25. 25. Ponce, F., Magnol, J.P., Ledieu, D., et al. (2004) Prognostic significance of morphological subtypes in canine malignant lymphomas during chemotherapy. Vet J 167:158–166.
  26. 26. Rebhun, R.B., Lana, S.E., Ehrhart, E.J., et al. (2008) Comparative analysis of survivin expression in untreated and relapsed canine lymphoma. J Vet Intern Med 22:989–995.
  27. 27. Ruslander, D.A., Gebhard, D.H., Tompkins, M.B., et al. (1997) Immunophenotypic characterization of canine lymphoproliferative disorders. In Vivo 11:169–172.
  28. 28. Aresu, L., Martini, V., Rossi, F., et al. (2013) Canine indolent and aggressive lymphoma: clinical spectrum with histologic correlation. Vet Comp Oncol DOI: 10.1111/vc0.12048
  29. 29. Ito, D., Frantz, A.M., and Modiano, J. (2014) Canine lymphoma as a comparative model for human non‐Hodgkins lymphoma: recent progress and application. Vet Immunol Immunopathol 159:192–201.
  30. 30. McCaw, D.L, Chan, A.S, Stegner, A.L., et al. (2007) Proteomics of canine lymphoma identifies potential cancer‐specific protein markers. Clin Cancer Res 13:2496–2503.

T‐cell‐rich large B‐cell lymphoma


Definition of the neoplasm


T‐cell‐rich large B‐cell lymphoma (TCRLBCL) is a B‐cell lymphoma but the diagnostic characteristic is a mixture of lymphocytes and other cells. Early TCRLBCL may have 80–90% small to intermediate‐sized non‐neoplastic T cells, with the rest of the cells being neoplastic large B cells, histiocytes, and connective tissue cells. The neoplastic large B cells gradually increase in proportion to the other cells. TCRLBCL is the or one of the most common lymphomas in horses.1–4 It is fairly common in cats5 but is uncommon in dogs (1% of all lymphomas).6


In cats this lymphoma has been called “feline Hodgkin’s disease” because of the way the tumor spreads and the presence of large, atypical, binucleated B‐cell lymphocytes that vaguely resemble the Reed–Sternberg cells of human Hodgkin’s lymphoma.5 The characteristic that most resembles the human tumor in cats is that the disease spreads only to contiguous lymph nodes and does not skip to nodes in other anatomic regions.5


Epidemiology, occurrence, and clinical presentation


TCRLBCL constitutes about 10% of all feline lymphomas. Most cats are mature, they are in good body condition, typically without loss of weight or appetite.5 Oddly, the presentation is almost always for a single enlarged node in the neck area, usually in a submandibular node and more often on the right side5 (Figure 7.23). Approximately 25% of cats will have multicentric tumors in almost any organ. In the dog it is seen in a peripheral node but may appear in any area of the body, even in the spinal canal.

Micrograph of T-cell-rich large B-cell lymphoma (TCRLBCL) in cat’s lymph node, displaying a mixture of large and small cells with the large cells having abundant cytoplasm and a single large nucleus and nucleolus.
Image described by caption.

Figure 7.23 T‐cell‐rich large B‐cell lymphoma (TCRLBCL), lymph node, cat. (A) There is a mixture of large and small cells with the large cells having abundant cytoplasm and a single large nucleus and nucleolus. Single cell necrosis is a characteristic of this lymphoma, as is the fine sclerosis dissecting through the tumor, which is why these tumors are solid and not soft on gross examination. (B) TCRLBCL cytological preparation from a lymph node of a different cat has primarily small and intermediate‐sized lymphocytes with a few larger cells that have large nuclei, prominent nucleoli, and abundant cytoplasm. Cytologic and histologic patterns for TCRLBCL are a heterogeneous mix of large and small lymphocytes and other cells, which is not a pattern typical of lymphomas. Therefore the correct diagnosis may be missed on initial exam.


In horses, TCRLBCL is the most common lymphoma, accounting for approximately 40% of equine lymphomas.1–4 The mean age is 10 years and horses appear in good health with minimal loss of weight or appetite.1 This lymphoma shares the slow progression seen in cats but has very different clinical presentations.1–5,7,8 In horses this lymphoma is unique in that the most common problem is subcutaneous nodules (see Figure 7.25). There may be a few or up to 100 or more discoid shaped tumors irregularly distributed over the entire body but most are concentrated at the lower area of the neck.1 These lesions are not in the skin but appear to arise along lymphatics in the subcutis. The subcutaneous tumors are typically 2–3 cm in diameter but some may become large, ulcerate the skin and form a mass weighing up to 30 pounds. The skin lesions have regressed during pregnancy and recurred following parturition attributed to progesterone receptors on the neoplastic cells.7,8 The other presentation is multicentric lymphoma with tumors in the gastrointestinal tract and many other organs. If only a few nodules are found and they are excised fully this has proven an effective treatment in some cases.4 Those cases in which the tumor does not recur post excision have survival times of up to 10 years.4


Pathology


Blood and bone marrow

Bone marrow was only noted to be involved in 1 of 200 equine cases.3 Similar results seem to be the case in cats and dogs, with only minor and focal involvement of the marrow. There may be mild to moderate anemia in advanced cases but this may be due to anemia of chronic inflammatory diseases rather than direct infiltration in the marrow.


Lymph nodes or skin nodules

The diagnosis rests on the recognition of a mixed population of cells with a few characteristic large binucleated cells with large central nucleoli and abundant cytoplasm (Figure 7.23). Unlike most lymphomas, TCRLBCL is not a homogeneous population of uniform lymphoid cells. Early lesions have more small cells and are T’cell‐rich. As the tumor progresses the larger B cells increase disproportionately (Figures 7.237.27). B cells may predominate in some regions of the tumor. The increase of B cells is usually accompanied by increased numbers of macrophages. Histiocytic cells are very obvious in some of the equine subcutaneous tumors and they will be accompanied by multinucleated giant cells (Figure 7.25). Some giant cells may have thin, crystalline‐like cytoplasmic inclusions.

Micrograph of T-cell-rich large B-cell lymphoma (TCRLBCL) in cat, with CD20. The large cells are selectively marked and the small T cells are unlabeled.
Micrograph of TCRLBCL with CD3 featuring numerous and strongly positive small and intermediate sized T cells.

Figure 7.24 T‐cell‐rich large B‐cell lymphoma (TCRLBCL), cat. (A) CD20: The large cells are selectively marked and the small T cells are unlabeled. (B) CD3: Small and intermediate‐sized T cells are numerous and strongly positive. The relatively few cells with large nuclei and the vascular and connective tissue cells are negative. When slides labeled with CD20 or CD3 were examined grossly, both stains appeared to be strongly positive. However, on light microscopic evaluation each antibody labeled different cell populations.

Micrograph of TCRLBCL in horse’s lymph node, displaying tumor composed of predominantly small cells, T lymphocytes with darkly stained nuclei and fewer large cells. Inset: TCRLBCL with multinucleated giant cells.

Figure 7.25 T‐cell‐rich large B‐cell lymphoma (TCRLBCL), lymph node, horse. The tumor is a mixture of predominantly small cells, T lymphocytes with darkly stained nuclei and fewer large cells, and B lymphocytes with abundant cytoplasm and large, prominent nuclei. Most of the large cells have nuclei with single or multiple nucleoli. Inset: Multinucleated giant cells are a feature of TCRLBCL in horses. This is the most common lymphoma of horses.

Micrograph of T-cell-rich large B-cell lymphoma (TCRLBCL) in horse’s lymph node, with CD79a.
Micrograph of T-cell-rich large B-cell lymphoma (TCRLBCL) in horse’s lymph node with CD3, featuring uniform and diffuse staining of the small non-neoplastic T lymphocytes.

Figure 7.26 T‐cell‐rich large B‐cell lymphoma (TCRLBCL), lymph node, horse. (A) CD79a: The large cells are variably marked and the small cells are universally unlabeled. CD20 is a better B‐cell marker in the horse; consider using multiple antibodies to clarify immunophenotype or flow cytometry. (B) CD3: There is uniform and diffuse staining of the small non‐neoplastic T lymphocytes. In many cases the T lymphocytes may predominate.

Micrograph of the T‐cell‐rich large B‐cell lymphoma (TCRLBCL) in a dog's lymph node.

Figure 7.27 T‐cell‐rich large B‐cell lymphoma (TCRLBCL), lymph node, dog. The tumor is a mixture of intermediate and small cells with a few very large cells and large nuclei. The pyknotic large cell (left) is a tumor cell undergoing single cell necrosis. Although this is a B‐cell lymphoma, T cells may be the predominant lymphoid cell, at least in different stages of the tumors progression.


An unusual, yet characteristic part of this lymphoma are the large B cells that can have marked anisokaryosis (Figures 7.23, 7.27, and 7.28). The nuclei in these cells will range from 2 RBC in diameter to 3–5. The chromatin of the large B cells is peripheralized, which accentuates a single large central nucleolus. The large cells are often binucleated and a few cells will have 3–4 nuclei. Some large cells are necrotic and have contracted brightly pink cytoplasm and a pyknotic nucleus. Necrosis of individual tumor cells should be looked for as this feature is always present (Figure 7.23).

Micrograph of T-cell-rich large B-cell lymphoma displaying large cells and cells of intermediate size strongly marked and assumed to be tumors.
Micrograph of T-cell-rich large B-cell lymphoma displaying the small and large neoplastic cells marked and unlabeled, respectively.

Figure 7.28 T‐cell‐rich large B‐cell lymphoma (TCRLBCL), lymph node, dog. (A) CD20: Tlarge cells are strongly marked as well as cells of intermediate size that are assumed to be of the tumor population. (B) CD3: The smaller neoplastic cells are lightly to moderately marked and the large neoplastic cells (center and left) are unlabeled. This is the typical immunostaining pattern (signature) for this type of lymphoma.


In all species there is fine sclerosis which renders the tumor cohesive (Figure 7.23). The nodes are firm to palpation and cut surfaces are solid, the typical soft white appearance of lymphoma is not present.


Spleen and other organs

There may be focal tumors in the spleen of cats but the spleen is rarely involved in other species. TCRLBCL is a lymphoma primarily of lymph nodes but in late stages liver and kidney may be affected, at least in cats. Subcutaneous tumors are unique to horses.


Cytochemistry and immunohistochemistry


The patterns seen with IHC are unusual because of the heterogeneity of cells within a tumor and different tumors from the same animal may vary in the proportion of B‐ versus T lymphocytes. The large B cells mark consistently with CD20 and less frequently with CD79 alpha (Figures 7.24, 7.26, and 7.28). CD20 is preferred for horses. The smaller T cells are positive with CD3 and in some cases, or at some stages, CD3‐positive cells will be predominant (Figure 7.24). IHC for the small T cells accentuates the large nuclei of the unstained B cells. IHC is useful to identify cell types but it does not distinguish neoplastic and non‐neoplastic cells. If histochemical stains are done for reticulin there are diffuse fine fibrillar fibers in all areas of nodal and subcutaneous tumors.


Differential diagnosis


The main differentials for TCRLBCL are hyperplasia or an infectious agent. Some cases, especially in horses will have so many multinucleated giant cells and histiocytic cells that stains for acid‐fast organisms or fungi are performed. These will be negative. Tumors with foci of ischemic necrosis can look inflammatory. Awareness that TCRLBCL is a mixture of cells combined with the large nuclei and anisokaryosis that is present in the neoplastic large cell population helps focus the diagnosis on lymphoma. A high index of suspicion for this lymphoma in cats and horses is helpful, especially if the lesion is a subcutaneous nodule from a horse. The large nuclei, binucleation, and prominent nucleoli in large cells are further indications of neoplasia. Clonality detection by PCR should be able to identify clonal antigen receptor genes for B cells if the clonal expansion is greater than 1%. However, the sensitivity and specificity of these assays vary with the expertise of the laboratory. This is a tumor in which expertise with reagents and interpretation is needed.


Evaluation of treatment and survival


Because of the slow rate of progression and the fact that in the cat and dog most of the early cases involve a single node, excision is a consideration. However tumors tend to recur in the same area as the excision. If there is only one or a few subcutaneous tumors then excision may be effective in horses. As stated earlier, some horses respond well to surgical excision and will have long survival times.4 The tumors may be so numerous that excision is not possible. Clinical oncology texts and references provide treatment recommendations. In the ACVP study of canine lymphoma there were 9 cases of TCRLBCL.6 These cases had a surprisingly short survival with a range of 89–105 days and a mean of 97 days; however, the follow‐up data in this study need to be confirmed.


Angiocentric B‐cell lymphoma with reactive T cells (lymphomatoid granulomatosis)


Lymphomatoid granulomatosis was the name first used for this unusual lymphoma. This name is still used and it is also classified as angiocentric B‐cell lymphoma. The latter name captures the characteristic histologic distribution of the tumor and part of the oncogenesis of the disease. The neoplasm has angiocentric and angiodestructive histologic patterns.4,9–12 The primary location of this tumor is lung, which is exceedingly unusual for lymphoma. However, it can be found in almost any tissue. The tumor is heterogeneous morphologically, cellularly, and phenotypically. The large B cells are believed to be the neoplastic population and the numerous T cells a cytotoxic response. Morphologically, it resembles TCRLBCL and the tumor may progress to large cell lymphoma.


In humans, the disease is associated with EBV infection. There are human cases in which immunosuppressive treatments caused latent virus to express and induce the tumor, as well as regression of the tumor following removal of the immunosuppressive drugs.4 The disease in dogs has been compared to lymphomatoid granulomatosis and the pulmonary form of Hodgkin’s disease in humans.11 Regardless of name(s), it is a lymphoid neoplasm of humans and animals that needs to be characterized further.


Grossly and histologically it is well described.9–12 It forms one or more discrete masses in the lung of dogs that can be up to 10 cm or greater in size. It is often only in one lobe, usually a caudal lobe, but there can be multiple masses through the lung and it may be found in other organs. The natural surface is smooth to lobulated. Cut surfaces are mottled and blood vessels or airways can be seen trapped in the tumor. In addition to the grossly visible tumor(s), other lesions will be found surrounding blood vessels on histologic sections taken from lung that appeared normal.


Histologically, the diagnostic feature is an angiocentric pattern of heterogenous cells that surround or invade blood vessels (Figures 7.297.32). There is an irregular mixture of large and small lymphocytes of mixed phenotype accompanied by binucleate, Reed–Sternberg‐like cells and atypical multinucleated giant cells that have long cytoplasmic tendrils (Figures 7.30 and 7.31). Nuclei in giant cells will be crowded together and in some cells there appears to be hundreds of nuclei. Eosinophils will be numerous in some cases and if histochemical stains are applied many mast cells will be found scattered through the tumor. Histiocytic cells and plasma cells are common. Somewhat like TCRLBCL, the mixed cellular infiltration in this disease looks inflammatory on first examination. A distinguishing feature from inflammation is that there are no areas of intensity of the inflammatory cells to form microabscesses or granulomas. Despite the mixture of cell types, the infiltration is uniform and the lesion dissects through tissues the way tumors do. There can be large areas of ischemic necrosis. It is seen primarily in the lung but can occur in the abdominal cavity.

Micrograph of an angiocentric B‐cell lymphoma with reactive T cells in the lymphomatoid granulomatosis of a dog's lung. Large foci of necrosis are evident in the center and lower left.

Figure 7.29 Angiocentric B‐cell lymphoma with reactive T cells, Lymphomatoid granulomatosis, lung, dog. The tumor formed multiple large, non‐encapsulated masses that compressed surrounding alveoli. There are large foci of necrosis in center and lower left. Clear foci within the tumor are blood vessels encased by tumor cells: angiocentric pattern.

Micrograph of a region of a tumor where the giant cells are numerous. Irregular cytoplasmic borders are evident as well as fine sclerosis and a mixture of undifferentiated mononuclear cells in surrounding tissue.

Figure 7.30 This is a region of the tumor where the giant cells were numerous. They are highly atypical and have very irregular cytoplasmic borders. Nuclei are crowded together, some cells will have what appears to be hundreds of nuclei. The surrounding tissues have fine sclerosis and a mixture of undifferentiated mononuclear cells scattered between reticulin fibers.

Micrograph of angiocentric B-cell lymphoma in a dog. The round to oval and undifferentiated nuclei of most of the mononuclear cells are evident.

Figure 7.31 Angiocentric B‐cell lymphoma, dog. Most of the mononuclear cells have round to oval nuclei and are undifferentiated. Two abnormal mitotic figures are subjacent to the giant cell. Eosinophils were not frequent in this case but can be numerous. Histochemical stains for mast cells will reveal a surprising number scattered through the tumor.

Micrograph of angiocentric B‐cell lymphoma in a dog. The characteristic angiocentric distribution of heterogeneous cells surrounding or invading the walls of blood vessels is evident in the tumors.

Figure 7.32 Angiocentric B‐cell lymphoma, lung, dog. These tumors have a characteristic angiocentric distribution of heterogeneous cells surrounding or invading the walls of blood vessels. The angiocentric and angiodestructive pattern may result in ischemia and necrosis to adjacent tissues. Most of these tumors are seen in the lungs of dogs but they can be present in visceral organs of animals and humans.


In humans this tumor is associated with EBV‐positive large B cells but no type of causation has been defined for dogs. T cells are present in the human lesion, they are not clonal and are believed to be a cytotoxic response. In humans, the large B cells have been shown to have rearranged Ig genes and some may progress into large B‐cell lymphomas. The lesions are graded in humans based on the number of large EBV‐positive B lymphocytes and the higher grades could be a variant of TCRLBCL.12 The immunophenotypic characteristics of the canine disease have been reported and need further clarification.10,11 The cellular infiltrates are positive for B’cell and T‐cell antigens and the large Reed–Sternberg‐like cells were positive for CD15 and CD30, which is a feature of Hodgkin’s disease.11


References



  1. 1. Kelley, L.C. and Mahaffey, E.A. (1998) Equine malignant lymphoma: morphologic and immunohistochemical classification. Vet Pathol 35:241–252.
  2. 2. Meyer, J., DeLay, J., and Bienzle, D. (2006) Clinical, laboratory, and histopathologic features of equine lymphoma. Vet Pathol 43:914–924.
  3. 3. Durham, A.C., Pillitteri, C.A., San Myint, M., and Valli, V.E. (2012) Two hundred three cases of equine lymphoma classified according to the World Health Organization (WHO) classification criteria. Vet Pathol 50:86–93.
  4. 4. Miller, C.A., Durham,A.C., Schaffer, P.A., et al. (2015) Classification and clinical features in 88 cases of equine cutaneous lymphoma. J Vet Diagn Invest 27:86–91.
  5. 5. Valli, V.E. (2007) T‐cell rich large B‐cell lymphoma. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 260–273.
  6. 6. Valli, V.E., San Myint, M., Barthel, A., et al. (2011) Classification of canine malignant lymphomas according to the World Health Organization criteria. Vet Pathol 48:198–211.
  7. 7. Henson, K.L., Alleman, R., Kelley, L.C., and Mahaffey, E.A. (2000) Immunohistochemical characterization of estrogen and progesterone receptors in lymphoma of horses. Vet Clin Pathol 29:40–46.
  8. 8. Henson, K.L., Alleman, A.R., Cutler, T.J., et al. (1998) Regression of subcutaneous lymphoma following removal of an ovarian granulose‐theca cell tumor in a horse. J Am Vet Med Assoc 9:1419–1422.
  9. 9. Jaffe, E.S. and Wilson, W.H. (1997) Lymphomatoid granulomatosis: pathogenesis, pathology and clinical implications. Lymphoma 30:233–248.
  10. 10. Smith, K.C., Day, M.J., Shaw, S.C., et al. (1996) Canine lymphomatoid granulomatosis: an immunophenotypic analysis of three cases. J Comp Pathol 115:129–138.
  11. 11. Park, H.‐M., Hwang, D.‐N., Kang, B.‐T., et al. (2007) Pulmonary lymphomatoid granulomatosis in a dog: evidence of immunophenotypic diversity and relationship to human pulmonary lymphomatoid granulomatosis and pulmonary Hodgkin’s disease. Vet Pathol 44:921–923.
  12. 12. Song, S.Y., Pittaluga, S., Dunleavy, K.,.et al. (2015) Lymphomatoid granulomatosis – a single institute experience: pathologic findings and clinical correlations. Am J Surg Pathol 39:141–156.

T‐CELL LYMPHOMAS


Precursor T‐cell lymphoblastic leukemia/lymphoma


Defining the neoplasms


Acute lymphoblastic leukemia (ALL) is a malignancy arising in the bone marrow. The neoplastic cells can be of T‐cell (T‐ALL) or B’cell (B‐ALL) origin. They look identical cytologically, therefore immunophenotyping is needed to identify the correct cell of origin.1–9 Lymphoblastic lymphoma (LBL) is a lymphoid malignancy that arises in peripheral nodes and it also can be of T‐cell (T‐LBL) or B‐cell (B‐LBL) origin. Canine B‐cell (<1%) and T‐cell (3%) LBL are uncommon.2,7,9,10 They are also uncommon in cats and horses, at less than 1%. See the section on B‐cell lymphoblastic tumors in this chapter for comments that overlap with T‐ALL and T‐LBL.


Lymphoblasts can be defined by functional, immunologic, or morphologic parameters. Lymphoblastic is a classification in the WHO/REAL system for leukemia and lymphoma. It is a term we have used imprecisely in veterinary pathology and very likely it has different connotations for different individuals. By using the term in a diagnosis it helps to classify the tumor but it does not mean the cell lines function as true lymphoblasts. In veterinary pathology it has been used generically to describe large immature lymphoid cells with large nuclei and nucleoli that appeared “blastic,” and which formed tumors that effaced lymph nodes. Many of these were likely DLBCL or at least large cell lymphomas, B‐ or T‐cell, and not lymphoblastic tumors. The designation lymphoblastic based on morphology should be reserved for cells that are intermediate in size, larger than a mature lymphocyte (nuclei about 1 RBC), and smaller than an immature lymphocyte (nuclei 2–3 RBC).


Lymphoblasts have minimal cytoplasm, the nuclei are approximately 1.5 RBC in diameter, and they have uniformly dispersed nuclear chromatin such that nucleoli are not obvious. Lymphoblasts are not large lymphoid cells with open chromatin and prominent nucleoli. Throughout this chapter the acronyms ALL and LBL are often used as B‐ versus T‐cell origin is either not clarified or there are conflicting data in the literature.


Leukemia

ALL originates in the bone marrow, more than 30% of the marrow is lymphoblasts, neoplastic cells are in circulation, and there are cytopenias. LBL is in lymph nodes and tumor cells should not be found in the peripheral blood or bone marrow, at least not when the tumor develops. These are the dividing lines (somewhat arbitrary) created to classify the diseases and identify where the two neoplasms arise, but they are both neoplasms of lymphoblasts. Furthermore, this is biology and as each disease progresses they start to overlap in their distribution throughout the body. ALL will progress into lymph nodes and other organs and some cases of LBL will enter the bone marrow, spleen, or liver, and neoplastic cells can be found in the peripheral blood (secondary leukemia, leukemic phase of lymphoma, stage V lymphoma). When diagnostic procedures intersect, the progression of these diseases is variable, and it may be difficult to distinguish whether neoplastic cells in peripheral blood are due to ALL or LBL with secondary leukemia. Nevertheless, at these stages, and probably earlier, the prognosis for both diseases is poor. In general, if the greatest tumor burden is in lymph nodes, the disease favored is LBL, whereas the greater the number of neoplastic cells in circulation and bone marrow, the disease favored is ALL.


In some reports the majority of canine ALL (>90%) were of B’cell origin8 and in other reports the majority were T‐cell.11 These latter authors preferred to designate the neoplasms as lymphoid rather than lymphoblastic but stated the cells “appeared as blast cells.”11 A study of canine leukemia reported approximately 30% B‐ALL and 13% T‐ALL, 35% AML, and 24% acute undifferentiated leukemia.12


Lymphoma

LBL originates in lymph nodes and/or lymphoid tissues outside the bone marrow. Most canine LBL are of T‐cell origin and there seems to be agreement on this pattern.9,10 However, if neoplastic cells are in circulation from a dog with LBL, stage V lymphoma (secondary leukemia), the neoplasms were reported to be of B‐cell origin (63%).8 Percentages of distribution of various lymphomas and leukemias will vary between laboratories due to expertise, antibodies used, and how each tumor is defined but hopefully the percentages will be somewhat similar. In the study that reported that approximately 60% of LBL with secondary leukemia were of B‐cell origin it is not clear if the high‐grade lymphoma category seen in 65 of 210 (31%) dogs were a mixture of LBL, DLBCL, or possibly B‐anaplastic large cell or even Burkitt’s‐like lymphomas.8 Criteria to distinguish these diagnoses are published and are in this chapter, but some of the distinguishing features are subjective and therefore the umbrella term of high‐grade lymphoma has practical utility.13


Classifications that divide lymphomas into high, intermediate, or low grade may also add the designation T‐cell or B‐cell, and then under each of these general classifications attempt to list specific lymphomas.7 Therefore, under the general heading of high‐grade T‐cell are: peripheral T‐cell, PTCL‐NOS, and T‐LBL high‐grade, and under high‐grade B‐cell lymphoma could be: DLBCL, LBL, anaplastic large cell, and Burkitt’s‐like. Specific types of diagnoses are further complicated by the current lack of markers to definitively identify each of these tumors. Molecular characterization should help define lymphomas in animals. Each tumor has defining morphologic, cellular, biologic behavior, and immunophenotypic characteristics, but some of the parameters are subjective, and no consensus has been reached. There may be “interpathologist variation,” therefore, as to the type of lymphoma. Nevertheless, some of the most aggressive lymphomas seen in animals are ALL or LBL (T‐cell or B‐cell) and have a rapid onset of signs and short survival times of <4 months. Small cell lymphomas and leukemia are indolent, and animals with these should survive for several years. Identifying these diseases has practical implications.


Morphology


Cytologically, the cells of ALL and LBL are of intermediate size with nuclei 1.5 RBC in diameter (Figures 7.337.35). One of the diagnostic features is dispersed chromatin without aggregation (Figure 7.33B,D). This tends to obscure the constant presence of multiple small nucleoli. They have mild anisokaryosis. The cytoplasm is moderate in volume and highly basophilic. These are not large cell lymphomas with large nuclei and prominent nucleoli.

Micrograph of a lymphoblastic lymphoma (LBL) in the lymph node of a dog. It displays a fairly uniform population of lymphoid cells that replaced nodal architecture and fine sclerosis in a 7-year-old poodle.
Micrograph of a lymphoblastic lymphoma (LBL) in the lymph node of a dog. It displays an intermediate size nuclei, 1.5 RBC, with anisokaryosis with chromatin is dispersed. Eosinophils (arrows) are also prominent.
Micrograph of a lymphoblastic lymphoma (LBL) in the lymph node of a dog, CD3 (left) and CD79a (right). It displays tumors positive in CD3 but negative with CD79a, though a few scattered B cells stain positively.
Micrograph of a lymphoblastic lymphoma (LBL) in the lymph node of a dog. It displays one mature lymphocyte nucleus in the center, nuclei are approximately 1.5 RBC in diameter, and chromatin is dispersed and dense.

Figure 7.33 Lymphoblastic lymphoma (LBL), lymph node, dog. (A) A 7‐year‐old standard poodle presented with sudden onset of depression, moderate lymphadenopathy, and good body condition. On Tru‐Cut biopsy a fairly uniform population of small to intermediate‐size lymphoid cells replaced nodal architecture. There is fine sclerosis with scattered eosinophils. (B) Nuclei are of intermediate size, 1.5 RBC, with mild anisokaryosis. Chromatin is dispersed and multiple small nucleoli are only occasionally apparent. The cytoplasm is minimal in volume and poorly defined. Mitoses are ill‐defined; nucleus in metaphase is in center of field; these tumors usually have a high mitotic count. Eosinophils (arrows) were prominent in this case, probably due to IL‐3 or IL‐5 production by the neoplastic cells. (C) CD3 left, CD79a right: The tumor is uniformly and strongly positive with CD3 but negative with CD79a (a few scattered B cells stain positively). (D) Cytologic preparation: The entire population consists of homogeneous medium‐sized immature lymphocytes. Cytologic diagnosis of lymphoma is straightforward: there is one mature lymphocyte nucleus in the center. Characteristics of LBL are that nuclei are approximately 1.5 RBC in diameter, and nuclear chromatin is dispersed and dense, which makes nucleoli inconspicuous. Tumor cells have relatively little cytoplasm. T‐cell versus B‐cell LBL requires immunophenotyping via IHC, ICC, or flow cytometry of suspended cells.

Micrograph of lymphoblastic lymphoma in dog's lymph node. It displays irregularly shaped nuclei with varying sizes and sharply indented nuclear membranes. Chromatin is dispersed and nucleoli inconspicuous.
Micrograph displaying CD3 stain of node. The cells are strongly and uniformly marked.

Figure 7.34 Lymphoblastic lymphoma, lymph node, dog. (A) The nuclei are irregular in shape and variable in size with some sharply indented nuclear membranes. Nuclear chromatin is dispersed and nucleoli inconspicuous, which is a feature of LBL. These are not large cell lymphomas. (B) CD3 stain of node in Figure 7.34: The cells are strongly and uniformly marked. T‐cell LBL is more common than B‐cell, and both have a poor prognosis.

Micrograph of acute lymphoblastic leukemia (ALL) in a cat. It displays lymphocytes the size of immature leukocyte with cytoplasm and increased WBCC. White blood cells are shifted left with cytoplasmic granules.
Micrograph of acute lymphoblastic leukemia (ALL) in a cat. It displays a buffy coat preparation, CD3 stain: almost all of the lymphocytes are marked for T cells. Inset: Tumor cells are CD79a negative.
Micrograph of acute lymphoblastic leukemia (ALL) in a cat. It displays a CD3 stain, splenic aspirate. Lymphoid cells are positive and the extramedullary hematopoietic cells are negative.
Micrograph of acute lymphoblastic leukemia (ALL) in a cat. It displays CD79a stain, splenic aspirate. Lymphoid cells are negative but the megakaryocyte is stained. Megakaryocytes are also positively stained.

Figure 7.35 Acute lymphoblastic leukemia (ALL), peripheral blood, cat. (A) The lymphocytes are the size of immature leukocytes, they have visible cytoplasm and were markedly increased in the WBCC. The white blood cells are shifted left (immature), they have cytoplasmic granules and the cytoplasm and nuclei exhibit asynchronous maturation (cytoplasm is more mature than the nuclei). This is a thick area of the blood film. (B) Buffy coat preparation, CD3 stain: almost all of the lymphocytes are marked for T cells. Inset: Tumor cells are CD79a negative. (C) CD3 stain, splenic aspirate: The lymphoid cells were strongly positive and the extramedullary hematopoietic cells which are numerous in this field negative. (D) CD79a stain, splenic aspirate: The lymphoid cells are negative but the megakaryocyte is stained. Positively stained megakaryocytes with CD79a is surprising but is seen in most animal species.


Histologic preparations of lymph nodes appear to be deeply stained when examined at low or intermediate magnifications (Figure 7.33). This is due to high cellularity, nuclear crowding, and the uniform staining of dispersed nuclear chromatin. The nuclei are of intermediate size and they are round to oval. A high proportion may be cleft but it is not known if clefting has prognostic or therapeutic significance (Figure 7.34). Each nucleus will have a nucleolus, but these are only irregularly apparent in histology because of the stained nuclear chromatin that does not have areas of “clearing” from chromatin aggregation. Larger cells are also present, some of which have oval nuclei; cytoplasm is minimal. The mitoses of LBL are not sharply defined and mitotic counts may be underestimated. Mitotic figures are numerous, up to 10/400× field, and rarely some cells in mitosis may be found in the peripheral blood. A unique feature of lymphomas of LBL is that despite numerous mitotic figures and apoptotic cells there are relatively few tingible body macrophages.


The morphology of cells is a useful means to separate immature lymphoid malignancies (LBL) from more mature lymphoid malignancies, such as CLL, but is not 100%. Morphology should not be relied on to distinguish immature lymphoid tumors from tumors of myeloid or megakaryocytic origin as the cells of each can look similar.14 Immunophenotyping may be needed to distinguish neoplasms of these cell lines and if leukemia is present then flow cytometric immunophenotyping of the peripheral blood with a battery of antibodies is the preferred technique to characterize the neoplastic cells in circulation. This can be and is preferably performed on the same sample that was used for the CBC and in which morphology of the neoplastic cells was evaluated. Some labs also use cells obtained from FNA of lymph nodes to evaluate the cells morphologically and to characterize the cells via flow cytometry. Ideally, future studies will use cytology, immunophenotyping, and histology to classify the neoplasms and to correlate these data with accurate clinical outcomes so we can provide prognostic information and guide therapeutic selection(s).


Occurrence and clinical presentation


Approximately 3% of canine lymphomas are T‐LBL: 12 of 300 in one study, 30/1000 and 17/608 in another.7,9,10 In 1000 canine lymphomas 30% were of T‐cell immunophenotype: 11% were T‐zone lymphoma, 13% were peripheral T‐cell lymphoma, 3% were T‐LBL, and the remaining 1–2% were cutaneous, enteric, or anaplastic large cell T lymphomas.7 In 600 dogs with lymphoma 35% were T‐cell lymphoma: 3% were T‐zone lymphoma, 13% were peripheral T‐cell lymphoma, 3% were T‐LBL, and 12% were cutaneous.10 Apart from the figures for cutaneous lymphomas, the relative percentages given in these studies are fairly close. Authors speculated that the seemingly high prevalence of cutaneous lymphomas may be due to sample bias created by using surgical pathology material that had a high proportion of skin biopsies.10


Animals with T‐ALL are often in good body condition and owners report a recent illness, which suggests an acute onset or at least a rapidly progressing disease. A routine CBC is diagnostic in many canine cases as the total white blood cell count is markedly increased to 50,000–400,000 cells/μL, 90% of which are neoplastic lymphoid cells. Differentiation of T‐ALL and B‐ALL requires immunophenotyping and perhaps with a battery of antibodies to ensure differentiation of membrane versus cytosol reaction. In cats, T‐ALL tends to occur in younger animals. Horses with T‐ALL have fewer neoplastic cells in circulation but the marrow is usually heavily infiltrated. The onset of clinical disease is also reported to be sudden in horses and can be seen in horses in training.


Pathology


Blood and bone marrow

The majority of animals with ALL have a lymphocytosis, which may exceed 100,000 neoplastic cells/μL. One study reported a mean of approximately 64,000/μL for ALL, 109,000/μL for 4 dogs with T‐ALL, and 53,000/μL for 47 dogs with B‐ALL.8 See the B‐ALL section of this chapter and original references for details about the erythroid, leukocytic, and platelet counts in dogs with ALL.8,11,12,14 Anemia was present in 98%, thrombocytopenia in 90%, and neutropenia in 80% of 51 dogs with ALL, but 90% of these were diagnosed as T‐cell.8 Cytopenias are expected in dogs with acute leukemias and not with CLL and this parameter can be used to help differentiate the tumors. Cytopenias occur as a result of myelophthisis and/or cytokines produced by the neoplastic cells. T‐cell ALL is not characterized by marrow fibrosis and most marrow aspirations will yield a large population of tumor cells. The histologic pattern of marrow involvement is not consistent within a bone and between bones. Hemorrhage and displacement of fat cells is a feature to look for at low magnifications. At higher magnifications in these regions and others the neoplastic cells can be found replacing normal marrow elements.


Lymph nodes

In T‐LBL there are multiple enlarged lymph nodes, peripherally and in body cavities. Aspirational cytology or biopsy of an enlarged node is diagnostic of lymphoma (Figure 7.33). There is a monomorphic population of immature lymphoid cells throughout the node. The neoplastic cells are as described earlier and phenotyping is necessary to differentiate B versus T cells. There may be several layers of lymphoma cells outside of the capsule and infiltrating perinodal fat. Extension of cells through and beyond the capsule is much more common in lymphoma than in hyperplasia. If uniform lymphoid cells infiltrate adjacent nodal tissues the diagnosis is lymphoma until proven otherwise. DLBCL will have a similar presentation as T‐LBL and immunophenotyping is a more objective means to differentiate them than cell size and morphology. All these diagnoses have a poor long‐term prognosis. Examine cells near the surface where fixation has been rapid to interpret subtle details such as dispersed chromatin and nucleoli. Within the node the tumor is diffuse and germinal centers and other normal structures are replaced by lymphoma. Despite numerous mitotic figures and pyknotic nuclei, tingible body macrophages are not common. These macrophages are common with DLBCL, Burkitt’s lymphoma, and some other lymphomas. Tingible body macrophages are rarely found cytologically with T‐LBL and may be found in cytologic preparations from DLBCL.


Hyperplastic (reactive) lymph nodes do not have a monomorphic population of lymphoid cells. They generally have less than 50% immature lymphoid cells, the majority of the cells are mature lymphocytes and there will be plasma cells. The more heterogeneous the population of cells and the more plasma cells seen the more likely it is a reactive (hyperplastic) node and not lymphoma.


Spleen and other organs

In T‐LBL the spleen and liver may not be involved when the tumor is present in lymph nodes. The spleen is almost always involved in T‐ALL, with paratrabecular infiltration and involvement of periarteriolar lymphoid sheaths. Neoplastic cells become confluent from these locations and splenomegaly may be marked. As both diseases progress there will be hepatic involvement and hepatosplenomegaly. T‐LBL tends to be perivascular and periportal and T‐ALL tends to be sinusoidal. There will be overlap of these patterns, depending on how advanced the diseases are. In some cases of T‐ALL, particularly in younger animals, the thymus is involved and markedly enlarged.


Cytochemistry, immunohistochemistry, and differential diagnosis


Differential diagnoses that look similar are T‐ALL, B‐ALL, T‐LBL, B‐LBL, DLBCL, large T‐cell lymphoma, AML, and AUL (acute undifferentiated leukemia). CLL is a differential but cell size (small), dense nuclei, an indolent course, and absence of cytopenias should distinguish CLL. There is no single test specific enough to differentiate these neoplasms. Differentiation requires the integration of organ distribution, morphology, immunophenotype, and histochemistry. Many tumors can be diagnosed by the patterns for each of these tests, but there will be some cases in which the results overlap such that a diagnosis cannot be definitive or the best we can do is favor a diagnosis. In these latter cases, if finances and expectations are high then employ as many techniques as possible: histology, cytology, histochemistry, IHC, ICC or flow cytometry, and multiple antibodies. Some cases will defy classification, in which case consider referral to a hematopathologist. The following guidelines are just that and there will be exceptions. The more differentiated the neoplasm the easier the identification. See the myeloid section of this chapter for additional information on these cancers and their identification.


Organ distribution

This is used to differentiate lymphoma versus leukemia. LBL should not be leukemic: the tumor is in lymph nodes, possibly spleen and liver. If there are neoplastic cells in blood and bone marrow but the tumor is absent or only mildly present in nodes, the diagnosis is ALL. ALL patients almost always have cytopenias of other hematopoietic cells. A greater tumor burden in lymph nodes, liver, and spleen and a lower burden in marrow favors LBL. Greater involvement of bone marrow and more evidence of cytopenia favor ALL. Next, results of histochemistry and immunophenotyping can be integrated to help differentiate other lymphomas and AML, AUL, and CLL.


Immunophenotyping

Differentiation between T‐ versus B‐cell uses the usual battery and expected profiles: CD79a, CD20, CD21 positive and CD3 negative is B‐cell; CD79a, CD20 negative with CD3 positive is T‐cell. Therefore T‐ALL and T‐LBL are strongly positive with CD3 and are distinguished by organ distribution and leukemia but both are neoplasms of T lymphoblasts.


To differentiate T‐ALL versus AML: AML profile should be negative to CD79a, CD20, CD3 and positive to CD34 as well as to one or more CD markers for specific cell lines (see Table 7.1), such as CD4, CD11b, CD11c, and/or CD14 (neutrophilic), CD11c, CD14 (monocytic), CD9, CD41, CD61 (megakaryocytic). CD34 immunoreactivity is considered a stem cell marker so it may be positive with ALL. However, if CD34 is positive and lymphoid markers negative the diagnosis favored is AML, especially if one or more myeloid markers is positive and if ALP is present in the neoplastic cells. Positive reaction to CD34 favors AML; however, there are undifferentiated types of AML that are CD34 negative and immunoreactivity of CD34 with lymphoid tumors can be variable. CLL is CD34 negative.


Table 7.1 Categories of myeloid neoplasms

























































Category Definition Subcategory Markers a
Acute myeloid leukemia (AML) Cytopenia and ≥20% blast cells in blood or bone marrow According to cell morphology and/or immunophenotypic features of cells:




  • Acute undifferentiated leukemia (AUL)
CD34




  • AML with neutrophilic differentiation
CD4, CD11b, CD18, CD34




  • AML with myelomonocytic differentiation
CD4, CD11b, CD14, CD18, CD34, MHCII




  • AML with megakaryoblastic differentiation
CD9, CD34, CD41, CD61
Myeloproliferative neoplasms (MPN, old term = chronic leukemia) Cytosis of mature appearing cells in blood, hypercellular bone marrow, <5% blasts According to cell morphology:


  • Polycythemia vera (PV)


  • Essential thrombocythemia (ET)
NAcNA




  • Chronic neutrophilic leukemia (CNL)
NA




  • Chronic monocytic leukemia (CMoL)
NA




  • Mastocytosis
NA
Myelodysplastic syndrome (MDS) Persistent anemia, neutropenia or thrombocytopenia, 5–20% BM blast cells ± dysplasia in blood or BM ± myelofibrosis According to the most pronounced cytopenia:


  • Refractory anemia with excess blasts (RAEB)
NA


  • Refractory neutropenia with excess blasts (RNEB)
NA


  • Refractory thrombocytopenia with excess blasts (RTEB)
NA

a Cell markers that can indicate cell type.


b CD11a, CD11b, and CD11c are expressed on most leukocytes, and generally in decreasing order of intensity on monocytes, granulocytes, and lymphocytes.


c Not applicable; cells are recognizable by morphology.


Histochemistry

LBL (T‐ or B‐cell) is negative for almost all the histochemical stains used to identify a myeloid neoplasm: myeloperoxidase, chloroacetate esterase, anaphthyl butyrate esterase, and Sudan black B. Nonspecific esterase is the exception, it may be positive with LBL and AML. Try to perform histochemistry on cytologic rather than histologic preparations so that cytoplasmic details can be better seen. A recent study reported using ALP to help identify AML, acute myelomonocytic and monocytic leukemia.11 ALP activity was seen in 20/20 AML tumors and there was strong activity in approximately two‐thirds of these. No ALP activity was seen in 49/49 lymphomas and 7 CLLs, but weak ALP activity was seen in approximately one‐third of 14 ALL. The authors concluded that ALP staining of leukemic cells was useful to identify monocytic forms of AML. ALP is not specific for AML as positive immunoreactivity can also be seen in T‐ALL, particularly LGL leukemic cells.11


Differentiation and prognosis


ALL and AML should both have >30% blast cells in bone marrow and tumor burdens are greater in marrow and less in lymph nodes and organs. They can look similar or be identical in blood films and on cytology, especially if AML is poorly differentiated and does not demonstrate maturation. The more differentiated the AML the easier the distinction should be. Granules seen in the cytoplasm of neoplastic cells favor AML. However, granules may be scarce to absent in poorly differentiated AML and granules may be found in LGLs. The granules in LGL may also be ALP positive but they should be granzyme B positive and the immunophenotypic profile should be that of lymphoid, T‐cell. Positive reaction for ALP favors AML.


The less differentiated the ALL or AML and the wider the organ distribution the more reliant we are on IHC, flow cytometry, and ICC (see section on Immunophenotyping above). The belief is that when ALL or AML has a wide distribution in organs (disseminated) it has progressed and is in a more advanced state. However, AML or ALL confined to bone marrow versus wide organ distribution could be different diseases that have different molecular profiles; one may be programmed for a more aggressive course (wider distribution) than those confined to bone marrow. These are interesting hypotheses, but from a practical view all of these tumors are aggressive and shorten the life of animals with survival times measured in days, and generally less than 4 months for each. CLL is one of the few leukemias with survival times that approach 3 years.


The literature provides contrasting data for the prevalence of T’cell versus B‐cell ALL in dogs.8,11,12 The antibodies selected and the expertise and methodologies used in different laboratories can result in different conclusions. The parameters that researchers use to define B‐ALL, T‐ALL, and AML influence final diagnoses. Until parameters are standardized and definitions agreed upon there will be discrepancies. A battery of antibodies, techniques, and histochemical stains should be used on cases that seem ambiguous.


Another approach is to avoid traditional diagnoses and instead define tumors that cause lymphocytosis by an antibody profile, size of the cells in blood, and whether the number of neoplastic cells in blood are greater or lesser than 30,000/μL.15 Briefly, dogs with CD34‐positive lymphocytosis had survival times of approximately 2 weeks, those CD8 positive (T‐cell) with >30,000 lymphocytes/μL approximately 4 months, those CD8 positive (T’cell) with <30,000 lymphocytes/μL approximately 3 years, and dogs with CD21‐positive (B‐cell) lymphocytosis composed of large cells approximately 4 months versus those with smaller circulating cells in which the MST was not reached.15 See the section on B‐cell ALL, LBL for similar summary and the appendix on lymphomas (p. 961).


MST values reported were 10 days (range 4–120 days) for dogs with T‐ALL and 8 days for B‐ALL (range 5–46 days), but the numbers per group were small.12 Dogs with AML had an MST of 10 days (range 3–73 days) and dogs with AUL had an MST of 7 days (range 1–90 days).12


References



  1. 1. Coustan‐Smith, E., Mullighan, C.G., Onciu, M., et al. (2009) Early T‐cell precursor leukaemia: a subtype of very high‐risk acute lymphoblastic leukaemia. Lancet Oncol 10:147–156.
  2. 2. Fournel‐Fleury, C., Ponce, F., Felman, P., et al. (2002) Canine T‐cell lymphomas: A morphological, immunological, and clinical study of 46 new cases. Vet Pathol 39:92–109.
  3. 3. De Keersmaecker, K., Marynen, P., and Cools, J. (2005) Genetic insights in the pathogenesis of T‐cell acute lymphoblastic leukemia. Haematologica 90:1116–1127.
  4. 4. Pui, C.H. (2005) Quest for effective agents to combat T‐cell acute lymphoblastic leukemia. Eur J Cancer 41:1243–1245.
  5. 5. Thomas, X., Le, Q.H., Danaila, C., et al. (2002) Bone marrow biopsy in adult acute lymphoblastic leukemia: morphological characteristics and contribution to the study of prognostic factors. Leuk Res 26:909–918.
  6. 6. Valli, V.E. (2007) Precursor T‐cell lymphoblastic lymphoma and lymphoblastic leukemia. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 275–286.
  7. 7. Valli, V.E., Kass, P., San Myint, M., and Scott, F. (2013) Canine lymphoma: The effect of age, stage of disease, subtype of tumor, mitotic rate and treatment protocol on overall survival. Vet Pathol 50:738–748.
  8. 8. Tasca, S., Carlil, E., Caldin, M., et al. (2009) Hematologic abnormalities and flow cytometric immunophenotyping results in dogs with hematopoietic neoplasia: 210 cases (2002–2006). Vet Clin Pathol 38:2–12.
  9. 9. Valli, V.E., San Myint, M., Barthel, A., et al. (2011) Classification of canine malignant lymphomas according to the World Health Organization criteria. Vet Pathol 48:198–211.
  10. 10. Ponce, F., Marchal, T., Magnol, J.P., et al. (2010) A morphological study of 608 cases of canine malignant lymphoma in France with a focus on comparative similarities between canine and human lymphoma morphology. Vet Pathol 47:414–433.
  11. 11. Stokol, T., Schaefer, D.M., Shuman, M., et al. (2015) Alkaline phosphatase is a useful cytochemical marker for the diagnosis of acute myelomonocytic and monocytic leukemia in the dog. Vet Clin Pathol 44:79–93.
  12. 12. Novacco, M., Comazzi, S., Marconato, L., et al. (2015) Prognostic factors in canine acute leukaemias: a retrospective study. Vet Comp Oncol DOI: 10.1111/vc0.12136
  13. 13. Munasinghe, L.I., Kidney, B.A., MacDonald‐Dickinson, V., et al. (2015) Evaluation of lymph node aspirates at diagnosis and relapse in dogs with high‐grade multicentric lymphoma and comparison with survival time. Vet Clin Pathol 144:310–319.
  14. 14. Vernau, W. and Moore, P.F. (1999) An immunophenotypic study of canine leukaemias and preliminary assessment of clonality by polymerase chain reaction. Vet Immunol Immunopathol 69:145–164.
  15. 15. Williams, M.J., Avery, A.C., Lana, S.E., et al. (2008) Canine lymphoproliferative disease characterized by lymphocytosis: immunophenotypic markers of prognosis. J Vet Intern Med 22:596–601.

T‐cell chronic lymphocytic leukemia/small cell lymphocytic lymphoma and large granular lymphocyte types


Defining the neoplasm


Chronic lymphocytic leukemia (CLL) originates in bone marrow or spleen and small cell lymphocytic (SLL) lymphoma starts in solid tissues, lymph nodes most commonly. These are defining features but these neoplasms are derived from similar cell lines and are likely manifestations of the same disease. Molecular profiles would help clarify this possibility. They can be of T‐ or B‐cell or NK origin or have atypical profiles.1–9 In humans over 95% of CLL are B‐cell, whereas in dogs and cats most cases of CLL are T‐cell; approximately 70–90% in dogs and 95% in the few cases reported in cats.2–6,11 Some tumors express T‐ and B‐cell antigens and some have aberrant expression patterns, but these types are uncommon.2–6,9 The lymphocytes in CLL may be granulated or not in dogs, but granular lymphocytic CLL was not seen in a study of 18 cats.10 Fifty‐four of 61 dogs with CLL had a T‐cell phenotype (88%), 49 of 54 (90%) were LGL subtype, and 5 were nongranular.4 Another study reported 73% of 73 canine CLL were T‐cell, 54% were LGL, and 26% were B‐cell.5 Cats with CLLs are usually FeLV negative, as opposed to FeLV positive, which is the trend for cats with ALL. FeLV‐associated lymphomas are usually multicentric or mediastinal and found in young cats. FIV‐associated tumors in cats are usually high‐grade B‐cell lymphomas.


Some authors have defined the following diagnostic criteria for CLL in cats. These are admittedly somewhat arbitrary but useful: lymphocytosis >9000/μL, clonal proliferation or homogeneous cells via phenotyping, and either bone marrow >15% mature lymphocytes or concurrent cytopenia of at least one cell line. The cytopenia identified most frequently was anemia (50% of 18 cats).10 These authors excluded cats with large cell lymphocytes or if multicentric lymphoma was deemed more likely. Similar criteria have been used to study dogs: lymphocytosis >6000/μL, clonal proliferation or homogeneous cells via phenotyping, negative CD34 and negative testing for ehrlichiosis, leishmania, and other causes of lymphocytosis (Addison’s, epinephrine surge, post vaccine).9


T‐CLL in cats and dogs is an indolent neoplasm composed of small lymphocytes that look cytologically “normal” in blood, but they cause a marked lymphocytosis, often >100,000/μL.4,6,9,10 The lymphocytosis has a wide range, likely dependent when diagnostic techniques intersect the disease, 5000 to >1,000,000/μL. The lymphocytes may be small, medium or large, and granulated or not. In cats the lymphocytes are not granular but in dogs granular lymphocytes (LGL) are expected in approximately 50–80% of CLL cases.4 CLL composed of LGL have larger cells, pale cytoplasm, larger, round or reniform nuclei, and are more open (chromatin is not as dense). Cytoplasmic granules can be inconspicuous but when found could be confused with myeloid lineage (AML). In general, cytoplasmic granules in LGL cells are easy to see in cats and horses but are smaller and less obvious in dogs. The cytoplasmic granules are easy to see in larger cells and are inconspicuous in small cell types (see Figures 7.40 and 7.41). Visibility is enhanced by Wright–Giemsa stains that have a methanolic base, oil‐immersion objectives, or PTAH.


Over 90% of canine CLLs composed of LGL are T‐cell neoplasms and they appear to originate in spleen rather than the marrow. If subtyped, the majority of canine cases are CD8 positive (cytotoxic), a few may be NK cell. The marrow may be involved but usually only late in the progression of the disease. Both dogs and cats can have LGL lymphomas (not CLL) that may or may not have leukemia. In cats, LGL lymphomas usually originate in lymphoid tissue in the gastrointestinal tract.11 These are aggressive but tend to be less so in dogs.


CLL composed of LGL is similar to a common and fatal leukemia in Fischer 344 rats.12 Rats with this leukemia typically have white blood cell counts >100,000/μL, severe icterus, massive splenomegaly, red blood cell agglutination, erythrophagocytosis, and severe anemia that is immune‐mediated. The leukemia and agglutination of red blood cells will produce mean corpuscular volumes (MCV) >120 fL.


The solid form of this disease is a peripheral lymphoma (SLL). The neoplasm is composed of small lymphocytes, B‐cell or T‐cell, which are indistinguishable until they are phenotyped (Figures 7.36 and 7.37). The majority of SLL are T‐cell but B‐cell types occur. Some cases will have neoplastic cells in blood and the number of lymphocytes can be marked. The line between CLL and SLL becomes blurred when there are neoplastic cells in circulation, bone marrow, spleen, and lymph nodes. Regardless of nomenclature, cases with wider organ distribution and T‐cell lymphocytosis >30,000/μL have a worse prognosis.7 Phenotype needs to be determined by IHC or flow cytometry, but in general, if a monoclonal gammopathy is identified it suggests the tumor is of B‐cell origin, if the tumor markedly infiltrated the bone marrow it is more likely B‐cell, and if granulated lymphocytes are seen they indicate T‐cell. CLL and SLL should be distinguished from ALL and AML as these latter neoplasms have a much worse prognosis and more aggressive course. CLL and SLL should be distinguished from intestinal LGL lymphoma in cats as many of these have a more aggressive course.

Micrograph of small cell lymphocytic lymphoma (SLL) in a dog's lymph node. It displays nodal architecture replaced by uniformly small lymphocytes. Nuclei are slightly larger than red cells and have cytoplasm.
Micrograph of small cell lymphocytic lymphoma (SLL) in a dog's lymph node. It displays higher magnification of neoplastic cells having little cytoplasm. Nuclei are the size of red blood cells. Chromatin is dense.

Figure 7.36 Small cell lymphocytic lymphoma (SLL), lymph node, dog. (A) Nodal architecture is replaced by a population of uniformly small lymphocytes. (B) Higher magnification: The neoplastic cells have little cytoplasm, nuclei are approximately the size of red blood cells, chromatin is dense, some nuclei have clefts, and nucleoli are not seen. A mitotic figure is present but overall the mitotic counts in SLL are low.

Micrograph of small cell lymphocytic lymphoma (SLL) in a dog's lymph node. It displays neoplastic cells positive with germinal center at top unstained. Inset: Lymphoma cells negative. Germinal center positive.
Micrograph of small cell lymphocytic lymphoma (SLL) in a dog's lymph node. In the cytology, the Nuclei are only slightly larger than the red blood cells.

Figure 7.37 Small cell lymphocytic lymphoma (SLL), lymph node, dog. (A) CD3: Neoplastic cells are heavily and uniformly positive. There is a fading germinal center at top that is not stained. Inset: CD79a: The lymphoma cells are negative and the non‐neoplastic cells in the fading germinal center are strongly positive. (B) Cytology, lymph node: Cellular and nuclear details are easier to evaluate in cytological preparations than in H&E. Nuclei are only slightly larger than the red blood cells. SLLs generally lack nucleoli and have a narrow envelope of basophilic cytoplasm. Spleen is often neoplastic and splenic or bone marrow involvement may result in a markedly increased, >50,000/μL lymphocytosis. CLL is a tumor of the same cell type but originates in bone marrow and causes moderate to marked lymphocytosis.


Epidemiology, occurrence, and clinical features


Approximately 50% of dogs and cats are asymptomatic and are diagnosed at annual examinations after blood is analyzed and a lymphocytosis is identified.1,6,10 Most affected animals are in good condition but weight loss is reported. Animals presented because of an illness are lethargic with reduced appetite and may have splenomegaly, lymphadenopathy, and fever, and very likely the disease has gone undetected for many months. The disease is seen in mature animals and is uncommon in dogs less than 5 years4,6 or cats less than 10 years old.5,6,10 The median age in cats was reported to be 12.5 years, with a range of 5–20 years.10 The median age in 61 dogs was 10, with a range of 5–19 years.4 CLL is identified less often in cattle and horses but likely occurs in most mammals. There do not appear to be any breed‐related tendencies or common causative exposures. Cats with CLL are FeLV negative.


Pathology


Blood and bone marrow

The diagnosis of CLL is made from examination of blood with total lymphocyte counts of 50,000–400,000/μL, but incredible ranges of 5,000 to >1,000,000/μL are sometimes seen. If the lymphocytosis is >50,000/μL the diagnosis is fairly easy. In cats, the total median lymphocyte count reported was 34,000/μL.10 In dogs median lymphocyte counts range from 36,000 to 166,000/μL. The lymphocytosis is persistent and usually sustained over an extended period of time, >3 months. Lymphocytosis can also be due to non‐neoplastic causes, such as tick‐borne diseases in dogs, bartonella in cats, hypoadrenocorticism and epinephrine‐induced physiologic response. Physiologic lymphocytosis is more common in cats and young horses but rarely in the range of CLL and it is transient, gone in less than one day if rechecked. The magnitude of lymphocytosis in canine ehrlichiosis, however, can be marked and in the range of CLL.13 The greater the lymphocytosis and the longer it persists the more likely the cause is CLL rather than one of these non‐neoplastic causes. See Differential diagnosis in this section.


Approximately 50–75% of dogs and 50% of cats with CLL will have varying degrees of non‐regenerative anemia, but it is not as severe as with ALL or AML.4,10 Neutropenia was not seen in approximately 180 dogs with CLL4,5,9 and is not reported in cats.4,10 Thrombocytopenia ranged between 10% and 25% in these three studies.4,5,9 Cytopenias are not a feature of CLL but they are expected with ALL and AML.


Morphologically, the tumor cells may be granulated or not. Granulated cells are LGL and are T cells. Beyond this, morphology cannot be relied on to identify B‐CLL or CLL or atypical phenotypes. The cells of CLL are mature and morphologically they look the same as in the non‐neoplastic causes of lymphocytosis. CLL cells have little cytoplasm and small nuclei, 7–9 μm in diameter, which is approximately the size of a canine red blood cell. The chromatin is densely stained without clear areas. Nucleoli are inconspicuous or not visible. The cytoplasm is minimal and lightly stained (Figure 7.37). They look like normal lymphocytes, therefore the key to the diagnosis is the marked lymphocytosis. If CLL is LGL type then the cells are larger, cytoplasm more abundant and pale, and cytoplasmic granules are present (described earlier). Neoplastic cells in ALL and AML can look similar, especially AML without maturation.


In CLL the bone marrow is usually neoplastic, >15% lymphocytes; however, CLL of LGL cells likely originates in the spleen and bone marrow is not infiltrated. Core biopsy will identify tumor in bone marrow more frequently than aspirational cytology.10 When involved, the marrow is hypercellular, fat is replaced, and 15–90% of the marrow will be CLL cells.1,2 However, the mitotic count is low and dividing cells are not found or are few. The disease is noted to be one of accumulation rather than proliferation, with the neoplastic cells having upregulation of the Bcl2 gene that blocks the apoptotic process. There can be focal areas of “reaction centers” in the solid areas of tumor where the dividing cells are of slightly larger type.


In the nongranulated form of CLL the marrow is “almost always” involved. Metaphyseal regions of appendicular bones and any axial bone will contain tumor. A bone marrow aspirate or core from the proximal femoral area will have high cellularity and provide a diagnostic sample. LGL CLL may not be in the bone marrow and samples from the spleen or an enlarged lymph node should contain tumor. LGL leukemia of a larger cell type can arise in the spleen of dogs and LGL lymphoma arises in the intestinal tract of cats. Many of these cases will also have neoplastic cells in blood.11


Lymph nodes

The lymph nodes of SLL are diffusely neoplastic. Germinal centers are absent or are reduced to a few fading clusters of mantle cells. The medullary cords are filled and expanded and the medullary sinuses are compressed. Nodes appear dense because the neoplastic cells have little cytoplasm and nuclei are crowded together.3 Nuclei are densely stained and only slightly larger than red cells, with nucleoli not apparent (Figure 7.36). The mitotic count is low, with none in most 400× fields. In general, the nodes have thin capsules and the subcapsular sinus is compressed.


Lymph nodes in CLL are often not enlarged and may even be atrophic. Other cases will have enlarged lymph nodes and neoplastic cells but the nodes are not as severely involved as with SLL. If CLL is marked, look in the lymphatics of the node as they will be packed with neoplastic cells identical to those in the blood. Megakaryocytes may be present in the collapsed medullary sinuses.


The diagnosis of SLL with cytology can be difficult because the lymphocytes are mature. An anatomic pathologist should consult with a clinical pathologist or consider immunophenotyping (flow cytometry or IHC) to assess homogeneity of the cells or PARR for clonality. The cells are uniform, small, mature, and are not the prototypical neoplastic large lymphocytes. There should be no or very few plasma cells and no neutrophils. If granules are seen or suspected, consider staining with a methanolic‐based Wright–Giemsa stain to enhance their prominence.


Spleen and liver

Splenomegaly is expected in dogs and cats with CLL.2,6,10 The spleen may not be neoplastic in nongranulated CLL as the tumor originates in bone marrow. If the spleen is involved the tumor is usually in the sinus areas, which may be partially or fully occupied by neoplastic lymphocytes. Tumor cells separate the resident smooth muscle trabeculae to varying degrees, depending on the magnitude of the tumor infiltrate. There usually is partial to complete atrophy of the thymic‐dependent periarteriolar lymphoid cuffs, and germinal centers are reduced in size. Usually there are no areas of extramedullary hematopoiesis and no sinus histiocytes bearing hemosiderin.


The liver is infiltrated in later stages of CLL. Periportal areas and hepatic sinusoids may be filled with CLL cells. In severe cases, neoplastic cells can be found in any tissue.2


Cytochemistry and immunohistochemistry


Most CLL and SLL in dogs and cats will have strong cytoplasmic labeling with CD3 and are negative with antibodies for B‐cell markers, CD79 alpha, CD20, and CD21 (Figure 7.37). If subtyped, 75–90% of canine T‐cell CLL are CD8‐positive cytotoxic lymphocytes. Of 61 canine cases, 54 (89%) were of T‐cell origin (CD3 positive, CD21 negative), and of these, 49/54 (91%) were LGL (CD3 positive/CD4 negative/CD8 positive), 4 (7%) were CD3 positive/CD4 negative/CD8 negative, and 1 (2%) was CD3 positive/CD4 positive/CD8 negative.4 Seven of the 61 (12%) canine CLL were of B‐cell origin (CD21 positive).4 A report of 43 CLL indicated 19 were T‐cell and 14 of these were LGL, 17 were B‐cell, and 7 had atypical phenotype, 4 of which were LGL.9 Seven cases had atypical profiles and these had short survival times.9 A study of 73 dogs reported 73% were T‐cell and 23% B‐cell.5 Forty of the T‐CLL were LGL; all were CD3‐positive and 90% CD8‐positive cytotoxic lymphocytes. These authors also detailed leukointegrin and other CD profiles, including 3 LGL that were double negative. Non‐LGL T‐CLL profiles were also characterized.5 No cases were CD34 positive and this is consistent with other reports.5


Almost all of the reported CLL in cats are of T‐helper lymphocytes.10 Seventeen of 18 cats with CLL were T‐cell phenotype and 16 of these were CD3‐positive/CD4‐positive/CD8‐negative T‐helper lymphocytes.10 One cat was CD21‐positive B‐cell CLL. Feline CLL cases are CD4‐positive/CD8‐positive T‐helper type which is also consistent with their nongranulated cell type.10


The patterns of immunophenotyping are varied and cannot be predicted from morphology; however, LGL CLL is considered T’cell and animals with macroglobulinemia B‐cell. Predicting B’cell, non‐LGL T‐cell, NK, or aberrant phenotypes is not possible from morphology.


Differential diagnosis


AML, ALL, and non‐neoplastic lymphocytosis are the main differentials for CLL. The higher the lymphocytosis and the longer it persists the more likely the diagnosis is CLL. Most antigenic substances, such as a vaccine, stimulate a relatively mild and transient lymphocytosis, <6000/μL. However, a moderate to marked lymphocytosis that is in the range of CLL can be induced in dogs with tick‐borne diseases, especially canine ehrlichiosis,13 and some cases of CLL or SLL will have lymphocytosis <20,000/μL. Furthermore the lymphocytes in ehrlichiosis can be LGL, and rare cases have a monoclonal gammopathy or cytopenias, further confounding the final diagnosis and forcing additional tests. Cats with bartonella or an epinephrine surge may have lymphocytosis and the magnitude is in the range of CLL. It would be unusual for a cat with an epinephrine response for the lymphocytosis to exceed 10,000/μL but if it did the lymphocytosis is transient. The higher the lymphocyte count and the longer it persists (>3 months) the more likely the disease is CLL. These differentials can be ruled out through serology, visualization of organisms, repeat blood work, PCR for specific infectious diseases, response to doxycycline and, if needed, immunophenotyping and clonality determination via PCR for gene rearrangement (PARR).


Immunophenotyping does not identify neoplastic cells but can be used to identify homogeneous lymphoid populations, which are indicative of neoplasia, and to aid prognosis.2,7 In a study of 31 dogs with lymphocytosis, homogeneity of phenotypes or aberrant patterns of lymphocytic antigens was used to identify neoplasia in all 31 dogs.2 Homogeneity of antigen receptor gene rearrangement analysis (clonality) was present in 27/31 (87%).2 The four cases that did not express clonality via PCR were negative with two different primer sets. These cases may have had lymphoid neoplasms with rearrangement using undefined V and/or J gene segments. Results like these exemplify the need to use multiple tests in gray zone cases and not to rely on only one test.


ALL and AML are differentials; both can have cells that are morphologically similar to CLL but both should be CD34 positive and CLL CD34 negative. AML cells may have fine azurophilic cytoplasmic granules similar to LGL and AML cases that do not demonstrate maturation look similar to CLL. However, AML will be positive, with ALP and CLL negative. ALL and AML have cytopenias of several cell lines that can be severe. CLL does not have neutropenia and many cases of AML or ALL are neutropenic. LGL CLL will be granzyme B positive.


Hepatosplenic lymphoma can look similar to CLL via histopathology and may also be CD3 and CD11d positive. However, leukemia is not a feature of hepatosplenic lymphoma. Erythrophagocytosis by the lymphoid cells, erythrophagocytosis by macrophages in spleen and liver, and marked involvement of spleen are identifying features of hepatosplenic lymphoma. The liver is more severely involved in hepatosplenic lymphoma and dogs with this form of lymphoma survive for less than 1 month versus years for CLL. Bone marrow is more severely involved with CLL. Hepatocytotropic lymphoma has characteristic infiltration of hepatic cords and an emperipolesis‐like pattern of neoplastic lymphocytes and hepatocytes (see sections of this chapter on these neoplasms).


LGL of large cell type can also produce an acute leukemia and this should be differentiated from LGL CLL as the outcomes are very different. CLL is indolent and dogs with LGL acute leukemia survive less than 6 months.5 The LGL cells in acute leukemia are larger, the cytoplasm more basophilic, and they have immature nuclei. Nuclear chromatin is more open, nucleoli are prominent, and the nuclei have irregular contours producing reniform or convoluted shapes. They are CD34 negative, as is CLL. Cats with LGL intestinal lymphoma may become leukemic and have very short survival times, less than 3 months.11


Survival and prognosis


CLL SLL are indolent lymphoid neoplasms and some dogs will live for 3 years or more following diagnosis even without treatment.5,6 In general, CLL is the least aggressive leukemia or lymphoma seen in animals but it is persistent and if the pet does not die of other causes CLL will likely cause or contribute to death of the pet. However, within the umbrella diagnosis of CLL different survival patterns can be predicted. Dogs with T‐cell CLL were reported to have a median survival of 930 days (indolent), those with B‐cell CLL 480 days, and dogs with aberrant or atypical phenotypes only 22 days.9 Treatments did not affect survival. Fewer cases have been followed in cats. The median overall survival for 17 cats was approximately 14 months with a range of 9–25 months, and almost 18 months for cats that responded to treatment.10 One study reported that older dogs with B‐cell CLL survived significantly longer than younger dogs, and dogs with T‐cell CLL and no anemia survived longer than dogs with anemia.9


Factors that are associated with survival in dogs with lymphoproliferative diseases and lymphocytosis include total T‐cell lymphocytosis at presentation, CD34‐positive lymphocytosis (should not be CLL), and if lymphocytosis is B‐cell type then the size of B cells.7 This study did not distinguish dogs with CLL, ALL, or stage V lymphoma. There was no significant difference in survival for dogs with T‐ versus B‐cell lymphoproliferative diseases, but within each group of T‐ or B‐cell diseases there were factors associated with prognosis. Dogs with total T‐cell lymphocyte counts >30,000 lymphocytes/μL at presentation had significantly shorter median survival (130 days) than dogs with <30,000 lymphocytes/μL (3 years).7 Dogs with CD34‐positive lymphocytosis had the shortest survival times, median of 16 days.7 CD34 positivity is a hematopoietic marker of early lineage and is associated with ALL and AML, while most CLL and LGL leukemia are CD34 negative.


Dogs with B‐cell lymphocytosis (CD21‐positive) were separated by the size of the neoplastic lymphoid cells.7 Dogs with large cell B‐lymphocytosis had an MST of 130 days, whereas in those with small cell B‐lymphocytosis the MST was not reached.7 Cell size was determined by flow cytometry on the peripheral blood. Magnitude of lymphocytosis of B cells was not statistically significantly associated with survival, in fact the trend was for higher counts to survive longer. However, the number of dogs in the different groups were small. This study reported that immunophenotype profile, degree of lymphocytosis, and size of lymphocytes were useful to predict survival.7 Aberrant phenotypes, as defined, were not associated with different survival times7, although a different study identified aberrant phenotypes that were associated with survival as short as 3 weeks.9


A similar study attempted to distinguish dogs with leukemia versus dogs with stage V lymphoma and to exclude dogs with blast cell morphology or that were CD34 positive; therefore the majority of the dogs had CLL not ALL.9 This study found phenotype as determined by flow cytometry predictive of survival. Dogs with T‐cell CLL had median survival of 930 days, those with B‐cell CLL 480 days, and dogs with aberrant or atypical phenotypes only 22 days.9 Treatments did not affect survival. The greatest difference between the groups was for dogs with aberrant phenotypic profiles. Dogs with T‐cell CLL had an indolent course and long survival times. However, eventually these neoplasms relapse (even if treated) or the neoplasm transforms to a more aggressive lymphoma – Richter’s‐like syndrome. CLLs that had atypical phenotypes were very aggressive.9 Aberrant or atypical phenotypes were CD3 negative/CD8 positive, CD3 positive/CD4 negative/CD8 negative, CD3 positive/CD4 positive/CD8 positive, or CD3 positive/CD21 positive. There were only a few dogs in each group. “Typical” phenotypes were T‐cell CD3 positive/CD8 positive and B‐cell CD21 positive. Young age at time of diagnosis for B‐cell and anemia for T‐cell CLL were negatively associated with prognosis.9


CLL can become more proliferative and transform to an accelerated phase (Richter’s syndrome) with a higher proportion of larger cells in division and patients exhibiting fever, organomegaly, and rapid clinical deterioration.1,9,10 Although considered rare in dogs and humans, if CLL underwent a Richter’s‐like transformation to become a large cell lymphoma it would be difficult to separate these diagnoses unless CLL was established at an earlier time.


In dogs, LGL lymphoma and leukemia are considered indolent. In cats, LGL lymphoma with or without neoplastic cells in blood is aggressive and causes death in days to a few months post diagnosis.11


References



  1. 1. Valli, V.E. (2007) Chronic lymphocytic leukemia. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 287–294.
  2. 2. Yagihara, H., Uematsu, Y., Koike, A., et al. (2009) Immunophenotyping and gene rearrangement analysis in dogs with lymphoproliferative disorders characterized by small‐cell lymphocytosis. J Vet Diagn Invest 21:197–202.
  3. 3. Valli, V.E., Jacobs, R.M., Norris, A., et al. (2000) The histologic classification of 602 cases of feline lymphoproliferative disease using the National Cancer Institute working formulation. J Vet Diagn Invest 12:295–306.
  4. 4. Tasca, S., Carlil, E., Caldin, M., et al. (2009) Hematologic abnormalities and flow cytometric immunophenotyping results in dogs with hematopoietic neoplasia: 210 cases (2002–2006). Vet Clin Pathol 38:2–12.
  5. 5. Vernau, W. and Moore, P.F. (1999) An immunophenotypic study of canine leukaemias and preliminary assessment of clonality by polymerase chain reaction. Vet Immunol Immunopathol 69:145–164.
  6. 6. Workman, H.C. and Vernau, W. (2003) Chronic lymphocytic leukemia in dogs and cats: the veterinary perspective. Vet Clin Small Anim 33:1379–1399.
  7. 7. Williams, M.J., Avery, A.C., Lana, S.E., et al. (2008) Canine lymphoproliferative disease characterized by lymphocytosis: immunophenotypic markers of prognosis. J Vet Intern Med 22:596–601.
  8. 8. McDonough, S.P. and Moore, P.F. (2000) Clinical, hematological, and immunophenotypic characterization of canine large granular lymphocytosis. Vet Pathol 37:637–646.
  9. 9. Comazzi, S., Gelain, M.E., Martini, V., et al. (2011) Immunophenotype predicts survival time in dogs with chronic lymphocytic leukemia. J Vet Intern Med 25:100–106.
  10. 10. Campbell, M.W., Hess, P.R., and Williams, L.E. (2012) Chronic lymphocytic leukaemia in the cat: 18 cases (2000–2010) Vet Comp Oncol 11:254–264.
  11. 11. Roccabianca, P., Vernau, W., Caniatti, M., and Moore, P.F. (2006) Feline large granular lymphocyte (LGL) lymphoma with secondary leukemia: primary intestinal origin with predominance of a CD3/CD8αα phenotype. Vet Pathol 43:15–28.
  12. 12. Stromberg, P.C., Rojko, J.L., Vogtsberger, L.M., et al. (1983) Immunologic, biochemical, and ultrastructural characterization of the leukemia cell in F344 rats. J Natl Cancer Inst 71:173–181.
  13. 13. Weiser, M., Thrall, M., Fulton, R., et al. (1991) Granular lymphocytosis and hyperproteinemia in dogs with chronic ehrlichiosis. J Am Anim Hosp Assoc 27:84–88.

T‐cell prolymphocytic leukemia


T‐cell prolymphocytic leukemia (T‐cell PLL) is a rare subtype of T‐cell leukemia. It has an aggressive course in humans, dogs, and cattle.1–5 The diagnosis is relatively easy as neoplastic cells in peripheral blood are numerous and in the range of 50,000–100,000/μL. Bone marrow is the site of origin. Immunophenotyping is required as B‐cell prolymphocytic leukemia is morphologically identical. Prolymphocytic leukemia is a cell type rather than a disease entity.


The diagnosis is based on morphology. Neoplastic cells in blood are prolymphocytes, they look immature and are characterized by nuclei of intermediate size, 1.5–2 RBC in diameter (Figure 7.38). The nuclear chromatin is distinctive and key to the diagnosis. It consists of multiple irregular aggregates of heterochromatin (chromocenters) that are joined by short narrow chromatin bands. The impression on looking at the nuclei at high power is that there is parachromatin clearing that extends in bands across the nucleus (Figure 7.38). These events are the same whether B‐cell or T‐cell prolymphocytic and are better appreciated in cytological preparations. In histological preparations the chromatin pattern is not as apparent. Nucleoli are not visible. A clear perinuclear region is characteristic. Cells have a broad rim of lightly stained cytoplasm. There are a few larger cells present with a nucleus fully 2.0 RBC or greater in diameter. These are felt to be the dividing population and these cells will occasionally have nucleoli. The diagnosis is best made with cytology of blood or films of bone marrow. In histologic sections the diagnosis is lymphoma and it would be difficult to proceed further.

Micrograph of prolymphocytic leukemia in a dog's blood. Lymphoid cells are immature and neutrophils size; neoplastic lymphocytes nuclei are 1.5–2 RBC in diameter with chromocenters joined by chromatin bands.
Micrograph of leukemic lymphoma in a cow. The neoplastic lymphocyte is of large cell type, smaller non-neoplastic lymphocyte is adjacent to a neutrophil that is hypersegmented.
Micrograph of Lymph node imprint. Sample is cellular, containing a marked proliferation of large lymphocytes with little cytoplasm. Nuclei have chromatin and multiple small nucleoli, and mitoses are present.

Figure 7.38 Leukemia, prolymphocytic, blood, mature dog. (A) Problems: weight loss, decreased appetite, mild non‐regenerative anemia, and marked lymphocytosis. All the lymphoid cells are immature, they are the size of neutrophils; nuclei of the neoplastic lymphocytes are 1.5–2 RBC in diameter with prominent large chromocenters that are joined by fine chromatin bands. This pattern is most apparent in the cell at lower left (arrow). Prolymphocyte classification is determined from cytologic morphology and immunophenotyping is needed to distinguish B‐ versus T‐cell prolymphocytic leukemias. (B) Lymphoma, leukemic, cow. The neoplastic lymphocyte is of large cell type, smaller non‐neoplastic lymphocyte is adjacent to a neutrophil that is hypersegmented indicating senescence, likely due to splenic overload and inability of spleen to remove older cells. (C) Lymph node imprint. The sample is very cellular and contains a marked proliferation of large lymphocytes with little cytoplasm. Nuclei have fine chromatin and multiple small nucleoli, and numerous mitoses are present (6). It is uncommon to see this many mitotic figures in one field of vision in a cytologic sample. It may be due to high tumor replication rate or abnormal mitoses held at different stages. Small pieces of cytoplasm from the neoplastic cells can be seen in between cells. These are referred to as lymphoglandular bodies. A few small basophilic mature lymphocytes remain.


Epidemiology and occurrence


T‐cell PLL is most likely to be diagnosed from examination of peripheral blood from a mature cow. Blood samples are taken because of unexplained illness, drop in milk production, or possibly changes in gait due to involvement of the spinal cord. Spleen and bone marrow are usually involved and lymph nodes may not be. In cattle, PLL is associated with BLV. The disease is rare in dogs. Lymph nodes and spleen are usually enlarged and neoplastic.


The main differential diagnosis is CLL and the distinction, at least in dogs is important. In CLL the neoplastic cells are smaller, they have less cytoplasm, only a thin rim and nuclei have a more mature appearance, they are darker without chromocenters. CLL has an indolent course and PLL progresses rapidly. CLL may transform to PLL cell type.


References



  1. 1. Catovsky, D., Muller‐Hermelink, H.K., and Ralfkaier, E. (2008) T‐cell prolymphocytic leukaemia. In WHO Classification of Tumours of Haematopoietic and Lymphoid Tissue (eds. S.H. Swerdlow, E. Campo, N.L. Harris, et al.). IARC Press, Lyon, France, pp. 270–271.
  2. 2. Soma, L., Cornfield, D.B., Prager, D., et al. (2002) Unusually indolent T‐cell prolymphocytic leukemia associated with a complex karyotype: is this T‐cell chronic lymphocytic leukemia?. Am J Hematol 71:224–226.
  3. 3. Valli, V.E. (2007) T‐cell prolymphocytic leukemia. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 302–304.
  4. 4. Valli, V.E. (2007) B‐cell prolymphocytic leukemia. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 161–168.
  5. 5. Valbuena, J.R., Herling M, Admirand, J.H., et al. (2005): T‐cell prolymphocytic leukemia involving extramedullary sites. Am J Clin Pathol 123:456–464.

T‐cell granular lymphocytic leukemia/lymphoma


Defining the neoplasm


Large granular lymphocytes (LGL) are lymphocytes that have cytoplasmic granules visible in normal and neoplastic cells. The granules are usually near the nucleus, the number varies widely from 3 to 20, and they range from large (3 μm) and obvious to inconspicuous or dust‐like. The cytoplasm of an LGL is relatively abundant and clear, or at least not deeply basophilic. They can form solid tumors with or without leukemia. In the intestinal tract LGL are located between epithelial cells, and in cats and horses the granules in normal LGL can be so large they are seen occasionally at medium magnifications or even with H&E (Figure 7.39). In cats they can be found between biliary epithelial cells and in the thymus. They have been erroneously designated globular leukocytes or as neoplasms of globular leukocyte origin.

Micrograph of large granular lymphocytes (LGL) in a cat's intestine. The prominent eosinophilic cells are LGL.
Micrograph of large granular lymphocytes (LGL) in a cat's intestine. Globular eosinophilic intracytoplasmic inclusions (arrows), intraepithelial and lamina propria lymphocytes (arrowed) are positive.
Micrograph of large granular lymphocytes (LGL) in a dog's blood, displayed in aqueous Romanowsky automated stain. Intracytoplasmic granules stain poorly.
Micrograph of large granular lymphocytes (LGL) in a dog's blood, displayed in methanolic Romanowsky automated stain. Intracytoplasmic granules are prominent and easily identified.

Figure 7.39 Large granular lymphocytes (LGL), cat, intestine. (A) The prominent eosinophilic cells are LGL. These cells have been called globular leukocytes. Only in cats and horses are LGL this obvious in H&E. (B) Same section, CD3: The globular eosinophilic intracytoplasmic inclusions (arrows) seen in H&E are strongly positive, as are other intraepithelial lymphocytes and a few lymphocytes in the lamina propria (arrow head). (Image courtesy of Luke Borst, North Carolina State University.) (C,D) LGL, dog blood. Comparison of aqueous (C) and methanolic (D) Romanowsky automated stains. Intracytoplasmic granules stain poorly with the aqueous stain (C) and are not visible but are prominent and easily seen with the methanolic stain (D). Granules in canine LGL tend to be smaller and finer than in cats and horses but there are wide ranges of size and numbers per cell and per species. A few nuclei have characteristic irregular contours with indentations. (C,D Images courtesy of Robin Allison, Oklahoma State University.)


A few LGL may be seen in blood films from normal animals. They may increase in inflammatory and infectious diseases and some of these conditions can induce a lymphocytosis that is in the range expected with leukemias.1,2 LGL may be markedly increased to 50,000–200,000/μL in leukemia of CLL, acute leukemia of LGL, or LGL lymphoma with secondary (concurrent) leukemia.3–14 In general, the granules are difficult to see in histologic preparations but fairly easy to see in cytologic preparations. The larger the LGL the easier the cytoplasmic granules are to see. The granules are not obvious in normal LGL in blood but are relatively easy to see in neoplastic cells in cytologic preparations made from LGL tumors, effusions with neoplastic LGL, or blood with neoplastic LGL. They can be seen easier with Wright–Giemsa or PTAH than with H&E or Diff‐Quik. The type of cytologic stain used is important to visualize the cytoplasmic granules. It is better to use alcohol‐based stains (methanolic) and not aqueous stains (Figure 7.39C,D), even if methanol fixation is used.15 Oil‐immersion objectives will improve the visualization of granules.


Cytology and flow cytometry are better techniques to identify tumors of LGL than histopathology sections. Histopathology can easily diagnose lymphoma but does not allow LGL type to be ruled out based on H&E as the cytoplasmic granules can be imperceptible. LGL cell type is better appreciated in cytology preparations and with special stains such as granzyme B.


Although all LGL look similar morphologically they are heterogeneous in how they function and by phenotypes. Canine LGL express the leukointegrin αdβ2 regardless of lineage. There are two broad phenotypes: CD3‐positive LGL that are cytotoxic T lymphocytes, most of which are CD8 positive, and CD3‐negative LGL that are considered NK cells. Canine LGL cytotoxic T lymphocytes may express T‐cell receptors that are αβ (TCRαβ) (approximately 60%) or γδ (TCRγδ) (approximately 30%). CD3‐positive and CD3‐negative LGL types can become reactive or neoplastic, lymphomas or leukemia. The diagnosis of an LGL leukemia or lymphoma is sufficient for most diagnostic cases in animals but they can be further subtyped as described above and in dogs we can expect >90% will be T‐cell and a few are considered to be NK.2,3,5,14 They may be further subdivided by size of cells or a profile of leukocyte antigen expressions.5,8,14


The biologic behavior of LGL tumors in cats is the same regardless of subtypes. It is an aggressive lymphoma that shortens the cat’s life, and the diagnosis can be determined from cytology or histopathology. In dogs, subtyping LGL tumors has not been correlated with survival or treatments. In dogs, LGL CLL or lymphoma can be indolent, with lymphocytosis lasting years or, in some cases, progressing rapidly to cause death. The latter is associated with larger and more immature cells types. In humans, biologic behavior seems to follow phenotype, CD3‐positive/CD8‐positive types are cytotoxic T‐cell lineage and are indolent. CD3‐negative types are NK lineage and have an acute aggressive course with marked hepatosplenomegaly.


There are ample data to conclude that LGL leukemia originates in splenic red pulp.2,5,16 In all species intestinal LGL lymphoma is believed to arise from intraepithelial LGL. This is enteropathy‐associated intestinal T‐cell lymphoma (EATCL),3,5 which in humans is believed to originate in LGL cells that have undergone a clonal transformation from prolonged antigenic stimulation induced by a variety of inflammatory bowel diseases (IBD). Cats and dogs have IBD and intestinal lymphoma and the association of the two diseases is often suggested but has never been proven. The tumor in cats is aggressive but in some dogs it can have a more indolent course. LGL leukemia or lymphomas have been seen in dogs,2,4 cats,5,6 horses,7–9,17 cows, rodents,10–12 and birds.13


LGL can cause lymphoma or leukemia (acute or chronic) or both and they may increase in a variety of inflammatory or infectious diseases (e.g. canine ehrlichiosis).1 Retroviral particles have been reported in a cell line derived from a canine LGL leukemia,18 but they are likely not causative.


In dogs


Epidemiology and occurrence


LGL diseases in dogs occur in mature and older dogs with a mean age of 10 years, usually of large‐breed types. They present with a variety of signs that may involve the gastrointestinal system and may have a history of weight loss and depression of variable degree. Females are nearly twice as often affected as males.


Pathology


Most dogs with this disease present with an LGL lymphocytosis of 5000 or more/μL. Animals with a clonal LGL tumor likely had lymphocytosis for at least 3 months, which makes causes due to inflammation less likely, but duration of the lymphocytosis is usually not known. Neoplasms of LGL may produce a T‐cell CLL or T‐cell lymphoma (rarely NK) with or without neoplastic cells detected in the blood. The key to the diagnosis of LGL neoplasms is to recognize cytoplasmic granules in the neoplastic lymphocytes. These granules are usually near the nucleus, and vary in number from 3–20/cell. In general, they are not as obvious in dogs as they are in cats and horses (Figures 7.40, 7.41, and 7.54). Granules are easier to find in cytologic then histologic slides (see this section on cats). Cytologically, LGL cells are of intermediate to large size with nuclei about 1.5–3 RBC in diameter, round to oval or indented, reniform or infrequently with clefts (Figures 7.40 and 7.41). In histopathology the nuclei appear round and the reniform pattern is only seen well in cytologic preparations. The cytoplasm is fairly abundant and lightly stained in cytology and unapparent in histopathology.

Micrograph of large granular lymphocytic (LGL) leukemia in a cat, a lymphocytosis of 64,100/μL.
Micrograph of large granular lymphocytic (LGL) leukemia in a dog.  The leukocyte count was 243,000/μL and >90% were LGL, T-cell. The large neoplastic cells have granules (arrows) and basophilic cytoplasm.” src=”http://veteriankey.com/wp-content/uploads/2020/03/c07f040b.jpg”> <FIGCAPTION><br />
<P><SPAN class=figureLabel><A id=c7-fig-0040 href=Figure 7.40 (A) Large granular lymphocytic (LGL) leukemia in 6‐year‐old cat, lymphocytosis of 64,100/μL. Results of CBC prompted an abdominal ultrasound. A mass was found in the intestines and mesenteric lymph nodes were enlarged. Treatment was started but the cat declined rapidly and was euthanized. Cats with this combination of intestinal and leukemic LGL have short survival times. Nuclei are 3–4 RBC in diameter, the indentations are prominent, and the granules are juxtanuclear. (Image courtesy of Jessica Bailey, Auburn University.) (B) Large granular lymphocytic (LGL) leukemia, dog. Two‐year‐old English setter presented with intense pruritus. The mucous membranes were pale and there were numerous petechiae. The leukocyte count was 243,000/μL and >90% were LGL, T‐cell. The large neoplastic cells have pink granules (arrows) and basophilic cytoplasm. The cytoplasmic granules are diagnostic for LGL, but they could be confused with myeloid cells, especially since these nuclei have indentations and the granules are fine. In histopathology nuclei look round and the reniform pattern is only seen well in cytologic preparations. Cytoplasmic granules are also difficult to see in H&E. Note mitotic figure (lower left) which is extremely unusual to find in the peripheral blood, but this is a thick area of the film where cells are concentrated and the white blood cell count was incredibly high. Some dogs and rats with LGL leukemia will have concurrent immune hemolytic anemia, but in this image the stacks of red blood cells are rouleaux not agglutination.

Photo of a horse's mesenteric node.
Photo of a horse's intestine.
Micrograph of large granular lymphocyte (LGL) intestinal lymphoma in a horse, touch imprint. Non-neoplastic lymphocytes are smaller, the nuclei are darkly stained and no cytoplasm is evident.
Micrograph of granzyme B-positive cells in a horse with LGL. Inset: The same micrograph with greater magnification.

Figure 7.41 Large granular lymphocyte (LGL) intestinal lymphoma, horse. (A) Mesenteric node. (B) Intestine. (C) Touch imprint. LGL lymphoma in the horse typically affects abdominal organs, such as intestine, liver, and lymph nodes. This is a large cell LGL, the nuclei are large (2–4 RBC in diameter), and cytoplasm is abundant. Many cells contain obvious intracytoplasmic granules typical of LGL in horses and cats. Non‐neoplastic lymphocytes are smaller, the nuclei are darkly stained and no cytoplasm is visible. LGL lymphoma is a T‐cell neoplasm. They will be CD3 positive and granzyme B positive and were misinterpreted as globular leukocyte tumors. (A–C Images courtesy of Allison Boone and Jennifer Neel, NCSU.) (D) Granzyme B‐positive cells in a horse with LGL. Inset: Higher magnification.


Splenomegaly is usually present and may be palpable at initial examination or detected by imaging. Peripheral lymphadenopathy is unusual but may be present. The LGL cells may be present in the marrow but not causing phthisis of normal marrow cells. Unlike the human or feline LGL tumors, the peripheral blood neutrophil numbers in dogs stay within normal limits or there is a mild neutrophilia. Mild to moderate non‐regenerative anemia (PCV 20–30%) is present in about half of the dogs.


The LGL tumor cells or persistent lymphocytosis may be phenotyped via IHC, ICC, or flow cytometry. Depending on the expertise, funds, and how detailed a subtype category is desired, a broad panel of proven antibodies should be used. At this time subtyping LGL has not been correlated with clinical outcomes or different treatments in dogs.


Cytochemistry and immunohistochemistry


Almost all LGL tumors in dogs are positive with CD3 (>90%), and those that are negative are assumed to be of NK cell type (<10%). For most cases the diagnostic work‐up will likely stop at the diagnosis of LGL lymphoma/leukemia or after the neoplastic cells are shown to be CD3 positive and negative to B‐cell antibodies. If pursued further, the neoplasms are granzyme B‐positive and can be typed with antibodies to recognize antigens CD8, CD4, CD11d, or with broad panels to determine surface receptors, αβ (TCRαβ) γδ (TCRγδ) and other characteristics.


A study of 25 dogs with persistent (>3 months) LGL lymphocytosis characterized the immunophenotypic profile of the LGL cells.2 Diagnoses included LGL leukemia, ehrlichiosis, reactive lymphocytosis, and persistent lymphocytosis with or without anemia. Clonality was not determined. The original reference should be read for the details provided.2 More than 90% of the LGL were CD3 positive and all were CD21 negative, indicating T‐cell phenotype. The majority (60%) of the LGL cases had an αβ type of T‐cell receptor with approximately 32% having the γδ type of T‐cell receptor; the two negative cases were considered NK (8%).2 In this series all cases had LGL cells that were positive for CD18, CD11a, >90% were positive with CD45Ra, CD11d and CD11b was absent and CD11c present in about two‐thirds of cases with all of these antigens on the cell surface and not in cytoplasm.8 The pattern of integrin αdβ2 (CD11d) expression was distinctive. It was present in over 90% of cases, suggesting splenic origin. CD11d is the leukointegrin expressed by macrophages and T cells of the splenic red pulp and by peripheral blood LGLs. To put this level of expression in focus, only about 1% of peripheral blood lymphocytes in normal dogs have the αdβ2 + γδ leukointegrin. However, in the sinus areas of the spleen over a third of the lymphocytes present have this signature type of leukointegrin. The combination of CD11d expression, splenomegaly, and the marrow only lightly involved suggested the spleen was the source of the LGL cells.2


Recently a gamma/delta (TCRγδ) T‐cell LGL tumor was identified in a dog with lymphoma in the mediastinum and thoracic lymph nodes.16 An extensive panel of antibodies were used in flow cytometry on neoplastic cells aspirated from the mediastinal mass. The results were used to characterize this LGL lymphoma and document the first report of a γδ T‐cell LGL in the dog. The mediastinal tumor was CD11d negative, which is the cluster of differentiation antibody that is associated with cells in the splenic red pulp, macrophages, T cells, and LGL. Although the tumor was present in the spleen and the spleen is a common origin of γδ lymphomas, the negative CD11d result prompted the authors to suggest that this LGL tumor originated in the mediastinum or possibly liver.16 The dog survived less than a month with treatments. Too few cases of γδ T‐cell lymphomas in dogs have been followed to predict accurate survival data, but it appears γδ T‐cell lymphomas are aggressive with short survival times post diagnosis.


LGLs contain perforin‐ and granzyme B‐positive granules. Granzyme B enzyme is a serine protease that induces apoptosis and is found in the granules of LGL of both T‐cell and NK‐cell origin. Granzyme B is a useful marker for LGL.


Differential diagnosis


If the diagnosis of lymphoma is apparent, then search for granules in neoplastic cells, preferably in cytologic preparations with alcohol‐based stains.15 The detection of cytoplasmic granules is the key to diagnose LGL neoplasms and distinguish them from other lymphomas. If lymphocytosis is present, it is helpful to monitor or determine how long the LGL cells have been increased in the blood. In humans and in animals it is assumed that if the increase in peripheral blood LGL has been present for 3 months or longer, the condition is a clonal malignancy of LGL cells. PCR for T‐cell gene rearrangement can be performed to determine clonality.


The most important differential for LGL leukemia or LGL lymphoma with secondary leukemia is an inflammatory or antigenic reactive lymphocytosis.1 Dogs with tick‐borne diseases, especially canine ehrlichiosis, will have clinical signs and blood parameters similar to dogs with neoplasms of LGL. Further complicating the distinction is that the lymphocytes in canine ehrlichiosis can be LGL and rare cases will have a monoclonal gammopathy. Thrombocytopenia can be a feature of tumors and is common in ehrlichiosis. Hypoadrenocorticism may cause lymphocytosis, but electrolyte patterns and ACTH stimulation will rule in or rule out this differential. Cats with bartonella or an epinephrine surge may have lymphocytosis. These differentials can be ruled out through serology, visualization of organisms, repeat blood work, PCR for specific infectious diseases, response to doxycycline and, if needed, immunophenotyping to determine homogeneity of lymphocytes or clonality determination via PCR for gene rearrangement (PARR).


Hepatosplenic lymphoma looks similar to LGL lymphoma as they both can have enlarged mesenteric lymph nodes, hepatosplenomegaly, and the spleen can be the origin of both tumors. The neoplastic cells in both are CD3, CD11d, and granzyme B positive. Gamma/delta T‐cell is the subtype of lymphocyte in hepatosplenic lymphomas, but now there is one report of γδ T‐cell that was an LGL.16 The neoplastic cells in hepatosplenic lymphoma are not reported to be granulated and they exhibit erythrophagocytosis; LGL lymphomas rarely do so although they will in the F344 rat. The liver should be more severely involved in hepatosplenic lymphoma.


Survival


In dogs it is believed that tumors of LGL follow an indolent course, but too few cases have been followed to provide accurate and predictive data. Also, the diseases reported may include reactive lymphocytosis, leukemia, or lymphoma. In humans, most cases seem indolent, but phenotyping is used to help predict indolent versus malignant behavior. Long‐term follow‐up correlated with morphology and immunophenotyping has been done in humans but not in animals. There are cases of LGL leukemia and lymphoma in dogs that are rapidly progressive,8,16 but LGL lymphocytosis appears to follow an indolent course in most dogs. In cats LGL neoplasms have an aggressive course, especially if there is leukemia and/or transmural intestinal LGL lymphoma.5,19


Concurrent problems in dogs are anemia and splenomegaly. Cytopenias do not appear to be a complicating factor. Some LGL tumors respond to chemotherapy and/or steroids, and the dogs that respond are relatively long survivors. Infectious causes of lymphocytosis need to be ruled out.


In cats


Epidemiology, occurrence, and clinical signs

Cats with LGL diseases have a wide age range and the mean is approximately 10 years.5,6,19 There is almost a 3:1 female predominance. They present because they are sick; this is not an indolent tumor in cats. Common problems are weight loss, anorexia, abdominal masses, enlarged mesenteric lymph nodes, hepatosplenomegaly, and renomegaly. Less common problems are icterus, diarrhea, vomiting, body cavity effusions, and peripheral lymphadenopathy.5 Lymphoma in the abdomen is a common differential diagnosis for cats with these types of abnormalities. Some cats will have a history of IBD. Cats with LGL diseases are almost always FeLV and FIV negative.


A diagnosis of LGL lymphoma/leukemia can often be made from examination of a blood film. Lymphocytosis ranges from mild (<2000/μL) to marked (>100,000/μL) and some cases will be >300,000/μL in blood. In one study, 18 of 21 cats had neoplastic LGL cells in the blood, although sample bias is likely as the results of blood cell counts were used to find cases of LGL.5 Anemia is present in about one‐third and cytopenia of leukocytes is not expected. In fact, feline LGL diseases will have neutrophilia and a paraneoplastic mechanism has been speculated to produce this neutrophilia.5 Serum hepatic enzymes and bilirubin are increased in about half of the cases but there are no characteristic patterns. The neoplasm is not associated with FeLV or FIV; both are negative in almost all cases.5,6 FeLV‐associated lymphomas are usually multicentric or mediastinal and in young cats. FIV‐associated tumors in cats are usually high‐grade B‐cell lymphomas.


Pathology


LGL can produce lymphoma and some of these cases will have neoplastic cells in the circulation. LGL may also be a leukemia and some of these cases will have neoplastic cells in solid tissues. LGL lymphoma appears to arise most frequently in the intestinal tract in cats.19 Cats with LGL lymphoma will have thickened intestines, usually jejunum and ileum, but other regions can be involved. The size of the intestines, white discoloration imparted by the tumor and transmural involvement all vary with the severity of the neoplasm. Invariably, the lamina propria is filled with neoplastic LGL cells. In most cases, the tumor extends into the submucosa, in many it extends into muscle layers, and some will be transmural. The tumor has an epitheliotropic pattern that may be obvious or may need to be searched for and highlighted by immunostaining for CD3.19 Numerous neoplastic LGL cells will be seen between intestinal epithelial cells, but this pattern is never as obvious as the infiltration in the lamina propria. Mesenteric lymph nodes are enlarged and neoplastic in all cases. The tumors will efface nodes and fill cortex and medulla in about half of the cases. Liver was neoplastic in 12 of 13 cases and the spleen in 8 of 13.5 The kidneys will be involved bilaterally in severe cases, and if enough parenchyma is compromised the cats will be azotemic. Other sites that may be involved include the skin, liver, spleen, and bone marrow.


Less than 25% will have enlarged peripheral lymph nodes and a few will have effusions in thoracic or abdominal cavities. The effusions contain neoplastic LGL cells and cytofuge preparations of the fluid provides excellent visualization of the cytoplasmic granules. Bone marrow is only lightly infiltrated by the LGL cells in cases of lymphoma and myelophthisis by the neoplasm does not appear to be a clinical problem.


Gross and histologic sections will establish the diagnosis of lymphoma easily. However, the diagnostic cytoplasmic granules are not easily seen with H&E‐stained slides. Cytologic preparations of blood, effusions, or touch impressions of mesenteric nodes or intestinal lesions will reveal the granules and therefore establish the diagnosis of LGL lymphoma easier than histopathology. The identification of LGL is important clinically as nongranulated forms of gastrointestinal lymphoma, especially small cell types that are mucosal, have a more protracted course than the rapidly progressive LGL lymphoma.19 If a diagnosis of lymphoma is made from biopsy material of mesenteric lymph node or gastrointestinal specimens from a cat, consider examination of cytologic preparations from the same specimens and/or blood films to look for granules. Slides stained with methanolic‐based stains are superior to stains with an aqueous base (see Figure 7.39).15 Flow cytometry to search for and characterize neoplastic cells in circulation is helpful as well. If none of these are available then search for eosinophilic granules at 100× magnification in H&E‐stained sections or purple granules in PTAH‐stained sections.


Two cell types are recognized by morphology. Phenotypically there may be several cell types, but the biologic behavior of all is similar, comprising an aggressive course with survival measured in days to weeks post diagnosis.5,19 The smaller cell type of LGL has a cell diameter of 8–15 μm, and the cytoplasmic granules are small and inconspicuous. They have round nuclei, often with an indentation, and the chromatin is coarse and dense. Nucleoli are not usually visible. The cytoplasm and nuclei impart a mature cell appearance. The other cell type is larger, 15–35 μm in diameter, the cytoplasm is lightly stained, and the cytoplasmic granules range from small to large and are distinctive. These cells look immature, they have more cytoplasm, and the nuclei are larger; the chromatin is also less dense and nucleoli are usually visible. The granules are much larger than those of the small cell type. Some will be 3.0 μm and are obvious in cytologic preparations, especially if examined with oil immersion (Figure 7.41). The granules stain dark purple with Wright–Giemsa, not as well with Diff‐Quik, and in H&E‐stained slides the granules are much less obvious. They are eosinophilic in H&E and purple in PTAH but not all tumors will react with PTAH. They are seen most easily with oil immersion in cytologic preparations of blood, body cavity effusions, or touch imprints of tumors. All LGL tumors should be granzyme B positive.


Cytochemistry and immunohistochemistry


The neoplastic cells are positive with the leukocyte marker CD18. Most cells in an LGL lymphoma stain positively with CD3 (approximately 90%) and are negative with multiple B‐cell markers. The minority that are negative for both are assumed to be of NK‐cell type.5 The majority of the T‐cell lymphocytes are cytotoxic lymphocytes (CD8 alpha positive) and a few are helper T lymphocytes (CD4 positive). CD11d is associated with splenic cells and was only expressed by approximately 25% of feline LGL lymphomas, suggesting they are not of splenic origin in cats as they are believed to be in dogs. Intestinal origin is more likely in cats, especially given the gross and histologic distribution of the tumors and the biology of LGL.5,19 Original manuscripts contain details of leukointegrin profiles and their interpretation.5,16,19


References



  1. 1. Weiser, M., Thrall, M., Fulton, R., et al. (1991) Granular lymphocytosis and hyperproteinemia in dogs with chronic ehrlichiosis. J Am Anim Hosp Assoc 27:84–88.
  2. 2. McDonough, S.P. and Moore, P.F. (2000) Clinical, hematological, and immunophenotypic characterization of canine large granular lymphocytosis. Vet Pathol 37:637–646.
  3. 3. Chan, W.C., Foucar, K., Morice, W.G., and Catovksy, D. (2008) T‐cell large granular lymphocytic leukemia. In WHO Classification of Tumours of Haematopoietic and Lymphoid Tissue (eds. S.H. Swerdlow, E. Campo, N.L. Harris, et al.). IARC Press, Lyon, France, pp. 272–273.
  4. 4. Wellman, M.L., Couto, C.G., Starkey, R.J., and Rojko, J.L. (1989) Lymphocytosis of large granular lymphocytes in three dogs. Vet Pathol 26:158–163.
  5. 5. Roccabianca, P., Vernau, W., Caniatti, M., and Moore, P.F. (2006) Feline large granular lymphocyte (LGL) lymphoma with secondary leukemia: primary intestinal origin with predominance of a CD3/CD8αα phenotype. Vet Pathol 43:15–28.
  6. 6. Wellman, M.L., Hammer, A.S., DiBartola, S.P., et al. (1992) Lymphoma involving large granular lymphocytes in cats: 11 cases (1982–1991). J Am Vet Med Assoc 201:1265–1269.
  7. 7. Kramer, J., Tornquist, S., Erfle, J., and Sloeojan, G. (1993) Large granular lymphocyte leukemia in a horse. Vet Clin Pathol 22:126–128.
  8. 8. Grindem, C.B., Roberts, M.C., McEntee, M.F., and Dillman, R.C. (1989) Large granular lymphocyte tumor in a horse. Vet Pathol 26:86–88.
  9. 9. Herraez, P., Berridge, B., Marsh, P., et al. (2001) Small intestine large granular lymphoma in a horse. Vet Pathol 38:223–226.
  10. 10. Losco, P.E. and Ward, J.M. (1984) The early stage of large granular lymphocyte leukemia in the F344 rat. Vet Pathol 27:186–291.
  11. 11. Stromberg, P.C., Rojko, J.L., Vogtsberger, L.M., et al. (1983) Immunologic, biochemical, and ultrastructural characterization of the leukemia cell in F344 rats. J Natl Cancer Inst 71:173–181.
  12. 12. Miyajima, R., Hosoi, M., Yamamoto, S., et al.(1999) Eosinophilic granulated cells comprising a tumor in a fischer rat. Toxicol Pathol 27:233–236.
  13. 13. Patnaik, A.K. (1993) Histologic and immunohistochemical studies of granular cell tumors in seven dogs, three cats, one horse, and one bird. Vet Pathol 30:176–185.
  14. 14. Williams, M.J., Avery A.C., Lana, S.E., et al. (2008) Canine lymphoproliferative disease characterized by lymphocytosis: immunophenotypic markers of prognosis. J Vet Intern Med 22:596–601.
  15. 15. Allison, R.W. and Velguth, K.E. (2010) Appearance of granulated cells in blood films stained by automated aqueous versus methanolic Romanowsky methods. Vet Clin Pathol 39:99–104.
  16. 16. Ortiz, A.L., Carvalho, S., Leo, C., et al. (2015) Gamma delta T‐cell large cell granular lymphocyte lymphoma in a dog. Vet Clin Pathol 44:442–447.
  17. 17. Mastorilli, C., Cesar, F., Joiner, K., et al. (2015) Disseminated lymphoma with large granular lymphocyte morphology diagnosed in a horse via abdominal fluid and transtracheal wash cytology. Vet Clin Pathol 44:437–441.
  18. 18. Ghernati, I., Corbin, A., Chabanne, L., et al. (2000) Canine large granular lymphocyte leulemia and its derived cell line produce infectious retroviral particles. Vet Pathol 37:310–317.
  19. 19. Moore, P.F., Moore, P.F., Rodriguez‐Bertos, A., and Kass, P.H. (2012) Feline gastrointestinal lymphoma: mucosal architecture, immunopheotype, and molecular clonality. Vet Pathol 49:658–668.

Peripheral T‐cell lymphoma not otherwise specified


Peripheral T‐cell lymphoma not otherwise specified (PTCL‐NOS) is a classification used for T‐cell lymphomas that are not yet characterized or that cannot be fully classified.1 In human oncology the tumors under this umbrella have decreased as subtypes within PTCL‐NOS are defined as disease entities, so the designation NOS is often not used and the specific diagnosis provided. This is also happening in veterinary pathology as we recognize the different T’cell lymphomas.


Classifications in humans use IHC and molecular markers as well as known disease patterns, biologic behaviors, and histopathology to characterize the diseases. In veterinary pathology we do not have all the tools or economic resources needed to characterize these lymphomas and we lack accurate follow‐up data, which is a critical component to characterize the lymphoma as a disease entity. Despite these limitations many of the tumors in this classification have now been recognized and characterized, at least to the limit of our antibodies and resources, although the number of animals in some groups are small.


Subtypes of PTCL include: angioimmunoblastic T‐cell lymphoma (AILT), angiocentric T‐cell lymphoma, hepatosplenic, intestinal T’cell, subcutaneous panniculitis type, mycosis fungoides, mature T‐cell leukemia/lymphoma, anaplastic large cell lymphoma, and T‐zone lymphoma (TZL).1 TZL is an uncommon lymphoma in humans and is listed under PTCL‐NOS, however, it is described as a distinct diagnosis in this chapter and is listed individually in Box 7.1 because it is a common and well‐characterized lymphoma in animals, especially dogs. In addition, we know the biologic behavior of this lymphoma is indolent. All the other tumors listed are aggressive lymphomas. Hepatosplenic T‐cell lymphoma is also listed separately in Box 7.1 as it is a defined entity in humans and is characterized in dogs, although the total number of cases is low. Hepatocytotropic T‐cell lymphoma is not a recognized disease but it has been reported in dogs and cats and is partially characterized. It is described with hepatosplenic T‐cell lymphoma in this section because they share similar features and it allows comparisons between the two tumors.


If a tumor can be classified that diagnosis should be provided; the names PTCL or PTCL‐NOS should be reserved for those T‐cell tumors that cannot be classified, perhaps because the tumor lacked differentiating features, complete antibody profiles or molecular tests were not available, the organ distribution of the tumor was not known, or the owners or oncologists did not want or could not afford complete characterization. Some generic descriptors for these tumors are the following. Most importantly, they are extrathymic T‐cell tumors and most are high grade. Many will look like DLBCL and are only differentiated by phenotyping. Most are in lymph nodes and cause a paracortical expansion; compressed sinus and tumor cells will extend beyond the capsule. A few are extranodal, usually occurring in the skin or subcutis. They may be associated with vascular proliferation, vascular invasion, and necrosis. The cells can be calssified as pleomorphic small, pleomorphic mixed (small, medium and large cells), or pleomorphic large. One study reported the majority to be large cell2 and another indicated the majority were mixed.3 Eosinophils and macrophages may be noted. The mitotic rate will be variable and can be used to assign high grade or low grade. Nuclei are round to oval and nucleoli quite variable.


Cytologically, PTCL look like immature aggressive lymphomas and immunophenotyping is needed to distinguish B‐ versus T‐cell type. Large cell types and those that efface nodes with extranodal tumor will look like DLBCL and are only differentiated by phenotyping. The mixed cell types look like TCRLBCL and are also distinguished via phenotyping. Inflamed cutaneous non‐epitheliotropic T‐cell lymphoma resembles cutaneous histiocytosis or even inflammation.


Combining data from two studies of canine lymphoma produces almost 1600 cases, of which approximately 15% were classified as PTCL‐NOS and 3–11% were T‐zone lymphoma.2,3 Therefore these two T‐cell classifications accounted for about 20% of the canine lymphomas. Identification of TZL is crucial as its indolent behavior is in contrast to the other lymphomas in this group. The unique histopathology of TZL is diagnostic. Flow cytometry with broad panels of antibodies have also been used to help subdivide T‐cell tumors.4–6 By flow cytometry, CD21 positivity, CD25 positivity, CD45 negativity, and high expression of class II MHC was used to identify TZL and predict long survival times.7,8 CD21 negativity, CD25 positivity, CD45 positivity, and low expression of class II MHC identified other T‐cell lymphomas that have short survival times of approximately 160 days.9 Only 15 cases had histopathology, 10 of which were PTCL‐NOS and 5 LBL.


The following sections include specific diagnoses that are in the broad group of PTCL.


Mycosis fungoides and Sézary syndrome


Defining the neoplasm


Cutaneous lymphoma can be B‐ or T‐cell, epitheliotropic, or non‐epitheliotropic. The most common is T‐cell epitheliotropic; that is, mycosis fungoides. Mycosis fungoides affects the skin and/or mucous membranes of humans and many species of animals, including rodents.1–3,7–15 The diagnosis requires histology and the disease is characterized by linear infiltrations of the epidermis by small to intermediate‐size lymphocytes that typically have sharp shallow nuclear indentations. In the fully developed lesion there are clefts in the epidermis known as Pautrier’s microabscess that are filled with neoplastic cells (Figures 7.427.45). The disease has stages beginning with an interface dermatitis that progresses to a patch or plaque. In the untreated case the skin lesion will enlarge, and if tumor cells appear in the blood the condition is called Sézary syndrome (Figure 7.46). In the early stages of the infiltration or at the edges of an established area of tumor the neoplastic cells form a single row just inside of the basement membrane known as the “string of pearls.” As the disease progresses the cellular infiltration separates dermal collagen fibers and infiltrates the mural aspect of the hair follicles and glandular adnexa. In dogs the infiltration may completely fill the apocrine glands. Tumor extends to the level of the deeper dermal vessels.

Micrograph of mycosis fungoides in a dog's oral cavity. Epithelium is thick and has lymphoid cells through the basement membrane and in the submucosa. There are small cystic areas with edema and lymphocytes.
Micrograph of mycosis fungoides in a dog's oral cavity. Neoplastic cells in the epithelium are strongly positive, indicating T-cell origin and those in submucosa are moderately positive.

Figure 7.42 Mycosis fungoides, oral cavity, dog. (A) The epithelium is thickened and markedly infiltrated by lymphoid cells that extend through the basement membrane and are in the submucosa. There are small cystic areas in the epithelium filled with edema and lymphocytes (Pautrier’s microabscess). (B) CD3: Neoplastic cells in the epithelium are strongly positive, indicating T‐cell origin and those in submucosa are moderately positive. See Figure 7.43.

Micrograph of mycosis fungoides in a dog's oral cavity. Neoplastic lymphocytes have considerable cytoplasm, nuclei are 1.5 times the size of RBC, chromatin is dense and many nuclei are folded or convoluted.
Micrograph of mycosis fungoides in a dog's oral cavity. Neoplastic cells are 2–3 times as large as small non-neoplastic lymphoid cells. Cytoplasm of neoplastic cells are basophilic; mitoses (arrow) are present.

Figure 7.43 Mycosis fungoides, oral cavity, dog. (A) Neoplastic lymphocytes have considerable cytoplasm, nuclei are 1.5 times the size of RBC, chromatin is dense and several nuclei are folded or convoluted. (B) Cytology: Tumor cells exfoliated in large numbers. The neoplastic cells are 2–3 times as large as the dense small non‐neoplastic lymphoid cells. Cytoplasm of neoplastic cells is abundant and deeply basophilic; mitoses (arrow) were present. Most nuclei are round or oval, some have indented nuclear profiles, and a few horseshoe‐shaped nuclei are present.

Micrograph of mycosis fungoides in a cat's skin. Skin is intermittently thick and thin, cystic areas in epidermis have edema and lymphocytes. Neoplastic lymphoid cells go into the deeper dermis and into adnexa.
Micrograph of mycosis fungoides in a cat's skin. The infiltrating lymphoid cells are strongly and uniformly immunostained, including those in the epithelial cysts.

Figure 7.44 Mycosis fungoides, skin, cat. (A) A 12‐year‐old domestic short hair cat presented with hair loss and swollen skin over the tail. The skin is intermittently thick and thin with large cystic areas in the epidermis filled with edema and lymphocytes. Neoplastic lymphoid cells extended into the deeper dermis and infiltrated adnexa. (B) CD3: The infiltrating lymphoid cells are strongly and uniformly immunostained, including those in the epithelial cysts.

Micrograph of pagetoid reticulosis in a dog's skin. Affected skin has greatly thickened epidermis with a marked infiltration of lymphocytes in the epidermis and in the hair follicles (arrows).
Micrograph of pagetoid reticulosis in a dog's skin. The infiltrating cells within the thickened epidermis are uniformly and heavily marked. Intraepithelial infiltration in one of the hair follicles (arrow).

Figure 7.45 Pagetoid reticulosis, skin, dog. (A) A 6‐year‐old Schnauzer dog presented for foci of hair loss and depigmentation that were unresponsive to antibiotics. A biopsy of affected skin has a greatly thickened epidermis with a marked infiltration of lymphocytes in the epidermis and in the hair follicles (arrows). Some regions of the epidermis are devoid of melanocytes (left) and in other areas they are still present (right); there also are melanophages in the dermis. (B) CD3: The infiltrating cells within the thickened epidermis are uniformly and heavily marked. The contrast in colors makes the intraepidermal pattern easier to see. Note intraepithelial infiltration in one of the hair follicles (arrow). Some of the cells stained “brown‐black” in the dermis are melanophages (see A).

Micrograph of Sézary syndrome on a horse's buffy coat. The animal had brick red oral mucosa and atypical cells in circulation. A high population of intermediate and large cells and occasional “blast-type” cells (center).
Micrograph of Sézary syndrome on a horse's buffy coat. Cell in the center is typical of large cells in the marrow, blood, and pericardial effusion. The nucleus has a chromatin pattern and two large nucleoli.
Micrograph of Sézary syndrome on a horse's buffy coat. Bone marrow and lymph nodes contain mononuclear cells. The tumor cells are CD3 positive.
Micrograph of Sézary syndrome on a horse's buffy coat. Dense infiltration of neoplastic lymphoid cells in oral submucosa. Mucous membranes are hyperemic. Inset: Neoplastic cells dissect through myocardium.

Figure 7.46 Sézary syndrome, horse, buffy coat. (A) A thoroughbred 10 years old presented in late pregnancy with ventral edema and muffled heart sounds. The animal had brick red oral mucosa and atypical cells in circulation. A buffy coat preparation of blood was prepared, which had a high population of intermediate and large cells and occasional “blast‐type” cells (center). The white blood cell count in this horse ranged between 11,000 and 15,000/μL. The atypical cells were originally classified as monocytic but after review were identified as lymphocytic. (B) The cell in the center is typical of the large cells seen in the marrow, blood, and pericardial effusion. The nucleus has a fine chromatin pattern and two very large nucleoli. (C) Bone marrow and lymph nodes contained mononuclear cells similar to those found in the blood. The reniform nuclei are monocyte‐like but the tumor cells will be CD3 positive. (D) There is a dense infiltration of neoplastic lymphoid cells in the oral submucosa. The mucous membranes were markedly hyperemic, resulting in fiery red oral mucosa which is a clinical feature of this disease. Inset: Neoplastic cells are dissecting through the myocardium and pushing muscle fibers apart in a pattern typical of neoplasia. Neoplastic cells were present in the pericardial effusion.


Epidemiology and occurrence


This is not a common lymphoma in animals but is well described in dermatology and pathology texts. Depending on the type of accessions, the prevalence of the disease can be exaggerated.2 There were 13 cases of mycosis fungoides in a collection of 502 canine lymphomas11 and 53 (9%) in another series of 600.2 In cats, 15 cases of mycosis fungoides were present in a collection of 751 lymphomas.11 Only one feline case involved the gingiva, all the others were in haired skin. The mean age was 11 years and there was no gender predominance. In cattle, mycosis fungoides typically occurs at 2 years of age and is not associated with infection with BLV.11 The lesions in cattle are usually along the back and sides. They depilate and ulcerate and then heal spontaneously before appearing elsewhere. In the early stages of the disease cattle remain in apparent good health, but over a period of about a year the disease progresses to become systemic. In the advanced cases the tumor fills the abdomen and virtually all of the organs may be involved with tumor. When this severe, mycosis fungoides looks like the enzootic type of bovine lymphoma.


Mycosis fungoides is rare in horses but also progresses to involve viscera.8,15 There may be a prolonged period before the cells appear in the blood of horses (Figure 7.46A–E).11


A human neoplasm called lymphomatoid papulosis is a chronic infiltration of the skin with clonal T cells that has several forms or stages that resemble mycosis fungoides.1,10,11 It is not reported in animals.


Clinical presentation


Dogs present with thickened gingiva or skin.11–14 In the dog the presenting lesions may be singular within the mouth or lip, or may appear in multiple sites, usually on the head or ventral area of the body. The mean age was 10 years and there appeared to be a male predominance.8 In the cow, the lesion is very distinctive and should initiate biopsy to confirm the diagnosis and salvage the animal before there is internal progression. The disease in horses may appear first in the mouth as plaque‐like areas that appear reddened but do not blanche with pressure. There may be mild peripheral lymphadenopathy that is not apparent unless looked for. Similarly, atypical cells may not be noted in peripheral blood but can be found if they are searched for.11


Pathology


The lesions of mycosis fungoides are primarily in the skin and/or mucous membranes. The lesions progress from an interface dermatitis into plaques that become confluent and form distinct gross lesions. There is epidermal or epithelial hyperplasia and intraepidermal aggregates of neoplastic T lymphocytes (Pautrier’s microabscess) (Figures 7.42–44 ). A variant is pagetoid reticulosis, in which the neoplastic cells are confined above the basement membrane of the epidermis, epithelium, or adnexa (Figure 7.45).


The intraepidermal tumor cells react positively to CD3 and the lymphocytic inflammation subjacent to the basement membrane will have scattered T cells but is predominantly filled with B lymphocytes and other inflammatory cells. The skin is markedly thickened by the intraepidermal neoplasm and the dermal inflammation. In the dog, mycosis fungoides cells in the skin are CD8 positive, unlike mycosis fungoides lesions in humans where the skin infiltrates are CD4 type. An epitheliotropic lymphoma in a cat had CD3‐positive neoplastic cells that expressed perforin and there were LGL in the blood.7


The neoplasm can be widespread in multiple tissues (Figure 7.46). In cattle, the lesions are raised, circular tumors or thickened areas of skin that form plaques 5–50 cm in diameter. The affected areas are hairless and may be ulcerated and bleeding. In the cow the lesions have been misdiagnosed clinically as ringworm (Trichophyton).


Cytologically the infiltrating cells are small to intermediate with nuclei <1.5 RBC in diameter (Figure 7.43). Nuclei have sharp shallow nuclear indentations much like those of TZL. However the nuclei are larger and more vesicular and the morphology more cerebriform. Chromatin is coarsely aggregated and small central nucleoli are present in about half of the neoplastic cells. The cytoplasm is moderate in volume and lightly stained with cell boundaries indistinct. The mitotic count is low, usually none in most fields at 400× (Figure 7.43). Macrophages are not part of the infiltration unless there is focal ulceration of the overlying epidermis. The infiltrating cells are similar in cats (Figure 7.44).


Survival time and staging


A system of staging mycosis fungoides exists for human cases and appears applicable to animals.11 The staging system is a tumor–node–metastasis–blood system. Bone marrow is seldom involved and usually is not evaluated. In animals, as in humans, the prognosis for mycosis fungoides appears to depend on the stage of the disease at the time of diagnosis. Early cases have long survival and long remissions, whereas late cases are progressive despite therapy. Cases with neoplastic cells in blood have a poor prognosis. Mycosis fungoides may transform into large T‐cell lymphoma with increased mitotic counts, infiltration of organs, and a more aggressive course.2 Retinoid receptors are present in canine mycosis fungoides and retinoids are effective in treating human tumors.13


References



  1. 1. Armitage, J.O. (1997) NHL classification project. A clinical evaluation of the International Lymphoma Study Group classification of non‐Hodgkin’s lymphoma. By the Non‐Hodgkin’s Lymphoma Classification Project. Blood 98:3909–3918.
  2. 2. Ponce, F., Marchal, T., Magnol, J.P., et al. (2010) A morphological study of 608 cases of canine malignant lymphoma in France with a focus on comparative similarities between canine and human lymphoma morphology. Vet Pathol 47:414–433.
  3. 3. Valli, V.E., Kass, P., San Myint, M., and Scott, F. (2013) Canine lymphoma: The effect of age, stage of disease, subtype of tumor, mitotic rate and treatment protocol on overall survival. Vet Pathol 50:738–748.
  4. 4. Seelig, D.M., Avery, P.R., Webb, T., et al. (2014) Canine T‐zone lymphoma: Unique immunophenotypic features, outcome, and population characteristics. J Vet Intern Med 28:878–886.
  5. 5. Avery, P.R., Burton, J., Bromberek, J.L., et al. (2014) Flow cytometric characterization and clinical outcome of CD4+ T‐cell lymphoma in dogs: 67 cases. J Vet Intern Med 28:538–546.
  6. 6. Martini, V., Poggi, A., Riondato, F., et al. (2015) Flow‐cytometric detection of phenotypic aberrancies in canine small clear cell lymphoma. Vet Comp Oncol 13:281–287.
  7. 7. Gilbert, S., Affolter, V.K., Gross, T.L., et al. (2004) Clinical, morphological and immunohistochemical characterization of cutaneous lymphocytosis in 23 cats. Vet Dermatol 15:3–12.
  8. 8. Potter, K. and Anez, D. (1998) Mycosis fungoides in a horse. J Am Vet Med Assoc 212:550–552.
  9. 9. Neta, M., Naigamwalla, D., and Bienzle, D. (2008) Perforin expression in feline epitheliotropic cutaneous lymphoma. J Vet Diagn Invest 20:831–835.
  10. 10. Kim, Y.H., Advani, R., Harvell, J.D., and Hoppe, R.T. (2009) Mycosis fungoides and Sezary syndrome. In Non‐Hodgkin Lymphomas , 2nd edn. (ed. J.O. Armitage). Wolters Kluwer, Philadelphia, PA, pp. 381–392.
  11. 11. Valli, V.E. (2007) Mycosis fungoides and Sezary syndrome. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 331–339.
  12. 12. De Lorimier, L.P. (2006) Updates on the management of canine epitheliotropic cutaneous T‐cell lymphoma. Vet Clin Small Anim 36:213–228.
  13. 13. De Mello Souza, C.H., Valli, V.E., Selting, K.A., et al. (2010) Immunohistochemical detection of retinoid receptors in tumors from 30 dogs diagnosed with cutaneous lymphoma. J Vet Intern Med 24:1112–1117.
  14. 14. Fournel‐Fleury, C., Ponce, F., Felman, P., et al. (2002) Canine T‐cell lymphomas: A morphological, immunological, and clinical study of 46 new cases. Vet Pathol 39:92–109.
  15. 15. Staempfli, H.R. and McAndrew, K.H. (1988) An unusual case of lymphoma in a mare. Equine Vet J 20:141–143.

T‐zone lymphoma


Defining the neoplasm


T‐zone lymphoma (TZL) was used in the Kiel classification of 1974 but it does not appear as a disease entity in the WHO classification of human lymphomas. In human lymphomas it is placed in PTCL‐NOS and comprises about 1% of human lymphomas.1 However, TZL is included here and in Box 7.1 as a diagnosis because it comprises 5–15% of canine lymphomas, about 30–40% of indolent lymphomas, and it is an important lymphoma to identify as it is one of the least aggressive lymphomas in veterinary medicine. A diagnosis of TZL is dependent on histopathology to recognize the characteristic pattern of lymphoma cells that proliferate adjacent to germinal centers (eccentric) and push them to the side. MZL surrounds the germinal center (concentric) and is a B‐cell lymphoma. Extension of the cytoplasm from an end of the cells in cytological preparations (“hand‐mirror” or “cone‐head”) is suggestive for TZL. The cytoplasm is lightly basophilic and the tumor has been referred to as small clear cell.


Described recently is a flow cytometric pattern of CD4 positivity, CD21 positivity, CD45 negativity, and high expression of class II MHC, which is proposed as a diagnostic aid to TZL.2 CD4 positivity, CD21 negativity, CD45 positivity, and low expression of class II MHC has been associated with aggressive T‐cell lymphoma.3 Other profiles were noted but the number of dogs in each group was low. Another study reported similar observations on dogs with small clear cell lymphoma presumed to be TZL.4 Loss of CD45 in neoplastic cells but not in normal or reactive nodes was consistent and the authors indicated this may be a useful means to distinguish lymphoid hyperplasia and TZL.4


Recognizing TZL is important because of its indolent course and if cytology and flow cytometry were definitive it would avoid histopathology and possible general anesthesia.2–6 TZL is a lymphoma of lymph nodes composed of distinctive small T cells that have a low mitotic count and the lowest rate of progression of any canine lymphoma.2,3,7 It has been referred to as small clear cell lymphoma.4,7 The tumors progress slowly and by the time a diagnosis is established the dog may have had the lymphoma for 1–2 years.


Clinical presentation


Dogs generally exhibit few clinical signs and the diagnosis may come from evaluating a single enlarged lymph node. Characteristically, the dog is in good health with normal appetite and activity.8–10 There may be one or several nodes enlarged, often submandibular and these are always mobile and non‐painful. In a study of 75 dogs with indolent lymphoma the mean and median age was 9.4 years and 41% of these dogs had TZL.10 Another study reported a mean age of 10 years and there was no sexual bias, but approximately half of the dogs were golden retrievers.2 TZL is seen rarely in cats and is very unusual in cattle and swine.


Pathology


Blood and bone marrow

A proportion of cases of TZL become leukemic, usually with a low‐grade lymphocytosis in the range of 6000–15,000/μL. Surprisingly, 25% of dogs with indolent lymphoma (different types) had lymphocytosis that ranged from 5600 to 82,000/μL.10 Lymphocytosis did not portend a worse outcome, it was not associated with cytopenias, and dogs were not anemic. Lymphocytosis was seen in approximately half of 20 dogs with TZL, the mean was approximately 5000/μL, and the range was 1000–23,000/μL.2 Eleven percent of the dogs were anemic and none were hypercalcemic.


The neoplastic cells are monomorphic and small to intermediate size and may be misinterpreted as reactive rather than neoplastic. Lymphocytosis may be the reason to search for an enlarged lymph node, perform cytology or biopsy as well as serology for tick‐borne infectious agents. The leukemic state does not occur in all dogs and it does not appear to indicate an increased rate of tumor progression. As a rule, the marrow is not neoplastic. This is a nodal lymphoma, but for leukemia to be present the tumor should be in marrow or spleen.


Lymph nodes

The histologic architecture of TZL is diagnostic. The lymphoma expands the paracortex and medullary cords but does not efface nodal architecture. Germinal centers can usually be found. The capsule is thin and the peripheral sinus and cortex compressed, but the lymphoma is confined to the node and does not extend into perinodal tissues (Figures 7.47 and 7.48). In some cases literally hundreds of small fading germinal centers remain and are a distinctive feature. TZL pushes the germinal centers to the side and the tumor is located eccentric to the fading germinal center. This feature gives TZL its architectural signature. In some cases the non‐neoplastic B‐cell areas will be pushed to the outer rim of the lymph node. In contrast, MZL (B‐cell) symmetrically surrounds germinal centers. These patterns can be appreciated in H&E but are seen more distinctly in IHC‐stained sections. Both tumors may lose their nodular pattern and become diffuse but neither efface nodes or extend into perinodal adipose. An absence of tingible body macrophages, low mitotic rate, and few apoptotic cells are features of TZL.

Micrograph of T-zone lymphoma (TZL) in dog's lymph node. Neoplastic cells form distinct nodular patterns with fading germinal centers (arrows). Other lesions are mild hyperplasia of fibrovascular structures.
Micrograph of T-zone lymphoma (TZL) in dog's lymph node. The nodular areas of T-cell proliferation are stained intensely while the germinal centers (arrows) are not.
Micrograph of T-zone lymphoma (TZL) in dog's lymph node. CD79a has clearly marked the germinal centers. The neoplastic cells are eccentric to the germinal center rather than surrounding it.
Micrograph of T-zone lymphoma (TZL) in dog's lymph node. Neoplastic lymphocytes are small to medium size, with nuclei 1–1.5 RBC in diameter. A few cells have polar extensions of their cytoplasm.

Figure 7.47 T‐zone lymphoma (TZL), lymph node, dog. (A) Neoplastic cells arise from paracortical T‐cell‐rich areas and form distinct nodular patterns with fading germinal centers (arrows). Key to the diagnosis is the way the germinal center is pushed to the side by the expanding T‐cell neoplasm; in contrast, in MZL and MCL the neoplastic cells surround the central germinal center. Other lesions are mild hyperplasia of fibrovascular structures, dilated and empty sinuses, and irregular thickening of the capsule. (B) CD3: The nodular areas of T‐cell proliferation are stained intensely while the germinal centers (arrows) are not. (C) CD79a has clearly marked the germinal centers. Compare the H&E, CD3, and CD79a sections to see the characteristic pattern of TZL in which the neoplastic cells are eccentric to the germinal center rather than surrounding it, as in MZL. (D) Cytology: The neoplastic lymphocytes are small to medium size, with nuclei 1–1.5 RBC in diameter. A few cells have polar extensions of their cytoplasm, referred to as the “hand‐mirror” formation, which is a characteristic of TZL. This preparation is an imprint of a lymph node and the hand‐mirror formation is less evident than in FNA preparations (see Figure 7.48).

Micrograph of T‐zone lymphoma (TZL) in dog's lymph node. Cytoplasm is readily apparent in almost all of the neoplastic lymphoid cells. Inset: One round neoplastic cell and one mature lymphocyte for comparison.

Figure 7.48 T‐zone lymphoma (TZL), lymph node, FNA, dog. Comparing images illustrates the difference between imprint (Figure 7.47D) and FNA preparations (Figure 7.48) in the induction of “hand‐mirror” cytoplasmic extensions. In this FNA preparation the “hand‐mirror” attenuation of the cytoplasm is readily apparent in almost all of the neoplastic lymphoid cells. Inset has one round neoplastic cell and one mature lymphocyte for comparison. (Image courtesy of Mary Anna Thrall, Ross University.) TZL is one of the most indolent types of lymphoma and it would be advantageous to recognize it in FNA. The hand‐mirror feature is highly suggestive of TZL but not diagnostic and not present in all TZL. It should be present in the majority of the cells not just a few. TZL cells tend to have fairly abundant cytoplasm that is light blue or semi‐clear, hence the name small clear cell lymphoma.


The medullary fibrovascular structures are hyperplastic, compatible with a node that has had a long history of hyperplasia. There usually are dilated medullary sinuses (sinus ectasia), which are completely empty except for a few small lymphocytes with almost none traversing the vessel wall to enter the node. This is likely the reason there are areas of empty sinuses because the residual benign cells are dying faster than new cells are being added from the low proliferation rate of the lymphoma. The indolent lymphomas are considered neoplasms of accumulation as opposed to proliferation.


Cytologically there are two subtypes – small cells and intermediate cells – but they are not correlated with prognosis.8,11 Neoplastic cells with smaller nuclei are more common. The nuclei are only slightly larger than a red blood cell. About half of the nuclei have a sharp shallow nuclear indentation. Nuclear chromatin is fine and dispersed such that nucleoli are not seen or a single small central nucleolus may be noted (Figures 7.47 and 7.48). These cells have a relatively abundant, almost water‐clear cytoplasm with indistinct cell boundaries. This amount of cytoplasm means that the nuclei are not in contact and provides a distinctive grid‐like appearance at intermediate levels of magnification. The larger cell type has nuclei that are about 1.5 RBC in diameter and the same shallow nuclear indentations. Nuclear chromatin is not as dispersed and more nucleoli may be seen. This cell type has 0–2 mitotic figures/400× field. It is not known if the larger cell type is associated with a higher frequency of leukemia or a more rapid rate of tumor progression. Tingible body macrophages and apoptotic cells are absent.


An FNA is diagnostic of lymphoma but because the neoplastic cells are more mature interpretation is not as easy as in immature lymphomas. The tumor exfoliates large numbers of uniform small to medium‐sized cells. Nuclei are round to slightly oval, of small to intermediate size. The very shallow nuclear indentations must be looked for to be detected. The chromatin is densely stained and nucleoli are small and not apparent in most cells. There should be an absence of plasma cells and inflammatory cells. A feature of TZL lymphoma in some cytological preparations is lightly basophilic cytoplasm, drawn out in a tail to one side. This is referred to as a “hand‐mirror” or “cone‐head” configuration. It is distinctive when seen and is used to favor a diagnosis of TZL in cytologic preparations (Figures 7.47 and 7.48). These cytologic changes should be a predominant feature, not just found in a limited number of cells. Although helpful, it is not a constant feature and is likely an artifact of preparation. It is seen more frequently in FNA than touch imprints. This feature is not seen in histopathology.


Spleen and other organs

There are too few cases with accurate autopsy data to know if spleen, abdominal lymph nodes, or other organs are involved and to what extent.


Cytochemistry and immunohistochemistry


TZL is expected to be strongly CD3 positive via IHC. Some cases will have solid sheets of positively stained CD3 cells. In others the nodal pattern and eccentric proliferation of T cells next to germinal centers will be accentuated by IHC. Staining with B‐cell markers (CD79a, Pax5) demonstrates the small germinal centers that are often peripheralized in the outer cortex and the TZL is negative.


Results of flow cytometry using broad panels of antibodies has been reported and used to help subdivide T‐cell tumors.2–4 In flow cytometry CD21 positivity, CD25 positivity, CD45 negativity, and high expression of class II MHC has been proposed as a means to identify TZL and thereby predict long survival times.2,4 CD21 negativity, CD25 positivity, CD45 positivity, and low expression of class II MHC has been linked with other T‐cell lymphomas that have short survival times, approximately 160 days.3 Only a limited number of cases had histopathology, 10 of which were diagnosed as PTCL‐NOS and 5 LBL. These authors also subclassified other profiles that correlated with different survival times or progression‐free intervals but the numbers of dogs in each group was small.3 Another group reported that small clear cell lymphomas presumed to be TZL had a similar flow cytometric profile: CD21 positivity and CD45 negativity.4 They also indicated CD3 positivity in 72%, CD5 positivity in 90%, and CD79a positivity in 15% of cases tested.4 There was variable staining with CD4 and CD8.4 Some, perhaps 50% of TZL, will have positively stained cells with CD21.


The percentage of positive cells in a sample will vary and the results must be interpreted with caution. CD21 is considered a B’cell marker. It is a protein that in humans is encoded by the complement receptor type 2 gene and is expressed on mature B cells and, in some T‐cell lymphomas, naive T cells. It is upregulated on memory T cells.2 It has been demonstrated to label TZL cells in flow cytometry but we are unaware of IHC studies with CD21 on frozen or FFPE sections of TZL. A molecular study of canine lymphoma identified a group of genes, including CD21, that were expressed at higher levels in TZL than in high‐grade T‐cell lymphomas.5 Negative reaction to CD45 in 95% of cases is consistent between studies.2,4 If a broad panel of antibodies for B and T cells are used, the tumor is clearly B cell negative and T cell positive. Clonal rearrangement of T‐cell receptor genes is not consistently present in TZL.11


Differential diagnosis


The histologic signature of TZL lymphoma is characteristic and provides a firm basis for the diagnosis on H&E. The critical feature for TZL is the numerous small fading germinal centers crowded to the outside of nodes with neoplastic proliferations of T cells adjacent to the fading follicles (eccentric pattern). MZL surrounds a fading follicle in a concentric pattern. A nodal pattern is a feature of TZL, MZL, MCL, and follicular lymphoma. The latter three are all B‐cell lymphomas and IHC is the easiest means to differentiate them from TZL. Another differential is SLL of T‐ or B‐cell type. SLL do not have fading germinal centers or sinus ectasia of TZL. Nuclear indentations favor TZL, but they are not a feature of SLL. Peripheral T‐cell lymphoma of intermediate cell type is a differential but the cells are larger and the mitotic count greater. These lymphomas are usually high grade and aggressive. The flow cytometry patterns described in IHC are reported to be distinctive.2,3 T‐zone hyperplasia has a mixture of large and small lymphocytes, macrophages, and dendritic cells. Heterogeneity of cells always favors hyperplasia over lymphoma.


Typical survival time


TZL is the most indolent of the lymphomas. Dogs may survive for years and treatment does not appear to alter survival times.10,11 An MST of 33.5 months is reported for TZL and an overall survival of 4.4 years for all dogs studied with indolent lymphomas, confirming their indolent behavior.10 In a study with a few dogs, none were reported to die from TZL, the MST was not reached at almost 2 years, and two dogs with long survival times were untreated.11 Other reports confirm long survival time.2 We know of a case that survived for 7 years with no treatment. The presence of a leukemic state does not seem to alter the rate of progression. The usual recommendation for this lymphoma is to withhold treatment until quality of life becomes an issue and then to treat conservatively with chlorambucil and prednisone. Watchful waiting has been a phrase used to summarize treatment options and the behavior of indolent lymphomas.


References



  1. 1. Ralfkiaer, E., Muller‐Hermelink, H., and Jaffe, E.S. (2001) T‐zone lymphoma. In WHO Classification of Tumours. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues (eds. E.S. Jaffe, N.L. Harris, H. Stein, and J.W. Vardiman). IARC Press, Lyon, France, pp. 227–229.
  2. 2. Seelig, D.M., Avery, P.R., Webb, T., et al. (2014) Canine T‐zone lymphoma: unique immunophenotypic features, outcome, and population characteristics. J Vet Intern Med 28:878–886.
  3. 3. Avery, P.R., Burton, J., Bromberek, J.L., et al. (2014) Flow cytometric characterization and clinical outcome of CD4+ T‐cell lymphoma in dogs: 67 cases. J Vet Intern Med 28:538–546.
  4. 4. Martini, V., Poggi, A., Riondato, F., et al. (2013) Flow‐cytometric detection of phenotypic aberrancies in canine small clear cell lymphoma. Vet Comp Oncol 13:281–287.
  5. 5. Frantz, A.M., Sarver, A.L., Ito, D., et al. (2011) Molecular profiling reveals prognostically significant subtypes of canine lymphoma. Vet Pathol 50:693–703.
  6. 6. Aulbach, A.D., Swenson, C.L., and Kiupel, M. (2010) Optimized processing of fine‐needle lymph node biopsies for automated immunostaining. J Vet Diag Invest 22: 383–388.
  7. 7. Ponce, F., Marchal, T., Magnol, J.P., et al. (2010) A morphological study of 608 cases of canine malignant lymphoma in France with a focus on comparative similarities between canine and human lymphoma morphology. Vet Pathol 47:414–433.
  8. 8. Valli, V.E. (2007) T‐zone lymphoma. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 294–302.
  9. 9. Valli, V.E., Kass, P., San Myint, M., and Scott, F. (2013) Canine lymphoma: The effect of age, stage of disease, subtype of tumor, mitotic rate and treatment protocol on overall survival. Vet Pathol 50:738–748.
  10. 10. Flood‐Knapik, K.E., Durham, A.C., and Gregor, T.P. (2013) Clinical, histopathological and immunohistochemical characterization of canine indolent lymphoma. Vet Comp Oncol 11:272–286.
  11. 11. Valli, V.E., Vernau, W., de Lorimier, L.P., et al. (2006) Canine indolent nodular lymphoma. Vet Pathol 43:241–256.

Intestinal T‐cell lymphoma


Defining the neoplasm


Part of this section includes lymphoma of the gastrointestinal system and part focuses on enteric lymphoma in cats.1–4 Consult Chapter 13 for a wider perspective of gastrointestinal lymphoma in different species and read publications on enteric lymphoma in cats for the excellent details they contain.2,4


In the post‐FeLV testing and vaccination era, lymphomas that arise in lymphoid tissues outside of lymph nodes seem to be the most common source of feline lymphomas. Clearly the anatomic site annotated is because the principal signs are due to clinical problems in that system. In many cases a thorough clinical or post‐mortem search is not performed and the true distribution or origin of the lymphoma is not known. However, for lymphomas in the abdominal cavity of cats it is clear that the lymphoma can arise from lymphoid centers in the gastrointestinal tract. In fact, enteric lymphoma is probably the most common lymphoma in cats. Enteric lymphoma arises in extranodal mucosal lymphoid centers in the intestinal tract.1–6 Tumors may arise from a clonal expansion of lymphocytes in Peyer’s patches, lamina propria, or intraepithelial layers. In the mucosa of the small intestine there are lymphocytes in the lamina propria and intraepithelial layer, collectively referred to as mucosal‐associated lymphoid tissue (MALT). The intraepithelial lymphocytes are T‐cell type and most are LGL. Lymphocytes in the lamina propria are T‐ or B‐cell types, and most (90%) are T lymphocytes. Tumors can arise from lymphocytes in any of these locations and the tumors may be of T‐cell, B‐cell, or NK type. The tumor cells can be further subtyped by size, receptors7 and granules. The T‐cell tumors are also classified as enteropathy‐associated T’cell lymphoma (EATCL or EATL) type I or II.


The types of lymphoma in different anatomic regions of the gastrointestinal tract have predictable patterns.2,4 The majority of small intestinal lymphomas in cats are T‐cell and the jejunum is the most common site, whereas enteric lymphomas in the stomach or ileal–cecal region may be of T‐ or B‐cell phenotype and B‐cell will predominate.2,4 Additional classifications use transmural versus mucosal, size of neoplastic lymphocytes (small cell, large cell), and presence or absence of cytoplasmic granules.


LGL lymphomas arise from intraepithelial LGLs. They are cytotoxic T cells (CD3, CD8, and granzyme B positive), are often transmural, and if concurrent lymphocytosis is present the cats only live for a short time (19 days MST in one study).6 LGL lymphomas have previously been erroneously named globular leukocyte tumors (see Chapter 13). In contrast, small cell T‐cell lymphomas are the most common intestinal lymphoma in cats. The majority are confined to the mucosa, with 90% in the jejunum. They have survival times approaching 2 years. Obviously the differentiation of these two type of T‐cell lymphomas has great clinical significance. B‐cell lymphomas are large cell type. They have an aggressive course and survival is only a few months.2 If any lymphoma is transmural as opposed to mucosal, survival is shortened.


As a generalization, small cell lymphomas tend to be indolent and large cell lymphomas aggressive. Differences between reports on phenotype, diagnoses, nuclear morphology, and other characteristics exist. Multiple explanations for these discrepancies are possible and some of the more likely include different antibodies, panels of antibodies included, definitions, and expertise. As new parameters, antibodies, primers, and techniques are developed, our database will be modified. It is hoped that some generalizations will remain as new information is created. Results from diagnostic tests may vary as a disease progresses and therefore our ability to establish a diagnosis, the survival data generated, and, likely, the responses to treatment may be different at different time points.


A diagnosis of lymphoma in domestic animals is usually straightforward at autopsy. The pathologist sees all the tissues and selects samples and size of samples to establish a diagnosis. Collecting cytological samples at autopsy is encouraged, especially to diagnose LGL lymphoma, and in many cases the diagnosis of lymphoma can be confirmed immediately. If subtyping is desired it is accomplished with cytology, histopathology, and IHC. Biopsy samples are variable and confidence of diagnosis ranges with the quality and quantity of samples submitted as well as how advanced or early the disease is. The most problematic cases are endoscopic mucosal samples, which may have only a few pieces of mucosa. Sampling artifacts may be present, the lesion may not be not fully developed, and IBD will be present or suspected. For the problematic cases clinical information is helpful: age, treatments, lymphocytosis, serum calcium, lymph nodes, and other masses or potential organ involvement. In difficult cases multiple tests are needed: cytology, IHC with broad panels, flow cytometry, FeLV, BLV, clonality, or additional samples of gastrointestinal tract, lymph nodes, or liver. These tests are not needed for cases in which the diagnosis is certain. Oncologists are likely to request phenotyping on biopsies.


Some enteric lymphomas, primarily in dogs, cats, and humans, are believed to originate in lymphocytes that have undergone a clonal transformation from prolonged antigenic stimulation induced by a variety of IBDs. Cats and dogs have IBD and intestinal lymphoma and the association of the two diseases is often suggested but never proven. Human enteric lymphoma data come from hundreds or thousands of cases using somewhat unlimited diagnostic tools, including molecular profiles. Associations have been noted with immune stimuli, immune‐mediated diseases, celiac disease, and infectious agents such as Helicobacter pylori. 8 Presently we do not have comparable resources, number of cases, parameters assessed, or especially accurate clinical follow‐up to dissect lymphomas in animals. Reports in the veterinary literature do not always include autopsy data or follow‐up results and therefore we do not know if the gastrointestinal tumors are also in lymph nodes, spleen, liver, bone marrow, or if these factors influence survival. The biology of normal enteric lymphocytes, including trafficking signals and their neoplastic counterparts, is well described in a paper about enteric lymphoma in cats. Readers are encouraged to read this reference as well as Chapter 13.2


Lymphoma occurs in the oral cavity of all species. In cats and dogs it often is in the tonsil and part of multicentric lymphoma. Epitheliotropic lymphoma occurs in the lips of dogs and is usually located in other mucocutaneous sites or skin. Gastric lymphoma is common in dogs, cats, and cattle and occurs in the horse. The majority of gastric lymphomas are of B‐cell phenotype4 and the lymphoma is not confined to the stomach. It will be located in lymph nodes and other organs, and is part of generalized lymphoma. It is imperative for data collection to try to determine if lesions are truly confined to the stomach or any single organ. It is very rare that lymphoma is only in one anatomic location, and how thoroughly tumors are searched for will influence final conclusions. Many of these tumors are multicentric and are large cell B‐cell phenotype.


In all locations lymphomas are white to tan unless ulcerated or containing excess red blood cells. Lymphocytes are white blood cells and the more pure the neoplasm the more consistent this white color is at gross examination. Lymphomas that are red or pink contain red blood cells from hyperemia, congestion, hemorrhage, or all three.


The abomasum of cattle is a classical site for lymphoma. The lymphoma is usually generalized, and other characteristic sites are lymph nodes, uterus, heart, and periocular area. Infection with BLV is causative and in regions of the world where BLV is common so is lymphoma.9,10 Readers are referred to Chapter 13 for detailed information on gastrointestinal lymphoma in domestic animals, as well as references that describe infectious diseases (H. pylori) associated with enteric lymphoma and the translocation abnormalities used to diagnose and provide prognoses.8


Lymphoma of the intestinal tract of animals is usually part of multicentric lymphoma. In some cases the predominant signs are referable to the gastrointestinal system and that system is attended to the diagnosis, but rarely is the lymphoma confined to the gastrointestinal tract if other anatomic sites are searched. Enteric lymphomas may be of B‐cell, T‐cell, or NK origin. B‐cell lymphomas are usually of large cell type and the disease is also in mesenteric lymph nodes, other abdominal organs and it may be in peripheral lymph nodes. The cells are immature and the diagnosis is uncomplicated. B‐cell lymphomas are the most common type of enteric lymphoma in sheep and pigs and Payer’s patches or other focal lymphoid centers are a likely site of origin (MALT). B‐cell lymphomas (DLBCL) are common in dogs and may involve the gastrointestinal system, liver, and spleen. A 2015 report indicated that large cell lymphoma was more common in the intestinal tract of dogs than intraepithelial lymphoma.11 In one study in cats all gastric lymphomas were of B‐cell phenotype, and that large cell and immunoblastic subtypes predominated.4 Enteric B‐cell lymphoma occurs in cats and is usually in the stomach or ileal–cecal region. Feline lymphomas in the jejunum are primarily of T‐cell origin. They may be small or large cell and large cell lymphoma may be granulated (LGL) or not. The majority of enteric lymphomas in cats are small cell type, mucosal, and have an indolent course, whereas T‐cell large granular cell lymphomas occur in the gastrointestinal system of all species and are usually aggressive.


In any species phenotype can only be determined by IHC, ICC, or flow cytometry. Clonality should not be relied on for phenotype. For many diagnostic cases the diagnosis of lymphoma will suffice and this can readily be determined by cytology or histopathology.


Clinical presentation


Small T‐cell lymphoma of the small intestine is seen most often in cats1–4 and less frequently in dogs5,6,12 but occurs in most animals, including horses.13,14 Animals are presented for weight loss, accompanied by degrees of diarrhea and, in cats and dogs, vomiting. Cats and dogs are usually 10 years of age or older and in cats there is a slight male predominance. There is no association between age of cat and phenotype of enteric lymphoma.4 There often is a history of gastrointestinal disease treated with dietary modifications or various medications, including steroids. IBD is a common clinical differential diagnosis11 (Figure 7.49). In humans, there are forms of intestinal lymphoma linked with celiac disease and molecular signatures for these lymphomas are known.8 Horses and other species often have weight loss and some will be emaciated. Protein‐losing enteropathy occurs and animals are hypoproteinemic and hypoalbuminemic.13,14 Decreased serum calcium is most likely due to decreased albumin, but malabsorption is also contributory.

Micrograph of inflammatory bowel disease (IBD) in a cat's jejunum. There are numerous lymphocytes in the lamina propria of the villi and in the epithelium, creating an epitheliotropic pattern.
Micrograph of inflammatory bowel disease (IBD) in a cat's jejunum. Higher magnification illustrates the lymphocytes are also in the epithelium.
Micrograph of inflammatory bowel disease (IBD) in a cat's jejunum. The staining pattern of T lymphocytes emphasizes the extent of lymphocyte colonization of the villous epithelium and lamina propria.
Micrograph of inflammatory bowel disease (IBD) in a cat's jejunum. Lymphocytes extend from villus tips, through muscle layers and onto the serosa.

Figure 7.49 Inflammatory bowel disease (IBD), jejunum, cat. (A) There are numerous lymphocytes in the lamina propria of the villi and in the epithelium, creating an epitheliotropic pattern. Although present, the lymphocytes are less concentrated in the deeper mucosa and none penetrate the tunica muscularis. Confinement of the lymphocytes to the mucosa favors IBD. Pleocellular infiltration in each villus also favors IBD. When the lymphoid infiltration penetrates submucosa, muscle layers, and serosa, the lesion is more likely lymphoma (see D). However, this histologic distinction requires a full thickness biopsy. (B) Higher magnification illustrates the lymphocytes are also in the epithelium. When present, the variation in degree of epitheliotropism from one villus to another is a further indicator of lymphoma. (C) CD3: The staining pattern of T lymphocytes emphasizes the extent of lymphocyte colonization of the villous epithelium and lamina propria, even in IBD. IHC does not identify cells as neoplastic. (D) Lymphocytes extend from villus tips, through muscle layers and onto the serosa. This pattern is characteristic of lymphoma and, combined with a uniform population of immature lymphocytes, is the best histologic feature to differentiate lymphoma from IBD. (D Image courtesy of J. Munday, Massey University.)


Pathology


How widespread the lymphoma is depends on the type of lymphoma, species of animal, stage of the disease, and how thorough the clinical or post‐mortem examination is. In many cases the mesenteric lymph nodes are enlarged and neoplastic. A diagnosis of lymphoma is readily apparent at gross inspection or on histopathology as there is partial or complete loss of germinal centers and in some cases the entire node is effaced. Depending on duration and severity, the lymphoma may extend through the capsule and into mesenteric tissue. Many cases will have an effusion and a diagnosis of lymphoma is relatively easy if the fluid is analyzed, as this is one of the few neoplastic effusions in which tumor cells predominate in the nucleated cell population. Some cases will have >90% lymphoma cells in the abdominal fluid. These are usually large cells, with fairly abundant basophilic cytoplasm and immature large nuclei with prominent nucleoli. These events occur in cattle and horses as well as pets. If the lymphoma starts in the small intestine then involvement of nodes, liver, and spleen may not be severe. If the lymphoma is multicentric and an immature cell type then liver and spleen will be enlarged and neoplastic. Some of the peripheral nodes as well as mesenteric nodes will be neoplastic.


Enteric lymphoma in cats


This is a common disease in cats.1–4 It is well described and eloquently illustrated in a series of 120 cats with over 125 tumors (some tumors were multifocal).2 The tissue samples from this latter study came from surgical biopsies (n = 47), endoscopic biopsies (n = 35) or autopsy (n = 38). Lesions ranged morphologically from mild to marked and they were categorized by anatomic distribution in the gastrointestinal tract by histologic patterns of transmural versus mucosal (lamina propria and intraepithelial at the surface of intestines and in crypts), cell size (small, large), and by phenotype. Approximately 80% were T‐cell and 15% B‐cell.2 Another study reported that approximately 50% of feline enteric lymphomas were B‐cell.4 Both studies agreed that T‐cell lymphomas were more common than B‐cell in the small intestine and that some tumors were not immunoreactive with B‐ or T‐cell antibodies.2,4 The majority of T‐cell lymphomas were mucosal (80%) and 95% of B’cell were transmural.2 The jejunum was the most common site for T‐cell lymphomas. B‐cell lymphomas (n = 19) were distributed fairly evenly from stomach to colon but none were found in the duodenum.2 All the B‐cell tumors were large cell type2 which is similar to another report.4 The T‐cell tumors were subdivided: if they were mucosal then 80 of 84 were small to intermediate size and if transmural 11 of 19 were large cell type. Nine of these 11 large cell type were LGL and all 9 reacted positively for granzyme B.2 Only two of 84 small cell type were LGLs.


All B‐cell lymphomas in cats were large cell‐type and they were further divided by nuclear morphology into centroblastic (17/19) or immunoblastic (2/19).2 Immunoblastic types have a single large central nucleolus, whereas centroblastic types have multiple nucleoli, often adjacent to the nuclear membrane. In another study all gastric lymphomas in cats were of B‐cell type and large cell immunoblastic was most common.4 B‐cell lymphomas in the small or large intestine were situated close to lymphoid follicles, but this pattern was not apparent in the stomach.2 Eighteen of 19 were transmural and the tissues were markedly infiltrated, which made the diagnosis straightforward but the association with lymphoid follicles difficult to determine. Intraepithelial location was seen in one case.


As the tumors were divided into groups by phenotype, size of cells, and whether mucosal or transmural, the number of cats in each group was reduced, and this was accentuated by the small number of cats that had follow‐up data. Cats with follow‐up data that had mucosal T‐cell lymphoma (n = 54) had an MST of 29 months and cats with transmural T‐cell lymphoma (n = 13) had an MST of 6 weeks.2 Dividing groups by size of cells revealed similar survival predictions. Cats with small cell T‐cell lymphoma (n = 54) had an MST of 28 months versus 6 weeks for large T‐cell lymphoma (n = 13).2 Six of 19 cats with follow‐up data that had B‐cell lymphoma survived about 3.5 months. Tumors and survival data were not correlated with treatment regimens. Although the numbers of cats in the different groups were small and treatments could not be factored in, trends were apparent. Cats with small T‐cell lymphoma in the mucosa have an indolent course (>2‐year survival), whereas cats with large T‐cell lymphoma, B‐cell lymphoma, transmural lymphoma, or LGL cell lymphoma have much shorter survival times. The worst prognosis may be feline LGL lymphomas that have concurrent lymphocytosis (leukemia), and transmural pattern since they had an MST of 19 days.6


T‐cell lymphomas (mucosal and transmural) were almost exclusively localized to the small intestine, especially the jejunum, and were uncommon in the duodenum or stomach.2 Practical considerations of this anatomic distribution is that endoscopic biopsies taken from the duodenum would miss the most common location of T‐cell lymphomas in cats (jejunum) and B‐cell tumors are rare in the duodenum. There were only 7 instances (approximately 7%) of T‐cell lymphoma involving the stomach or large intestine.


Lesions in the lamina propria of cats with enteric lymphoma will range from mild infiltration to severe effacement.2–4 Histologically, mild lesions will have shortened or fused villi with variable infiltration in the lamina propria. Villi should be 2–4 times the height of crypts but it is preferential to examine villi in which the superficial epithelial cells align with subjacent crypts. Pathologists should search sections until this alignment can be assessed as these regions are perpendicular sections and are not cut in a tangential plane that will produce blunting or fusion. Avoid villi over Peyer’s patches as they are normally short and interspaced with domes of the lymphoid cells from the follicles beneath. The villi are increased in diameter to 2–3 times normal size due to neoplastic infiltration in the lamina propria. The neoplastic cells may fill the lamina propria or form aggregates (patches) at various levels of the lamina propria (Figure 7.50). Adjacent villi may be unaffected or markedly affected. This characteristic of “skipping” villi is a feature of lymphoma and if seen is an aid to differentiate lymphoma and IBD. The mucosal infiltration in IBD does not skip adjacent villi. Neoplastic cells are always in a greater concentration in the lamina propria than in an intraepithelial location. An epitheliotropic pattern is more common in surface epithelium than crypt and it is present in approximately 50–60% of mucosal or transmural T‐cell lymphomas.2,4 Intraepithelial lymphocytes can be singular or clustered. If they occur in aggregates of 4–6 cells then clonal proliferation is likely. The colonization of the epithelium by neoplastic lymphocytes may be so great that there are more nuclei of lymphocytes than those of the epithelial cells. The more intraepithelial lymphocytes present in a lesion favors lymphoma. Immunostaining with CD3 will enhance the intraepithelial pattern and is useful when there does not appear to be many lymphocytes in the epithelium.2 There is often a clear space around or close to the nuclei of the intraepithelial lymphocytes. Search these cells for eosinophilic granules which indicates they are LGL.

Micrograph of a lymphoma in a dog's duodenum. Villi are about twice normal width and there is lymphoid infiltration. There is distortion of the sample and sections taken over a Peyer’s patch.
Micrograph of a lymphoma in a dog's duodenum. Magnification: Surface epithelium is thin and numerous lymphocytes in intraepithelial areas and in lamina propria. There are also neutrophils and eosinophils.
Micrograph of a lymphoma in a dog's duodenum. Lymphocytes in the villus tips, in intraepithelial areas, and in solid area of lymphoid proliferations. Inset: Displays clusters of intraepithelial lymphocytes.
Micrograph of a lymphoma in a dog's duodenum. Few cells are positive and the cellular infiltration extends transmurally.

Figure 7.50 Lymphoma, duodenum, dog. (A) Villi are about twice normal width and there is lymphoid infiltration. However, there is distortion of the sample and sections taken over a Peyer’s patch usually have shorter villi. At this magnification inflammatory bowel disease (IBD) is a likely differential. This sample only contained mucosa, therefore infiltration into or past muscularis mucosa could not be evaluated to help differentiate IBD and lymphoma. (B) Higher magnification: The surface epithelium is thin and there are numerous lymphocytes in intraepithelial areas and in lamina propria; however, there are also neutrophils and eosinophils, which favor IBD. In cases like this, PCR for clonality and IHC may help differentiate IBD and lymphoma. Of equal value is to examine as many other sections as possible to determine the uniformity or heterogeneity of the infiltration. (C) CD3: Lymphocytes in the villus tips, in intraepithelial areas, and in the solid area of lymphoid proliferations in the deeper mucosa are all T‐cell type, supporting a clonal proliferation. Inset: Higher magnification of villus tips. Clusters of intraepithelial lymphocytes as in this image also support lymphoma. (D) CD79a: Relatively few cells are positive, in contrast to the numerous T cells seen in (C). The absence of a mixture of B and T cells in the mucosa favors the diagnosis of lymphoma. When the cellular infiltration extends transmurally, as in Figure 7.49D then this strongly favors lymphoma and PCR or IHC should not be needed. When intracytoplasmic granules are seen, the tumor is T‐cell type, but B‐ versus T‐cell requires phenotyping in lymphomas without cytoplasmic granules.


Of 84 mucosal T‐cell lymphomas in one study,2 80 were small to intermediate with nuclei about 1.5 RBC in diameter with a nuclear chromatin‐dense, mature appearance. Small intraepithelial lymphocytes were the most common T‐cell lymphomas in an earlier study.4 Mitoses may be present but are not frequently encountered, even where there is heavy lymphoid proliferation in the deeper mucosa and lymphoma is apparent.1,15 The crypts may be separated from the muscularis mucosa by a laminar band of small lymphocytes. A few eosinophils may be present in lymphomas but eosinophils will not be predominant. These changes may be accompanied by a secondary and nonspecific moderate or mild dilation of the villus lymphatics. When the diagnosis is small cell type, mucosal intraepitheliotropic lymphoma there is little to be gained by IHC for B‐ versus T‐cell since >95% will be T‐cell and IHC only identifies phenotype, it cannot identify neoplastic cells. However, visualization of the intraepithelial distribution is enhanced by staining with CD3 and the homogeneity of the proliferation is apparent. Therefore IHC is helpful to recognize these characteristic features. Mucosal T‐cell lymphoma of small cell type is analogous to EATCL type II.


Transmural types have the most severe lesions, with effacement of the lamina propria, moderate to marked infiltration of the submucosa, and tumor cells dissecting to the serosa or beyond. The diagnosis of lymphoma is easy in these cases and the majority will be large cell type. However, they can be either T‐cell or B‐cell phenotype and this is determined by IHC. The B‐cell types are large cells and are comparable to DLBCL. They have a median survival of approximately 3.5 months. Eleven of 19 transmural T‐cell lymphomas were large cell type and 9 of these were LGL.2 Large T‐cell types have nuclei greater than 2 RBC in diameter and nuclei appear more immature and pleomorphic; they may also have variable contours. Cytoplasmic granularity should be assessed, which is easiest to do in cytologic preparations using methanolic‐based stains rather than aqueous‐based ones16 (see Figures 7.39 and 7.54). If cytology slides are not available then consider oil‐immersion objectives and look for eosinophilic granules close to the nuclei or in sections stained with PTAH look for dark purple granules. LGL will be granzyme B positive and, if available, this is an easy method to identify LGL in histopathology preparations. Granzyme B labeling is cytoplasmic, usually juxtanuclear. See the section in this chapter on LGL tumors. Transmural T‐cell lymphoma of large cell type is analogous to EATCL type I, and many of these will be LGL.


Clonality for T‐cell receptor gene (TCRG) rearrangement will support the diagnosis in about 90% of mucosal and transmural T’cell lymphomas.2 Approximately 80% will be monoclonal and 10% oligoclonal. The other approximate 10% are polyclonal and clonality was undetected, which can happen if there are gene segments used in the rearrangement that are not detected by the primers used in the assay. Clonality was detected in 50% of the B‐cell lymphomas and pseudoclonality in about 40%; 10% were polyclonal. The authors discussed possible explanations.2 No test is 100% sensitive and specific; however, if morphology and IHC are not definitive than clonality testing may help these ambiguous cases.


Small intestine and differential diagnoses


Dogs, cats, and horses have IBD that mimic lymphoma clinically, grossly, and microscopically (Figures 7.49 and 7.50).11–15 In many cases the differentiation is clear from histology or gross examination, but in some the differentiation of these two diseases is difficult.11 Problematic cases are endoscopic biopsies that have mucosa only (depth of invasion cannot be assessed) or if there are limited number of samples submitted or the quality of the sections is not good. The distinction of these two diseases is often difficult and, although this is not proven, it is thought that lymphoma may arise in areas of IBD and that the two lesions may be present in the same patient. Difficult cases require additional parameters and integration of results from lymph node biopsy, possibly including samples of liver, cytology, PARR for clonality, IHC, flow cytometry, FeLV, and BLV. Clonality can be helpful but there are false‐negative and false‐positive results just like any test. IHC cannot identify neoplastic cells but it can help identify homogeneous versus heterogeneous populations and an intraepithelial pattern. We tend to emphasize or remember those cases that are exceptions or that were difficult to diagnose.


The primary histologic lesion in both diseases is in the mucosa of the small intestine, and lymphocytes are the most numerous cell type. Both diseases will have enlarged mesenteric lymph nodes. Endoscopic mucosal biopsies are taken from the duodenum, which is not the most common location of enteric lymphoma in cats and is a very uncommon location for B‐cell lymphoma.


The following are features to help differentiate IBD from lymphoma. In almost all cases of IBD the inflammation is confined to the mucosa, whereas lesions that enter the submucosa or extend further favor lymphoma. In general, all or the great majority of each villus is affected in IBD, but in lymphoma some villi may be skipped. If one villus has numerous cells in the lamina propria and the next villus is unaffected, this favors lymphoma. The inflammatory cell population is mixed and it is not monomorphic in IBD. If the inflammation is granulomatous then the distinction is easy and the task is to determine the cause (e.g., granulomatous enteritis of horses, M. avium, ulcerative colitis, etc.). The more mixed the cellular infiltration and the greater the plasma cell population the more likely it is IBD. IBD has areas in which the inflammation is intensified, if sections are searched carefully. Intensification may appear as foci of plasma cells, pockets of eosinophils or neutrophils, and the normal mucosal structures are destroyed rather than pushed aside.


Neoplasia should be uniform, monomorphic, it should infiltrate and push normal structures apart but it does not cause focal intensification with inflammatory cells or discrete areas of necrosis surrounded by inflammatory cells. Neoplastic lymphoid cells will dissect between muscle fibers rather than produce areas of necrosis or myositis (see Figure 11.24C,D). Examine the superficial epithelium for the type of cells and clustering of cells. The presence of more neutrophils here favors IBD, whereas more lymphocytes favors lymphoma, especially if they form intraepithelial aggregates of 4–8 cells. LGLs with large cytoplasmic granules that are scattered in the epithelium favors IBD. Numerous LGLs in the lamina propria and intraepithelial favors lymphoma. The greater the expansion of the lamina propria by lymphocytes, the more monomorphic this population and the more likely it is lymphoma. Mucosal lymphoma is accompanied by a loss or absence of plasma cells. If lesions are present in liver, spleen, and kidney then lymphoma is the working diagnosis. If white tumors are present in organs then lymphoma, another tumor, or one of the granulomatous diseases is present, not IBD. Peripheral lymphadenopathy or hypercalcemia are features of lymphoma.


TCRLBCL is an uncommon form of enteric lymphoma but its heterogeneous cell population can make distinction of inflammation versus neoplasia difficult.4 Polyclonal patterns in PARR and heterogeneous lymphoid populations in IHC are characteristic of IBD and monoclonal or oligoclonal patterns with homogeneous lymphoid cells are features of lymphoma.


References



  1. 1. Cesari, A., Bettini, G., and Vezzali, E. (2009) Feline intestinal T‐cell lymphoma: assessment of morphologic and kinetic features in 30 cases. J Vet Diagn Invest 21:277–279.
  2. 2. Moore, P.F., Rodriguez‐Bertos, A., and Kass, P.H. (2012) Feline gastrointestinal lymphoma: mucosal architecture, immunopheotype, and molecular clonality. Vet Pathol 49:658–668.
  3. 3. Mahony, O.M., Moore, A.S., Cotter, S.M., et al. (1995) Alimentary lymphoma in cats: 28 cases (1988–1993). J Am Vet Med Assoc 207:1593–1598.
  4. 4. Pohlman, L.M., Higginbotham, M.L., Welles, E.G., et al. (2009) Immunophenotypic and histological classification of 50 cases of feline gastrointestinal lymphoma. Vet Pathol 46:259–268.
  5. 5. Coyle, K.A. and Steinberg, H. (2004) Characterization of lymphocytes in canine gastrointestinal lymphoma. Vet Pathol 41:141–146.
  6. 6. Roccabianca, P., French, R.A., Seitz, S.E., and Valli, V.E. (1996) Primary epitheliotropic alimentary T‐cell lymphoma with hepatic involvement in a dog. Vet Pathol 33:349–352.
  7. 7. Woo, J.C., Roccabianca, P., van Stijn, A., and Moore, P.F. (2002) Characterization of a feline homologue of the αE integrin subunit (CD103) reveals high specificity for intra‐epithelial lymphocytes. Vet Immunol Immunopathol 85:9–22.
  8. 8. Bacon, C.M., Du, M.‐Q., and Dogan, A. (2007) Mucosa‐associated lymphoid tissue (MALT) lymphoma: a practical guide for pathologists. J Clin Pathol 60:361–372.
  9. 9. Burton, A.J., Nydam, D.V., Long, E.D., and Divers, T.J. (2010) Signalment and clinical complaints initiating hospital admission, methods of diagnosis, and pathological findings associated with bovine lymphosarcoma (112 cases). J Vet Intern Med 24:960–964.
  10. 10. Dukes, T.W., Bundza, A., and Corner, A.H. (1982) Bovine neoplasms encountered in Canadian slaughterhouses: a summary. Can Vet J 23:28–30.
  11. 11. Carrasco, V., Rodriquez‐Bertos, A., Wise, A.G., et al. (2015) Distinguishing intestinal lymphoma from inflammatory bowel disease in canine duodenal endoscopic biopsy samples. Vet Pathol 52:668–675.
  12. 12. Steinberg, H., Dubielzig, R.R., Thompson, J., and Dzata, G. (1995) Primary gastrointestinal lymphosarcoma with epitheliotropism in three shar‐pei and one boxer dog. Vet Pathol 32:423–426.
  13. 13. La Perle, K.M., Piercy, R.J., Long, J.F., and Blomme, E.A.G. (1998) Multisystemic, eosinophilic, epitheliotropic disease with intestinal lymphosarcoma in a horse. Vet Pathol 35:144–146.
  14. 14. Pinkerton, M.E., Bailey, K.L., Thomas, K.K., et al. (2001) Primary epitheliotropic intestinal T‐cell lymphoma in a horse. J Vet Diagn Invest 202:150–152.
  15. 15. Valli, V.E. (2007) Enteric T‐cell lymphoma. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 318–327.
  16. 16. Allison, R.W. and Velguth, K.E. (2010) Appearance of granulated cells in blood films stained by automated aqueous versus methanolic Romanowsky methods. Vet Clin Pathol 39:99–104.

Hepatosplenic and hepatocytotropic lymphoma (γδ T‐cell lymphoma)


These are two different lymphomas that share certain morphologic and phenotypic characteristics. Both tumors are rare; they are gamma/delta (γδ) cytotoxic T‐cell lymphomas that progress rapidly and dogs die or are euthanized because of the neoplasm in a matter of days or weeks after diagnosis.


Hepatosplenic lymphoma (HS‐TCL) is a recognized classification and has been reported in dogs,1–3 cats,4,5 and a horse.6 HS‐TCL is a rare lymphoma in humans associated with systemic signs of illness, anemia, thrombocytopenia, and a poor prognosis.7 There is infiltration of the spleen, liver, and bone marrow without lymphadenopathy and the spleen is suspected to be the origin of the lymphoma.1,7 Neoplastic cells in circulation are not a feature of HS‐TCL or hepatocytotropic T‐cell lymphoma (HC‐TCL). The cells in human and canine HS‐TCL have a γδ type T’cell receptor.1,7


Only a few cases of HS‐TCL in dogs have been reported and most were from retrospective material.1,3 Patterns were noted and they are summarized below but the original publications should be read for the details they provide. Dogs with HS‐TCL have T‐cell lymphoma in the liver and spleen. Some cases have tumor in mesenteric lymph nodes and other abdominal organs but not peripheral lymph nodes; however, autopsy data is incomplete therefore it is difficult to know how extensive these lymphomas are.


The diagnosis of HS‐TCL is made from histology of the liver which contains a lymphoma that is primarily located sinusoidal rather than periportal or perivascular. The tumor cells form a linear pattern along hepatic cords. They may compress and cause atrophy of hepatocytes. There will be neoplastic cells in portal tracts and around central veins, but this pattern is less pronounced than the sinusoidal distribution (Figure 7.51, see also Figure 14.19A,B). The infiltration is usually marked and easily seen at lower magnifications. Involvement of bone marrow is patchy, with some areas heavily infiltrated and other areas so mildly involved it may require IHC with CD3 to be certain the tumor is present (Figure 7.52). Mesenteric nodes are irregularly infiltrated. The colonization of nodes spares the outer cortex and germinal centers and involves the inner paracortex and medullary areas. Blood vessels in the lungs will contain tumor cells and this lesion ranges from mild to marked.1

Micrograph of hepatosplenic T-cell lymphoma in a dog. Pattern of infiltration in this liver is periportal and sinusoidal.
Micrograph of hepatosplenic T-cell lymphoma in a dog. In the portal area, bile duct (arrow), the sinusoids are dilated by neoplastic lymphocytes that are intermediate to large cell size.
Micrograph of hepatosplenic T-cell lymphoma in a dog. In the spleen, there is cuffing around arterioles by the infiltrating lymphocytes. Inset: Lymphoma cells are larger than the red blood cells.
Micrograph of hepatosplenic T-cell lymphoma in a dog. In the spleen, half of the tumor cells stain strongly, the rest appear unlabeled. Inset: Tumor cells are negative. Residual lymphocytes are non-neoplastic.

Figure 7.51 Hepatosplenic T‐cell lymphoma (HS‐TCL), dog. (A) The pattern of infiltration in this liver is periportal and sinusoidal. Although most lymphomas in the liver are primarily periportal or perivascular, hepatosplenic types of lymphomas are sinusoidal. Myeloproliferative diseases are also primarily sinusoidal. These patterns help distinguish the neoplastic processes but are not definitive. (B) Portal area, bile duct (arrow): The sinusoids are dilated by neoplastic lymphocytes that are intermediate to large cell size. (C) Spleen: There is cuffing around arterioles by the infiltrating lymphocytes. Inset: Lymphoma cells are much larger than the red blood cells. (D) CD3, spleen: About half of the infiltrating tumor cells stain strongly and the rest appear unlabeled. Hepatosplenic lymphomas are of gamma/delta type and therefore not all neoplastic cells will stain with CD3. Inset CD79a: The tumor cells are negative. A few residual lymphocytes near an arteriole are non‐neoplastic.

Micrograph of hepatosplenic T-cell lymphoma in a dog. In the spleen, cytological preparation: Erythrophagocytosis by neoplastic cells is prominent (arrows) and mitotic figure is present.
Micrograph of hepatosplenic T-cell lymphoma in a dog. Cytological preparation: Neoplastic cells in the spleen are strongly marked and the mature red blood cells and developing erythroid cells are not.
Micrograph of hepatosplenic T-cell lymphoma in a dog. In the liver, neoplastic lymphoid cells fill and dilate the hepatic sinusoids.
Micrograph of hepatosplenic T-cell lymphoma in a dog. CD3: Lymphoma cells in sinusoids are strongly marked. Inset: Bone marrow stained with CD11d illustrates positive staining of the tumor.

Figure 7.52 Hepatosplenic T‐cell lymphoma (HS‐TCL), dog. (A) Spleen, cytological preparation: Erythrophagocytosis by neoplastic cells is prominent (arrows); mitotic figure is present; note erythropoiesis with numerous metarubricytes. Tumor cells and macrophages are erythrophagocytic in HS‐TCL. (B) CD3, cytological preparation: The neoplastic cells in the spleen are strongly marked and the mature red blood cells and developing erythroid cells are not. (C) Liver: Neoplastic lymphoid cells fill and dilate the hepatic sinusoids. This distribution, along with tumor cells that form a linear pattern along hepatic cords, are characteristic of HS‐TCL. Many of them are of LGL origin. (D) CD3: Lymphoma cells in sinusoids are strongly marked. Inset: Bone marrow stained with CD11d illustrates positive staining of the tumor. CD11d is an integrin expressed by cells in the splenic red pulp, predominantly histiocytes and granular lymphocytes that traffic to the red pulp. The latter gives rise to some types of hepatosplenic lymphoma and splenic T‐CLL. CD11d is considered a marker of splenic red pulp origin. The combination of neoplastic lymphocytes in sinusoids of liver, lymphoma in the spleen, erythrophagocytosis by tumor cells, and positive staining pattern of tumor with CD3 and CD11d provides the diagnosis of HS‐TCL. CD11d will also mark granular lymphocyte T‐CLL and hemophagocytic histiocytic sarcomas (of splenic red pulp macrophage origin), but the more common histiocytic sarcomas of dendritic cell origin are usually CD11d negative. (Images courtesy of W. Vernau, UC Davis.)


Cytologically or histologically, a key to the diagnosis is erythrophagocytosis by neoplastic lymphoid cells and non‐neoplastic histiocytes. Erythrophagocytosis is more prominent in macrophages and it must be searched for in the neoplastic lymphocytes. The spleen contains the tumor in sinudoidal areas and red pulp. The white pulp may be atrophic. Erythrophagocytosis by macrophages is prominent in the spleen.


The neoplastic lymphocytes are intermediate to large, nuclei are 1.5–3 RBC in diameter.1 Cytoplasm ranges from a thin rim to abundant and sometimes is clear. Mitotic figures range from 0 to 2/10 HPF, which is fairly low for aggressive lymphomas. Six out of 7 dogs with HS‐TCL were dead within 1 month of their diagnosis.1 The immunophenotypic profile was CD3 positive, CD11d positive, γδ type T‐cell antigen receptor, and granzyme B positive.1 Taken together, these data support splenic cytotoxic T cells as a likely origin of this lymphoma. None of the tumor cells were identified as LGL, although CD11d is expressed on LGL (as well as macrophages and T lymphocytes in the splenic red pulp), the cytoplasm was clear, and LGL can be of γδ T lymphocytes. Clinical parameters indicated that dogs with HS‐TCL had a regenerative anemia, thrombocytopenia, hypoalbuminemia, variable hepatic enzyme increases, and an absence of leukemia and bilirubinemia.1


Dogs with HC‐TCL had sufficiently different clinicopathologic results, histologic lesions, and immunophenotype that they were considered a separate entity and not a subtype of HS‐TCL.1 There were only two dogs found with this tumor so data were limited. The principal lesions in each dog were centered on the liver and consisted of neoplastic lymphocytes that were not just sinusoidal but followed and invaded hepatic cords and which appeared to be intrahepatocytic (see Figures 7.53 and 14.20). This appearance was distinctive. Transmission electron microscopy revealed the lymphocytes were not truly in the cytoplasm of hepatocytes and the appearance was due to an invagination of the hepatocyte cell membrane as lymphocytes encroached.1

Micrograph of large granular lymphocyte-type lymphoma in a cat's liver. The tumor cells appears to be within the cytoplasm of hepatocytes providing an emperipolesis-like pattern.
Micrograph of large granular lymphocyte-type lymphoma in a cat's liver. Magnification demonstrates the pseudo-intracellular location of the lymphoma.

Figure 7.53 Large granular lymphocyte‐type lymphoma, liver, cat. (A) The tumor cells appear to be within the cytoplasm of hepatocytes providing an emperipolesis‐like pattern characteristic of hepatocytotropic T‐cell (HC‐TCL) lymphoma. (B) Higher magnification demonstrates the pseudo‐intracellular location of the lymphoma, but cytoplasmic granules in the lymphoid cells are difficult to see in H&E sections. Touch imprints or methanolic‐based stains would demonstrate the granules better. PTAH staining to visualize granules is not reliable but sometimes is helpful. Intracytoplasmic granules are difficult to see in histopathology.


Neoplastic lymphoid cells looked the same as in HS‐TCL. The phenotypic profile was the same, with the exception that the lymphocytes in HC‐TCL were negative for CD11d.1 Other differences were the absence of erythrophagocytosis, no anemia but pronounced icterus, hyperbilirubinemia (>20 mg/dL), and massive increases of serum ALP (>20,000 IU/L) and gamma glutamyltransferase.1 Both dogs died within days of diagnosis. The cell of origin for HC‐TCL is believed to be cytotoxic γδ T cells but perhaps the liver is the source of these cells rather than the more expected splenic source. Sinusoidal distribution of lymphoma cells in the liver is a diagnostic aid for HS‐TCL and the unique pattern following hepatic cords for HC‐TCL. Sinusoidal distribution of tumor cells in the liver is a characteristic of myeloid neoplasms or leukemic forms of lymphoma.


Clinical features and response to treatments of hepatic lymphoma were described in 18 dogs.2 These dogs had heterogeneous types of lymphomas that included B‐cell, multicentric, and mediastinal.2 Presumably a few were HS‐ or HC‐TCL but insufficient characterizations were performed. The MST in this group of dogs was approximately 2 months with a range of 2–402 days.2


The histologic signature of HC‐TCL has also been observed in the liver of cats with T‐cell lymphoma.4,5 The authors chose the term emperipolesis‐like to emphasize the distinctive histologic pattern of lymphoid cells that appeared to be in hepatocytes. The light microscopic appearance of an intracellular location was due to invaginations in the cell membrane of hepatocytes, and the tumor cells were not actually in the cytoplasm of hepatocytes.5 Four of 12 cats were of LGL cell type (Figure 7.54) and five had monoclonal proliferations of neoplastic cells as determined by PCR for the TCRG gene.5 Survival and prognostic data were not reported but they are probably similar to those in humans and dogs.

Micrograph of large granular lymphocyte (LGL) leukemia in a cat. Neoplastic cells are found in blood and there is peripheral lymph node enlargement.
Micrograph of large granular lymphocyte (LGL) leukemia in a cat, featuring a magnification of leukemic LGL cells. Granules in cell are densely stained and prominent. Nuclei are round to oval with uniform contour.
Micrograph of large granular lymphocyte (LGL) leukemia in a dog. Few pink granules are in the cytoplasm of the LGL, which is larger than the neutrophil. The nucleus has a characteristic indented portion.

Figure 7.54 Large granular lymphocyte (LGL) leukemia, cat (A,B) and dog (C). (A) Six‐year‐old male mixed‐breed cat with large mid‐abdominal mass. Fine‐needle aspirate of mass indicated LGL lymphoma, likely in intestine. Neoplastic cells were also found in blood and there was peripheral lymph node enlargement. Cats with intestinal LGL lymphoma and leukemia have very short survival times (<1 month). (B) High magnification of leukemic LGL cells. Granules in this cell and those in (A) are densely stained and prominent. Nuclei are round to oval and the contour uniform. LGL often have irregular nuclei and indentations. (A,B Images courtesy of John Harvey, University of Florida.) (C) Dog LGL leukemia. A few pink granules can be seen in the cytoplasm of the LGL, which is larger than the neutrophil. Granules in canine LGL tend to be less obvious than in feline LGL. They are faintly visible to the right of the nucleus in this large neoplastic cell. The nucleus has a characteristic indented portion.


References



  1. 1. Keller, S.M., Vernau, W., Hodges,J., et al. (2012) Hepatosplenic and hepatocytotropic T‐cell lymphoma: two distinct types of T‐cell lymphoma in dogs. Vet Pathol 50:281–290.
  2. 2. Dank, G., Rassnick, K.M., Kristal, O., et al. (2011) Clinical characteristics, treatment and outcome of dogs with presumed primary hepatic lymphoma: 18 cases (1992–2008). J Am Vet Med Assoc 239:966–971.
  3. 3. Fry, M.M., Vernau, W., Pesavento, P.A., et al. (2003) Hepatosplenic lymphoma in a dog. Vet Pathol 40:556–562.
  4. 4. Ossent, P., Stockli, R.M., and Posposchil, A. (1989) Emperipolesis of lymphoid neoplastic cells in feline hepatocytes. Vet Pathol 26:279–280.
  5. 5. Suzuki, M., Kanae, Y., Kagawa, Y., et al. (2011) Emperipolesis‐like invasion of neoplastic lymphocytes into hepatocytes in feline T‐cell lymphoma. J Comp Pathol 144:312–316.
  6. 6. Roccabianca, P., Paltrinieri, S., Gallo, E., and Giuliani, A. (2002) Hepatosplenic T’cell lymphoma in a mare. Vet Pathol 39:508–511.
  7. 7. Belhadj, K., Reyes, F., Farcet, J.P., et al. (2003) Hepatosplenic gammadelta T‐cell lymphoma is a rare clinicopathologic entity with poor outcome: report on a series of 21 patients. Blood 102: 4261–4269.

Intravascular large T‐cell lymphoma, subcutaneous panniculitis‐like T‐cell lymphoma, angioimmunoblastic T‐cell lymphoma, aggressive NK‐cell leukemia/lymphoma


Intravascular large T‐cell lymphoma


In this rare disease, neoplastic lymphocytes proliferate inside blood vessels rather than in hemolymphatic organs.1–7 Despite intravascular proliferation and numerous neoplastic cells in blood vessels on histopathology, leukemia is not present, at least not when assessed with routine means of detection. In approximately 5% of human cases neoplastic cells can be detected in circulation. One canine case documented a LGL intravascular large T‐cell lymphoma (IVL) that had neoplastic cells in circulation but the number of tumor cells were low (approximately 500/μL, 8%).6 The authors concluded that it was not a histiocytic tumor, there was erythrophagocytosis by tumor cells, and it was CD18/CD45 positive but E‐cadherin negative. Therefore it could be histiocytic. The cells certainly look histiocytic in published images.


IVL occurs in humans and is very rare in animals but is reported in dogs,2 cats,3 and a horse.7 The majority in humans are B‐cell (90%) but in dogs T‐cell is more common. B‐cell is reported in dogs, as is non‐T non‐B.2 The most common clinical signs in dogs are related to the brain or spinal cord, which are among the most common tissues to contain IVL.2 Veins and arteries from multiple organs have been found to contain IVL and the numbers of neoplastic cells per vessel vary widely from only a few to numerous. Some vessels are packed with neoplastic cells which distend and/or occlude the vessel (Figures 7.55, 7.56, and 19.38) and adjacent tissues are infarcted. There are a variety of secondary lesions in the walls of blood vessels. Vessels in any tissue may be affected, and there are examples in which vessels of the nasal turbinates were involved and the tumor bridged the cribriform plate and invaded the meninges of the forebrain (Figure 7.56).

Micrograph of intravascular lymphoma (IVL) in a dog's turbinates. Arteries are thrombosed with fibrin and neoplastic lymphocytes. Surrounding submucosa have edema with large numbers of inflammatory cells.
Micrograph of intravascular lymphoma (IVL) in a dog's turbinates. Veins and arterioles are nearly occluded by lymphoma cells.

Figure 7.55 Intravascular lymphoma (IVL), turbinates, dog. (A) This adult dog presented with epistaxis. The arteries are thrombosed with fibrin and neoplastic lymphocytes. There is edema of the surrounding submucosa with increased numbers of inflammatory cells but all the neoplastic cells are intravascular. (B) Veins and arterioles are nearly occluded by lymphoma cells.

Micrograph of intravascular lymphoma (IVL). Section of brain from the olfactory region of the dog is stained with CD3. It has many positively stained neoplastic lymphoid cells inside blood vessels in meninges.
Micrograph of intravascular lymphoma (IVL).Vessels are packed with neoplastic lymphocytes, heavily and uniformly positive. Inset: Neoplastic cells are completely unstained by the B-cell antibody.

Figure 7.56 Intravascular lymphoma (IVL). (A) Section of brain from the olfactory region of the dog described in Figure 7.55 stained with CD3 has numerous positively stained neoplastic lymphoid cells inside blood vessels in the meninges. (B) CD3: The vessels are packed with neoplastic lymphocytes that are heavily and uniformly positive. Most IVL in dogs are T‐cell and they seem to have a predilection for CNS. Inset: CD79a: The neoplastic cells are completely unstained by the B‐cell antibody. They are a very rare type of lymphoma and are aggressive.


Diagnosis of IVL in animals is made on histopathology based on seeing neoplastic lymphocytes in the lumen of blood vessels but without neoplasia in nodes or other tissues. The endothelium is still visible and muscular walls are unaffected. Intravascular neoplasia needs to be differentiated from tumor cells that form a cuff around a vessel and tumors that invade the wall of a vessel. The latter is seen in some histiocytic sarcomas. The lumen of the vessel is devoid of tumor cells but there are intramural tumor cells, and these are not IVL.3 Vaccine reactions can create a pattern that looks similar to IVL.3 Occasionally, a lymphoma may form a cuff or mantle between the vessel wall and the pericyte sheath that enlarges and constricts the lumen of the vessel, usually a vein. The lumen of the vessel may be a thin slit in the center but tumor cells are not inside the lumen and these are not IVL. In some examples of IVL areas of necrosis are adjacent to affected vessels, but usually there is a viable rim of inflammatory or neoplastic cells immediately around the blood vessel.


Subcutaneous panniculitis‐like T‐cell lymphoma


This subtype of lymphoma is one of the most uncommon lymphomas in humans and certainly in animals.1,8 It is a cytotoxic T’cell lymphoma that has a predilection for subcutaneous adipose tissue.1 The tumor has a high apoptotic index and tumor cells surround and infiltrate between adipose cells.7 It has been reported in a dog.3 The lesion was a subcutaneous mass that appeared to be panniculitis. There was extensive necrosis of fat and the cellular infiltration was pleocellular. There were large atypical lymphoid cells admixed in the reaction and some tumor cells formed discrete foci. The neoplastic cells had nuclei that were 2–2.5 RBC in diameter, with a vesicular appearance due to peripheralized chromatin. Nuclei contained one or two large central nucleoli. A large portion of the lesion appeared to be only panniculitis and it would be relatively easy to overlook the neoplastic cell population. The atypical lymphocytes were weakly positive with CD3 and negative with B‐cell markers (Figures 7.57 and 7.58).

Micrograph of large T-cell lymphoma in a dog's subcutis. There is infiltration of the subcutaneous adipose by lymphoma.
Micrograph of large T-cell lymphoma in a dog's subcutis. Neoplastic cells nuclei are >2 RBC in diameter, with irregular chromatin and large central nucleoli. Cell is in necrosis (arrow), below is a mitotic figure.” src=”http://veteriankey.com/wp-content/uploads/2020/03/c07f057b.jpg”> <FIGCAPTION><br />
<P><SPAN class=figureLabel><A id=c7-fig-0057 href=Figure 7.57 Large T‐cell lymphoma, subcutis, dog. (A) There is infiltration of the subcutaneous adipose by lymphoma. It is possible to overlook the neoplasm because of extensive necrosis and hemorrhage, as in this image. (B) Nuclei of neoplastic cells are >2 RBC in diameter, with irregular chromatin and multiple large central nucleoli. One cell is undergoing necrosis (arrow) and below this is a mitotic figure. It can be difficult to distinguish apoptosis and mitosis: eosinophilic cytoplasm and round nucleus (arrow) favor apoptosis, whereas hairy extensions from the metaphase aggregate indicate mitotic figure. Eosinophils are numerous in this example. Infrequently lymphomas, as well as other neoplasms (fibrosarcoma), can cause paraneoplastic eosinophilia.

Micrograph of large T-cell lymphoma in a dog's subcutis. A third of the cells are lightly positive and a few are strongly positive. Large oval nuclei were stromal cells.
Micrograph of large T-cell lymphoma in a dog's subcutis. Tumor cells are completely negative.

Figure 7.58 Large T‐cell lymphoma, subcutis, same dog as in Figure 7.57. (A) CD3: About a third of the cells are lightly positive and a few are strongly positive. Large oval nuclei were stromal cells. (B) CD79a: Tumor cells are completely negative. Interpretation: Most lymphomas are more uniform and have a higher percentage of positive cells than seen in this example. When CD79a is negative and CD3 is not conclusive, consider using other B‐cell antibodies (CD20). Loss of CD3 expression is a feature of some T‐cell lymphomas. When CD3 is positive in only a fraction of the neoplastic cells the T cells could be of gamma/delta type. Additional batteries of antibodies are needed for this diagnosis (CD11d). If results are still inconclusive, consider PCR for clonality, additional sections of subcutis, and examination of a lymph node. This is a good example of combining morphology, which provided the diagnosis of lymphoma with ancillary tests to help identify cell of origin.


In human cases the cells of this lymphoma have been shown to be of γδ receptor type.1,8 This may be true for the dog, and this could account for the light labeling with CD3 because the gamma/delta receptor, like the alpha/delta receptor, is co‐expressed with CD3.3


Angioimmunoblastic T‐cell lymphoma


Angioimmunoblastic T‐cell lymphoma (AILT) is described in humans as a rarely encountered neoplasm characterized by generalized lymphadenopathy with fever and polyclonal gammopathy. The tumor was formerly called angioimmunoblastic lymphoma with dysproteinemia (AILD).1,3 It is rare in dogs. The unique feature is a regular interlacing pattern of prominent vessels that each enclose about 50 lymphocytes. This forms a grid‐like pattern that is appreciated at low or medium magnifications (Figure 7.59). The vessels have deeply stained cytoplasm and vesicular endothelial nuclei. The enclosed lymphocytes are of variable type with round to oval nuclei of 1.5–2.0 RBC in diameter with branched chromatin and a single prominent central nucleolus. The cytoplasm is relatively abundant, deeply stained, and amphophilic. The amphophilic cytoplasm and eccentric nuclei impart a plasmacytoid appearance. The mitotic count is low. About one‐third to half of the cells are strongly positive with CD3 but many cells are unmarked, perhaps due to the γδ receptor configuration.3 The only cells positive with B‐cell markers are an occasional plasma cell.

Micrograph of angioimmunoblastic T-cell lymphoma (AILT) in lymph node of a dog at low magnification with clear spaces and the tumor cells being clustered or individualized creating a grid-like pattern.
Micrograph of angioimmunoblastic T-cell lymphoma (AILT) in lymph node of a dog, displaying small vessels within the tumor and the neoplastic cells having eosinophilic cytoplasm surrounded by halos (inset).
Two micrographs of angioimmunoblastic T-cell lymphoma (AILT) in lymph node of a dog, displaying CD3 to right and CD79a to left.

Figure 7.59 Angioimmunoblastic T‐cell lymphoma (AILT), lymph node, dog. (A) At low magnification there are clear spaces and the tumor cells are in clusters or are individualized, creating a grid‐like pattern. (B) There are small vessels within the tumor. The neoplastic cells have eosinophilic cytoplasm and are surrounded by clearly recognizable halos (inset). Neoplastic cells vary from intermediate to large size; chromatin is densely stained and the larger cells have irregular dispersion of chromatin with prominent central nucleoli. The cytoplasm is densely stained and the clear areas are not artifact. These tumors could be misinterpreted as mast cell tumors via H&E. (C) CD3 to right, CD79a to left: CD3 strongly marks about a third of the tumor cells and CD79a is negative (the nuclear staining is an artifact). AILT is a recognized entity in human medicine but it is poorly defined in veterinary medicine. This case resembles the disease seen in humans.


Aggressive NK‐cell leukemia and blastic lymphoma


The NK‐cell leukemias and lymphomas are rarely encountered in humans or animals and few labs in veterinary medicine have the capability to diagnose these tumors, so they are largely unknown. In the WHO classification these are considered separate diseases. There are no markers to recognize NK tumors in domestic animals, so this designation is extrapolated from all the data. It is important to note that they do not mark with surface receptors for B or T lymphocytes.1,4 There is no phenotypic specificity to NK tumors. They are germ‐line cells without rearrangement of the T‐cell receptor gene and therefore express neither the more common αβ nor the less common γδ type of T‐cell antigen receptor.1,4 These cells usually express the ε‐chain part of the CD3 complex and as a result may have cytoplasmic labeling but not membrane reactivity. The absence of T‐cell or B‐cell receptors can also be due to reasons other than NK lineage (e.g., methodologies, errors, prolonged storage).


In the human forms of this tumor the cells are consistently of intermediate size and of LGL cytoplasmic type, with moderately abundant clear cytoplasm that contains a few small eosinophilic granules.1,4 The cell type is the same for both the leukemic and solid forms of the neoplasm and in humans both are associated with EBV infection. The tumor cells are identified by positive immunoreactivity with CD56, CD2, T1A‐1, and granzyme B, with the T‐cell receptor remaining germ line (negative).1,4 The tumor cells exhibit their deleterious effect through major histocompatibility unrestricted cytotoxicity.


Aggressive NK‐cell leukemia


This tumor occurs in people of Asian origin with a mean age of approximately 40 years.1,4 The clinical presentation includes fever, hepatosplenomegaly, lymphadenopathy, and leukemia. Neoplastic cells in the blood vary from normal‐appearing LGL cells to larger and more atypical cells with lobulated nuclei and lightly basophilic and granulated cytoplasm. The bone marrow may be lightly to heavily invaded by the tumor. The condition is rapidly fatal with a median survival of approximately 2 months. The major differential diagnosis is the more common LGL leukemia that follows a more indolent progression in humans.


Blastic NK‐cell lymphoma


The extranodal NK‐cell blastic lymphoma has a higher prevalence in Asian, Mexican, and South American Indian populations.1,4 The presentation is almost always in adults, with a 2:1 male predominance. The syndrome includes the lymphoma previously known as lethal midline granuloma or nasal lymphoma. Other sites include the skin, gastrointestinal tract, lung, eye, and soft tissues. Skin lesions ulcerate and present with areas of necrosis.


The disease in dogs that may be similar is hepatosplenic lymphoma (see earlier section). The neoplastic cells have moderate cytoplasmic staining with CD3 but not membrane and are found on clonal examination to have germ‐line T‐cell receptor.4 The neoplastic cells are LGL or nongranulated lymphocytes that may be small or large. Some nuclei will be indented, reniform, or folded. Giemsa‐stained preparations with a methanolic base are recommended for visualizing the cytoplasmic granules. Involvement of the gastrointestinal tract is usually transmural and there may be coagulative necrosis. In deeper skin lesions the NK tumor resembles the panniculitis‐like T‐cell lymphoma.5,8


References



  1. 1. Cheuk, W. and Chan, J.C. (2011) NK‐cell neoplasms. In Hematopathology (eds. E.S. Jaffe, N.L. Harris, J.W. Vardiman, et al.). Saunders/Elsevier, Philadelphia, PA, pp. 473–491.
  2. 2. McDonough, S.P., Van Winkle, T.J., Valentine, B.A., et al. (2002) Clinicopathological and immunophenotypical features of canine intravascular lymphoma (malignant angioendotheliomatosis). J Comp Pathol 126:277–288.
  3. 3. Valli, V.E. (2007) Angioimmunoangioblastic lymphoma. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 312–316.
  4. 4. Valli, V.E. (2007) Aggressive NK‐cell leukemia. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 309–312.
  5. 5. Valli, V.E. (2007) Subcutaneous panniculitis‐like T‐cell lymphoma. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 327–330.
  6. 6. Lane, L.V., Allison, R.W. (2012) Canine intravascular lymphoma with overt leukemia. Vet Clin Pathol 41:84–91.
  7. 7. Raidal, S.L., Clark, P., and Raidal, S.R. (2006) Angiotrophic T‐cell lymphoma as a cause of regenerative anemia in a horse. J Vet Intern Med 20:1009–1013.
  8. 8. Sen, F., Rassidakis, G.Z., Jones, D., and Medeiros, J. (2002) Apoptosis and proliferation in subcutaneous panniculitis‐like T‐cell lymphoma. Mod Pathol 15:625–631.

Adult T‐cell leukemia/lymphoma


Defining the neoplasm


The current WHO classification gave this disease its name. The human neoplasms are caused by an endemic retrovirus (HTLV‐1 virus) on the south Island of Japan and in the Caribbean basin.1–8 The virus is highly cell associated and can be spread by cells in breast‐fed infants. This was also shown years ago as the mechanism by which BLV is spread from dam to calf.7 There is no direct animal counterpart of the HTLV‐1 virus of humans, but the mode of transmission is highly comparable to that of BLV.6,7 The conversion from a viral infection to a neoplastic disease in humans is low: approximately 1.5 males/5000 infected men occurred over a period of about 40 years. In cattle the rate of neoplastic conversion is estimated to be about 1% of infected animals and occurs over a 7‐ to 8‐year time line.7


Epidemiology and occurrence


The cells containing virus can be spread by sexual contact in humans and cattle but the disease in cattle is not spread by frozen semen. The presence of the virus and the lymphoma is much more common in dairy herds than in beef herds. This is partly due to the proximity of the animals in dairy herds and procedures such as dehorning or vaccinations if the equipment used becomes contaminated with blood containing BLV‐infected cells.


In cats FeLV is in the blood and is not cell associated so the mode of spread is more by aerosol and fighting.8 Testing and vaccinating for FeLV has switched the pattern of lymphomas from multicentric and mediastinal in young cats to extranodal in adult cats. An aggressive lymphoma developed in genetically SCID mice given intradermal and intraperitoneal injections of splenic cells from transgenic mice given the Tax gene of the human HTLV‐1 virus.9 These mice all died within 28 days with extensive lymphoma involving the spleen, nodes, bone marrow, liver, kidney, and lung. The tumor cells had the characteristic floral cells of the human lymphoma.


Pathology


Once the retroviral lymphoma has developed, there are many similarities in the tumors of humans, cats, and cattle. In the human lymphoma induced by HTLV‐1 the involved tissues include the nodes, marrow, spleen, liver, gastrointestinal tract and central nervous system. All of which can be involved in lymphomas caused by FeLV and BLV.6–8 The cells in the blood of humans with adult leukemia/lymphoma are intermediate to large with multilobated nuclei of a large atypical cell type. All of the cells in the human tumor are CD4 positive and strongly express IL‐2 and the IL receptor IL‐2R and CD25.1.3–5


The diagnosis in cattle may be made in an asymptomatic animal that has a high level of circulating lymphocytes (100 × 103/μL).6,7 In bovine lymphomas the cells are usually of large cell type (66%), about 20% are of intermediate size, and 10% are small cell type as determined on a review of over 1000 cases.6 The multilobated nuclei seen in the human adult T‐cell type of lymphoma are also present in the blood of cows with BLV‐positive lymphomas. In contrast to the retroviral‐induced human tumor, most of the large cell lymphomas in cattle are of B‐cell type.6.7 In 602 cats with lymphoma there were large cells in 54%, with 18% of intermediate cell type and 17% of cases with small cells.8 In cats the large cell lymphomas can be of either B‐ or T‐cell type.


References



  1. 1. Hasegawa, H., Sawa, H., Lewis, M.J., et al. (2006) Thymus‐derived leukemia‐lymphoma in mice transgenic for the Tax gene of human T‐lymphotrophic virus type I. Nature 12:466–471.
  2. 2. Jaffe, E.S. (2011) Adult T‐cell leukemia/lymphoma. In Hematopathology (eds. E.S. Jaffe, N.L. Harris, J.W. Vardiman, et al.). Saunders/Elsevier, Philadelphia, PA, pp. 521–531.
  3. 3. Semmes, J.O. (2006) Adult T cell leukemia: a tale of two T cells. J Clin Invest 116:858–860.
  4. 4. Taylor, G.P. and Matsuoka, M. (2005) Natural history of adult T‐cell leukemia/lymphoma and approaches to therapy. Oncogene 24:6047–6057.
  5. 5. Tobinai, K., Watanabe, T., and Jaffe, E.S. (2010) Human T‐cell leukemia virus type I‐associated adult T‐cell leukemia lymphoma. In Non‐Hodgkin Lymphomas , 2nd edn. Wolters Kluwer/Lippincott Williams & Wilkins, Philadelphia, pp. 404–414.
  6. 6. Vernau, W., Valli, V.E.O., Dukes, T.W., et al. (1992) Classification of 1,198 cases of bovine lymphoma. Vet Pathol 29:183–195.
  7. 7. Valli, V.E. (2007) Adult T‐cell lymphoma/leukemia. In Veterinary Comparative Hematopathology . Blackwell, Ames, IA, pp. 346–355.
  8. 8. Valli, V.E., Jacobs, R.M., Norris, A., et al. (2000) The histologic classification of 602 cases of feline lymphoproliferative disease using the National Cancer Institute working formulation. J Vet Diagn Invest 12:295–306.
  9. 9. Gessain, A., Mahieux, R., and de Thé, G. (1996) Genetic variability and molecular epidemiology of human and simian T cell leukemia/lymphoma virus type I. J Acquir Immune Defic Syndr 13:132–145.

Anaplastic large cell lymphoma


The lymphomagenesis of anaplastic large cell lymphoma (ALCL) is partially understood.1–7 The tumor is rare, and it may be T‐cell, B’cell, or NK. In humans it is most commonly T‐cell or NK. Some tumor cells will also express CD30 cytokine receptor for the tumor necrosis factor family. Additionally most cases also express the cytotoxic granule–associated proteins. A morphologic signature is horseshoe‐shaped nuclei in some of the large anaplastic cells, so‐called “hallmark cells” of ALCL. ALCL also has Hodgkin’s‐like cells admixed with numerous large undifferentiated cells.2 There will be variable mixtures of histiocytes, neutrophils, eosinophils, lymphocytes, and plasma cells. Inflammation in the tumor can make the diagnosis challenging. A genetic signature is t(2;5) chromosomal translocation that results in the anaplastic lymphoma kinase gene fusing with the nucleoplasmin gene.6 This fusion results in the production of the anaplastic lymphoma kinase (ALK) protein, which can be detected by IHC and is used for the diagnosis of ALCL in humans. In humans an ALK‐negative form of ALCL is seen in older patients.7 There are cutaneous and systemic forms of this disease. There is an animal model in mice1 and dogs have tumors that fit morphologic descriptions of ALCL.7


Cutaneous anaplastic large cell lymphoma T cell: Non‐epitheliotropic cutaneous lymphoma


In dogs the cutaneous form of ALCL has a diffuse infiltration of large lymphoid cells that are primarily in the superficial dermis, non‐epitheliotropic (NE).7 The neoplastic cells extend from the basement membrane of the epidermis into the dermis and may have irregular deep extensions into the panniculus but infiltration of epidermis or hair follicles is absent. When this pattern is homogenous in HE or CD3 the diagnosis is straightforward (Figures 7.60 and 7.61). However, cases with pleocellular infiltration by histiocytes, lymphocytes, eosinophils, and neutrophils can mimic cutaneous histiocytosis or even inflammation such that IHC or PARR are required for diagnosis.8 Inflammatory NE cutaneous T‐cell lymphoma is a diagnostic challenge because of its heterogeneous cellular infiltration, variable CD3 expression, and variable T‐cell antigen receptor gene rearrangement results (Figure 8.17). Neoplastic T cells often express CD18, which further confounds the interpretation of IHC patterns in tumors heavily infiltrated by histiocytes. Neoplastic lymphoid cells are of varying sizes, including large anaplastic cells with large nuclei 3–4 RBC in diameter. Binucleated cells are common. Characteristic horseshoe‐shaped nuclei can be found, along with nuclei that have complex foldings of nuclear membranes. Large nuclei of convoluted or reniform shape have prominent nucleoli. Neoplastic cells have relatively abundant cytoplasm that is lightly stained with indistinct cell boundaries. There are 1–5 mitoses/400× field.

Micrograph of anaplastic large cell lymphoma (ALCL), cutaneous T‐cell type, skin, dog, illustrating the dermis being filled by neoplastic cells.

Figure 7.60 Anaplastic large cell lymphoma (ALCL), cutaneous nonepitheliotropic T‐cell type, skin, dog. The dermis is filled by neoplastic cells. The epidermis, hair follicles, and adnexa are not invaded. The epidermis does not bulge and is not ulcerated. Some cases will have thrombosis and ischemic necrosis of the dermis and subcutis.

Micrograph of anaplastic large cell lymphoma in a dog displaying CD3 with the tumor positively stained, diffusely and strongly. Inset: Demonstrating higher magnification.

Figure 7.61 Anaplastic large cell lymphoma, dog. CD3: The tumor is positively stained, diffusely and strongly. Inset: Higher magnification. Inflamed types of NE cutaneous lymphoma will not be stained this homogenously. Epidermis and hair follicles are infiltrated in MF.


Part of the definition of this disease in humans is that extracutaneous disease is not recognized for at least 6 months following initial diagnosis. Immunoreactivity to CD30 is used in humans to assist diagnosis. Most NE‐cutaneous lymphomas in dogs are of T‐cell type, but intensity of CD3 expression is highly variable. When CD3 expression is partially “lost” it makes interpretation of the staining patterns difficult.8 Despite the aggressive nature suggested by the name of the tumor, the 5‐year survival in humans is 80–90%. Median survival time of 9 months has been reported in dogs with inflammatory form of NE cutaneous lymphoma.8


Systemic anaplastic large cell lymphoma of T‐cell type


The systemic form of ALCL is rare in dogs or is rarely recognized as a disease entity.7 It is seen in young, large‐breed dogs that present with severe systemic illness, generalized skin disease, enlarged nodes, and dependent edema.7 Neoplastic cells are as described for the cutaneous form. They are large cells with anaplastic nuclei that have reniform, horseshoe shapes, or multinucleation. That they are CD3‐positive helps differentiate them from histiocytic diseases. Mitotic figures are common, occurring at 10–20/400× field. Neoplastic involvement of nodes ranges from patchy to diffuse. The capsule is thinned and the subcapsular sinus compressed and irregularly obliterated. There may be vasculitis and thrombosis with hemorrhage and necrosis in surrounding tissue (Figure 7.62). Systemic ALCL in dogs has a rapid onset and progression.

Micrograph of lymph nodes from a different dog, depicting thrombosis and hemorrhagic necrosis of systemic tissues.

Figure 7.62 A lymph node from a different dog with ALCL is no longer recognizable due to marked ischemic necrosis, hemorrhage, and congestion due to arterial thrombosis. Thrombosis and hemorrhagic necrosis of systemic tissues, as depicted here in skin and lymph nodes, are characteristics of this very aggressive disease. ALCL is a high‐grade T‐cell lymphoma that typically affects young dogs.


References



  1. 1. Bittner, C., Feller, A.C., Renauld, J.C., et al. (2000) An animal model for anaplastic large cell lymphoma in the immunocompetent syngeneic C57BI/6 mouse. Lab Invest 80:1523–1531.
  2. 2. Falini, B. and Gisselbrecht, C. (2010) Anaplastic large cell lymphoma. In Non‐Hodgkin Lymphomas , 2nd edn. (eds. J.O. Armitage, B. Coiffier, P.M. Mauch, et al.). Wolters Kluwer Philadelphia, PA: pp. 415–432.
  3. 3. Kinney, M.C. and Kadin, M.E. (1999) The pathologic and clinical spectrum of anaplastic large cell lymphoma and correlation with ALK gene dysregulation. Am J Clin Pathol 111:56–67.
  4. 4. Krenacs, L., Wellmann, A., Sorbara, L., et al. (1997) Cytotoxic cell antigen expression in anaplastic large cell lymphomas of T‐and null‐cell type and Hodgkin’s disease: evidence for distinct cellular origin. Blood. 89:980–989.
  5. 5. Li, C., Takino, H., Eimoto, T., et al. (2007) Prognostic significance of NPM‐ALK fusion transcript overexpression in ALK positive anaplastic large‐cell lymphoma. Mod Pathol 20:648–655.
  6. 6. Delsol, G., Lamant‐Rochaix, L., and Brousset, P. (2010) Anaplastic large cell lymphoma, ALK positive and ALK negative. In Non‐Hodgkin Lymphomas , 2nd edn. (eds. J.O. Armitage, B. Coiffier, P.M. Mauch, et al.). Wolters Kluwer, Philadelphia, PA, pp. 564–579.
  7. 7. Valli, V.E. (2007) Anaplastic large cell lymphoma. In Veterinary Comparative Hematopathology . Blackwell, Ames. IA, pp. 339–345.
  8. 8. Moore, P.F., Affolter, V.K., and Keller, S.M. (2013) Canine inflamed nonepitheliotropic cutaneous T‐cell lymphoma: a diagnostic conundrum. Vet Dermatol 24:204–211.

MYELOID NEOPLASMS


Myeloid neoplasms are clonal cancers that originate in hematopoietic tissue from granulocytic, monocytic, erythrocytic, and megakaryocytic or mast cell precursors. Myeloid leukemia in most instances can be distinguished from lymphoid leukemia by tissue distribution, cell appearance, and cell markers. This distinction is important since the different leukemias have unique prognoses and responses to chemotherapy. Myeloid neoplasms are extremely heterogeneous, and may present with severe acute or very indolent illness, with hypo‐, normo‐, or hypercellular bone marrow, with marked leukocytosis or leukopenia, with marked cell dysplasia or with relatively normal cell morphology. Myelophthisis is a form of bone marrow failure with replacement or infiltration of normal hematopoietic tissue by malignant cells that release suppressive or destructive cytokines or fibroblast growth factors resulting in reduced hematopoiesis. Myelophthisis is a factor contributing to cytopenia in myeloid neoplasia.


Cases with overt leukemia characterized by a large number of abnormal cells in blood or bone marrow are straightforward to diagnose, but others require data derived from sequential blood counts, review of blood and bone marrow films, histopathology of bone marrow, and/or other hemolymphatic tissue, and immunohistochemistry and/or flow cytometry to establish the diagnosis. Consultation with cytopathologists is essential to assess hematopoietic cells in blood, when cell populations are heterogeneous or difficult to identify, or when dysplastic cells are present. Most leukemia cases are more complicated than looking at a single histopathology section of a solid tumor. Careful assessment of clinical history and morphological features of blood and bone marrow samples should enable a diagnosis of myeloid neoplasia to be made and then placement into one of the three main categories: acute myeloid leukemia (AML), myeloproliferative neoplasm (MPN, formerly called chronic leukemia), or myelodysplastic syndrome (MDS) (Table 7.1). The terms acute and chronic refer to the clinical course of disease entities and not the morphological descriptors.


The pathogenesis of myeloid neoplasms consists of proliferation of hematopoietic cells with release of a variable number of neoplastic cells into blood, myelophthisis, suppression of normal hematopoiesis, and variable infiltration of spleen, liver, lymph nodes, and other tissues. Most myeloid neoplasms in animals are similar to types recognized in people; therefore the classification of myeloid neoplasms in people is reasonably easy to adapt to animals.1 This classification utilizes nomenclature based on hematological and morphological features of the neoplasm.


Immunophenotyping of neoplastic cells with antibodies to specific antigens, and analysis of neoplastic cells for mutations and cytogenetic changes, are utilized to subcategorize myeloid neoplasms in people. Immunophenotyping of undifferentiated blast cells with a more limited range of antibodies is available at several North American and European academic veterinary laboratories, but genetic and cytogenetic analysis of cancer in animals is in its infancy. Furthermore, knowledge of the natural progression and response to therapy for different subtypes of myeloid neoplasms in domestic animals is largely lacking. Therefore, the clinician, clinical pathologist, and anatomic pathologist have to carefully integrate a thorough history, the hematology results, and bone marrow findings to derive a possible diagnosis of myeloid neoplasia. In most cases information from these sources is sufficient to make the diagnosis of leukemia and to categorize the neoplasm as AML, MPN or MDS. Immunophenotyping by flow cytometry or immunohistochemistry is useful to subcategorize undifferentiated neoplastic cells, which will eventually increase our knowledge about pathogenesis, response to therapy and prognosis.2,3


Myeloid neoplasia should be suspected if an animal has persistent anemia, thrombocytopenia, or neutropenia without obvious cause such as blood loss, inflammation, or immune‐mediated destruction of blood cells. Abnormal cells in circulation combined with cytopenia increase suspicion of myeloid neoplasia and indicate need for bone marrow biopsy. Animals with AML are often acutely ill with bleeding tendency from thrombocytopenia, or fever due to infection secondary to neutropenia. MPN and MDS have a more insidious onset with gradual weight loss and reduced energy levels. Retrospective analysis of CBCs may reveal anemia or other cytopenia that has existed for months or even years with gradual worsening, or slowly increasing neutrophil count in chronic neutrophilic leukemia, or slowly increasing hematocrit in polycythemia vera. Persistent cytopenia and dysplastic cells (giant neutrophils, rubricytes with asynchronous nuclear and cytoplasmic maturation, giant platelets) on the blood film are suggestive of MDS, and are an indication for bone marrow biopsy.


Myeloid neoplasms are probably underdiagnosed in veterinary medicine. This is because (a) biopsies and bone marrow aspirate films submitted to diagnostic laboratories are often of insufficient quality to establish a diagnosis, (b) it is difficult to make a diagnosis from a single snapshot of even a good‐quality bone marrow biopsy, (c) AML in some instances may be fatal before a bone marrow biopsy has been collected, and (d) animals are not consistently submitted for post‐mortem examination. Obtaining good‐quality bone marrow biopsies requires a general anesthetic, special needles, and substantial technical expertise. Aspirates for cytology are more readily obtained than trephine (core) biopsies, but in order to fully assess the bone marrow, in particular for a potential neoplasm, both cytological and histological evaluation is necessary. For histopathology, a trephine biopsy at least 1.5 cm, and preferably more than 2 cm in length, and free of hemorrhage, crushing, and other artifact should be collected. Therefore, biopsies need to be collected before bone marrow is aspirated at that site, and the small core of tissue needs to be handled with care to avoid crush artifact.


Good core biopsies are not easy to collect from small dogs and cats, but an increasing number of human pediatric biopsy needles and drivers are available, and can provide power‐assisted drilling and special mechanisms to retain the core in the needle during needle withdrawal.4,5 Appropriate sites for trephine biopsies are the pelvis (wing of the ilium or tuber coxae) and the head of the humerus. Bone marrow aspirates can also be obtained from the sternum.


The strength of cytology evaluation of aspirates is the qualitative assessment of cell features (granulation, vacuolation, hemoglobinization, dysplasia, stain intensity and chromatin patterns), while histology evaluation is essential to quantitatively evaluate overall cellularity, topography, mitotic rate, boney trabeculae and stroma. Identification of cell type is best accomplished by cytology plus flow cytometry and/or immunochemistry, if needed. Histopathology is used to assess overall composition of the marrow and other hemolymphatic organs, but identification of cells is more difficult. Percentage of undifferentiated precursor (“blast”) cells should be estimated on cytology and histopathology.


This chapter summarizes the approach to diagnosis of myeloid neoplasms in animals, the essential features of different tumors, and the utility of ancillary diagnostic methods. The following are steps to establish a diagnosis of myeloid neoplasia.


Diagnostic steps


Step 1. Rule out lymphoid neoplasia


By far the most common neoplasm of the hemolymphatic system in animals is lymphoma. Lymphoma is a solid tumor (with rare exceptions) of lymph nodes or extranodal sites, as detailed above. Most lymphomas consist of a homogeneous population of lymphocytes without other leukocytes. Lymphoma may be associated with hematological abnormalities, but most animals have no or only mild cytopenia or cytosis, and no abnormal cells in circulation. Lymphoid leukemia, on the other hand, is a neoplasm that arises in bone marrow or spleen, manifests with circulating neoplastic cells, and is also divided into acute and chronic forms. Animals with ALL are young to middle‐aged and have severe cytopenia and/or cytosis.


Morphologically, it is difficult or impossible to distinguish ALL from undifferentiated AML. Expression of lymphoid markers such as CD3, CD20, CD21, or CD79a, or PCR‐based detection of clonal lymphocyte antigen receptor gene rearrangement (PARR), is supportive of lymphoid rather than myeloid origin. CLL is a slowly progressive cancer of small to medium lymphocytes that originate in spleen (T‐CLL) or bone marrow (B‐CLL). CLL is readily differentiated from reactive lymphocytosis by flow cytometry or IHC. Thus, differentiating lymphoma and CLL from myeloid tumors is straightforward, whereas distinction between ALL and AML requires immunoassays.


Step 2. Maximize the complete blood cell count


The first component of assessing an animal for a potential myeloid neoplasm is a CBC. Cells in blood are maintained within tight limits in homeostasis, but virtually all animals with myeloid neoplasia have cytopenia and/or cytosis and are assessed because of these numerical abnormalities. Performing a CBC with modern hematology analyzers, such as the Advia 2120 or Sysmex XT‐2000iV, yields not only quantitative information on thousands of cells, but also qualitative information regarding neutrophil maturity, as indicated by the degree of nuclear lobulation, leukocyte peroxidase content relative to cell size, erythrocyte and reticulocyte morphology and hemoglobin concentration, platelet cytoplasmic component characteristics, size distribution of cells, and more (Figure 7.63A–C) These parameters are very helpful for initial characterization of cell types present in a blood sample.6,7

Schematic of leukocyte size versus peroxidase plot depicting a population of large cells having little peroxidase content (circled).
Schematic graph of cytogram of nuclear complexity versus size identifying increased proportion of large round nuclei (circled) consistent with blast cells and many non-lyzed large cells.
Schematic graph of analysis of erythrocytes for RNA content versus cell size indicating very few (0.4%) reticulocytes, consistent with a poorly regenerative anemia.
Micrograph of blood film review reduced erythrocyte and platelet density, displaying blast cells (arrowed) and few neutrophils.
Micrograph of bone marrow aspirate film displaying more than 80% blast cells.

Figure 7.63 Hematology analyzer (Advia 2120) cytograms of blood from an anemic and thrombocytopenic dog with AML. (A) Leukocyte size versus peroxidase plot shows a population of cells that are large and have little peroxidase content (circled). These large cells could be blast cells of any myeloid or lymphoid cell type. (B) Cytogram of nuclear complexity versus size identifies an increased proportion of large round nuclei (circled) consistent with blast cells and many non‐lyzed large cells (upper quadrant). (C) Analysis of erythrocytes for RNA content versus cell size indicates there are very few (0.4%) reticulocytes, consistent with a poorly regenerative anemia. (D) Blood film review confirms reduced erythrocyte and platelet density, shows blast cells (arrows) and few neutrophils. (E) Bone marrow aspirate film shows more than 80% blast cells.


Step 3. Review of the blood film


Following review of the hematology analyzer report, clinical pathologists should review a freshly prepared and well‐stained blood film. Slide review should address whether the cells on the slide correspond to the categories identified by the automated analysis. If not, a manual differential count of at least 100 leukocytes should be performed, and the report should be amended accordingly. If the leukocyte count is increased or there is much cell heterogeneity, 200 cells should be differentially counted. Cells are classified by conventional criteria with particular attention to enumerating immature and blast cells as separate categories. Blast cells are round cells with large, round to oval nuclei and one to three nucleoli among otherwise homogeneous nuclear chromatin (Figure 7.63D). Nucleoli are sometimes difficult to appreciate on slides stained with Romanowsky‐type stains, but with good‐quality staining and high magnification they can be identified. Rapid stains such as Diff‐Quik® have insufficient cell penetration and are not suitable to evaluate neoplastic leukocytes.


Most MPN (“chronic leukemia”) can be diagnosed from review of the CBC, blood film, and clinical history without need for bone marrow biopsy. Numerous relatively well‐differentiated cells predominate in blood, and the clinical history should rule out other causes of marked and persistent cytosis. An example of MPN is polycythemia vera, where there is an accumulation of mature erythrocytes and sometimes other cell types in an animal without dehydration, and hypoxic or paraneoplastic erythropoietin stimulation. The packed cell volume in polycythemia vera can exceed 70%, and animals may present with neurological signs due to ischemia resulting from blood stasis in small vessels of the brain.


Chronic neutrophilic leukemia (CNL) is characterized by marked leukocytosis of nondegenerate segmented neutrophils in the absence of inflammatory or paraneoplastic cytokine stimulation. In these cases, documenting repeated and consistent neutrophilia and ruling out other causes of neutrophilia is often sufficient for diagnosis of myeloproliferative neoplasia, and evaluation of a bone marrow biopsy may not be indicated or necessary.


Step 4. Aspirate bone marrow


Sample acquisition


Various sites and techniques for bone marrow aspiration have been described.5,8,9 Rosenthal needles are most widely used in small animals, and Jamshidi‐type needles in large animals. It is useful to rinse the needle with anticoagulant (2 mM EDTA in saline or 3% sodium citrate) and to retain a few drops of anticoagulant in the syringe for bone marrow aspiration in order to allow adequate time to prepare slides without clots, and to yield suspension samples suitable for subsequent automated hematology analysis and/or flow cytometry.10 Only 0.5–1 mL of marrow should be aspirated, since greater volumes result from induction of sinusoidal rupture and hemodilution, and slides should be prepared with utmost care to reduce the amount of blood and to concentrate bone marrow particles on the slides. For this purpose, slides with a drop of bone marrow placed near the frosted edge are tilted onto paper towels for the blood to run off and the bone marrow particles to “stick” to the glass slide. Then another glass slide is placed on top of the drop, the drop is allowed to spread, and the top slide is pulled along the length of the bottom slide for a thin “squash” preparation. Some films should also be prepared with a “feathered edge”, and all slides should be air dried rapidly.8 A “paint brush technique” for making bone marrow films from opened bone containing red marrow is suitable in rodents or other small animals at autopsy.11 Hematopoietic cells degenerate relatively quickly post mortem, and samples collected more than 6 hours after death yield limited cytological detail, but are still suitable for histopathological assessment of cellularity and cell proportions.


Obtaining a good bone marrow aspirate and preparing high‐quality films requires care and expertise. Poor‐quality films can destroy the diagnostic value of an otherwise good sample. Good communication between clinician and clinical pathologist can help improve specimen quality and film preparation. If bone marrow cannot be aspirated despite needle patency and proper placement, there may be fibrosis or increased cell adhesiveness, and cytology imprints should be prepared from rolling the trephine biopsy specimen on the glass slide.


Automated cell analysis


Bone marrow that has been aspirated into syringes containing EDTA, or quickly placed into an EDTA tube, is suitable for analysis in automated hematology instruments. Obvious clots should be removed prior to sample aspiration to avoid line occlusion. Automated bone marrow analysis can provide very useful information on sample cellularity, reticulocyte concentration and parameters, leukocyte differential and peroxidase content, and the number of large unstained cells (Advia) or cells with high fluorescence (Sysmex), which represent undifferentiated hematopoietic cells devoid of peroxidase activity.7,10 Leukemic samples are easily identified on scatter plots, the degree of erythrocyte and platelet regeneration (or lack thereof) is readily assessed, and the proportion of different leukocyte populations can be appreciated. Although this type of analysis is helpful for interpretation, the composition of bone marrow aspirates varies somewhat by operator, technique and lesion, and therefore automated analysis by no means replaces microscopic assessment.


Film evaluation


Good‐quality bone marrow films are essential for diagnosis of myeloid cancer and need to be evaluated systematically. Low‐power assessment of sample and film quality, particle cellularity, megakaryocyte number and morphology, and presence of iron stores should be followed by evaluation of maturation within each individual cell line, presence of dysplasia, and mitotic activity relative to cell maturity. Particle cellularity (referring to the amount of a particle composed of hematopoietic relative to adipose cells) corresponds well to overall bone marrow cellularity, as determined on histological sections.12 In normal bone marrow, megakaryocytes encompass a range of maturational stages, and there should be between 2 and 5 megakaryocytes per particle. Mitotic activity should be restricted to immature granulocytes and rubricytes.10 Bone marrow films should be assessed with a 500‐cell differential count. Such a count enumerates the proportion of blasts, promyelocytes, myelocytes, metamyelocytes, band and segmented neutrophils, early, intermediate and late stage rubricytes, rubriblasts, and any other hematopoietic cells. The “blast” category includes myeloblasts, monoblasts, megakaryoblasts, and unclassifiable blast cells.1 Blast cells comprise <5% of bone marrow cells in healthy adult dogs.10


Bone marrow stem cells are undifferentiated precursor cells able to reconstitute all hematopoietic cells after transplantation. Morphologically, stem cells are indistinguishable from other undifferentiated cells, and immunologically they are characterized by lack of expression of a panel of typically detected antigens. Thus, stem cells cannot be evaluated during routine bone marrow assessment.


Under conditions of prolonged increased granulopoiesis due to systemic inflammation, the proportion of myeloblasts increases only slightly, and differentiated post‐mitotic metamyelocytes, band and segmented neutrophils should continue to outnumber immature myeloblasts, promyelocytes, and myelocytes by a factor of at least 2.12 Recovery from acute bone marrow injury (i.e., parvoviral infection, radiation, myelotoxic drugs) or severe immune‐mediated hematological disease in cats can transiently manifest with more immature than mature granulocytes, but this shift normalizes as the animal recovers from the bone marrow injury or responds to immunosuppressive therapy, respectively. Hence, it is difficult to make a diagnosis of myeloid cancer from review of a single bone marrow biopsy, and a good history and concurrent CBC are essential. If myelotoxicity is a possible etiology, the animal should be monitored with daily or alternate day CBCs for resolution of cytopenia, and bone marrow should be reassessed after 4–6 days.


Microscopic interpretation of bone marrow films should summarize each finding in relation to the CBC, blood film, clinical history, therapy, and previous bone marrow findings. There should be indication of what, if any, additional tests may be helpful to rule in or rule out myeloid neoplasia, or to classify myeloid neoplasia. For example, presence of >5% blast cells in bone marrow of a dog with persistent non‐regenerative anemia should prompt careful review for possible toxic or infectious bone marrow insult, assessment of serial CBCs and blood films for worsening anemia or appearance of abnormal leukocytes, and rule out other causes of non‐regenerative anemia (such as iron deficiency). Lack of improvement in cytopenia increases the likelihood of myeloid neoplasia, which should then be investigated by bone marrow trephine biopsy, and possible flow cytometric analysis of blast cells.


Step 5. Bone marrow trephine biopsy


Section evaluation


Bone marrow histology is essential for diagnosis and monitoring of AML, myelofibrosis, myeloma, and nonleukemic infiltrative diseases, and for determining the degree of myelophthisis in MPN. As mentioned above, it is important to assess a good‐quality biopsy, and definitive diagnoses should not be based on specimens less than 1.5 cm in length or containing extensive artifact. Three inter‐trabecular spaces free of crush and hemorrhagic artifact at minimum are needed for appropriate assessment. Collecting a core biopsy at the site of prior aspiration is a common cause of hemorrhage artifact, and should be avoided. The core biopsy should be obtained before the aspirate, or separate sites should be used for aspirate and biopsy.


Trephine biopsies need to be briefly decalcified through incubation in EDTA or acidic solution, and should be sectioned at 2–3 μm rather than the usual 5 μm to allow identification of individual hematopoietic cells. Granulopoiesis proceeds from myeloblasts concentrated in peritrabecular regions toward band and segmented neutrophils at the center of the marrow cavity, erythropoiesis from the central part of the intertrabecular region toward central sinuses, and megakaryocytes are mostly located beside sinusoids. Microscopic review of bone marrow sections should be systematic and comprehensive, as outlined in Box 7.2.

Mar 30, 2020 | Posted by in INTERNAL MEDICINE | Comments Off on Tumors of the Hemolymphatic System

Full access? Get Clinical Tree

Get Clinical Tree app for offline access