Canine inflamed nonepitheliotropic cutaneous T-cell lymphoma: a diagnostic conundrum
Background – Cutaneous T-cell lymphoma (CTCL) in dogs is a heterogeneous disease complex, which consists of nonepitheliotropic (NE) and epitheliotropic forms. These lymphomas are readily recognized by the presence of dominant populations of cytologically atypical lymphocytes.
Objective – The objective of this study was to introduce the key features of inflamed NE-CTCL, which is easily confused with reactive, inflammatory histiocytic disease.
Animals – Twenty-four dogs (mean age 7.5 years) presented with inflamed NE-CTCL. Lesions presented as nodules, plaques or masses. An initial diagnosis of cutaneous reactive histiocytosis (11 dogs) or histiocytic neoplasia (three dogs) was made by primary pathologists.
Methods – Lesions were assessed by histology and immunohistochemistry to detect canine leukocyte antigens. Lesional genomic DNA was extracted and gene rearrangement analysis of the T-cell receptor γ locus was assessed.
Results – The cutaneous lesions consisted of pleocellular infiltration of the dermis with variable extension into the subcutis. The lesions often surrounded vessels and adnexae. Epitheliotropism was minimal or lacking. Small lymphocytes, plasma cells and intermediate to large, cytologically atypical lymphocytes were scattered between prominent histiocytic infiltrates. Atypical lymphocytes often had marked variation in the intensity of CD3 expression. Molecular clonality analysis of the T-cell receptor γ locus revealed clonal expansion of T cells in 22 of 23 dogs tested.
Conclusion – The recognition of inflamed NE-CTCL and its differentiation from cutaneous reactive histiocytosis depends on careful assessment of lymphocyte morphology and immunostaining patterns. Confirmation of the diagnosis is best accomplished by T-cell antigen receptor gene rearrangement analysis.
Cutaneous T-cell lymphoma (CTCL) is a heterogeneous group of disorders, with a variety of clinical presentations and morphological features in both humans and dogs. Epitheliotropic CTCL, also known as mycosis fungoides (MF), accounts for the majority of cases of CTCL, and is readily recognized by virtue of the prominent epitheliotropism of the T-cell infiltrates for epidermis and adnexal structures.1–4 Nonepitheliotropic CTCL (NE-CTCL) is an uncommon but well-recognized entity in dogs.2,5–7 The World Health Organization (WHO) classification of tumours of haematopoietic and lymphoid tissues considers most NE-CTCL as peripheral T-cell lymphomas (PTCL), unspecified.8 However, the extreme heterogeneity of diseases included in this classification is becoming increasingly recognized.9,10 In time, a number of specific lymphoma entities will be extracted from the PTCL unspecified category. For example, an indolent form of NE-CTCL was recently described in dogs.5 Also, primary cutaneous γδ T-cell lymphoma and two other provisional CTCL entities have been added to the WHO lymphoma classification for humans.9,11
In dogs, NE-CTCL are readily recognized when they occur as relatively homogeneous dermal infiltrates of large, cytologically atypical lymphocytes with clear CD3 expression revealed by immunohistochemistry.2 However, diagnosis of NE-CTCL is more difficult when the dermal infiltrate is heterogeneous with respect to cell lineage, or if CD3 expression by neoplastic T cells is diminished or absent. The hierarchy of pan-T-cell antigen loss in human PTCL is well documented.10,12 The goal of the present study was to present the morphological, immunohistochemical and molecular characteristics of canine NE-CTCL in which the neoplastic T cells are associated with a dominant pleocellular inflammatory response.
Materials and methods
Four dogs were presented to the Veterinary Medical Teaching Hospital of the University of California at Davis for evaluation and treatment of their lesions. The remaining 20 dogs were patients at private veterinary hospitals, and the original surgical biopsy specimens from skin were submitted to private pathology laboratories (IDEXX Laboratories, USA; ANTECH Diagnostics, USA; or VDx Veterinary Diagnostics, Davis, CA, USA). These latter cases were submitted for second opinions regarding the original diagnosis and/or for molecular diagnostic investigation to confirm suspected lymphoma. Surgical biopsy specimens of skin lesions were obtained from 24 canine patients. Multiple accessions were received for five dogs, and multiple skin sites were sampled from all but four dogs. Specimens consisted of formalin-fixed tissue samples, paraffin blocks or unstained slides. All specimens were obtained with the informed consent of the pet owners. Dogs had not received chemotherapy for their lesions prior to initial surgical biopsy sampling.
Accurate survival data were obtained for 19 of 24 dogs with inflamed NE-CTCL. The Kaplan-Meier survival curve and median survival time were estimated using the survfit function in R version 2.15.0 (R Development Core Team 2012; R: a language and environment for statistical computing; R Foundation for Statistical Computing, Vienna, Austria).
Tissue handling and immunohistochemistry
The tissues were collected over an 8 year period from 2004 until the present. Tissues were fixed in 10% neutral buffered formalin. Formalin-fixed tissue was embedded in paraffin, and 5-μm-thick sections were stained with haematoxylin and eosin (H&E). For immunohistochemistry, sections were stained by a streptavidin–horseradish peroxidase method according to the manufacturer’s instructions (Zymed, South San Francisco, CA, USA) using previously described methods.13 Monoclonal antibodies (MAb) or polyclonal antibodies specific for canine leukocyte antigens were applied to sections as diluted tissue culture supernatants. Negative controls consisted of substitution of specific MAb with isotype-matched irrelevant MAb (FE1.7B12, mouse IgG1, specific for feline CD4) or omission of the primary antibody.
Antibodies specific for canine leukocyte antigens included MAb specific for CD3 (CD3–12, rat IgG1, AbD Serotec, Kidlington, UK), CD79a (HM57, mouse IgG1, AbD Serotec, Kidlington, UK), CD18 (CA16.3C10, mouse IgG1), CD11d (CA18.3C6, mouse IgG1) and CD45 (CA12.10C12, mouse IgG1). Polyclonal antibodies specific for conserved epitopes on human CD20 (rabbit IgG; Lab Vision, Freemont, CA, USA) and human granzyme B (rabbit IgG; Spring Bioscience, Pleasanton, CA, USA) were used in some instances.
Lymphocyte antigen receptor gene rearrangement analysis
Genomic DNA was extracted from two (25-μm-thick) unmounted sections of formalin-fixed, paraffin-embedded lesional tissue using the manufacturer’s recommended protocol (DNeasy tissue kit; Qiagen, Valencia, CA, USA). Extracted genomic DNA was quantified spectrophotometrically using an Ultraspec 2100 pro spectrophotometer (Amersham Pharmacia Biotech, Uppsala, Sweden). All PCRs were performed on GeneAmp PCR system 2700 thermocycler or a Veriti thermocycler (Applied Biosystems, Foster City, CA, USA). Two primer sets were used to assess T-cell receptor γ locus (TRG) rearrangements. All dogs (except dog 5) were analysed with a simplex PCR primer set as previously described.14,15 Thirteen dogs (dogs 11–22 and dog 24) were analysed with a multiplex PCR primer set that detects all rearranged variable and joining genes.16 The PCR products were size separated by conventional polyacrylamide gel electrophoresis as previously described, or by capillary electrophoresis using an eGene HDA-GT12 capillary electrophoresis analyser (renamed QlAxcel; Qiagen). All PCRs were run in duplicate or triplicate. Duplicate or triplicate PCRs were used to assist distinction between true clonal samples and pseudoclonal samples. A sample was regarded as clonal if one or two sharp bands of the appropriate and same size were present in duplicate or triplicate samples run in adjacent lanes (conventional electrophoresis) or if clonal peaks were at least twice the height of the polyclonal background (capillary electrophoresis). If the bands differed in size in duplicate/triplicate samples, the sample was considered pseudoclonal.
Twenty-four dogs, 14 males and 10 females, with a mean age of 7.5 years (range 2–15 years) were included in this study. No breed predilection was observed. Demographic details are summarized in Table 1. Lesions presented as nonpruritic nodules, plaques or masses up to 4 cm in maximal dimension; some lesions were associated with alopecia, ulceration and crust formation. Lesions occurred throughout the skin. Frequently affected sites included the face (lips, nasal planum, eyelids), lower extremities (paws, interdigital folds), neck and trunk (Table 1 and Figure 1).
Treatments and clinical outcome are listed in Table 1. Owing to progression of the lesions, 11 dogs were euthanized and two died within 1–24 months. The median survival time was 9 months as determined by the Kaplan-Meier method (based on 19 dogs; Figure 2). A broad range of treatments were administered to the dogs with inflamed NE-CTCL (Table 1) and, as such, limits the value of the survival data in this retrospective study. Seven dogs are currently alive. Six of these are free of lesions (5 months to 3 years). The remaining dog (dog 22) was treated with Lomustine (CCNU) and developed new lesions within 2 months, and currently receives palliative treatment only. Clinical follow-up was not available for four dogs, and the date of euthanasia was not available for dog 11.