2 Paul C. Stromberg1 and Donald J. Meuten2 1The Ohio State University, USA 2North Carolina State University, USA Increases in the complexity of oncology patient management, growing sophistication of veterinary medicine, and elevated client expectations have combined to place more demands on veterinary pathologists than just making a diagnosis. What is a simple matter of tumor diagnosis in autopsies would, in many cases, be inadequate in the surgical pathology arena. Add to this the limited or often complete lack of communication between the clinician collecting the tumor tissue, the technician trimming it into cassettes, and the pathologist reading the cases and formulating the diagnosis and the potential for an unsatisfactory outcome is substantial. Although the most important item of information in any assessment of tumors is the diagnosis, in the surgical biopsy arena it is often equally important to note additional information that may impact the case management, such as tumor grade (if one exists) and completeness of the resection.1 To ensure a diagnosis is possible, representative tissue must be collected, properly preserved and processed. With small masses that fit into a standard 2 cm × 2.5 cm × 5 mm processing cassette, this is not a problem. However, many tumor specimens exceed this size and must be trimmed to fit within those dimensions. It is this critical step which ensures that both diagnostic tissue and the surgical margins are included in the biopsy and properly oriented in the cassette so that the pathologist can complete the evaluation, make a diagnosis and accurately assess the surgical margin so the clinician can formulate a prognosis and the appropriate treatment plan. The surgical margin can be defined and reported in several ways. The goal is to provide the clinicians with the information they need to decide what, if any, additional treatments will be recommended. In general, the likelihood of getting a diagnosis from a biopsy is directly proportional to the amount of the tumor tissue evaluated. It is highest when 100% of the tumor can be examined. The optimal specimen is one that fits into a single or several tissue‐processing cassettes, permitting evaluation of almost all the tissue in the sample. This is common in small skin tumors. But often the entire tumor cannot be sampled by the surgeon or trimmed in by the pathologist. Tumor sampling by the surgeon can be divided into those that completely remove the entire mass (excisional biopsy) (Figure 2.1) and those that remove only a portion of the mass (incisional biopsy). The choice is the surgeon’s and is often driven by the first rule of medicine (“First do no harm”), client preferences, and cost. Complete removal (total excisional biopsy) is the most desirable option because both diagnosis and treatment can be accomplished in one procedure. A desirable metric of excisional biopsies is a margin evaluation to assess the completeness of the excision. When tumor location and patient concerns dictate only a portion of the mass can be sampled, an incisional biopsy (Figure 2.2) is collected so a diagnosis can be obtained in advance of the treatment decision. Incisional biopsies collect a variable percentage of the tumor depending on the collection technique. Figure 2.1 Excisional biopsy. The entire mass is removed by the biopsy. Margin evaluation can be performed. Figure 2.2 Incisional biopsy. Only a portion of the mass is removed by the biopsy. Margin evaluation is not possible. Minimally invasive techniques are employed to collect a small sample and obtain a diagnosis. Endoscopic and Tru‐cut needle biopsies generally collect a very small percentage of the tumor. Laparoscopic wedges and punch biopsies collect more and subtotal wedges often provide a substantial amount of tissue. Minimally invasive procedures are also used when a discrete mass cannot be visualized. The goal of these incisional techniques is diagnosis alone with no concern about margins. The disadvantage is that sometimes the sample does not contain diagnostic tissue. This commonly occurs in tumors possessing substantial areas of necrosis, inflammation, reactive stroma, or collection artifacts. The small size of the sample compounds the effect each of these has on the diagnosis. Samples in which no lesion is visible because of artifact or inadequate diagnostic tissue are termed “Nondiagnostic biopsies.” When very small percentages of a tumor are collected, the risk of a nondiagnostic biopsy is highest. In all of these cases, diagnosis is the most important outcome of the biopsy. By definition, margin evaluation is not an issue in these samples because complete removal was not attempted. Specimen trimming and orientation in the cassette are less important because of the limitations associated with tiny specimens. Indeed, most endoscopic samples are too small for further manipulation. But these samples must be protected from damage or loss during processing. Endoscopic and needle specimens should be processed in small screened containers such as Cell Safe™ cassettes that fit right into the standard processing cassettes and protect delicate tissues from damage and loss (Figure 2.3). Figure 2.3 Cell Safe™ cassette containing endoscopic intestinal samples. Samples this small are difficult to orient. Tissues in FFPE block may need to be rotated to obtain correct orientation. Large tumors present a different set of problems. The entire tumor or organ containing the tumor (spleen, lung lobe, limb amputation) may be completely removed by the surgeon and submitted for evaluation and diagnosis. Although the pathologist has 100% of the tumor it is often far too large to evaluate all or even a substantial portion of it. The pathologist must decide what portions to trim and examine. Because the surgeon desires a margin evaluation in addition to the diagnosis, the specimens must be trimmed and oriented in the tissue cassettes so as to ensure the most accurate assessment of the margin possible. In many cases, the diagnosis has already been made on a previous small incisional or needle specimen so completeness of excision is now the important information desired from the biopsy procedure. This is a very different issue from trimming tumors from autopsies. What and how many portions of the tumor should be trimmed and how they are faced in the tissue‐processing cassette becomes critical to ensure that diagnostic tissue is available on the slides and that the correct margins are identified, included, and oriented so as to permit the most complete evaluation of the surgical margin possible. This is a critical step. It is subjective and will vary somewhat with the pathologist or technician. If treatment will go beyond the present excision then margin evaluation will help the oncologist decide what treatment plan is optimal for the pet and its owner. In such cases the trimming technician should consult with the pathologist for assistance in proper selection and orientation. The most definitive indication of the surgical margin is made by the surgeon on the specimen at the time of collection.2 Measurement of the margins taken by the surgeon before, during, and immediately after removal of the mass will be greater than the histologic margin reported by the pathologist. Shrinkage of specimens can be up to 40% and the majority of this occurs immediately after removal and before fixation.3,4 Although suture tags have been a traditional method, inking or dyeing the margin of interest is the superior method because the ink is applied by the surgeon who knows which margins are to be evaluated. The ink is visible to the pathologist at trimming and while reading the finished slide. This is a form of direct communication between the surgeon and pathologist. If properly applied by the surgeon, the pathologist can be confident that the inked margin depicts the real margin(s) of interest. The surgeon can color code for regions they are most concerned about. If the surgeon trimmed off fat and or connective tissue around the tumor then this needs to be reported. Applying ink to the regions after adipose or fascial layers are trimmed off may be unnecessary or misleading. Although tissue can be inked by the trimmer during preparation, they must translate the clinician’s notes or other form of designation, which may be imprecise and so create doubt in the pathologist’s mind about the accuracy of the designation. Inking can be done before or after formalin fixation. Surgical ink is commercially available in multiple colors. Black, green, and yellow colors are optimal because they provide high contrast to the blue and orange colors of H&E‐stained slides. The tissues should be blotted dry and inked with a cotton swab (rather than immersed in the ink) and then allowed to dry for 5–10 minutes. Alternately it can be rinsed with a dilute glacial acid solution to set the ink. A specimen can be painted with one or more colors if needed. When properly inked, the surgical margin can be easily seen in the histological section by the pathologist and the relationship of the margin to the tumor assessed and measured. Suture tags can establish a margin but are not observed in the finished slide. Whether the surgical margin is inked by the surgeon or the tissue trimmer, the process should be noted on the referral form, clearly indicating which color indicates which margin. Inking establishes a clear line of communication between the surgeon and the laboratory about precisely which is the real margin(s) of interest and guides the tissue trimmer about how to orient the specimen for placement in the processing cassette. Specimens that are not inked have a substantial probability of incorrect margin assessment. Once the true surgical margin has been established and marked, the trimmer orients the tissue in the processing cassette so that the margins can be properly evaluated by the pathologist. The margin principle states that to adequately assess the completeness of the resection you should “examine as much surface area on the margin as possible.” If you can examine 100% of the margin (and the client can afford this many sections), you should do so. However, this is impractical and common sense combined with a few principles will determine how many sections need to be trimmed. In general, two orientations that produce two to four sections are commonly used and their use is dictated by the conditions around each case. Also called radial sectioning, in this orientation the tissue is cut perpendicular to the surgeon’s plane of section (Figure 2.4A). This is the conventional orientation employed most commonly and familiar to pathologists. The advantage of radial sectioning is that the relationship between the tumor and the surgical margin can be observed and the histologic distance measured or estimated (Figure 2.4B). The disadvantage of this orientation is that very little of the circumference of the margin is actually examined; just the 2–5 µm thickness of the tissue section in that radius of the tumor and therefore the potential for false‐negative rates is high. Because of this, some feel this is the least favored method for margin evaluation.1 The percentage of margin area can be increased by serially sectioning or “bologna slicing” (“bread loafing”) the specimen, but even this will only examine 1–5% of the margin (Figure 2.5), which could be prohibitively expensive in large specimens for many veterinary clients. A variation of this makes two perpendicular cuts through the specimen at right angles (“points of the compass” cut) so that the lateral margin in four directions can be assessed (Figure 2.6A). Although there is only a trivial increase in surface area examined, the pathologist can get some sense of the symmetry of the tumor and its relationship to the lateral margins in four directions. In large specimens these cuts may not include the tumor and the margin in the trimmed section so the margin width cannot be observed on the glass slide (Figure 2.6B). Figure 2.4 (A) Perpendicular margin. The tissue is cut perpendicular to the surgeon’s plane of cut. (B) When placed on the glass slide, the histologic tumor‐free margin (HTFM) and mass can both be visualized and the distance between edge of the tumor and non‐neoplastic tissue estimated or measured and reported as M1–M4. Figure 2.5 Perpendicular margin, pictogram. Serial sections or “bread/bologna slicing” of the entire mass. This technique will assess approximately 1–5% of the circumferential margin depending on the size of the tumor and the number of sections taken at specified intervals, none of which is standardized for animal tumors. Figure 2.6 (A) Perpendicular margin. Two cuts are made at right angles to evaluate five components of the margin: four lateral (“points of the compass” cut) and the ventral margin. (B) Perpendicular margin. Modified “points of the compass” cut in an oversized mass to evaluate the margin in five directions. In this modification the specimen is too large to include the margin and mass in the same section so the histologic tumor‐free margin (HTFM) cannot be observed directly and measured or estimated on the glass slide. These type of sections can be reported as M4, >5 mm. Standard histopathology cassettes are 3.0 cm × 2.5 cm. Also called tangential, “orange peel” or “shave” margins, in this orientation the tissue is cut parallel to the surgeon’s plane of section. The principal advantage is that the amount of surface area evaluated can be hundreds of times greater than in a perpendicular margin. In some cases, such as digits, ear canal ablations, bowel resections, tail amputations, etc., nearly 100% of the surface area can be examined, fulfilling the criteria for optimal margin evaluation (Figure 2.7A). The principal disadvantage of this is that the relationship between the tumor and the margin is lost from the slide and the margin width cannot be measured (Figure 2.7B). This can be critical, especially when looking at round cell tumors and soft tissue sarcomas. A second disadvantage is that in large specimens, such as limb amputations, the cost of trimming and evaluating the entire en face margin could be prohibitive and impractical. Figure 2.7 (A) Parallel or en face margin, pictogram. The tissue is cut parallel to the surgeon’s plane of cut. This technique prioritizes examination of the outermost tissues supplied by the surgeon. A second or third section must be obtained from the tumor to establish the diagnosis. The cuts (sections) in this pictogram are made in a vertical plane; en face sections in a horizontal plane are seldom done in animal tumors but will yield tumor and peripheral margin until the deep margin is reached. (B) Parallel or en face margin placed on the glass slide. More surface(s) on the margin are examined but the relationship between the mass and the margin is not visible on the slide. It is important for the tissue trimmer to note what type of margin orientation was faced in the cassette because it may not be apparent from just looking at the slide. En face margins should be inked by the trimmer. The presence of any tumor cells on an en face margin indicates an incomplete resection. Large, complicated specimens may present a combination of both types of trim margins. The most effective way to communicate this is to draw a trimming diagram on the submission form indicating the shape of the surgical specimen, the direction and type of trim margins, and in which cassette each was placed. This communication between trimmer and pathologist is critical to ensuring the surgical margins are correctly interpreted. A modification is to cut on a horizontal plane rather than perpendicular plane. This will yield tumor and lateral margins but junction of the tumor and deep margin may be lost depending on the number of sections sampled. This is the principle of the Mohs technique, which can evaluate 100% of the surgical margin.5,6 Complete circumferential peripheral and deep margin assessment allows examination of the entire surgical margin and has a lower false‐negative rate than “bread loafing.” In human medicine this is employed on frozen sections of basal cell tumors while the patient and surgeon wait for results. The object of trimming neoplasms is to obtain a diagnosis and to maximize the accuracy of margin evaluation so that completeness of excision can be determined. The histologic diagnosis is the most important factor to influence treatment plans and provide a prognosis. For many tumors the diagnosis predicts outcome (e.g., sebaceous adenoma, fibroma, circumanal gland adenoma, histiocytoma, osteosarcoma, urothelial cell carcinoma, beta cell neoplasms, etc.). Margin evaluation, grading, and mitotic count (MC) are not needed for these tumors or objective data that correlates margins, MC, or grades with outcome is not known. In a recent survey, margin evaluation was considered the most important component of the biopsy report, presumably after the diagnosis.1 The true margin between tumor and non‐neoplastic tissues is in the patient. The surgical margin is created by the surgeon and the histologic margin is created by trimming, paraffin embedding, and histological sectioning. The margin of neoplastic versus non‐neoplastic tissues in vivo can only be extrapolated via assessment of a specimen that has been removed from the body, fixed in formalin, embedded in paraffin (FFPE), and sectioned for microscopy. Tissues retract or shrink significantly following surgical excision.3,7–9 Additionally, only a small percentage of the histological margins are assessed. If “bread loaf”, perpendicular vertical cuts are used it is estimated that 1–5% of the margin is examined depending on the size of the lesion and the number of sections removed.5,6 Despite the inherent inaccuracies, the determination of the “margins” during routine histological examination may influence treatment plans or the prognosis offered to owners. There is some shrinkage caused by retraction of elastic tissues immediately after removal.3 Studies of oral and cervical carcinomas in humans have shown significant shrinkage (30–50%) can be caused by formaldehyde fixation and tissue processing.7–9 Shrinkage in length and width of 20–30% and increase in thickness up to 75% in normal dog skin caused by processing has been documented.10 The presence of muscle blunted this affect whereas fascia did not. There were different patterns seen in different anatomic locations of the body. Other reports have indicated the majority of tissue shrinkage occurs before fixation in formalin and is in the range of 20% for length, 10–15% for width and area.3 Some have proposed pinning the resected specimen to a solid object to reduce shrinkage. These factors explain the differences in the tumor margin described by the surgeon versus the margins reported by pathologists. The width of the margin between tumor and non‐neoplastic tissue is greater in vivo than that observed in histologic sections. Studies to compare the in vivo margin to the histologic margin in animal tumors and tissues are reported.4,10,11 Retraction and shrinkage up to 40% occur immediately post removal4,11 and likely will vary between tissues and possibly among species or even tumors, adding to the difficulty in arriving at a consensus on what constitutes a “safe margin.” Regardless, the true margin is in the patient and with present techniques we do not know what that is. We can only provide an estimate. It is likely that other parameters are more predictive of recurrence, disease‐free interval, and survival times than the margin width, no matter how carefully it is established or defined. Most important are the inherent biologic characteristics of the tumor and the host. We try to extrapolate tumor behavior from traditional histological features such as diagnosis, invasion, anaplasia, and MC. Immunohistochemical determination of proliferation indices (Ki67, PCNA, AgNOR) and gene expressions or mutations may also help predict the behavior of certain tumors. Molecular characterization of tumors is used in human medicine to predict prognosis and select treatments. These tools should prove useful if we are asked by an oncologist for information that cannot be obtained from H&E sections. It is likely that these characteristics will vary from tumor to tumor. For instance the MC for some tumors14,15 may have prognostic significance (mast cell tumor) while for others it does not (histiocytoma). Evidence‐based medicine is needed for each type of tumor to identify what the important prognostic characters are. However, for many cases in veterinary medicine the diagnosis and the information that can be provided from H&E sections or cytology are sufficient. For the diagnostic pathologist communication from the clinical veterinarian is essential; if an owner does not want to consider treatment beyond a primary surgery then applying costly and time‐consuming techniques are not justified. Many techniques can be applied to archive material should this information be desired at a later time. Applying new techniques to samples of tumors is relatively fast compared with gathering accurate follow‐up data over years. Too often our follow‐up data are based on imprecise clinical assessments of recurrence (visualization, palpation) and nonstandardized assessment of metastases via imaging, survival and cause‐of‐death data based on phone interviews with owners or primary care veterinarians. We can retrieve archived material if a new technique is developed but we need follow‐up data (outcome assessment) if we want to determine whether the technique has clinical applicability. This requires partnerships between clinicians, molecular biologists, and pathologists. Methods that help determine the biology of the tumor and the host may prove better predictors of prognosis12,13 and theranostics than changing how we enumerate margins and MCs.14 In veterinary medicine, margin assessment has been predominantly applied to cutaneous neoplasms. The premise of measuring margin widths is simple: by determining if tumor cells are present in histologic sections at the margin or how far they are from the margin we should be able to predict local recurrence. The true distance between the tumor and normal tissues is in the patient and that is not known. The distance visualized in histologic sections between the tumor and the inked margin is the histologic tumor‐free margin (HTFM). If desired it can be reported for cranial, caudal, and both lateral and deep margins. In human oncology there is emphasis to determine the HTFM width that will prevent local recurrence. The closest histologic margin to a tumor is referred to as the histologic safety margin (HSM).5,16 The HSM is considered the microscopic distance that will prevent local recurrence. HSM is not known for animal tumors, and it is more appropriate to report HTFM because HSM implies we know the distance that will prevent local recurrence. Determination of the HSM in certain human tumors is accomplished by the Mohs technique which can assess 100% of the margin.5,6 Some refer to the acronym, CCPDMA – complete circumferential peripheral and deep margin assessment. CCPDMA allows complete examination of the entire surgical margin and therefore has a lower false‐negative rate than “bread loafing” in a perpendicular plane, which only provides a few slides from selected areas. The Mohs surgical technique is essentially serial frozen sections taken in a horizontal plane while the patient and surgeon wait for results. Sometimes the surgeon also examines the frozen sections. The advantages are immediate knowledge of margin, 100% assessment of surgical margins and maximum preservation of normal tissues. The latter is important for tumors like basal cell carcinomas located on the head, eyelids, and other sites where the preservation of tissue is highly desirable.5 Some disadvantages are the high costs (>US$10,000), frozen sections, large tumors, deeply infiltrative neoplasms, and expertise needed. In human medicine, Mohs micrographic surgery is performed by board‐certified individuals. It is used for dermatologic specimens, when no metastases or micrometastases are expected and preservation of normal tissues is needed. However, its use and advantages have also been reported for tumors in a limited number of animals, most notably when 100% assessment of the margin is desired.6 That degree of certainty has a cost that may not be supported by veterinary medicine. The HSM differs between tumors or within subtypes of the same tumor, presumably because of differences in how the tumor grows, whether infiltrative or discrete, aggressive, biologic behavior, etc. There are numerous factors that contribute to local recurrence or metastases that go beyond the HSM. Even the type of anesthesia used for the surgical removal of tumors has been associated with increased risk of recurrence and metastases.17 The biology of the tumor and the host are likely more important determinants of recurrence than how close tumor cells (microscopic assessment) are to the inked margins. Until recently there has been little research in veterinary medicine about this and no standardization of how to trim tumors, measure margins or report the results.2,4,6.10,16 The distance of the closest approach of the tumor cells to the histologic edge (ink) should be measured or estimated in the field of vision. This can be assessed along the cranial, caudal, both lateral and deep margins and reported as the HTFM. HTFM should be reported and not the tumor‐free margin. The HTFM varies among the histologic sections examined, is influenced by the pattern of tumor growth (discrete, infiltrative), the type of excision performed, the trimming protocol and total number of sections evaluated. The total number of sections examined is not standardized. HSM has not been determined for animal tumors, and for canine mast cell tumor (MCT) it could not be determined.16 In some human breast cancer studies the standard is “the tumor should not touch the margin ink.”18 In veterinary oncology some have proposed defined distances for canine MCT and soft tissue sarcoma (STS) from 2 mm to 5 cm and/or one fascial plane deep.19–22 Given the variable nature of the periphery, the asymmetry of some tumors, the minimal amount of the total margin actually examined, and shrinkage induced by removal and fixation we recommend pathologists report the HTFM as a range: To measure the HTFM more precisely (e.g., “The margin width is 2.6 mm”) implies a greater precision than is real from an FFPE specimen and that the entire margin was examined. The shortest microscopic distance to a margin should be reported, along with any specific regions requested by the surgeon. Gross measurement of margins should be taken by the clinician during surgery, if this is desired, and not performed on a fixed specimen. Immediately at or after removal there is shrinkage of the specimen. For tumors that are M1 or M2 the report should also approximate the histological amount of tumor present at or near the margin: focal: few foci of tumor cells are seen or diffuse: large number of tumor cells present. Many of us do not use a micrometer, grid, or reticule to determine measurement(s) and estimate a distance based on prior determination of how many millimeters are present in the field of view of our objectives. This can be determined with a stage micrometer, a ruler or by this simple calculation:
Trimming Tumors for Diagnosis and Prognosis
Specimen sizing
Orientation of trimmed specimens
Perpendicular margins
Parallel margins (en face)
Margin evaluation
Skin tumors
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
