Scar revision for incompletely or narrowly excised cutaneous mast cell tumors in dogs

Abstract

Objective

Describe clinical features and outcomes of dogs undergoing scar revision for incompletely or narrowly excised cutaneous mast cell tumors without gross disease.

Animals

52 dogs undergoing 54 scar revisions.

Procedures

Retrospective record review with information collected on signalment, tumor type/location, pre-surgical diagnostics, surgical and pathologic findings for the first excision and scar revision surgeries, and follow-up. Descriptive statistics were generated.

Results

Prior to initial excision, cytology was performed on 38.9% (21/54) of tumors and the initial surgery report rarely described surgical resection margins [14.6% (7/48) of surgeries]. Residual tumor was identified pathologically in 29.6% (16/54) of scars. Local recurrence following scar revision occurred in 3.7% (2/54) of all scars [median follow-up 881.5 d (range: 0 to 3317 d)]; both scars had either complete excision of residual mast cell tumor or no evidence of mast cell tumor on scar revision and neither dog received radiation therapy.

Conclusions

Identification of residual disease was uncommon, and local recurrence was less common than previously reported for incomplete/narrow mast cell tumor excision.

Clinical relevance

Scar revision of unplanned primary excisions appears to yield a high likelihood of durable tumor remission in patients. First opinion practitioners are encouraged to avoid unplanned excisions by informing the surgical plan with cytology of the primary tumor and detailing surgical margin excision which may facilitate revision, if necessary.

Introduction

Mast cell tumor (MCT) is the most common tumor of the skin in dogs, with a reported prevalence of 16 to 21% for all canine skin tumors (1). Surgical excision of the primary tumor is the recommended treatment, and in the case of low-grade tumors without metastasis, complete surgical excision may be curative. For surgical excision, wide surgical margins are recommended, including circumferential margins of 2 cm and deep margins of one uninvolved fascial plane (1,2). Marginal resection without wide circumferential and deep margins is highly likely to lead to incomplete excision of the tumor due to microscopic extensions of the tumor (3). A common location for MCTs is the limbs and up to 28% of MCTs occur in this location (4), which likely increases the prevalence of incomplete margins due to lack of skin available for closure following a wide surgical resection. Local recurrence after incomplete resection in dogs with MCT is reported in 23 to 38% of dogs and studies have shown that recurrence is more common with higher grade tumors, even with complete excision (5–7).

Although surgical margins of 2 cm circumferentially and one deep fascial plane are recommended, surgical margins and histologic margins do not always correlate due to tissue shrinkage (8). Histologic margin status is broadly categorized into complete or incomplete; but within the complete margin group, an additional classification of narrow is often discussed depending on the distance of the margin to the tumor. In human medicine, the TNM Residual Tumor Classification system is used and 3 categories are recognized: no residual tumor (R0), microscopic residual tumor (R1), and macroscopic residual tumor (R2) (9). There is no consensus on the cut-off for narrow versus complete margins for canine cutaneous MCTs. Although guidelines have recommended against the use of the terms clean, dirty, close, and narrow in pathology reports in favor of reporting a numerical margin distance (10), use of these terms persists in the literature. Previous veterinary studies have defined these terms with a numerical margin length with complete defined as > 0 mm (3,6,11,12), whereas other studies included margins of < 1 mm (3,5,12), < 2 mm (11), or < 3 mm as narrow (6,13). Other groups have defined incomplete as any margin < 1 mm (5,13). This lack of consensus makes it difficult to apply results of these studies to patients as they frequently have different definitions of narrow margins.

One group reported a higher rate of local recurrence with incompletely and narrowly excised MCTs (67% and 50%, respectively) compared to completely excised MCTs (9%), and another group identified a local recurrence rate of 38% for narrowly excised MCTs (6,11). Because local recurrence is also considered a negative prognostic indicator (6,14), scar revision surgery is often recommended when incomplete or narrow margins are identified after resection of MCTs, with the intent of obtaining a microscopically tumor-free surgical margin. Despite this recommendation, other studies have shown that the histologic tumor free margin was not a factor for recurrence (7). Grade has been shown to be a strong predictor of recurrence with lower grade tumors being less likely to recur, independent of margin status (7,15). The decision to proceed with scar revision is complicated, particularly with low-grade tumors. However, scar revision is frequently recommended in the authors’ practices in the event of narrow excision, regardless of tumor grade. When surgical revision is not possible, radiation therapy can be used to treat microscopic disease (1,6). In some cases, active surveillance is chosen to monitor the site for recurrence so that any tumor recurrence is identified early when additional treatment may remain an option.

There are 2 prior studies which reported residual MCT in 27% and 48% of scar revision specimens (6,13). These findings are quite disparate, with one study finding almost twice the prevalence of residual MCT than the other study. The authors of the present study subjectively appreciate that it is uncommon to identify residual tumor in the scar revision tissues after scar revision for incompletely or narrowly excised MCT in the absence of gross disease. The objective of this study was to describe clinical features and outcomes of dogs undergoing scar revision for incompletely or narrowly excised MCTs without gross disease. One focus of this was to determine the rate of residual MCT identification in the scar revision tissues to add further data to the currently conflicting literature. The hypothesis was that neoplastic cells would be identified on scar revision in less than 33% of cases.

Materials and methods

A retrospective medical record review was performed at 2 veterinary teaching hospitals (University of Georgia and Oregon State University) to identify dogs undergoing a scar revision for a previously incompletely or narrowly excised cutaneous MCTs. All dogs had an initial excision performed in first opinion practice. Incomplete excision was defined by a histologic tumor-free margin (circumferential and/or deep margin) of 0 mm. Narrow excision was defined by a histologic tumor-free margin (circumferential and/or deep) of ≤ 3 mm for Grade II MCT and ≤ 5 mm for Grade III MCT. Dogs were excluded if gross disease was present at the scar, if the MCT was determined to be subcutaneous in origin on pathologic evaluation, if a scar revision had been previously performed, if the original excision was not performed in first-opinion practice, or if the pathology report from the initial excision was unavailable. Data collected from the medical record included signalment, tumor grade [Patnaik and Kiupel (16,17)], tumor location (head/neck, limb, trunk), diagnostics performed prior to the first excision, date of first excision, surgical information and pathologic findings related to the first excision, date of scar revision, staging results, length of the scar, surgical information and pathologic findings related to the scar revision, adjuvant therapy, date of last follow-up, and presence of recurrence at time of last follow-up. When multiple tumor-free margin lengths were reported, the narrowest margin was chosen for analysis. Local recurrence was suspected when a mass was present at the site of the scar and was confirmed when cytologic or histopathologic evaluation was consistent with MCT. Follow-up was obtained by evaluation of records from the primary care veterinarian or by conversation with the owner.

Data were tested for normality by visual inspection of the normal quantile plot. Normal data are expressed as mean ± SD, non-normal data are expressed as median (range).

Results

Fifty-two dogs with 54 scar revisions were included. Mean age at time of scar revision was 7.2 ± 3.0 y. There were 34 females (33 spayed, 1 intact) and 18 males (16 neutered, 2 intact). The most represented breeds were boxer (8/52, 15.4%), Labrador retriever (8/52, 15.4%), and mixed breed dog (7/52, 13.5%). Mean body weight was 23.7 ± 13.2 kg. Tumor locations were trunk (23/54, 42.6%), limb (23/54, 42.6%), and head/neck (9/54, 16.7%). The tumor was reported to be present for a median of 70 d (range: 7 to 728 d) prior to the first excision. Mass size at the time of initial excision was reported for 31 masses (31/54, 57.4%) with a median diameter of 1.8 cm (range: 0.3 to 15 cm). Cytology was performed on 21 tumors (21/54, 38.9%) prior to the first excision, with results being indicative of MCT in 14 dogs (14/21, 66.7%), concerning for MCT in 3 dogs (3/21, 14.3%), and non-diagnostic, inconclusive, concerning for sarcoma, and report unavailable in 1 dog each (1/21, 4.8%).

Initial excision

Medical records were available for review for 48 initial excision surgeries (48/54, 88.9%). Surgical margins were not reported for 41 initial excision surgeries (41/48, 85.4%). In 4 initial excision surgeries (4/48, 8.3%), only deep margins were reported and in 3 initial excision surgeries (3/48, 6.3%), both circumferential and deep surgical margins were reported. Circumferential surgical margins were 2 cm (n = 1) and 0 cm (n = 2). Deep surgical margins were subcutaneous tissue (n = 3), fascia (n = 1), or reported as 0 cm (n = 3).

The pathology report from the initial excision was available in all dogs. All tumors were assigned a grade on the Patnaik or Kiupel system (Table 1). Circumferential pathologic margins were reported for 50 initial excisions (50/54, 92.6%), and deep pathologic margins were reported for 52 initial excisions (52/54, 96.3%), with all initial excisions having at least 1 margin type reported (Table 2). When considered together, 34 dogs (34/54, 63.0%) had at least 1 incomplete margin and 20 dogs (20/54, 37.0%) had at least 1 narrow margin.

Table 1.

Grades for 54 mast cell tumors surgically excised from 52 dogs according to the Patnaik system (n = 54) (16), Kiupel system (n = 39) (17), or both (n = 39) based on the initial excision.

Patnaik grade		Kiupel grade		Combined grades
I	0 (0)	Low	29 (74.4)	II/low	28 (71.8)
II	44 (81.5)	High	10 (25.6)	II/high	4 (10.3)
III	10 (18.5)			III/low	1 (2.6)
				III/high	6 (15.4)

Data are reported as n (%).

Table 2.

Results of circumferential and deep margins as assessed by pathology for completeness of excision in 54 dogs after initial mast cell tumor excision.

Circumferential margins	Deep margins	Number (%)
Complete	Narrow	2 (3.7)
	Incomplete	3 (5.6)
Narrow	Complete	4 (7.4)
	Narrow	13 (24.1)
	Incomplete	1 (1.9)
	Not reported	1 (1.9)
Incomplete	Incomplete	24 (44.4)
	Narrow	1 (1.9)
	Not reported	1 (1.9)
Not reported	Incomplete	4 (7.4)

Scar revision

Length of time between the initial excision and the scar revision was a median of 28 d (range: 6 to 232 d). Thoracic imaging was performed in 15 dogs and abdominal imaging in 45 dogs with no evidence of metastasis in any dog. Regional lymph nodes were evaluated for metastasis via cytology (33 scars) and histopathology (26 scars), with 21 scars having both cytologic and histopathologic evaluation of the regional lymph node. Cytologic results were negative (20 lymph nodes), positive (11 lymph nodes), concerning (1 lymph node), or inconclusive (1 lymph node) for metastasis. Histopathology results were positive for mast cells (19 lymph nodes), negative for metastasis (6 lymph nodes), or inconclusive (1 lymph node). In lymph nodes that contained mast cells on histopathologic evaluation, final diagnosis was HN0 (n = 2), HN1 (n = 3), HN2 (n = 11), and HN3 (n = 3). Scar length was reported in 33 scars (33/54, 61.1%) with a mean length of 4.5 ± 2.4 cm.

For the scar revision surgery, circumferential surgical margins were described in 45 surgeries (45/54, 83.3%), with a median of 2 cm (range: 0.5 to 3 cm) and margins of 3 cm (12/45, 26.7%), 2.5 cm (3/45, 6.7%), 2 cm (15/45, 33.3%), 1.5 cm (4/45, 8.9%), 1 cm (9/45, 20.0%), and 0.5 cm (2/45,4.4%). The deep surgical margin was reported in 48 scar revision surgeries (48/54, 88.9%) and was fascia or the entire compartment (42/48, 87.5%) or subcutaneous tissue (6/48, 12.5%).

Residual disease was identified in 16 scars (16/54, 29.6%), and no residual disease was identified in 38 scars (38/54, 70.4%). Correlation between findings of the initial excision and the presence of residual disease on scar revision pathology were evaluated (Table 3). For the 16 scars with residual MCT, circumferential margins in the scar revision pathology report were evaluated as incomplete (4/16, 25.0%), complete (10/16, 62.5%), or not reported (2/16, 12.5%). Deep margins in the scar revision pathology report were evaluated as incomplete (4/16, 25.0%), narrow (1/16, 6.3%), complete (8/16, 50.0%), or not reported (3/16, 18.8%). Seven dogs (7/16, 43.8%) had complete circumferential and deep margins, 6 dogs (6/16, 37.5%) had at least 1 incomplete margin, 2 dogs (2/16, 12.5%) had no reported margins, and 1 dog (1/16, 6.3%) had complete circumferential margins with unreported deep margins. A Patnaik grade was assigned to the scar revision specimen in 9 scars. A grade consistent with the grade of the first excision was reported in 7 scars (all Grade II), a grade that was higher than the first excision was reported in 1 scar (Grades II to III), and a grade lower than that of the first excision was reported in 1 scar (Grades III to II). With the Kiupel system, a grade was assigned for 6 scar revision specimens, the grade was consistent with the first excision in 5 scars (all low grade), and the sixth scar had not been previously graded on the Kiupel system.

Table 3.

Correlation between presence of mast cell tumor or no evidence of disease on scar revision with the margins from the original excision.

Initial excision margin status	Scar revision pathology findings	Number (%)
Incomplete (34 scars)	MCT	14 (41.2%)
	NED	20 (58.8%)
Narrow (20 scars)	MCT	2 (10.0%)
	NED	18 (90.0%)

MCT — Mast cell tumor; NED — No evidence of disease.

Follow-up

Median length of follow-up from the date of scar revision for all dogs was 881.5 d (range: 0 to 3317 d). Thirty scars (30/54, 55.6%) had follow-up > 2 y (median: 1316 d, range: 847 to 3317 d), 10 scars (10/54, 18.5%) had follow-up < 2 y but > 1 y (median: 499 d, range: 373 to 785 d), and 14 scars (14/54, 25.9%) had follow-up < 1 y (median: 142 d, range: 0 to 331 d). Twenty-four dogs received adjunctive chemotherapy, and 4 dogs received radiation therapy of the scar following scar revision. In 3 dogs receiving radiation therapy, the scar revision had incomplete margins, whereas in 1 dog there was concern over a deep tumor-free margin of 3 mm for a Grade II tumor. No dogs receiving radiation therapy developed recurrence with follow-up times of 666, 976, 1022, and 1316 d. In scars that did not receive adjunctive radiation therapy (n = 50), 2 scars developed local tumor recurrence (2/50, 4.0%). The overall recurrence rate was 3.7% (2/54). Both dogs with recurrence had received vinblastine and prednisone following scar revision. In 1 dog, the MCT recurred at 217 d; this MCT was graded as a Grade III on the initial excision and a Grade II, completely excised on the scar revision. In the second dog, recurrence of a Grade II MCT was confirmed via pathology at 463 d post-scar revision. The scar revision pathology report had revealed no evidence of residual MCT in this dog.

Discussion

Residual disease was discovered in the scar in 29.6% of dogs; therefore, the hypothesis that < 33% of dogs would have residual disease was accepted. The authors often inform owners that histopathologic evaluation of the scar revision tissues may not identify tumor cells, which was confirmed in this study. Lack of identification of tumor cells in the scar revision tissues may be due to a combination of factors, including low residual disease volume and method of pathologic evaluation. Although recommended guidelines for veterinary pathologic evaluation have been published (10), a consensus on an ideal technique for evaluating margins was not reached. Radial sectioning evaluates a very small amount of the specimen, so combining this technique with parallel or tangential sectioning is recommended to increase the amount of tissue evaluated for margin assessment, although these techniques increase the time and cost associated with histopathologic analysis of the specimen. The method of sectioning used in this study was not evaluated and likely varied among pathologists. In addition, the rate of identification of residual disease was lower in dogs with a narrow margin from the initial excision (10.0%) compared to those with an incomplete margin from the initial excision (41.2%). In a recent review on margin assessment in veterinary medicine, the treatment of narrow and incomplete excisions as being the same was questioned (18). The lack of consensus on margins in canine MCT makes it difficult to interpret the difference between incomplete and narrow margins across studies, as some studies report a high rate of recurrence with narrowly excised tumors (38 to 67%) (6,11). A prospective, controlled study with tumors of multiple grades would be required to determine if narrowly excised tumors behave like completely or incompletely excised tumors.

Local tumor recurrence rate was 4.0% (2/50) for dogs that did not receive adjuvant radiation therapy after scar revision. For the 2 dogs with tumor recurrence, scar revision resulted in complete excision in 1 dog and in the other dog, the pathologist was unable to identify residual MCT in the scar. Although no evidence of neoplasia was seen, this scar experienced MCT recurrence; thus scar revision may still be useful to decrease disease burden even if neoplastic mast cells are not identified on histopathology. It is also important to recommend active surveillance of the scar even when apparently complete margins are obtained, as recurrence is still possible because the entire resection margin is not evaluated histologically. Although completeness of resection is a known risk factor for recurrence (11), the impact of grade on tumor recurrence must also be considered. Increasing grade has been shown to increase the recurrence rates for MCTs (6), and this is one reason the definition of a narrow excision was different for Grade III tumors in this study. It is likely a combination of tumor grade and completeness of resection that affect risk of tumor recurrence. Only a single residual tumor cell is required for recurrence, although grade likely influences the speed of recurrence with higher grade tumors recurring more quickly than lower grade tumors even with a complete resection (7). One group reported that aspiration of the surgical site scar in the case of incomplete resection could provide information on the risk of local recurrence with 89% accuracy (19). This could be considered to help guide the decision to pursue scar revision or active surveillance in the case of an incompletely excised tumor; this was not performed in the dogs of this study.

The findings reported here are consistent with previous reports of local recurrence after scar revision for incompletely or narrowly excised MCTs: in 1 of these studies, there was an 8% local recurrence rate with a median follow-up time of 403 d (13); in the other study, there was a 13% local recurrence rate with a median follow-up time of 431 d (6). The present study had a lower recurrence rate with a longer follow-up time (4.0% local recurrence with a median follow-up time of 881.5 d). All these studies evaluating scar revision for inadequately excised MCTs have lower recurrence rates than have been historically reported for incompletely excised MCTs (23 to 37%) (5,6). Previous studies evaluating MCT recurrence following surgical excision have recommended a minimum 2 y of follow-up to evaluate for recurrence at the scar due to a long recurrence-free interval (11,20). Recurrence-free intervals for MCTs have been reported to range from 167 ± 136 d, including dogs with complete, narrow, and incomplete margins (11), to 399 d in dogs with incomplete or narrow (< 3 mm) excision (6). For incompletely excised MCT undergoing subsequent scar revision, one group reported a median time to local recurrence of 2390 d in 3 dogs experiencing local recurrence (6), whereas another group identified local recurrence at 212, 555, and 993 d in 3 dogs experiencing local recurrence (13). In the present study, local recurrence occurred at 217 and 463 d, which is consistent with the latter study; it is unclear why the one study (6) had a longer local recurrence time of 2390 d.

General knowledge of the biology of the tumor is helpful to plan surgery. The more information that is obtained about the dog and the tumor prior to definitive excision, the more likely the surgery is to be successful (21). Samples for cytology were taken at the first surgical excision in only 38.9% of cases, and biopsies were not taken in any case. Cytology should be a first-line diagnostic test in evaluation of skin and subcutaneous masses. Fine-needle aspiration with cytology is a straightforward, minimally invasive test that can provide immediate results. It is particularly helpful for MCT diagnosis and was consistent with MCT in 81.0% of dogs in the present report. Multiple cytologic grading schemes have been developed for MCTs (22–24). Currently, these systems require additional improvements prior to widespread use, as high-grade tumors seem more likely to be inappropriately diagnosed as low grade with these systems (22–24). Biopsy prior to definitive resection is uncommon for MCT due to risk of degranulation and healing complications with residual disease, especially since they are typically readily identifiable with cytology. More frequently, excisional biopsy without cytology was performed for the initial excision for the dogs in this study. This is highly likely to lead to incomplete resection; thus, cytology is strongly recommended prior to any form of surgery, and submitting slides for evaluation by a boarded clinical pathologist may improve the success rate of diagnosis. Cytology of MCTs provides information on the tumor biology, allowing the surgeon to appropriately plan the extent of surgery required. Without this information, a higher proportion of incomplete excisions is expected (25), which is consistent with the findings of this study in which most masses were not thoroughly investigated prior to surgical resection.

It was surprising to discover that medical records at the time of first excision rarely included information on surgical margins, with 85.4% of surgeries having no information on surgical margins and only 6.3% of surgeries having information on both circumferential and deep surgical margins. When evaluating a scar from an incomplete resection to determine if scar revision is possible or recommended, information on the first excision is helpful for surgical planning. If the surgical margins from the initial excision are unknown, the surgeon performing the scar revision does not know what they will encounter during the scar revision procedure. Cross-sectional imaging can be helpful to assist with defining the extent of the tumor or scar; however, despite the usefulness of such imaging, it is often difficult to discern scar tissue from tumor regrowth in the case of scar revision, thereby complicating the assessment. One study reported that although computed tomography and magnetic resonance imaging identified peritumoral nodules in cats with injection site sarcoma, many of these lesions were non-neoplastic on histopathologic evaluation (26). Even if cross-sectional imaging is to be performed, a clear, thorough explanation of the first surgical procedure can provide a wealth of information to the surgeon performing the scar revision to assist with proper planning of a revision procedure.

Despite circumferential margins being potentially less complicated than deep margins, tissue shrinkage occurs throughout the resection and fixation process, with the circumferential margin reducing to 50% of the grossly normal surgical margin with MCTs (8). The deep margin can be difficult to assess histopathologically. When a surgeon obtains fascia as the deep margin, there is a level of confidence that the deep margin will be complete since MCT generally respect fascial boundaries. Barriers to tumor invasion have been defined in human oncologic surgery, and such barriers, often fascia or bone, are classified as either thick or thin barriers (27). These barriers are assigned an equivalent normal tissue width, indicating their resistance to tumor invasion regardless of their actual thickness. In veterinary medicine, fascia is not always identifiable on pathologic evaluation and it is typically thin (28). If a tumor extends to the fascia, but not through it, a narrow deep margin of fascia would still be considered complete and not narrow, although there are no equivalent normal tissue widths assigned in veterinary medicine. Despite this, the pathologist often does not know if fascia was taken, and deep margins are instead interpreted based on width. Given the issues with interpreting margins based on width in terms of narrow excision, this is not ideal (18). Prospective studies with long-term follow-up are needed to more fully evaluate deep and circumferential margins based on width and tumor grade to determine risk of recurrence.

Limitations of this study largely relate to its retrospective nature. Attempts were made to obtain records from the first excision, and although the pathology report was available for every case that was included, medical records for the first excision were not obtained in 11.1% of surgeries. Even for surgeries for which records were available, there was frequently either no surgery report or a surgery report that lacked details regarding the specific excision technique used. Pathology reports from the first excision also varied in their detail in reporting margin assessment, none included information on how the specimen was trimmed, and the pathologists evaluating the specimens differed. However, this is consistent with clinical practice and prospective studies would be necessary to have a uniform evaluation of the samples in terms of trimming and pathologist assessment. Follow-up obtained from owners may have been inaccurate. Records obtained from referring veterinarians may have been misinterpreted; if no mention of local recurrence was made and a normal physical examination was reported, the dog was considered to not have local recurrence, which may have underestimated the true rate of recurrence. In this study, the definition of a narrow margin differed depending upon tumor grade. This was chosen to reflect the clinical applicability of the data in which a scar revision may be recommended for a different margin for a high-grade versus a low-grade tumor. As discussed, recurrence of MCT is likely multifactorial and depends upon both the grade and the number of residual cells in the scar, thus this clinical difference is important for distinguishing where high-grade tumors may have a higher risk of recurrence.

In summary, although residual disease may not be identified, the recurrence rate following scar revision is lower than what has been previously reported for incompletely excised MCTs indicating a clinical benefit of scar revision. Pre-surgical cytology should be used before the first excision of any mass to assist in surgical planning. Surgical reports are an important component of the medical record and should include details specific to the surgical margins and tissue taken as the deep surgical margin. This information will be helpful to both the pathologist and the surgeon performing scar revision, if indicated. Scar revision should be considered, when possible, for incompletely or narrowly excised MCTs. CVJ