Predictive ability of fine-needle aspirate cytology for incompletely resected mast cell tumor surgical sites

Abstract

The goal of this study was to evaluate whether fine-needle aspirate cytology of a previous surgical site was predictive of recurrence for incompletely excised mast cell tumors (MCTs). Electronic medical records were searched for dogs diagnosed with MCTs; those with histologically confirmed, incompletely resected MCTs evaluated by scar aspiration cytology within 60 days after surgery were included for analysis. Variables were compared between groups using Fisher’s exact test and logistic regression. Twenty-nine cutaneous and 7 subcutaneous tumors were evaluated. Local recurrence, confirmed by either histopathology or cytology, occurred in 13.8% of cases. No significant differences were identified for any variables other than surgical site cytology status. The negative predictive value of surgical site aspirate cytology without residual mast cell tumor was 93.5%, with an overall predictive accuracy of 88.9%. For the dogs evaluated in this report, surgical site aspiration cytology was predictive of local disease control for incompletely resected MCTs.

Introduction

Mast cell tumors (MCTs) are the most common cutaneous tumors in dogs, accounting for 7% to 21% of all cutaneous neoplasms (1,2). Surgical excision is usually considered front-line treatment for cutaneous and subcutaneous MCTs (2). Historically, 1 cm lateral and 1 fascial plane deep margins, and 2 cm lateral and 1 fascial plane deep margins were the recommendations for Grade I and Grade II MCTs, respectively (3,4). Less is known about the surgical margins needed for Grade III MCTs, likely because they occur less frequently (1). A recent study recommended MCT excision with modified proportional margins and this approach had relevance, regardless of grade, using both the 2- and 3-tier grading schemes (5).

The local recurrence rate of incompletely resected MCTs has been reported to range from 18% to 38% (6–9). Some studies reported no difference in local recurrence rates for Grade II MCTs excised with histologically tumor-free margins versus those with incomplete margins (7,10). In a study by Kry and Boston (6), 11 of 23 dogs (48%) treated with scar revision surgery had evidence of residual microscopic disease in the second biopsy section. Moreover, none of the dogs without residual mast cell tumor in the second biopsy experienced recurrence in the study period. Additionally, Vincenti and Findji (10) reported that 84% of incompletely excised MCTs that underwent revision surgery did not have evidence of microscopic disease in the second histopathologic sample. However, in another study, local recurrence was associated with progression to multicentric regrowth and/or metastasis in nearly 60% of dogs (11).

There are limitations to consider when deciding if additional surgery is warranted. Some examples include additional cost to the owner and morbidity to the patient. Other factors in this decision include the comfort level and experience of the surgeon, access to specialty care and availability of additional local tissues for wound reconstruction. It is reasonable to conclude that some animals with histopathologically confirmed residual disease would likely benefit from additional treatment, whereas others have achieved local control and would not. A simple, low morbidity diagnostic test to evaluate the surgical sites from incompletely excised MCTs would seem useful in determining dogs that could benefit from additional treatment.

The objective of this study was to determine if fine-needle aspirate cytology of the surgical site is predictive of local recurrence for incompletely resected MCTs. The hypothesis of this study was that fine-needle aspiration cytology is predictive of whether local control (i.e., no residual tumor) was achieved.

Materials and methods

The electronic medical records from dogs presented to the oncology service at Auburn University from 2004 to 2016 and diagnosed with MCTs were reviewed. Dogs treated with surgical resection of cutaneous or subcutaneous MCTs, either at Auburn University or at their primary care veterinary hospital, with histopathologically confirmed incomplete margins, and aspiration of the surgical site within 60 days following surgery, were included in this study. Exclusion criteria included a revision procedure after surgical site fine-needle aspiration and dogs without recurrence having < 6 mo of follow-up information.

Hospital records for all animals with a diagnosis of MCT were reviewed. Information retrieved included signalment, weight, date of diagnosis, date of surgery, anatomic location of the MCT, presence of metastasis, histologic margin status, date of surgical site aspiration, cytology results, presence of recurrence, date of recurrence, complications following surgery (minor or major), use of adjuvant therapy, and time to last follow-up. Metastasis to lymph nodes, liver, spleen, and bone marrow was based on cytologic or histopathologic diagnoses. When necessary, follow-up was obtained from primary care veterinarian records and/or telephone calls to owners.

Time to local recurrence was defined as the interval between the initial surgical resection and the time to detection of gross disease verified by either cytology or histopathology. When re-evaluation was only available through telephone communication with the owner, presence of a mass at the surgical site was considered local recurrence. Follow-up time was defined as date from initial diagnosis to date of last follow-up or death.

Minor complications were defined as complications not requiring additional surgical intervention to achieve healing. Major complications were defined as issues requiring surgery to resolve.

This was a retrospective study and surgical site fine-needle aspirate technique could not be standardized. In general, fine-needle aspirates were taken adjacent to and deep within the scar of the previous excision site, in a number of locations, as deemed appropriate by the supervising clinician for the dimension of the scar, taking care to ensure the entire surgical site could be evaluated to the best of their ability.

Fisher’s exact tests were used to investigate associations of sex, alteration status, weight, age, adjunctive therapy, anatomic location, histologic grade, cutaneous or subcutaneous location, wound complications, and surgical site aspiration cytology results with respect to recurrence. Predictive values were calculated based on cytology results. The association between cytology and recurrence was evaluated with a simple logistic regression model. All statistical analyses were performed using commercial software (SAS 9.2; SAS Institute, Cary, North Carolina, USA) and P < 0.05 was considered significant.

Results

Records from 503 dogs with 665 mast cell tumors were evaluated. Sixty-nine dogs had fine-needle aspiration of scars within 60 d after surgery. Thirty-six dogs were excluded from the study for the following reasons: complete surgical margins (n = 18), scar revision surgery (n = 6), radiation therapy of the scar (n = 6), or were lost to follow-up (n = 6). Thirty-three dogs with 36 tumors met the inclusion criteria. Breeds represented included Labrador retriever (n = 11), boxer (n = 4), mixed breed (n = 3), American Staffordshire terrier (n = 2), and 1 each of Boston terrier, Jack Russell terrier, Chinese crested, Chihuahua, Cocker spaniel, German shepherd, golden retriever, Great Pyrenees, Lhasa Apso, miniature schnauzer, rat terrier/feist terrier, Siberian husky, and Yorkshire terrier. Two Labrador retrievers and 1 Chihuahua had 2 tumors each that were included in the analysis. There were 19 spayed females, 11 neutered males, and 3 intact males. Median age at diagnosis was 9.3 y (range: 3.8 to 14.0 y). Median body weight was 27.1 kg (range: 4.4 to 63.4 kg). Median time of surgical site fine needle aspiration following surgery was 18 d (range: 5 to 56 d). Median follow-up time was 470 d (range: 120 to 2882 d).

Overall, local recurrence occurred in 5 of 36 tumors. None of the subcutaneous tumors (n = 7) had local recurrence. Cutaneous tumors recurred in 5 of 29 tumors. Median time to recurrence was 89 d (range: 15 to 285 d). Recurrence was more frequent in males than in females (20% versus 9.5%, respectively) but this difference was not significant (P = 0.63). Neither age nor weight was statistically related to recurrence.

Anatomic location of the tumor was stratified into 4 groups. These groups included extremity, (limb or tail, n = 21, 58%), genitoperineal (n = 3), head/neck (n = 3), and thoracoabdominal (n = 9).

There were no local recurrences in the genitoperineal or head/neck regions. There were 3 recurrences of extremity tumors and 2 of the thoracoabdominal tumors. Anatomic location was not related to recurrence (P = 0.86).

At least 1 of the histologic grading systems was used to classify 29 tumors. There were 10 MCT’s graded with only the Patnaik system and there were 1 Grade I, 7 Grade II, and 2 Grade III tumors (2). The Grade I MCT did not recur. Recurrence was observed in 1 each of the Grade II and Grade III MCTs. There were 5 MCTs graded with only the Kiupel system and there were 3 low-grade and 2 high-grade tumors (2). One of the high-grade tumors recurred and none of the low-grade tumors recurred.

Fourteen tumors were graded according to both systems. There were 10 MCTs categorized as Kiupel low-grade/Patnaik Grade II. There was 1 recurrence in this group. There were 2 MCTs graded as Kiupel high-grade/Patnaik Grade II. There were no recurrences in this group. There were 2 MCT’s graded as Kiupel high-grade/Patnaik Grade III. There was 1 recurrence in this group.

The frequency of recurrence was more likely as the histologic grade increased for the Patnaik system (Grade I = 0%, Grade II = 10.5%, Grade III = 50%), but statistical significance was not reached (P = 0.27). Seven tumors in 5 dogs without histologic grading were all subcutaneous tumors and none of these had local recurrence.

Metastatic disease was identified at the time of initial diagnosis or surgery in 4 dogs and was considered probable in 1 dog. The 4 confirmed metastatic sites were to lymph nodes and the probable metastasis was to the spleen (Table 1). Two of the 5 dogs with metastatic disease had local recurrence of the primary tumor compared to 3 of 28 without metastatic disease. The presence of metastasis was not significant for recurrence (P = 0.08).

Table 1.

Tumor characteristics in dogs with metastatic disease at time of initial diagnosis or surgery.

Site of metastasis	Anatomic location of primary tumor	Kiupel grading system	Patnaik grading system	Recurrence
Inguinal lymph node	Extremity	high	n/a	Yes
Axillary lymph node	Thoracoabdominal	low	II	No
Superficial cervical lymph node	Thoracoabdominal	high	III	No
Inguinal lymph node	Thoracoabdominal (subcutaneous)	n/a	n/a	No
Spleen (probable)	Thoracoabdominal	high	III	Yes

n/a — Not applicable.

Wound healing complications occurred in 11 of 36 cases (31%). Minor complications occurred in 7 cases and major complications occurred in 4 cases. Three dogs with major wound complications had local recurrence of their disease.

Surgical site fine-needle aspiration cytology was divided into 3 groups, based on cytologic results and included Group 1, absence of mast cells (n = 28, 77%), Group 2, rare-to-few mast cells (n = 3), and Group 3, residual mast cell tumor (clustering of mast cells inconsistent with wound healing, n = 5). Recurrence occurred in 1 of 28 dogs in Group 1, 1 of 3 dogs in Group 2, and 3 of 5 dogs in Group 3 (P = 0.006). However, recurrence rates were not different between Groups 1 and 2 (P = 0.56), or between Groups 2 and 3 (P = 1), but were higher for Group 3 compared to Group 1 (P = 0.02).

For comparison purposes, Groups 1 and 2 were combined and designated non-residual disease and Group 3 was considered residual MCT. Recurrence was identified in 2 of 31 dogs in the non-residual disease group compared to 3 of 5 dogs in the residual MCT group. The negative predictive value of surgical site aspirate cytology without residual mast cell tumor was 93.5%, with an overall accuracy of 88.9%. Dogs with cytologic evidence of residual MCT were 21.8 times more likely to have local recurrence than those that had absence of, or rare to few mast cells in their scar cytology (P = 0.001).

Adjuvant chemotherapy was pursued for 17 of 36 tumors (47%). Four of 17 dogs treated with adjuvant therapy had recurrence. One of 19 tumors not treated with adjuvant therapy had recurrence. Recurrence was less frequent in dogs that did not have adjuvant therapy compared to those that did (5.2% versus 23.5%, respectively, P = 0.17).

Discussion

To the authors’ knowledge, this is the first study evaluating the predictive ability of surgical site fine-needle aspiration cytology for incompletely resected MCTs. In this study, scar aspirate cytology without residual mast cell disease was significantly predictive of local disease control.

In this study population, dogs with metastatic disease identified at the time of referral evaluation had a proportionally higher chance of local recurrence than those without metastatic disease. This trend in the data may have been due to sample bias, because 3 of 5 dogs with metastatic disease had high-grade or Grade III tumors, which are more likely to have local recurrence than low-grade tumors (12).

Wound complications were not infrequent in this study population. Most complications were minor, but of those dogs with major complications, 75% had local recurrence of MCT. Local recurrence was not detected in dogs with minor wound complications. It is likely that the high rate of recurrence in dogs with major complications was secondary to residual tumor effects. Furthermore, dogs with minor complications likely had low residual disease burden, if any. This may have resulted in loss of remaining tumor autonomy or the healing environment may have been cytotoxic to the residual neoplastic mast cells. To the authors’ knowledge, this concept has not been evaluated in veterinary oncology and studies are needed to determine if this occurs.

In this study, increasing histologic grade with both the 2-tier Kuipel grading system and the 3-tier Patnaik system was positively associated with an increasing likelihood of local recurrence. This was not statistically significant, however, and the authors encourage caution interpreting this variable, as the cohorts in each group were relatively small, and the result may represent a type-II error.

Surgical site fine-needle aspiration cytology was the only variable to reach statistical significance in this study. The negative predictive value of fine-needle aspirate cytology for incompletely resected MCTs was 93.5%, with an overall accuracy of 88.9%. The authors frequently use this technique to assess need for adjuvant local therapy for incompletely excised MCTs. This is an appealing concept, as the technique is simple, relatively inexpensive, and is widely applicable in both primary care and specialty practice settings. Additionally, based on our data, aspirate cytology can provide valuable information to help guide veterinarian and client decision-making.

Limitations inherent to this study largely stem from its retrospective nature and sample size, consistent with other veterinary studies. Surgical site fine-needle aspiration could not be standardized among patients or clinicians. Adjuvant chemotherapy protocols were not standardized among dogs but were also not a focus of this study. There was a higher proportion of recurrences in dogs treated with adjuvant chemotherapy compared to those that were not. This finding was likely a selection bias for chemotherapy use in dogs with more advanced/aggressive disease. Dogs treated with chemotherapy were not excluded from analysis for this study. This could be considered a limitation. However, chemotherapy has been previously shown to not impact recurrence or impart local disease control (10,13). The main objective of this report was to evaluate the predictive ability of fine-needle aspirates from the surgical site and therefore chemotherapy administration was not a criterion for exclusion.

Not all samples were read by the same cytopathologist. Additionally, there were several cases in which telephone calls to owners were made to assess for recurrence. In an effort to not bias the study in a falsely favorable direction, any mention by the owner of tumor formation at or near the previous surgical site was deemed to be recurrence. This may have overestimated the recurrence rate.

Based on the results of this study, the presence of cytologically confirmed residual MCT in the surgical site increased the probability of recurrence and demonstrated that cytology can be used to investigate surgical sites. This can help the clinician and owner in guiding treatment decisions.