Vinblastine and Prednisolone as Adjunctive Therapy for Canine Cutaneous Mast Cell Tumors

Introduction

Mast cell tumors (MCT) are among the most common cutaneous neoplasms of dogs^1–7 and have a highly variable biological behavior. Histopathological grades are usually good predictors of survival.^8–11 Recommended management options for the various histopathological grades and clinical stages of canine MCT have recently been published.¹² The optimum management of local disease of any grade requires wide surgical excision (i.e., 2 to 3 cm of grossly normal tissue around the tumor) with histopathological confirmation of complete excision. When complete excision is not achieved, adjunctive radiotherapy may provide good local disease control.^13–15 Chemotherapy has generally been reserved for animals with measurable or systemic disease.^16–18 There is limited information in the veterinary literature on the use of chemotherapy as an adjunct to other therapies. One report evaluating the efficacy of vinblastine and prednisolone for MCT¹⁹ included dogs with intermediate- or high-grade tumors with residual microscopic disease on histopathological assessment of tumor margins (modified World Health Organization [WHO] clinical stage 0²⁰). In seven dogs given adjunctive chemotherapy, disease-free survival times were considered improved over expected results with surgery alone, and the protocol was proposed as an alternative treatment for microscopic MCT when radiotherapy is not possible.¹⁹

The purpose of the study reported here was to further evaluate the local tumor control achieved with a standardized vinblastine and prednisolone chemotherapeutic protocol in a population of dogs with intermediate or high-grade cutaneous MCT.

Materials and Methods

Dogs presented to Murdoch Animal Cancer Care Unit (MACCU) for management of canine MCT between November 1999 and December 2000 were eligible for inclusion in the study. Inclusion criteria for the study group were a histopathological diagnosis of grade II or III cutaneous MCT, inadequate surgical excision defined by histopathological assessment of surgical margins, and no concurrent chemotherapy. Animals with systemic disease, known metastatic disease, or multifocal MCT were not included in this study. As this study was performed without external financial support, complete clinical staging was routinely recommended prior to treatment, but was not always performed.

A chemotherapeutic protocol based on that reported by Thamm and others¹⁹ was used. Eight treatments of vinblastine at 2.0 mg/m² body surface area were given by rapid intravenous (IV) bolus injection over a 12-week period. Oral prednisolone was administered concurrently at an initial dose of 2.0 mg/kg per day, tapering over the 12-week treatment period [Table 1]. The schedule was altered at the discretion of the attending clinician if significant toxicity or other adverse effects developed.

A veterinary anatomical pathologist (Jardine), who was unaware of clinical outcome, reviewed the slides of eligible animals to confirm histopathological grade and categorize the surgical margins as “incomplete” or “clean-but-close.” Histopathological grading conformed to the scheme described by Patnaik et al.,⁹ with grade 3 representing poorly differentiated (i.e., high grade) tumors and grade 2 representing moderately differentiated (i.e., intermediate grade) tumors. Surgical margins were defined as “incomplete excision” (i.e., neoplastic cells extending to the surgical margins or to within 1 mm of the surgical margins) or “clean-but-close” margins (i.e., neoplastic cells within 1 to 10 mm of the surgical margins).²¹

Animals undergoing chemotherapy received a full physical examination, including thorough surgical site and regional lymph node evaluation, and total white blood cell count with differential count prior to each treatment. Toxicity was graded utilizing the Eastern Cooperative Oncology Group common toxicity criteria,^a as modified for canine patients^b [Table 2].

Follow-up examinations were performed by one examiner (Davies) and continued for at least 12 months from the commencement of chemotherapy in 20 dogs. Follow-up consisted of a full physical examination; inspection and palpation of the surgical site and all lymph nodes; fine-needle aspiration cytopathology of any cutaneous mass, subcutaneous mass, or enlarged lymph node; and cytopathological examination of a buffy-coat smear.¹² All cytopathological examinations were performed by a veterinary clinical pathologist. Other relevant information was obtained by personal interview with the owners or by telephone interview with referring veterinarians. Follow-up procedures were approved by the Murdoch University Animal Ethics Committee.

At the time of last follow-up, dogs were classified as a complete remission (i.e., no evidence of local, distant cutaneous, lymph node, or visceral MCT) or as a treatment failure (i.e., local, distant cutaneous, lymph node, or visceral recurrence of MCT). Within the treatment failure group, local recurrence was defined as MCT growth within 2 cm of the original surgical site, and distant cutaneous recurrence was defined as any cutaneous MCT >2 cm from the original surgical site. Dogs with known treatment failure before the time of follow-up examination were recorded according to the date of laboratory confirmation of MCT recurrence. Dogs that had died before follow-up were recorded according to cause of death and status of MCT at the time of death, if these were known. Only dogs with known MCT status at the time of death or follow-up examination were included in the final data analysis.

Data analysis was performed using a statistical software package.^c Discrete or categorical data were analyzed with the Fisher’s exact test or the chi-square test for independence using the Yates correction factor. Continuous data were analyzed with an analysis of variance (ANOVA) after testing for equality of variances with the Levene’s statistic. Breed risk in the study group was assessed by calculation of the odds ratios (OR) and the 95% confidence intervals (CI) relative to mixed-breed dogs, using the total hospital dog population for 1999 and 2000 as the reference population. A P value of ≤0.05 was considered to be statistically significant. The power of the analyses was determined with the statistical software package, G-Power.^d

Results

Twenty-seven dogs presenting to MACCU for management of cutaneous MCT during the 14-month study period were eligible for inclusion in the study. Within the study group, age at time of first treatment ranged from 3 to 12 years (mean, 7.1±2.8 years; median, 7 years). There were four intact males (15%), six castrated males (22%), four intact females (15%), and 13 spayed females (48%). Of the breeds in which multiple individuals were present, boxers (5/27; OR, 11.3; CI, 3.4 to 37.6), Staffordshire bull terriers (5/27; OR, 6.1; CI, 1.9 to 20.3), and Labrador retrievers (4/27; OR, 4.2; CI, 1.2 to 14.9) were significantly over represented. Thirteen (48%) tumors were excised from the limbs; six (22%) were excised from the inguinal or perineal region; seven (26%) were removed from the trunk; and one (4%) was excised from the facial area. Twenty-four (89%) tumors were of intermediate grade, and three (11%) were of high grade. Surgical margins were incomplete in 19 (70%) dogs and were clean-but-close (2 to 5 mm) in eight (30%) dogs, comprising two distinct subgroups. Analysis of age (P=0.21), breed (P=0.31), boxer versus breeds other than boxer (P=0.09), gender (P=0.53), histopathological grade (P=0.36), and site of MCT (P=0.31) showed no significant differences between the incomplete and clean-but-close groups.

A total of 206 doses of vinblastine were given to the dogs in the study group. One animal that died during treatment and one that withdrew from the study did not complete the 12-week protocol. Twenty-six (13%) doses were associated with adverse effects, but only one dog required hospitalization for chemotherapy-related toxicity [Table 3]. Ten (5%) doses were delayed by 5 to 8 days (mean, 6.4±1.0 days; median, 7 days) in nine dogs because of neutropenia. In one dog with an incompletely excised, intermediate-grade MCT, the dose of vinblastine for subsequent treatments was reduced by 10%, because grade 2 neutropenia developed after the first treatment. This dog remained in complete clinical remission when evaluated at 476 days postinception of therapy. One animal with an incompletely excised, high-grade MCT developed grade 4 neutropenia after the second treatment and died of sepsis 12 days after commencing chemotherapy. Another animal with a high-grade MCT and clean-but-close margins (4 mm), developed grade 4 neutropenia and diabetes mellitus after the third treatment; prednisolone was discontinued at this time, and insulin therapy was commenced. The neutropenia was not clinically significant, and the dog completed vinblastine therapy but was euthanized by the referring veterinarian 361 days after starting chemotherapy, because of acute vomiting, nodular hepatic disease, and hemoabdomen. Postmortem histopathological examination of liver tissue, including toluidine blue staining, did not confirm the presence of hepatic mastocytosis. Another dog with an incompletely excised, intermediate-grade MCT died suddenly of an unknown cause 268 days from the commencement of chemotherapy. The referring veterinarian did not report recurrence of MCT at this time, but postmortem examination was not performed. One dog with an incompletely excised, intermediate-grade MCT was withdrawn from the study by its owners after four treatments, because of the development of vestibular disease.

Twenty dogs that completed the protocol were available for follow-up after a minimum of 12 months from the date of commencing chemotherapy. Time from onset of chemotherapy to final examination ranged from 400 to 730 days (mean, 550±108 days; median, 537 days). Buffy-coat smears were negative for mast cells in all animals. No dogs had palpable enlargement of lymph nodes, liver, or spleen or other evidence of systemic mastocytosis (such as vomiting or melena) on follow-up examination. Additional clinical staging procedures were not performed.

Six dogs were classified as treatment failures. One dog with an intermediate-grade MCT and clean-but-close surgical margins (2 mm) had local recurrence of MCT confirmed by cytopathology at 392 days. One dog with an incompletely excised, intermediate-grade tumor had a local recurrence at 154 days and a distant cutaneous recurrence at 286 days, both confirmed by excisional biopsy. Three dogs with intermediate-grade MCT (two with incomplete excisions and one with a clean-but-close excision [2 mm]) developed distant cutaneous recurrence of MCT; follow-up times for these dogs were 534, 675, and 712 days. One dog, a boxer, had a confirmed distant cutaneous recurrence at 228 days and developed a distant cutaneous recurrence at a third site at 638 days. Dogs with confirmed distant cutaneous recurrence included three Staffordshire bull terriers and two boxers.

Fourteen (70%) dogs were in complete remission at the time of the last follow-up examination, including the only surviving dog with a high-grade tumor (499 days) that was removed with incomplete margins. Follow-up periods for these dogs ranged from 403 to 730 days (mean, 542±92 days; median, 499 days).

Three dogs that had completed treatment and were still alive were unavailable for follow-up examination, but they had not suffered any known recurrence when their owners were telephoned >12 months from the onset of chemotherapy. As noted above, three dogs died prior to follow-up examination, and one withdrew without completing chemotherapy. The deaths could not be attributed to MCT. These last seven dogs were not included in final data analysis.

No significant difference in outcome existed between the two surgical margin subgroups (P=0.61). No significant associations were found between outcome and age (P=0.97), gender (P=0.55), breed (P=0.41), boxer versus breeds other than boxer (P=0.55), histopathological grade (P=1.0), and site of MCT (P=0.35). Power estimates for all Fisher’s exact and chi-square tests were >0.80; however, the estimated power of the ANOVA (assessing possible differences in age) was only 0.35.

Table 4 shows outcomes for each subgroup and the total study population with respect to local tumor recurrence. Local tumor control rate was 90% at the time of follow-up. Median survival time and median disease-free interval were not determined.

Discussion

Results of this study show that the rate of local MCT recurrence 12 to 24 months after incomplete or marginal surgical excision and followed by adjunctive vinblastine and prednisolone chemotherapy, is low and is comparable with studies evaluating complete surgical excision²²²³ or incomplete surgical excision with additional radiotherapy.^13–15 Recurrence of MCT within 2 cm of the surgical margin occurred in two (10%) of 20 dogs available for follow-up examination and was not reported in three treated dogs still alive but not evaluated further. Distant cutaneous recurrence rates were higher (25%) and were in accordance with other studies of MCT treatment.¹⁹²²²³ No evidence of systemic mastocytosis was seen in any of the animals that were reevaluated. Toxicity associated with the protocol was generally mild and self-limiting, although one dog died of neutropenia and sepsis.

The frequency of recurrence of incompletely resected canine cutaneous MCT is uncertain and is an area of current controversy in veterinary oncology. Studies comparing therapies for incompletely resected MCT (with small control groups) found local recurrence rates for dogs without adjunctive treatment to be 66%²⁴ to 83%²⁵ for intermediate-grade tumors and 100%²⁴²⁵ for high-grade tumors. A recent retrospective study²⁶ compared outcomes for dogs with histopathologically complete or incomplete MCT excision and no further therapy. In that study, five of 11 dogs with incomplete excision had no local recurrences over a follow-up period of 19 to 60 months. All local recurrences developed within 12 months of surgery. Recent studies evaluating surgical excision alone (histopathologically complete excision) had local recurrence rates of 5%²² to 11%²³ for intermediate-grade MCT. Clearly, in some cases, apparent complete surgical excision may still result in local recurrence, while incomplete excision does not invariably lead to local recurrence. The present uncertainty over the true rate of local recurrence of microscopic MCT after surgical excision significantly limits assessment of the protocol reported here. True recurrence rates can only be addressed with the inclusion of a matched, untreated control group. With the current prevailing opinion that residual microscopic MCT should be treated,¹² inclusion of untreated controls is difficult because of ethical concerns. A prospective study directly comparing the results of chemotherapy with standard therapy in a similar cohort of animals may provide additional information.

The recurrence of MCT at a distant cutaneous site is difficult to categorize, and because of that difficulty, local and distant cutaneous recurrences of MCT were considered separate treatment failure events in this study. It is currently unknown whether these tumors represent metastases of the primary tumor or development of a new MCT. Séguin, et al.,²² reported that six (11%) animals with completely excised, intermediate-grade cutaneous MCT later developed distant cutaneous MCT, but that only one of these was associated with visceral metastasis. Studies have shown that dogs with multiple cutaneous MCT do not have a worse prognosis than dogs with single tumors,^16171927 which supports the argument that each mass may represent a discrete primary skin tumor rather than metastatic disease. The results of the authors’ study suggest that adjunctive chemotherapy does not have an appreciable effect on the subsequent development of distant cutaneous MCT (25% of dogs evaluated developed new MCT). Thamm, et al.,¹⁹ reported a similar outcome in that 26% of dogs with adequate control of their primary MCT developed MCT at a distant cutaneous site after vinblastine and prednisolone chemotherapy. Chemotherapeutic agents are in themselves mutagenic, and the relatively high number of dogs with new tumor development after treatment may represent an adverse effect. However, no new tumors of other types were noted in any of the dogs reported here. Alternatively, these distant cutaneous tumors may have arisen after treatment, because there was a long time delay (>150 days) between the final dose of vinblastine and the earliest detection of the mass in the authors’ study. It is the authors’ belief that distant cutaneous recurrence in a population of dogs with an apparent predisposition to the development of MCT (all distant recurrences were in boxers or Staffordshire bull terriers) does not necessarily imply failure of therapy for the primary tumor. However, if microscopic disease were present at the termination of chemotherapy, then the recurrence would indicate a treatment failure. Because of the uncertainty surrounding the pathogenesis of sequential cutaneous MCT, dogs with preexisting multiple MCT were not included in this study.

None of the dogs in the authors’ study developed evidence of regional lymph node metastasis or systemic mastocytosis. The dog that was euthanized at 361 days was suspected of having hepatic mastocytosis; however, postmortem histopathological examination of liver tissue did not confirm this suspicion. Only one of three dogs with a high-grade primary tumor survived to follow-up, and the dog was judged to be in complete remission at 499 days. High-grade tumors are usually associated with a poor prognosis, with reported median survival times of 14 weeks⁷ and 1 month.²⁸ The natural behavior of intermediate-grade tumors is less certain. Large studies have shown that tumor-related deaths in this latter category are lower than with high-grade tumors and higher than with low-grade tumors.⁹²⁹ In Séguin, et al.,²² three (5%) dogs with completely excised, intermediate-grade cutaneous MCT were diagnosed with metastatic disease between 143 and 266 days. Metastatic rates ranged from 3%¹⁵ to 12%²⁰ following radiation therapy of intermediate-grade MCT.

Radiotherapy of incompletely resected MCT has become the treatment of choice and has reported 12-month local control rates of 86% to 97% for intermediate-grade MCT.^13–15 The current lack of radiotherapy facilities for pets in Western Australia led the authors to investigate alternative means of adjunctive therapy. Chemotherapy has only been previously evaluated for modified WHO clinical stage 0 MCT (i.e., residual microscopic disease) in seven dogs.¹⁹ Previously reported vinblastine and prednisolone therapy led to a 12-and 24-month overall disease control rate of 67% and 50% for intermediate- and high-grade tumors, respectively.¹⁹ The authors’ results in a larger group of animals suggest that vinblastine and prednisolone chemotherapy can be considered as a possible alternative to radiotherapy or radical surgery when surgical margins are considered inadequate.

Although the results of this study appear promising, there are several limitations that must be considered. No control group of dogs was included, and the authors can only report their observations without testing the hypothesis that vinblastine and prednisolone significantly decreased the local tumor recurrence rate in the dogs reported here. Examination of the literature shows that studies evaluating the current standard of care also suffer from this limitation, and future studies should be designed to address this weakness. The relatively small size of the study population, although comparable with other studies of MCT therapy, may have been too small to allow a discernable benefit with respect to infrequent events, such as metastasis of an intermediate-grade tumor. The inclusion of the clean-but-close group may confound overall interpretation of the data, as there is very little available historical data regarding this subset of animals. It is recognized that with MCT, clumps of tumor cells may remain in situ despite the microscopic appearance of a “clean-but-close” margin;²¹ therefore, the authors chose to treat and include this subpopulation of dogs in this study. Comparison of the two subgroups showed no significant differences in prognostic indicators or outcome, although the power for detection of significant age differences was low. Complete clinical staging was not performed in this study in all animals, and it is possible that dogs were misclassified with respect to clinical stage either at their entry into the study or on follow-up evaluation. The use of buffy-coat examination as an indicator of systemic mastocytosis is considered by some to be an insensitive and nonspecific screening test;³⁰³¹ however, normal dogs do not have visible mast cells in buffy-coat smears.³² Hepatic and splenic fine-needle aspirates have been shown to be the more sensitive in the staging of MCT,^e but their use may have decreased owner compliance with follow-up evaluations. The authors’ staging protocol also included cytopathology of palpably enlarged lymph nodes only, and it is possible that lymph node metastasis was not detected in dogs with palpably normal lymph nodes. A recent study³³ has shown that the sensitivity of lymph node palpation (compared to excisional biopsy) for the detection of tumor metastasis was 93% when all nodes considered mildly or substantially enlarged were found to be positive. Given the nature of the authors’ study, any lymph node enlargement was viewed with suspicion, and it is expected that few false-negative results occurred. The specificity of fine-needle aspirate cytopathology of palpably normal lymph nodes is questionable, because 24% of normal dogs have mast cells in their popliteal lymph nodes.³¹ The follow-up period of this study (minimum, 13 months; median, 18 months) may have been insufficient to allow detection of recurrence in all individuals; however, this period was chosen with reference to previous reports and to allow comparison with historical data for local tumor control. It is anticipated that few dogs suffered local recurrence or metastasis after termination of the study. Although further cutaneous distant recurrence may occur in some dogs still alive from this study, monitoring for these tumors was not a primary aim of the study, given the authors’ opinion that these MCT are likely to represent de novo tumors rather than true recurrences of the primary tumor.

Conclusion

The use of vinblastine and prednisolone adjunctive chemotherapy in this study of dogs with stage 0 cutaneous MCT provided 12-month local control rates that were similar to previous rates reported for intermediate-grade MCT treated by complete surgical excision or excision and adjunctive radiotherapy. The protocol reported here was well tolerated by most animals. The subsequent development of new cutaneous tumors may not be prevented by the protocol. The low incidence of high-grade tumors in this population and the study design do not allow conclusions to be made with respect to prevention of metastatic disease. Further controlled prospective studies to ascertain the true benefit of adjunctive MCT therapy are warranted.

Acknowledgments

The cooperation of the referring veterinarians and the expertise of clinical pathologists Drs. Philip Clark and Jenny Mills and anatomical pathologist Dr. Amanda O’Hara are gratefully acknowledged. Thanks are also due to Dr. Gina Michels and her coauthors for generously sharing their data with the authors prior to its publication in this journal.