Mast cell tumors (MCTs) represent the most common

Transcription

1 J Vet Intern Med 1999;13: Prednisone and Vinblastine Chemotherapy for Canine Mast Cell Tumor 41 Cases ( ) Douglas H. Thamm, Elizabeth A. Mauldin, and David M. Vail Forty-one dogs with mast cell tumors (MCTs) were treated with oral prednisone and injectable vinblastine (VBL), both in the adjuvant setting (23 dogs) and in dogs with gross disease ( dogs). Adverse effects were noted in 20% (8/41) of the patients, usually after the 1st dose of VBL. Adverse effects were considered mild in 6, and severe, necessitating treatment discontinuation, in 2 (5%). Overall response rate in the evaluable dogs with gross disease was 4% (/15), consisting of 5 complete responses and 2 partial responses. Median response duration was 154 days (24 to 645 days). As adjuvant therapy to incomplete surgical resection, prednisone and VBL conferred a 5% 1- and 2-year disease-free rate. Median survival time (MST) for the entire patient population was not reached with a median follow-up of 53 days; however, the MST for dogs with grade III MCT was 331 days, with 45% of dogs alive at 1 and 2 years. This is an apparent improvement over historical survival data employing surgery alone. Upon univariate analysis, significant prognostic factors (P.05) for survival included presence of a locally recurrent tumor, presence of gross disease, argyrophilic nucleolar organizer region frequency, lymph node status, histologic grade, previous chemotherapy, and ulceration of the tumor. Similar criteria were significant when analyzed for time to treatment failure. Response to therapy was also predictive of survival in the gross disease group. Upon multivariate analysis, histologic grade (P.012) and presence of a locally recurrent tumor (P.001) were significant factors for survival. Key words: Dog; Mastocytoma; Retrospective study; Treatment; Vinca alkaloid. Mast cell tumors (MCTs) represent the most common malignant cutaneous tumors in the dog, comprising 16 21% of all cutaneous tumors. 1 A large degree of variation occurs in the histologic appearance and biologic behavior of canine MCTs, ranging from histologically and behaviorally benign to histologically and behaviorally malignant. High-grade or undifferentiated MCTs (Patnaik grade III) comprise 29 40% of all MCTs. 2 4 In addition to being locally infiltrative, they have a metastatic rate of 55 96%, and are more likely to result in death as a result of tumor than are low- or intermediate-grade tumors. 1 Thus, aggressive local therapies alone may be insufficient for optimum tumor control. Aggressive surgery with wide margins remains the mainstay of treatment for most canine MCTs. However, a variable local recurrence rate occurs after surgery, ranging from 11 to 64%, with higher-grade tumors typically exhibiting a higher rate of recurrence. 5,6 Survival after surgery alone for high-grade MCTs has been reported as 15% survival at months, 6% survival at 48 months, and a median survival time of 13 weeks. 2,4, Previous studies have identified many prognostic factors associated with MCTs, including age, 8 breed, 2 sex, 9 histologic grade, 2,4,9 11 clinical stage, 9 11 location, 10 recent rapid growth, 2,12 argyrophilic nucleolar organizer region (Ag- NOR) frequency,,13,14 and proliferating cell nuclear antigen assessment. 14 Other factors that anecdotally carry prognos- From the Department of Medical Sciences (DHT, DMV) and Pathobiological Sciences (EAM), School of Veterinary Medicine, and the Comprehensive Cancer Center (DMV), University of Wisconsin Madison, Madison, WI. Presented in abstract form at the th Annual Veterinary Cancer Society Meeting, Estes Park, CO, October Reprint requests: David M. Vail, DVM, Department of Medical Sciences, School of Veterinary Medicine, University of Wisconsin Madison Linden Drive West, Madison, WI 5306; vaild@svm.vetmed.wisc.edu. Submitted December 8, 1998; Revised March 1, 1999; Accepted March 16, Copyright 1999 by the American College of Veterinary Internal Medicine /99/ /$3.00/0 tic significance include the presence of edema or ulceration, recurrence after previous surgery, and gastrointestinal signs (anorexia, vomiting, melena, hematemesis). 1 Systemic corticosteroids have been considered beneficial in the treatment of MCTs, but durable efficacy is lacking in controlled clinical trials. 15 A study investigating singleagent vincristine for dogs with measurable MCT yielded a % partial response rate. However, 32% of dogs had treatment discontinued prematurely because of toxicity or perceived decreased quality of life. 16 A cyclophosphamide vinblastine prednisone combination demonstrated a 8% response rate (all partial responses) in animals with measurable metastatic MCTs. 1 Combination therapy with cyclophosphamide, vincristine, prednisone and hydroxyurea in dogs with measurable MCTs resulted in an objective response rate of 60%, with a median response duration of 53 days. 9 Vinblastine (VBL) is a potent plant alkaloid that has been used in the treatment of hemolymphatic neoplasia in dogs, and various human malignancies including testicular cancer, Hodgkin s disease, Kaposi s sarcoma, and carcinomas of the bladder and breast. 19 VBL has been described anecdotally as an active agent for canine mast cell neoplasia. 20,21 The current study was undertaken to assess the safety and efficacy of oral prednisone combined with injectable VBL for the treatment of canine MCTs, both in the adjuvant and in the gross disease settings. Materials and Methods Patient Selection A search of the University of Wisconsin Veterinary Medical Teaching Hospital (UW-VMTH) patient database identified dogs with MCTs who had received prednisone and VBL. Criteria for inclusion included absence of severe concurrent disease, tissue or slides available for histologic evaluation, and no concurrent systemic therapy other than prednisone and VBL. Previous chemotherapy was allowed if more than 3 weeks had elapsed since the last treatment. All patients underwent clinical staging, consisting of physical examination, complete blood count, serum chemistry panel, abdominal ultrasound, palpation and/or needle aspiration cytology of the regional lymph node, and bone marrow aspirate and/or buffy coat smear.

2 492 Thamm, Mauldin, and Vail Patients were divided into 3 populations based on the adequacy of local tumor control. Gross disease was defined as the presence of a measurable tumor. Microscopic disease was defined as the absence of measurable disease but histologic evidence of tumor cells extending to the margins of surgical excision. Adequate local therapy (ALT) was defined as appropriately aggressive surgery with no histologic evidence of tumor cells at the surgical margin, or marginal excision followed by radiotherapy (RT). The RT consisted of either 15 daily 3.2-Gray (Gy) treatments given over 3 consecutive weeks (Monday Friday) to a total dose of 48 Gy (4,800 rads), or 4 once-weekly 8-Gy treatments given to a total dose of 32 Gy (3,200 rads). RT treatments were performed using a 60 Co teletherapy unit (Theratron 80, Atomic Energy of Canada Ltd, Kanata, Ontario, Canada), with 0.5-cm tissueequivalent bolus material over skin incisions. Chemotherapy Administration Prednisone was administered orally at an initial dosage of 2 mg/kg daily, and this dosage was tapered and discontinued over weeks. VBL was given as a rapid intravenous bolus at 2 mg/m 2 every 1 2 weeks. Although some variation occurred in the schedule of treatment, 3 of 41 of patients received a protocol consisting of weekly treatments for 4 weeks, followed by 4 biweekly treatments. In the context of macroscopic disease, biweekly treatment was continued for as long as it was believed to be effective based on objective response criteria. Histology/AgNOR All histology slides were reviewed by a single pathologist (EAM), blinded as to treatment outcome and tumor location. Histologic grade was assigned according to Patnaik s criteria. 4 AgNOR frequency was assessed from 5- m histologic sections according to a recently published protocol. 13 Briefly, after deparaffinization in xylene and rehydration in various ethanol solutions and distilled water, a silver nitrate solution was pipetted onto each section of tumor. Slides were incubated in darkened, humidified chambers for 45 minutes and then rinsed. The slides were coverslipped and evaluated by a single individual blinded as to treatment outcome and histologic grade. AgNOR frequency per mast cell nucleus was determined by counting AgNOR in 100 cells in random representative areas of the tumors viewed under oil immersion (1,000 ) by means of light microscopy. Patient Assessment Antitumor responses were assessed according to standard criteria. Briefly, a complete response (CR) was defined as the disappearance of all detectable tumor for at least 28 days, and a partial response (PR) was defined as a reduction of 50% or greater in the sum of all tumor volumes for at least 28 days. Stable disease was defined as less than 50% reduction, or less than 25% increase in total volume. Progressive disease was defined as a greater than 25% increase in total volume, or the appearance of new masses or metastasis. Follow-up information was obtained through recheck examinations at the UW-VMTH, or through telephone communication with the referring veterinarians or owners. Time to treatment failure (TTF) was defined as the time interval between the initiation of chemotherapy and the development of recurrence, metastasis or new MCTs. Survival time (ST) was defined as the time interval between the initiation of chemotherapy and death. Statistical Analysis Table 1. Characteristic Median age (years) (range) Median weight (kg) (range) Sex Spayed female Castrated male Intact male Intact female Breed Labrador Retriever Golden Retriever Mixed breed Shar-Pei Poodle Other Previous chemotherapy Previous prednisone Local recurrence at presentation Positive lymph node Multiple cutaneous tumors Local treatment Gross disease Microscopic disease Adequate local therapy Patient characteristics.. 8 (3 13) 2 (3 56) Variables assessed for value as predictors of outcome (ST and TTF) included age, sex, breed (retriever breed versus other), weight, tumor location ( historically unfavorable, that is, oral cavity, mucous membranes, perianal, preputial, scrotal, inguinal, or subungual, versus other), number of cutaneous tumors (1 versus multiple), lymph node status, locally recurrent tumor at time of treatment initiation, tumor ulceration, previous chemotherapy, previous prednisone therapy, histologic grade, AgNOR frequency, type of local therapy, concurrent RT, and eosinophil count. These variables were selected based on previous reports and anecdotal evidence. Univariate analysis to assess for prognostic value of the different variables was performed by the Kaplan Meier product limit method. 22 Differences in TTF and ST between dogs grouped in different covariates were analyzed using the generalized Wilcoxon test. 23 This was then followed by a multivariate model in which significant variables identified in the univariate model were included. 24 The AgNOR frequency was compared with histologic grade using a nonparametric equivalent of Student s t-test. 25 Breed representation was compared with the hospital population during the study time using 2 analysis. 26 Statistical calculations were performed using commercial statistical software packages (Prism, GraphPad Software Inc, San Diego, CA; BMDP, Statistical Solutions Inc, Boston, MA). A P value of.05 was considered significant for all statistical analyses. Results The initial database search revealed 55 patients with MCT that had received VBL and prednisone. Fourteen patients were excluded from analysis for the following reasons: other concurrent chemotherapy given (n 8), histology not available for review (n 5), incomplete staging (n 1). Thus, 41 dogs were available for study (Table 1). A trend toward significant overrepresentation occurred in Labrador Retrievers and Golden Retrievers (P.08), and Shar-Peis were significantly (P.001) overrepresented when compared to the general VMTH population. Five dogs had previous chemotherapy, consisting of oral prednisone and vincristine (n 3), liposome-encapsulated doxorubicin (n 1), or prednisone and asparaginase (n 1). Eighteen dogs were classified as having gross disease, as having microscopic disease, and 16 as having ALT. Lymph node metastasis was documented cytologically or

3 Treatment of Canine Mast Cell Tumor 493 Fig 1. Survival curves, according to response to therapy, for 15 dogs with macroscopic mast cell tumors receiving prednisone and vinblastine chemotherapy; P.005. histologically in 22 patients, and lymph node excision was performed in of these 22, before or concurrent with the initiation of chemotherapy. Three dogs had systemic disease, as evidenced by demonstration of inappropriately large numbers of mast cells in bone marrow aspirate (n 1) or buffy coat preparation (n 2). One dog had signs consistent with gastrointestinal ulceration (melena, inappetance, vomiting). Eighteen dogs had more than 1 cutaneous tumor at presentation. Eighteen patients had intermediately differentiated (Patnaik grade II) MCTs, and 23 had poorly differentiated (Patnaik grade III) tumors. AgNOR frequency was assessed in 2 patients. The median AgNOR frequency was 3.04 per mast cell nucleus, with a range of statistical association was found between histologic grade and AgNOR frequency in our patient population (P 0.48). The median number of VBL doses administered per patient was 8, with a range of 1 0. Four patients treated in the adjuvant setting received VBL every other week for 6 or 8 treatments, in place of the standard protocol described above. Adverse effects were noted in 20% (8/41) of patients, usually after the 1st dose of VBL. These were considered mild in 6 patients, and severe in 2. Mild adverse effects included self-limiting vomiting in 2 patients, neutropenia without evidence of sepsis in 5 (-day post-vbl neutrophil count less than 2,000/ L), and lethargy or soft stool in 1. Severe adverse effects consisted of refractory vomiting and severe neutropenia (neutrophil count less than 500/ L) with fever after the 1st VBL dose in 2 patients. Both owners whose pets experienced severe adverse effects elected to discontinue VBL after the 1st dose. However, these 2 dogs were included in survival and TTF analysis. Fifteen of patients with measurable disease were evaluable for response. Overall response rate in this population was 4% (/15), consisting of 5 CRs and 2 PRs. Median response duration was 154 days (28 to 645 days). Three of 5 (60%) intermediate-grade tumors responded, and 4 of 10 (40%) high-grade tumors responded. statistical association was found between histologic grade and duration or completeness of response. Median ST (MST) in the responding group was significantly longer than in the nonresponders (median not reached versus 0 days, P 0.005) (Fig 1). As adjuvant therapy after incomplete surgical resection (microscopic disease group), VBL and prednisone treatment conferred a 5% 1- and 2-year disease-free rate. Ten dogs received RT immediately before, or in conjunction with, prednisone and VBL. RT was performed both on animals with gross disease (n 4) and animals with microscopic disease (n 6). Seven patients received 48 Gy, and 3 patients received 32 Gy. statistical difference was found in outcome based on type of RT protocol employed. Animals with gross disease receiving concurrent RT were considered ineligible for assessment of response. However, 1 patient receiving RT had a 2nd lesion outside the irradiated field that was considered evaluable. Fifteen patients received additional therapy after having failed prednisone and VBL, including surgery (n ), RT (n 3), and chemotherapy (n 8) (additional prednisone and VBL, vincristine, cyclophosphamide, chlorambucil, asparaginase, chloroethylcyclohexylnitrosourea [lomustine], cyclophosphamide vincristine doxorubicin prednisone combination, liposome-encapsulated doxorubicin, and dolastatin-10, an investigational drug). Several dogs received more than 1 additional chemotherapy drug, or combinations of chemotherapy and additional surgery or RT. Of the study population, 22 patients were censored from survival analysis for the following reasons: alive without evidence of disease (n 13), dead due to unrelated causes (n 4), lost to follow-up without evidence of disease (n 2), alive with disease (n 2), lost to follow-up with disease (n 1). Nineteen patients were also censored from TTF analysis. Prognostic factors that were subjected to univariate analysis as predictors of survival are shown in Table 2. Presence of a locally recurrent tumor at treatment initiation, type of local therapy, AgNOR frequency, lymph node status, tumor ulceration, histologic grade, and the administration of previous chemotherapy were all significant predictors of survival. Identical factors were significant for predicting time to treatment failure, with the exception of the administration of previous chemotherapy (data not shown). Table 3 shows survival data of different patient subsets when analyzed based on both type of local treatment and histologic grade. It must be stressed that these survival data are based on observations made in a small number of cases. Upon multivariate analysis, the overall median TTF was 31 days, with 49% progression-free at 2 years. After accounting for all other variables, statistically significant prognostic variables for treatment failure were histologic grade (P.005), recurrent tumor (P.001), and presence of gross disease (P.001). Seven of 23 patients (30%) whose disease was controlled at least grossly by surgery (microscopic and adequate local therapy populations) developed progression of MCTs. Recurrence at the site of previous tumor resection occurred in 1 patient, 5 developed disease elsewhere, and 1 had both local recurrence and distant disease. The overall MST was not reached, with a median follow-up of 59 days. Sixty-three percent of patients were alive at 1 year, and 56% were alive at 2 years. Statistically significant prognostic factors for survival after multivariate analysis were recurrent tumor (P.001) and histologic grade (P.012) (Figs 2, 3).

4 494 Thamm, Mauldin, and Vail Table 2. Univariate analysis of prognostic variables for effect on survival time for 41 dogs with mast cell tumors treated with prednisone and vinblastine. Variable Recurrent Local treatment Gross Micro/ALT AgNOR frequency* Histologic grade III II Lymph node status Positive Negative Ulceration Previous chemotherapy Concurrent RT Age 8 y 8 y Breed n-retriever Retriever. of Tumors 1 1 Sex Female Male Previous prednisone Location Unfavorable Other Eosinophil count b 284/ L 284/ L. Dogs Median Survival Hazard Ratio (Days) P Value a (95% CI) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) CI, confidence interval;, median not reached; ALT, adequate local therapy; RT, radiotherapy. a P values were calculated with the generalized Wilcoxon test. b Values not assigned for all dogs.

5 Treatment of Canine Mast Cell Tumor 495 Table 3. Survival time associated with histologic grade and type of local therapy in 41 dogs with mast cell tumors treated with prednisone and vinblastine.. Cases OYS (%) TYS (%) MST (Days) P value Gross, grade II Gross, grade III Micro, grade II Micro, grade III ALT, grade II ALT, grade III , OYS, one-year survival; TYS, two-year survival; MST, median survival time; Gross, gross disease; Micro, microscopic disease;, median not reached; ALT, adequate local therapy. Discussion The adverse effects of prednisone and VBL at the doses used in this study were acceptable, with only 5% of dogs developing serious toxicity necessitating hospitalization or treatment discontinuation, and no treatment-related deaths. This is in contrast to the reported adverse effects of singleagent vincristine in dogs with MCTs, where 32% of patients required premature discontinuation of treatment or hospitalization because of toxicity. 16 Although the overall response rate in patients with measurable disease of 4% is slightly lower than those reported employing more complex multiagent chemotherapy protocols, 9,1 the percentage of CR in this study was higher than in the studies cited above. In addition, the median response duration of 153 days is longer than the 53 day response duration reported by Gerritsen et al, 9 and the MST of 245 days in our patient population with gross disease exceeds that reported by Elmslie 1 for a similar patient population (5 months). Prednisone and VBL provided longer survival in patients with grade III MCTs than has surgery alone has in previous reports, with an MST of 331 days, and 45% of patients with grade III MCTs alive at 1 and 2 years. In addition, a majority of the patients (5%) with grade III tumors did not have surgical excision of their tumor at the time of treatment, and therefore constitute a less favorable population than the patient populations in previous studies. 2,4, It must be acknowledged that comparison with historical controls is a suboptimal method of assessing treatment effect, and a randomized phase III study with a control arm is necessary to convincingly demonstrate a survival advantage. The demographics of the patients in this study are similar to those reported elsewhere, 4,8 with the exception that Shar- Peis have not previously been identified at increased risk, and that Boxers were not overrepresented in our population. It is interesting to note that 39% (/) of the patients with grade II MCTs had evidence of lymph node metastasis. However, dogs with evidence of lymph node metastasis were more likely to receive adjuvant chemotherapy than dogs without metastasis at our institution, and thus this group of animals likely does not represent the patient population with grade II MCTs as a whole. Previous studies have identified histologic grade, Ag- NOR frequency, and clinical stage as prognostic factors for dogs with MCTs. Tumor ulceration, which carried prognostic significance in our study, could be a surrogate marker for large tumor size, given that ulcerated tumors might be larger than nonulcerated tumors. Alternatively, ulceration might correlate with tumors that are not amenable to surgical resection. However, 33% (2/6) of ulcerated tumors in our study were resected before chemotherapy. Unfortunately, tumor size was often not discernible in our patient population because of the large number of dogs that presented having had surgery performed elsewhere. The significant difference in survival between the gross Fig 2. Survival curves, according to presence of recurrent tumors at time of chemotherapy administration, for 41 dogs with mast cell tumors receiving prednisone and vinblastine chemotherapy; P Fig 3. Survival curves, according to histologic grade, for 41 dogs with mast cell tumors receiving prednisone and vinblastine chemotherapy; P.0124.

6 496 Thamm, Mauldin, and Vail disease category and other categories of local treatment underscores the importance of surgery as a mainstay of treatment for canine MCTs. In addition, the data suggest that adjuvant therapy such as prednisone and VBL is best employed after the initial tumor resection, rather than at the time of recurrence. The risk of metastasis and death from tumors has been documented to increase in human patients with soft tissue sarcoma whose tumors are recurrent, 2 and it is logical that the same finding would exist in tumors of veterinary patients. The fact that no apparent difference existed in survival between dogs with single cutaneous tumors and multiple cutaneous tumors lends credence to the theory that multiple cutaneous MCTs may represent independent tumors arising de novo, rather than an aberrant form of metastasis. Thus, early and aggressive therapy for each cutaneous MCT that occurs may still confer a favorable outcome. In addition, the current World Health Organization staging criteria, 28 in which animals with multiple cutaneous tumors are assigned a higher clinical stage, may not be a reliable system for predicting outcome given the current findings. Our inability to detect an association between prognosis and MCTs in locations anecdotally associated with a poor prognosis (subungual, inguinal, preputial, scrotal, perianal, oral) 1 may reflect a negation of this effect by the addition of prednisone and VBL therapy, or may simply be a function of inadequate case number. Our failure to detect a significant difference in outcome between patients that had adequate local therapy and patients with microscopic disease may represent a similar phenomenon, because only patients with microscopic disease were identified. The long-term (2-year) disease-free survival rate in the population of patients with microscopic disease, that is, incomplete surgical resection, was 5%. Although variable recurrence rates have been reported after surgery alone, our clinical experience suggests that the majority of dogs with incompletely excised MCTs will develop local tumor recurrence, and thus the addition of prednisone and VBL represents an improvement over marginal excision alone. Although this seems less effective than the 85 95% 2-year disease-free rate conferred by surgery plus RT, 8,29,30 prednisone and VBL may represent a reasonable therapeutic option for some owners for whom RT is not possible. In addition, data reported regarding RT for MCTs refer mostly to dogs with intermediate-grade tumors. The fact that the majority of dogs with microscopic disease in this study had high-grade tumors (4/) thus may make comparisons with published results of RT difficult to interpret. It should be pointed out that this cohort of patients represents a small case number, and that results may vary with a larger sampling of patients with microscopic disease. In summary, we have demonstrated that the combination of oral prednisone and injectable VBL is an active regimen for the treatment of MCTs in dogs, and has a favorable adverse effect profile. Upon multivariate analysis, histologic grade, recurrent tumor, and presence of gross disease were prognostic indicators for treatment failure, and histologic grade and recurrent tumor were prognostic indicators for survival. The conclusions drawn must be accepted with understanding of the limitations inherent in a retrospective study, and the fact that a small degree of variation occurred in the protocol of drug administration. Again, the importance of early, aggressive local therapy (ie, surgery and/or RT), in conjunction with systemic therapy if indicated, cannot be overstressed as the first and most important treatment for canine MCTs. Acknowledgment We thank Ms Lynn Volk for assistance in performing the AgNOR staining. References 1. Vail DM. Mast cell tumors. In: Withrow SJ, MacEwen EG, eds. Small Animal Clinical Oncology, 2nd ed. Philadelphia, PA: WB Saunders; 1995: Bostock DE. The prognosis following surgical removal of mastocytomas in dogs. J Small Anim Pract 193;14: Hottendorf GH, Neilsen SW. Survey of 300 extirpated canine mastocytomas. Zentralbl Veterinarmed A 196;14: Patnaik AK, Ehler WJ, MacEwen EG. Canine cutaneous mast cell tumor: Morphologic grading and survival time in 83 dogs. Vet Pathol 1984;21: Hottendorf GH, Nielsen SW. Canine mastocytoma A review of clinical aspects. J Am Vet Med Assoc 1969;154: Grier RL, DiGuardo G, Myers R, et al. Mast cell tumour destruction in dogs by hypotonic solution. J Small Anim Pract 1995;36: Bostock DE, Crocker D, Harris K, et al. Nucleolar organiser regions as indicators of post-surgical prognosis in canine spontaneous mast cell tumours. Br J Cancer 1989;59: LaDue T, Price GS, Dodge R, et al. Radiation therapy for incompletely resected canine mast cell tumors. Vet Radiol Ultrasound 1998; 39: Gerritsen RJ, Teske E, Kraus JS, et al. Multi-agent chemotherapy for mast cell tumours in the dog. Vet Q 1998;20: Turrel JM, Kitchell BE, Miller LM, et al. Prognostic factors for radiation treatment of mast cell tumor in 85 dogs. J Am Vet Med Assoc 1988;193: Ayl RD, Couto CG, Hammer AS, et al. Correlation of DNA ploidy to tumor histologic grade, clinical variables, and survival in dogs with mast cell tumors. Vet Pathol 1992;29: Allan GS, Gilette EL. Response of canine mast cell tumors to radiation. J Natl Cancer Inst 199;63: Kravis LD, Vail DM, Kisseberth WC, et al. Frequency of argyrophilic nucleolar organizer regions in fine-needle aspirates and biopsy specimens from mast cell tumors in dogs. J Am Vet Med Assoc 1996;209: Simoes JPC, Schoning P, Butine M. Prognosis of canine mast cell tumors: A comparison of three methods. Vet Pathol 1994;31: McCaw DL, Miller MA, Ogilvie GK, et al. Response of canine mast cell tumors to treatment with oral prednisone. J Vet Intern Med 1994;8: McCaw DL, Miller MA, Bergman PJ, et al. Vincristine therapy for mast cell tumors in dogs. J Vet Intern Med 199;11: Elmslie R. Combination chemotherapy with and without surgery for dogs with high grade mast cell tumors with regional lymph node metastases. Vet Cancer Soc Newsl 199;20:6.. Golden DL, Langston VC. Uses of vincristine and vinblastine in dogs and cats. J Am Vet Med Assoc 1988;193: Rowinsky EK, Donehower RC. Antimicrotubule agents. In: DeVita VT, Hellman S, Rosenberg SA, eds. Cancer: Principles and Practice of Oncology, 5th ed. Philadelphia, PA: Lippincott-Raven; 199: Rosenthal RC. Clinical applications of vinca alkaloids. J Am Vet Med Assoc 1981;19:

7 Treatment of Canine Mast Cell Tumor Theilen GH, Madewell BR. Clinical application of cancer chemotherapy. In: Theilen GH, Madewell BR, eds. Veterinary Cancer Medicine, 2nd ed. Philadelphia, PA: Lea & Febinger; 198: Kaplan EL, Meier P. nparametric estimation from incomplete observations. J Am Statist Assoc 1958;53: Breslow N. A generalized Kruskal Wallis test for comparing k samples subject to unequal patterns of censorship. Biometrika 190; 5: Cox DR. Regression models and life tables. J R Statist Soc Ser B 192;34: Conover WJ. Practical nparametric Statistics. New York, NY: John Wiley; 1980: Shott S. Statistics for Health Professionals. Philadelphia, PA: WB Saunders; 1990: Benjamin RS, Rouesse J, Bourgeois H, et al. Should patients with advanced sarcomas be treated with chemotherapy? Eur J Cancer 1998;34: Owen LN, ed. World Health Organization TNM classification of tumors in domestic animals, 1st ed. WHO, Geneva, Switzerland; 1980: Frimberger AE, Moore AS, LaRue SM, et al. Radiotherapy of incompletely resected, moderately differentiated mast cell tumors in the dog: 3 cases ( ). J Am Anim Hosp Assoc 199;33: Al-Sarraf R, Mauldin GN, Patnaik AK, et al. A prospective study of radiation therapy for the treatment of grade 2 mast cell tumors in 32 dogs. J Vet Intern Med 1996;10:36 38.