Mast cell tumors (MCTs) are the most commoncutaneous tumors in dogs, accounting for 7% to

21% of all skin tumors.1,2 These tumors are highly vari-able in regard to location, biological behavior, and treat-ment response. Factors including histologic grade,World Health Organization stage, duration of disease,argyrophilic nucleolar organizer region count, com-pleteness of excision, DNA ploidy, tumor location,breed, and abnormal expression of p53 tumor suppres-

sor gene have been evaluated as prognostic indicatorsfor MCTs in dogs.3-15 Tumor location affects prognosisfor dogs with incompletely resected MCTs treated withradiation therapy.5 Tumors of the limbs are associatedwith lower metastatic rate, longer duration of remis-sion, and longer duration of survival than tumors locat-ed on the trunk.5 There have been reports that dogswith a tumor of high grade or in a particular body loca-tion (ie, perineum, pinna, prepuce) have poor responseto local-regional radiation therapy.16-19 The purpose ofthe study reported here was to analyze retrospectivelythe efficacy of local-regional radiation therapy with analternating-day protocol in dogs with incompletelyexcised histologic grade-III solitary MCTs withoutregional lymph node involvement (stage 0).

Criteria for Selection of CasesThe medical records for all dogs admitted to our

hospital for irradiation of localized, incompletelyresected, histologically confirmed MCTs of the skin(any grade or location) between 1987 and 1998 werereviewed. Medical records of those dogs with histo-logic grade-III tumors (stage 0) and prior excision ofthe localized tumor without prior chemotherapy wereincluded in the study. To appropriately evaluate theradiotherapy protocol, records of dogs that receivedconcurrent or post-irradiation chemotherapy (includ-ing the use of prednisone) were excluded from analy-ses.

ProceduresMedical records were reviewed for signalment, lab-

oratory data (including CBC, serum biochemical pro-file, urinalysis, buffy coat smear, and regional lymphnode aspiration or biopsy), clinical staging (2-view sur-vey radiographs of the thoracic and abdominal cavities,and abdominal ultrasonography), growth characteris-tics of the tumor, changes in regional lymph nodes,concurrent diseases, treatment methods and associatedcomplications, and outcome. Regional lymph nodemetastasis was defined as a large population of mastcells that infiltrated the lymph node, poorly differenti-ated mast cells in the lymph node, or both. For all cases,biopsy specimens were reviewed to confirm the diagno-sis, grade, and stage of the tumors.a Tumors were stagedaccording to a modified version of the classification sys-tem developed by the World Health Organization.5,20

Tumors were graded according to the system describedby Patnaik et al4 in which grade I has the best prognosisand grade III has the worst prognosis.4

Treatment methods—Dogs were treated with radi-ation therapy via an external-beam megavoltage tech-

JAVMA, Vol 224, No. 1, January 1, 2004 Scientific Reports: Retrospective Study 79

SMALLANIMALS

Efficacy of radiation therapy for incompletely resected grade-III mast cell tumors

in dogs: 31 cases (1987–1998)

Kevin A. Hahn, DVM, PhD, DACVIM; Glen K. King, DVM, DACVR; Janet K. Carreras, VMD, DACVIM

Objective—To determine the efficacy (durations ofremission and survival) of an alternating-day radiationprotocol for incompletely excised histologic grade-IIIsolitary mast cell tumors (MCTs) in dogs. Design—Retrospective study.Animals—31 dogs.Procedure—Radiation (52 Gy in an 18-fraction alter-nating-day protocol) was delivered to an area bor-dered by margins ≥ 3 cm around the surgical scar andto the associated local-regional lymph nodes. Dogswere not given chemotherapeutic agents concurrent-ly or after radiation. Information on signalment, dura-tion of remission, and survival time was obtainedfrom medical records. Results—Median and mean durations of remissionwere 27.7 and 17.0 months, respectively (range, 1 to47 months). Median and mean durations of survivalwere 28 and 20 months, respectively (range, 3 to 52months). Dogs with tumors located on the skin of thepinna, perineum, and prepuce had a median durationof remission greater than dogs with tumors located atother sites (27.7 and 14.4 months, respectively). Dogswith tumors ≤ 3 cm in maximum diameter beforesurgery survived longer than dogs with tumors > 3cm (31 and 24 months, respectively). The remissionrate was 65% and survival rate was 71% at 1 yearafter treatment. Sixteen dogs that were euthanatizedhad complications associated with local-regionaltumor progression. Systemic metastases to liver,spleen, intestine, and bone marrow were detected in1 dog. Conclusions and Clinical Relevance—Without fur-ther treatment, incompletely excised grade-III mastcell tumors have high local-regional recurrence; local-regional treatment with radiation may effectively beused to manage many such tumors. (J Am Vet MedAssoc 2004;224:79–82)

From Gulf Coast Veterinary Oncologists 1111 W Loop South, Ste150, Houston, TX 77027.

The authors thank Ginger Haug and Sherry Stephens for technicalassistance.

Address correspondence to Dr. Hahn.

03-03-0088.qxd 12/9/2003 2:20 PM Page 79

nique with a cobalt 60-unit or 4-MeV linear accelera-tor, with a source-to-skin distance of 80 cm. Two meth-ods were used to calculate the radiation dose for thetumor. Prior to 1993, tumor dose was calculated by useof a single point along the central axis at the base of thesurgical scar. Tumor invasion and size were determinedby use of survey radiography and physical examina-tion. Starting in 1993, patients received a computedtomography scan, and a 2-dimensional treatment-plan-ning computer was used to calculate the dose-to-tumorvolume. Doses to surrounding critical structures werecalculated to ensure that the dose to those tissues didnot exceed tolerance levels. The radiation field includ-ed the region of the excisional scar and regional lymphnodes, with 3-cm margins. The dose to the lymph nodewas equivalent to the dose to the primary tumor. Dogswere treated with 18 fractions of 2.9 Gy on an alter-nate-day (Monday, Wednesday, Friday) schedule for atotal dose of 52.2 Gy. To establish our protocol, weassumed a biologically effective dose for acute- andlate-affected tissues (ie, α/β) of 3 Gy for late-respond-ing tissues. Conventional daily treatment of 2 Gy/frac-tion (total dose, 50 Gy) has a biologically effective dose(ie, E/α) for late-responding tissues equal to 83 Gy. Foran alternate day protocol with 3 Gy/fraction (totaldose, 54 Gy dose), the E/α is 88 Gy. Our alternate-dayprotocol with 2.9 Gy/fraction (total dose, 52.2 Gy) hada biologically effective dose (E/α) of 86.6 Gy, whichwas similar to other conventional treatment protocolvalues.21

Data analyses—Follow-up information concern-ing response to treatment and outcome was obtainedthrough hospital visits or by telephoning the referringveterinarian or the owner. Survival time was calculatedas the interval from the start of treatment to death or tothe date on which the dog was last known to be alive.22

Because all dogs had microscopic tumor involvementonly within the surgical site (ie, stage 0), duration ofremission was calculated as the interval from the startof treatment to time of visible local-regional tumorregrowth (eg, within the surgical scar or draininglymph node) or distant metastasis. Local tumor pro-gression was confirmed cytologically or histologically.Regional or systemic metastasis was confirmed histo-logically, cytologically, and radiographically. Differencesin distribution of dog and tumor characteristics wereanalyzed by use of a t test for continuous variables andPearson χ2 statistics for categoric variables. Variablesexamined as indicators of prognosis included sex, age,presurgical tumor size, tumor stage, and tumor loca-tion. Survival and local control rates were computed byuse of the product-limit method.23 Actuarial estimatesof survival and local control distributions were testedfor statistical differences by use of log-rank statistics.Dogs were censored in the survival analysis if they werelost to follow-up, death was not caused by MCT ortreatment, or relapse had not occurred before the end ofthe study period. The Cox proportional hazards regres-sion model was used to determine independent prog-nostic factors for remission duration and survivaltime.24 The relative risk of tumor recurrence was esti-mated by use of the hazard rate ratio. Differences were

considered significant at P < 0.05. Statistical analyses ofdata were performed with computer software.b

ResultsPatient characteristics—Medical records of 31 dogs

met study inclusion criteria. Twenty-one dogs were pure-bred; Labrador Retrievers (n = 7), Golden Retrievers (6),and Boxers (6) were most often represented. Fifteen dogswere spayed females, 10 were neutered males, 3 weresexually intact females, and 3 were sexually intact males.Dogs ranged from 3 to 18 years old (mean, 9.4 years;median, 9 years) and from 5 to 59 kg (11 to 129.8 lb) inbody weight (mean, 15.2 kg [33.4 lb]; median, 16 kg[35.2 lb]). Median duration from surgical removal toirradiation was 17 days (range, 10 to 42 days; mean, 22 days). Although alopecia was observed in most dogs,treatment delays associated with complications of radia-tion therapy (edema, moist desquamation, or erythema)were not observed in any dog.

Tumor behavior—Median tumor size prior to exci-sion was 3.9 cm at the widest diameter (range, 1 to 6 cm;mean, 3.1 cm). After radiation therapy, 16 of the 31 dogshad developed enlarged ipsalateral regional lymph nodescaused by metastasis. One dog had antemortem evidenceof systemic tumor involvement (hepatosplenomegaly,gastrointestinal metastasis, and tumor cells in bone mar-row aspirate but not buffy in coat smear). A correlationwas found between small presurgical tumor diameter (≤3 cm vs > 3 cm) and increased duration of remission (χ2

test; P = 0.034). A correlation was also found betweentumor location (high-grade location vs truncal location)and increased duration of survival (χ2 test; P = 0.043).None of the dogs had tumor cells in a buffy coat smearor bone marrow core aspiration biopsy specimen whenfirst evaluated.

Prognostic factors—Seventeen dogs died or wereeuthanatized because of clinical signs associated withtumor recurrence or progression (16 from ipsalaterallocal-regional lymph node enlargement, 1 from systemicmetastasis). Three dogs died from unrelated causes(pancreatitis, disk disease, or traumatic injuries). Elevendogs were alive with no evidence of disease progressionat last follow-up (follow-up range, 12 to 52 months;median, 25 months). Median and mean durations ofremission were 27.7 and 17.0 months, respectively(range, 1 to 47 months). The product-limit estimates ofthe 1-year and 2-year tumor remission rates (mean ±SEM) were 65 ± 8% and 26 ± 11%, respectively. Medianand mean durations of survival were 28 and 20 months,respectively (range, 3 to 52 months). The product-limitestimates of the 1-year and 2-year survival rates (± SEM)were 71 ± 5% and 39 ± 9%, respectively. Univariateanalysis of duration of remission revealed that tumorslocated at the pinna, perineum, and prepuce (previous-ly reported high-grade locations)16-19 had a longer (log-rank analysis, P = 0.045) duration of remission thandogs with tumors located at other sites (Fig 1).Univariate analysis of duration of survival revealed thatpresurgical tumor size was a significant (log-rank analy-sis, P = 0.003) prognostic indicator; dogs with maxi-mum tumor diameter ≤ 3 cm lived longer than dogswith larger tumors (Fig 2).

80 Scientific Reports: Retrospective Study JAVMA, Vol 224, No. 1, January 1, 2004

SMALLANIMALS

03-03-0088.qxd 12/9/2003 2:20 PM Page 80

DiscussionCanine mast cell tumors are histologically graded

on a scale of I, II, and III or they are assessed as beingwell-differentiated, intermediate, or poorly differentiat-ed.3,4 There is a correlation between the grade, ordegree of differentiation, and survival time.3,4 In 1study,4 93% of dogs with well-differentiated tumorswere alive at 1,500 days, but only 6% of dogs withpoorly differentiated tumors were alive at 1,500 days.4

When complete excision is not obtained or cannotbe attempted, radiotherapy is a good treatment optionto consider.5,25-29 If possible, surgery should be per-formed before radiotherapy to reduce the tumor to amicroscopic volume. Median durations of remissionafter local radiation therapy, regardless of grade orcompleteness of excision, range from 2 to 125 months(median, 60 months).5,25,27-29 The response of grade-Iand -II canine MCTs is regarded as excellent after radi-ation, but response is poor for grade-III tumors.27,28

Frimberger et al27 reported that radiation therapy forincompletely excised moderately differentiated cuta-neous MCTs in 37 dogs resulted in 3% recurrence at 1year and 7% recurrence at 3 years after treatment.27

LaDue et al28 used a similar protocol and reported thatradiation therapy for incompletely excised MCTs in 56dogs had median duration of remission of 32.7 months(95% confidence interval, 19 to 70 months).28 In astudy by Turrel et al5 that evaluated 85 dogs, medianduration of remission was 17 months.5 Percentages ofdogs that were tumor-free at 1 and 2 years were 78.8%and 77.0%, respectively. Factors significantly associat-ed with duration of remission were clinical stage (P < 0.001) and neoplasm location (P = 0.019).Prognostic factors significantly associated with sur-vival rates were clinical stage (P < 0.001), neoplasmgrade (P = 0.006), and neoplasm location (P = 0.034).The prognosis after radiation treatment for grade-IIIMCTs, with or without regional lymph node metasta-sis, has been regarded by many as poor, compared withthat of dogs with grade-I or grade-II tumors.5,18,28,29

However, few studies have reported the response toradiation therapy in dogs with grade-III MCTs treatedwith a uniform radiotherapy protocol. In a study of 18dogs with and without grossly evident tumors, totaltumor-delivered radiotherapy doses ranging from 36 to48 Gy in 9 to 12 equal fractions were given over 19 to27 days. In some instances, local-regional nodes werenot irradiated. The 1- and 2-year survival rates for alldogs regardless of grade (n = 85) were 78.8% and77.0%, respectively, and were comparable to the resultsof our study. The 1-year tumor-free rate was 55%,lower than the dogs of our study.

Our study confirmed the predictive value ofpresurgical tumor size of MCTs for duration of remis-sion.5,7,30 Dogs with tumors ≤ 3 cm in maximum diam-eter before surgery survived longer than dogs withtumors > 3 cm (31 and 24 months, respectively).Staging is not the same as grading. Grade is the assess-ment of the degree of cellular anaplasia and potentialfor growth. Staging is the determination of actualmetastasis or lack thereof.1,20 In addition, the growthrate of the tumor but not the size may be an importantprognostic indicator.3,7 Dogs with MCTs that grow > 1 cm/wk have a 25% chance of living an additional30 weeks.7

In our study, duration of survival was associatedwith tumor location. Dogs with MCTs located in high-grade locations had median times to recurrences thatwere longer than dogs with tumors in other locations(27.7 and 14.4 months, respectively). Canine MCTs inhigh-grade regions are considered by many investiga-tors to be a negative prognostic feature and associatedcommonly with metastasis to local and deep lymphnodes.5,16-19 The inconsistency between the findings ofothers and those of our study is not unexpected.Analysis of other predictive factors in dogs with MCTs,such as chromosome nucleolar organizer regionsstained with silver, matrix metalloproteinase activities,and c-kit tandem duplication mutations, has been eval-uated and correlates well with histologic grade but failsto substantially improve prognostic accuracy for treat-ment response or metastatic potential.2,6-9,11,12 Tumor

JAVMA, Vol 224, No. 1, January 1, 2004 Scientific Reports: Retrospective Study 81

SMALLANIMALS

Figure 1—Estimated duration of remission for dogs with incom-pletely excised grade-III mast cell tumors (MCTs) treated withradiation therapy. Dogs with tumors in low-grade locations (ie,the trunk) had significantly (P = 0.045) shorter duration of remis-sion than dogs with tumors in high-grade locations (ie, pinna,prepuce, and perineum).

Figure 2—Estimated duration of survival for dogs withincompletely excised grade-III MCTs treated with radiationtherapy. Dogs with tumors ≤ 3 cm in diameter beforesurgery lived significantly (P = 0.003) longer than dogs withlarger tumors.

03-03-0088.qxd 12/9/2003 2:20 PM Page 81

location may represent another inconsistent prognosticfactor for dogs.

Results of our study support the value of evaluat-ing tumor location, presurgical tumor size, and tumorgrade when considering radiation in the managementof grade-III MCTs. Our results confirm that grade-IIIMCTs are biologically aggressive tumors with highlocal-regional metastatic rates. However, many of thesetumors may be effectively managed with the local-regional alternating-day radiation therapy protocoldescribed in this report. One year after radiation ther-apy, the remission rate was 65% and the survival ratewas 71%.

aAntech Laboratories, Irvine, Calif.bGraphPad Prism, version 3.02, GraphPad Software Inc, San Diego,

Calif.

References 1. Dorn ER, Taylor D, Schneider R, et al. Survey of animal

neoplasms in Alameda and Contra Costa counties, California. J Natl Cancer Inst 1968;40:307–318.

2. Cohen D, Reif SS, Brodey RS, et al. Epidemiological analy-sis of the most prevalent sites and types of canine neoplasia observedin a veterinary hospital. Cancer Res 1974;34:2859–2868.

3. Bostock DE. The prognosis following surgical removal ofmastocytomas in dogs. J Small Anim Pract 1973;14:27–40.

4. Patnaik AK, Ehler WJ, MacEwen EG. Canine cutaneousmast cell tumor: morphologic grading and survival time in 83 dogs.Vet Pathol 1984;21:469–474.

5. Turrel JM, Kitchell BE, Miller LM, et al. Prognostic factors forradiation treatment of mast cell tumors in 85 dogs. J Am Vet Med Assoc1988;193:936–940.

6. Kravis LD, Vail DM, Kisseberth WC, et al. Frequency ofargyrophilic nucleolar organizer regions in fine-needle aspirates andbiopsy specimens and biopsy specimens from mast cell tumors indogs. J Am Vet Med Assoc 1996;209:1418–1420.

7. Bostock DE, Crocker J, Harris K, et al. Nucleolar organiserregions as indicators of post- surgical prognosis in canine sponta-neous mast cell tumors. Br J Cancer 1989;59:915–918.

8. Ayl RD, Couto CG, Hammer AS, et al. Correlation of DNAploidy to tumor histologic grade, clinical variables, and survival indogs with mast cell tumors. Vet Pathol 1992;29:386–390.

9. Simoes JPC, Schoning P, Butine M. Prognosis of canine mastcell tumors: a comparison of three methods. Vet Pathol 1994;31:637–647.

10. Miller DM. The occurrence of mast cell tumors in youngShar Peis. J Vet Diagn Invest 1995;7:360–363.

11. Ginn PE, Fox LE, Brower JC, et al. Immunohistochemicaldetection of p53 tumor-suppressor protein is a poor indicator ofprognosis for canine cutaneous mast cell tumors. Vet Pathol 2000;37:33–39.

12. Jaffe MH, Hosgood G, Taylor HW, et al. Immunohistochemicaland clinical evaluation of p53 in canine cutaneous mast cell tumors. VetPathol 2000;37:40–46.

13. Fan TM, Kitchell BE. Retrospective analysis of oral mast celltumors in dogs, in Proceedings. 19th Annu Vet Cancer Soc Conf1999;40.

14. Takahashi T, Kadosawa T, Nagase M, et al. Visceral mast celltumor in dogs: 10 cases (1982–1997). J Am Vet Med Assoc 2000;216:222–226.

15. Michels GM, Knapp DW, DeNicola DB, et al. Prognosis fol-lowing surgical excision of canine cutaneous mast cell tumors withhistologically tumor-free versus nontumor-free margins: a retrospec-tive study of 31 cases. J Am Anim Hosp Assoc 2002;38:458–466.

16. McManus PM. Frequency and severity of mastocytemia indogs with and without mast cell tumors: 120 cases (1995–1997). J Am Vet Med Assoc 1999;215:355–357.

17. Rassnick KM, Moore AS, Williams LE, et al. Treatment ofcanine mast cell tumors with CCNU (lomustine). J Vet Intern Med1999;13:601–605.

18. Ogilvie GK, Moore AS. Mast cell tumors. In: Ogilvie GK,Moore AS, eds. Managing the veterinary cancer patient: a practice man-ual. Trenton, NJ: Veterinary Learning Systems, 1995;503–514.

19. Thamm DJ, Mauldin EA, Vail DM. Prednisone and vinblas-tine chemotherapy for canine mast cell tumor—41 cases (1992–1997).J Vet Intern Med 13;5:491–497.

20. Consultation on the biological behavior and therapy of tumorsof domestic animals. Geneva: World Health Organization, 1978.

21. EJ Hall. Time, dose, and fractionation in radiotherapy. In:Hall EJ, ed. Radiobiology for the radiologist. 4th ed. Philadelphia: JBLippincott Co, 1994;211–230.

22. Freidman LM, Furberg CD, DeMets DL. Survival analysis.In: Freidman LM, ed. Fundamentals of clinical trials. 2nd ed.Littleton, Mass: PSG Publishing Co Inc, 1985;191–200.

23. Kaplan E, Meier P. Nonparametric estimation from incom-plete observations. J Am Stat Assoc 1958;53:457–481.

24. Cox DR. Regression models and life-tables. J Roy Stat Soc1972;34:187–220.

25. Al-Sarraf R, Mauldin GN, Patnaik AK, et al. A prospectivestudy of radiation therapy for the treatment of grade 2 mast celltumors in 32 dogs. J Vet Intern Med 1996;10:376–378.

26. Weisse C, Shofer FS, Sorenmo K. Recurrence rates and sitesfor Ggrade II canine cutaneous tumors following complete surgicalexcision. J Am Anim Hosp Assoc 2002;38:71–73.

27. Frimberger AE, Moore AS, LaRue SM, et al. Radiotherapy ofincompletely resected, moderately differentiated mast cell tumors in thedog: 37 cases (1989–1993). J Am Anim Hosp Assoc 1997;33:320–324.

28. LaDue T, Price GS, Dodge R, et al. Radiation therapy forincompletely resected canine mast cell tumors. Vet Radiol Ultrasound1998;39:57–62.

29. Chaffin K, Thrall DE. Results of radiation therapy in 19dogs with cutaneous mast cell tumors and regional lymph nodemetastasis. Vet Radiol Ultrasound 2002;43:392–395.

30. Gerritsen RJ, Teske E, Kraus JS, et al. Multi-agent chemother-apy for mast cell tumors in the dog. Vet Q 1998;20:28–31.

82 Scientific Reports: Retrospective Study JAVMA, Vol 224, No. 1, January 1, 2004

SMALLANIMALS

03-03-0088.qxd 12/9/2003 2:20 PM Page 82