Abstract
Feline cutaneous mast cell tumours (MCTs) are the second most common skin tumour in cats; but, unlike in dogs, there is currently no histological grading system for this type of tumour. This study recorded the signalment and anatomical location from a total of 287 records from MCTs submitted to a UK commercial diagnostic laboratory. Questionnaires to submitting practices were used to obtain follow-up data, and the histological features of 86 tumours were evaluated from 69 cats with a known outcome. Twelve of the 69 cats (17.4%) died of MCTs, with significantly lower survival times. The median age of cats presenting with MCTs was 11 years (range 5 months–19 years), with no sex or neutered status predilection. Some pedigree breeds were more susceptible to MCTs, particularly the Siamese, Burmese, Russian Blue and Ragdoll. The head was the most common site in younger cats, compared with the trunk in older cats. The number of tumours had no effect on survival. A new subcategory of well-differentiated MCTs with prominent multinucleated cells is described, and three of the five cats with this novel form died from MCT-related disease. There was an association between mitotic index and survival time. However, there was no significant association between histological type and survival.
Introduction
Mast cell tumours (MCTs) are one of the most common skin tumours encountered in feline practice, accounting for up to 21% of all cutaneous neoplasms in this species.1–3 Cutaneous MCTs are typically seen in older cats, 4 but the age at presentation ranges widely from younger than 12 months up to 19 years.1,5–8 The precise aetiology of feline MCTs is unknown. However, a genetic predisposition in Siamese cats has been cited in some studies.1,9
There is currently no grading system for feline cutaneous MCTs, and accurate prognostication is challenging. Tumours are classified as one of three histological types: well-differentiated mastocytic, pleomorphic mastocytic and atypical (formerly referred to as histiocytic).6,7,10 Although the typical feline cutaneous MCT is a well-differentiated and benign mass cured by complete surgical excision, a small but important proportion of these tumours are more biologically aggressive and will recur and/or spread to local lymph nodes or more distant sites within months of excision. 11 Cats presenting with single tumours are reported to have longer survival times than those developing multiple skin lesions. 7 Cutaneous tumours can also be secondary to primary visceral masses, when there is a much more guarded prognosis. 4 Histologically, the most reliable prognostic marker currently available is mitotic index (MI),2,7,10 with a high MI associated with a worse outcome. 7 However, this association is not consistent across the different histological types.7,10 Evaluation of clinical presentation together with other histological features is necessary to identify potentially aggressive tumours, including the histological type, the presence of histologically complete surgical margins, local recurrence, metastatic spread and the presence of primary visceral masses.
The aim of this study was to utilise the large number of feline cutaneous MCTs submitted to a commercial diagnostic laboratory in order to examine the signalment of affected cats and the anatomical location of their tumours. A retrospective study of these factors, together with histological features of the tumours and known clinical outcomes, was then made in an attempt to identify potentially useful prognostic indicators.
Materials and methods
Records from a large commercial diagnostic laboratory, Finn Pathologists (Diss, UK), were searched for all feline MCTs diagnosed based on fixed tissue samples submitted to the laboratory during the period from April 2006 to May 2011. These dates were selected based on the availability of archived material and to allow a minimum of 2 years’ follow-up for any given case. Clinical details, including sex and neuter status, age, breed and anatomical location, were recorded for all submissions diagnosed with cutaneous MCTs (n = 287 cats with 308 tumours).
The breed of cats in the study population was compared with the breed prevalence of the control population (n = 3771) using a χ2 test; the control population was based on cats from which fixed tissue samples were received by the laboratory throughout the study period and with any diagnosis, including MCT.
Anatomical location was specified for 228 tumours, and these were categorised as one of the following: (1) head (including the ear), (2) neck and trunk or (3) extremities (including limbs, paws and tail). Anatomical location was then further subcategorised as head, ear, neck, trunk, limbs, paws or tail.
Further information was sought from these 287 cases and was acquired for 69 cases using a questionnaire to the submitting practices, approved by the Royal Veterinary College Ethical Committee. Data requested included whether the cat had succumbed to MCT-related disease, non-MCT-related disease or was still alive, together with the date of death where relevant and whether the cat had developed further MCTs. For each of these 69 cases, haematoxylin and eosin (HE)-stained sections of each tumour were blindly reviewed by a pathologist (MJD), with some cases additionally reviewed by a second pathologist (KCS). Each individual tumour was assessed for histological type, MI, degree of eosinophil infiltration and the presence of lymphoid aggregates. MI was calculated as the number of mitotic figures per 10 high power fields (× 400). The number of eosinophils and lymphoid aggregates were each subjectively categorised as none, low, low–moderate, moderate, moderate–high or high.
Survival time was assessed as time from first diagnosis until either the date the questionnaire was returned by the submitting practice or the date that death/euthanasia occurred, whether due to MCT-related disease or other causes. The Kaplan–Meier method was used to calculate median survival times, with 95% confidence intervals. Log rank test was used to compare survival times between cats dying of MCT-related disease and those which did not succumb to MCT-related disease within 2 years. Five cases were excluded from survival analysis; three cases had abdominal masses and one had a thoracic mass, none of which were investigated further and therefore may or may not have represented MCT spread. The fifth case died from non-MCT-related causes, but the date of death was not specified. Data were analysed using IBM SPSS Statistics for Windows v19.0. Two categorical variables were analysed using χ2 (breed prevalence, sex), or Fisher’s exact tests if there were two binary variables (sex and neuter status). The association between a categorical and numerical variable was analysed using one-way ANOVA. If groups were not normally distributed, then the non-parametric Wilcoxon rank–sum test was used. A P value of <0.05 was considered significant.
Results
Breed
Of the 287 cats with MCTs diagnosed during the study period, 231 (80.5%) were non-pedigree (including cats denoted as domestic shorthair (DSH), domestic longhair (DLH), domestic cat and as simply ‘crossbreed’ on the original submission form, with no further indication as to breed) and 56 (19.5%) were pedigree breeds. Three cats had no breed stated on the form and were included in the non-pedigree group. Cats with multiple incidences of MCTs were included only once in this data set. There were 204 (71.1%) DSH cats (including ‘domestic cat’ and ‘crossbreed’), 24 (8.4%) DLH cats, 15 (5.2%) Siamese, seven (2.4%) each of Burmese and Persian/Persian cross, six each (2.1%) of Ragdolls and Russian Blues, and five (1.7%) Maine Coons, with the remaining breeds each represented by fewer than five individuals. Siamese, Burmese, Russian Blues and Ragdolls were present in greater numbers than expected within the study population (Table 1; P <0.05).
Prevalence of cutaneous mast cell tumours (MCTs) in different cat breeds
The breed prevalence of cats in the study population (n = 287) is compared with the breed prevalence of the control population (n = 3771), based on fixed tissue sample submissions to the laboratory throughout the study period. Those breeds indicated with * are present in greater than expected numbers within the study group compared with the control population (P <0.05). DSH includes domestic shorthair, domestic cat and crossbreed categories. Persian includes Persian cross. Rex includes both Cornish Rex and Devon Rex
DSH = domestic shorthair; DLH = domestic longhair; Norwegian FC = Norwegian Forest cat; BSH = British Shorthair; Exotic SH = Exotic Shorthair; European = European Shorthair
Age and sex
The median age for the 274 cats with their age specified on the submission form was 11 years, with a mean of 10.4 years and a range of 5 months to 19 years (Figure 1). Four of the cases were younger than 4 years of age. The median age of non-pedigree cats presenting with MCT was 11 years (mean 10.6 years), while the median age for pedigree cats was 10 years (mean 9.2 years). Of the 287 cats, 124 were female (44%; 19 entire, 105 neutered), 156 were male (56%; 27 entire, 129 neutered) and seven cats had no sex specified.
Age distribution of cats presenting with cutaneous mast cell tumour(s) (MCT[s]), for the 272 cats for which the age was stated on the submission form. Grey bars represent numbers of tumours occurring in pedigree cats, while black bars represent those arising in non-pedigree cats (domestic shorthair [DSH], domestic longhair [DLH], domestic cat and crossbreed)
Anatomical location
The anatomical location was recorded on the original submission forms for 228 tumours, with a further 80 tumours from unspecified sites. The head, including the ear, was the most common site overall with 87 lesions (38.1%), closely followed by the neck and trunk (76 lesions, 33.3%). Extremities, including the limbs, paws and tail, accounted for the remaining 65 tumours (28.5%). The head was by far the most common (82%) region affecting the 11 Siamese cats in the study population. Dividing the anatomical location into one of seven categories demonstrated the areas least likely to develop MCTs were the tail (nine lesions, 4%) and paws (nine lesions, 4%), followed by the neck (13 lesions, 6%) and ears (29 lesions, 13%). Young or juvenile cats (⩽7 years old) were more likely to have a MCT arising on the head than elsewhere (not significant [ns], P = 0.07); with older cats (⩾8 years old), the trunk was most often affected (including the neck; P = 0.04).
Signalment and outcome
Of the 69 cats with follow-up information, 12 (17.4%) died of MCT-related disease (Table 2). Five cases (7.2%) were excluded from the survival analysis; three of these cases had abdominal masses and one had a thoracic mass, which were not investigated further and therefore may or may not have represented MCT metastasis. The fifth case died from non-MCT related causes, but the date of death was not specified. The median survival time for these 12 cats was 150 days from the initial diagnosis, with a range of 7–598 days (95% CI 56, 445 days; Figure 2). The median survival time for the remaining 52 cats that did not succumb to MCT-related disease was not reached within 2 years, with a range of 4 days (from a non-related mass) to 1856 days (including cats who were still alive at the time of the study; P <0.001). The majority of cats in both groups were DSH cats. Of the 12 cats dying from MCT-related disease, 10 were DSH cats, one was recorded as ‘crossbreed’ and one was Siamese. There was no statistically significant difference between the percentage of cats alive 2 years after initial diagnosis when comparing pedigree vs non-pedigree cats (P = 0.3) or cats of different breeds. There was no correlation between sex and outcome, or neuter status and outcome. The median age of all 69 cats was 11 years, with a range of 3–17 years. For those cats dying from MCT-related disease within 2 years of diagnosis, the median age was 11.5 years, ranging from 4–16 years, while for the 52 cats who did not succumb to MCT-related disease, the median age was 11 years, varying from 3–17 years; ns, P = 0.6).
Signalment and histological features of cases with follow-up information
Cats that did not die of mast cell tumour (MCT)-related disease (cats may have died of an unrelated disease; eg, chronic renal failure or fibrosarcoma); †cats that have died from MCT-related disease within 2 years of the initial submission of a MCT; ‡non-pedigree cats includes domestic shorthair, domestic cat, domestic longhair and crossbreed; §from the initial population of 69 cats with follow-up information, five are excluded from further analysis due to inadequate diagnostic work-up of an abdominal or thoracic mass or no date of death being supplied
HPFs = high power fields
Kaplan–Meier survival curves comparing cats that did not succumb to mast cell tumour (MCT)-related disease within a minimum of 2 years (solid line, including those succumbing to non-MCT-related diseases) with those who died from MCT-related disease (black dashed line). The median survival time for those cats dying from MCT-related disease was 150 days from diagnosis. 95% confidence intervals are indicated (grey dashed lines) for the survival curve for cats dying of MCT-related disease
Histological type and outcome
Eighty-six cutaneous MCTs from 69 cats were evaluated histologically. Fifty-seven (66.3%) of these MCTs were classified histologically as well-differentiated, mastocytic type (Table 2); within this group, the median MI was two, with a range of 0–40. The degree of eosinophilic infiltration was most often low but varied considerably between tumours. The presence of lymphoid aggregates varied in a similar fashion, most often present in low numbers. Five (5.8%) of the tumours were classified as a novel subcategory of well-differentiated MCT with prominent multinucleated cells; these cells were larger, most often present towards the periphery of the mass and contained more than two nuclei per cell. The nuclei themselves demonstrated minimal anisokaryosis, and mitotic figures were not noted (Figure 3). Seventeen (19.8%) MCTs were classified histologically as pleomorphic, mastocytic type (Table 2), and the MI in these tumours ranged from 0–32, with a median of one. The degree of eosinophilic infiltration was variable but most often low, while lymphoid aggregates were typically present in moderate numbers. Seven (8.1%) of the MCTs were classified histologically as atypical type. The median MI in these tumours was six, ranging from 0–21. The degree of eosinophilic infiltration was most often high, and lymphoid aggregates were typically present in moderate numbers.
A well-differentiated mastocytic-type cutaneous mast cell tumour (MCT) with multiple prominent multinucleated cells. Five of the tumours in the study had similar histological characteristics and were subclassified as this novel histological type. The multinucleated cells were typically larger and contained more than two nuclei per cell. The nuclei themselves demonstrated minimal anisokaryosis and mitotic figures were not noted within these cells (haematoxylin and eosin stain, x 400)
From this initial population of 69 cats, five were excluded from further analysis; three of these cases had abdominal masses, and one had a thoracic mass, none of which were investigated further and therefore may or may not have represented MCT spread. The fifth case died from non-MCT-related causes, but the date of death was not specified. Of the 65 tumours occurring in cats that did not die of MCT-related disease, 45 (69.2%) were well-differentiated mastocytic-type tumours. Of the 16 tumours occurring in cats who succumbed to MCT-related disease within 2 years from submission of the mass, 11 (68.8%) were well-differentiated mastocytic-type tumours (Table 2). For pleomorphic mastocytic tumours, 15 (23.1%) of the 65 tumours present within cats that did not die of MCT-related disease were of this histological type, while one (6.3%) of the 16 tumours associated with death from MCT-related disease within 2 years was pleomorphic mastocytic (Table 2). Four (6.2%) atypical MCTs were present in cats that did not die of MCT-related disease (out of a total of 65 tumours), while one of the 16 tumours (6.3%) associated with MCT-related death was of this histological type (Table 2). For the five tumours in the novel subcategory (well differentiated with multinucleated cells), adequate follow-up was available for four cats. One cat (1.5%) was alive 1 year post diagnosis. However, the remaining three (18.8%) had died of MCT-related disease within 2 years. There was no significant difference in the prevalence of different histological types between those cats that survived and those that died of MCT-related disease (P = 0.6).
MI and outcome
Of the 69 cats with follow-up information, the median MI for all 86 tumours was 1.0, ranging from 0–40 (Table 2). In the tumours arising in the 12 cats that died from MCT-related disease, the median MI was 3.5 with a range of 0–32, compared with a median MI of 1.0 in the 52 cats who did not die of MCT-related disease (range 0–40); the difference in MI between these two groups of cats was statistically significant (P <0.002; Figure 4).
Mitotic index from cats dying from mast cell tumour-related disease within 2 years of diagnosis, compared with those who survived, based on 81 tumours from cats with a known clinical outcome. Outliers are indicated by a cross and represent either a single or a double observation at that value
Multiple tumours and outcome
Multiple cutaneous MCTs, either arising simultaneously, sequentially or a combination of both, occurred in 10 of the 69 cats (Table 2). Of these, three cats died of MCT-related disease within 2 years of diagnosis, and the remaining seven cats survived. One cat presented with a single cutaneous MCT four times in 3 years affecting different parts of the body, with no apparent adverse outcome. Several further cats were presumed to have had multiple tumours over time. However, not all of these masses were submitted for histopathology but were mentioned instead in the questionnaire responses; these cases were not included within the analysis for multiple tumours and outcome.
Discussion
This study confirms the findings of several other reports that feline cutaneous MCTs have a tendency to arise in older cats, most often affect the head region and that Siamese cats are at increased risk.1,4 Not currently reported in the literature are the other feline pedigree breeds which also appear at increased risk, namely the Burmese, Russian Blue and Ragdoll, which were present in greater numbers than expected within the current study population compared with the control population. Pedigree cats in the present study were more likely to develop MCTs at a younger age compared with non-pedigree cats, but were statistically no more likely to succumb to MCT-related disease. Some studies have reported that Siamese cats have a tendency to develop atypical MCTs;7,11 this was not evident in the present study, although the number of Siamese cats in the study was small.
Within the literature, the reported mean age at presentation varies from 8–10.5 years, with MCTs occurring in cats ranging from younger than 1 year of age up to 20 years.1,6,7,8 The findings in the current study are consistent with this, with a mean age of 10.4 years and a range of 5 months to 19 years. Furthermore, for those cats dying from MCT-related disease, the mean age was higher than those that survived and, although this did not reach statistical significance, this finding is in agreement with a previous study. 7
The majority of MCTs in this study were well-differentiated mastocytic type (66.3%), consistent with other studies.7,8 Pleomorphic mastocytic-type tumours represented 19.8% of the cases, while atypical MCTs represented only 8.1% of cases, compared with 26% 8 and 11% 7 reported in other studies. These differences may reflect the way diagnoses are recorded and retrieved from the database utilised in the current study, since only MCTs with a definitive diagnosis recorded in the diagnosis field would have been identified by the search. A novel but relatively rare histological subtype of MCT is described in this study, composed of well-differentiated mast cells together with prominent multinucleated cells. This subtype occurred in older cats (8–16 years old) and sometimes in conjunction with other MCTs of different histological types. Of the three cats with tumours of this type with sufficient follow-up information available, two had died of MCT-related disease within 2 years of diagnosis, and this novel type of MCT clearly warrants further study with a larger sample size. The new two-tier grading system for canine MCTs where the presence of three or more multinucleated mast cells per 10 high power fields indicates a high-grade tumour. 12
In one study, the malignant MCTs were all either pleomorphic or atypical. 7 However, the current study did not find any difference in the prevalence of the histological types between those cats who succumbed to MCT-related disease and those who did not, suggesting that feline MCTs have a wide spectrum of biological behaviour and that it is not possible to predict the degree of malignancy based on the histological type alone. Several studies have shown a correlation between higher mitotic rates and poor outcomes,7,10,13 and this was confirmed in the present study; the MI was significantly higher in those cats dying from MCT-related disease compared with those that survived. However, the range of MI for both groups was very large with significant overlap, meaning that MI cannot reliably be used as a sole prognostic indicator.
The nature of this study means there are several limitations, including the lack of detailed information regarding any staging of clinical disease performed (although in some cases this additional information was kindly supplied by the responding veterinary surgeon), or information regarding any potential additional therapies used. Clinical outcomes were only available for a relatively small proportion of the original 287 cases identified, and the cause of death was largely based on clinical findings rather than necropsy.
In this study, a significant number of cats presented with multiple lesions either simultaneously or sequentially. Of the 10 cats with a known outcome, only three died from MCT-related disease within 2 years of diagnosis. However, previous studies have shown a correlation between multiple tumours and a poor outcome;6,7 75% of such cases in one study had an unfavourable outcome, as compared to 10% of cases with a solitary mass, 7 and in some cases, multiple cutaneous lesions may represent metastatic spread from an primary visceral mass. This study and others 13 have found that there is no correlation between the number of tumours and outcome. However, this number may be underestimated in the current study due to the relative lack of staging and/or necropsy examinations.
Conclusions
The present study provides further evidence that the biological behaviour of cutaneous MCTs in cats is difficult to predict based on clinical and histological parameters alone and that there is a need for more accurate prognostic markers in these cases.
Footnotes
Acknowledgements
The authors would like to thank all of the veterinary practices that kindly responded to the questionnaires, supplying the necessary follow-up information for their cases. The authors would also like to thank Finn Pathologists for allowing access to their database and archived materials and for preparing the HE-stained sections of the tumours.
Conflict of interest
The authors do not have any potential conflicts of interest to declare.
Funding
This research was performed as part of a final year research project (KM) supported by the Royal Veterinary College.