Breed predilections and prognosis for subungual squamous cell carcinoma in dogs

Abstract

Objective

To better document the prevalence, breed predilections, and clinical behavior of subungual squamous cell carcinomas in dogs.

Procedure

Retrospective analysis of records from 278 812 canine biopsy submissions including 1518 subungual squamous cell carcinomas from dogs in Canada between the years 2003 and 2021.

Results

In agreement with previous studies, giant schnauzers [odds ratio (OR): 56.7], standard schnauzers (OR: 20.3), Gordon setters (OR: 18.3), black standard poodles (OR: 11.1), Kerry blue terriers (OR: 9.4), Rottweilers (OR: 7.0), and several other breeds of large black dogs had a strong predilection for development of subungual squamous cell carcinomas. In giant schnauzers and standard poodles specifically, the risk of developing additional tumors on additional digits was 56%. There were no local postoperative recurrences, and the risk of detecting metastatic disease within 5 y after initial diagnosis was very low at 4%.

Conclusion

Moderately large black, or black and tan, dogs have a marked increase in the prevalence of subungual squamous cell carcinomas. At least in giant schnauzers and black standard poodles, the risk of developing additional similar tumors on additional digits is high, but the metastatic risk is very low.

Clinical relevance

Veterinarians receiving a histologic diagnosis of subungual squamous cell carcinoma in a large black (or predominantly black) dog should advise the owners of a substantial risk that the dog will develop similar tumors on other digits in 2 or 3 y following initial diagnosis, but that the risk of local recurrence or metastatic spread is extremely low.

Introduction

Osteolytic invasive squamous cell carcinomas arising from the nail bed epithelium of dogs are the most prevalent digital neoplasms in dogs, representing a mean of 36% of all digital neoplasms in 3 studies totaling 459 dogs. In those same studies, squamous cell carcinoma represented 54, 65, and 70% of those tumors specifically arising at or near the nail bed (1–3). In a larger recent study, squamous cell carcinoma accounted for 758 of 1528 (49.6%) of all digital neoplasms in dogs (4). They occur primarily in medium- and large-breed dogs with a black or predominantly black haircoat, with a mean and median age of 10 y at the time of first diagnosis. Breeds identified in multiple studies as having a particularly high risk include giant schnauzers, standard schnauzers, black standard poodles, Rottweilers, Gordon setters, briards, and Kerry blue terriers. In these multiple studies, the prevalence in giant schnauzers has consistently been substantially higher than in any other breed (4–7). In data from the records of the Laboratory of Pathology and Toxicology, University of Pennsylvania, the probability for development of subungual squamous cell carcinoma in giant schnauzers was 14.6 times that of dogs in general (6). In a more recent study based upon submissions from veterinarians in 24 European countries, the predilection for giant schnauzers was 13.6 times that of dogs in general (4). The other studies did not adjust their data to account for wide variation in the overall prevalence of these breeds within their sample population, so true breed predilections cannot be determined from those data (5–8).

In these predisposed breeds, there are several reports of dogs developing multiple subungual squamous cell carcinomas on multiple digits, but only 2 of those studies involve enough dogs to allow estimates of the actual risk of having multiple digits involved at the time of presentation or in the 2-year postoperative interval. In 1 study, 11 of 49 dogs (22%) had multiple digit involvement (5). In the second study, 7 of 79 dogs (9%) had multiple digit involvement (8).

Estimates of the metastatic risk of subungual squamous cell carcinoma have varied from 0 to 24% (Table 1), with those estimates based on follow-up information from relatively limited numbers (19 to 79 dogs) (1–3,5,7). Some of the variation may be related to the duration of the follow-up interval, intermingling of data from subungual tumors with those arising elsewhere on the digit, imprecise discrimination of benign subungual keratoacanthomas from true subungual squamous cell carcinomas, and failure to define whether “metastatic disease” included just spread to the regional lymph node or whether it required compelling evidence of more distant metastasis. In virtually all cases, the diagnosis of metastatic disease was based on evidence derived from radiography or more advanced imaging but without histopathology. Additional variation might stem from the source of the cases (diagnostic laboratories serving primary care practices versus cancer referral centers). The latter probably see more advanced or more aggressive disease, and probably use more sophisticated techniques to detect metastatic spread.

Table 1.

Postoperative behavior of subungual squamous cell carcinomas in dogs.

Reference	Location	Number of dogs with follow-up	Local recurrence postoperatively	Initial or eventual involvement of additional digits Number (%)	Metastasis Number (%)
(7)	Italy	79	3	18 (23%)	14 (18%)
(5)	France	49	0	11 (22%)	4 (8%)
(3)	USA/Canada	42	1	7 (6%)	10 (24%)
(1)	USA	33	0	3 (9%)	0
(2)	USA	19	0	NG	1 (5%)
(8)	USA	21	0	0	0

NG = Not given.

The primary goals of this study were to determine the magnitude of the predilection (odds ratios) for subungual squamous cell carcinomas in black giant schnauzers, standard poodles, and other predisposed dog breeds in Canada, and to determine (in giant schnauzers and black standard poodles) the risk of affected dogs developing multiple such tumors, the rapidity of such development, and the metastatic risk.

Materials and methods

Case selection

The records of Histovet Surgical Pathology were searched for the years 2003 to 2021 to identify all cases of canine digital squamous cell carcinomas confirmed by histologic assessment of amputated digits. This laboratory receives cases from primary care veterinary practices across Canada with a small proportion (3%) from the United States, Bermuda, and Asia.

All submissions were formalin-fixed amputated digits. Slides were prepared by routine histology techniques from sagittal sections taken specifically to capture the nail bed. All diagnoses were made by the same pathologist (BW) based upon standard diagnostic criteria: destruction of nail bed architecture by branching invasive cords of disorganized stratum spinosum-like squamous epithelium accompanied by substantial reactive fibrosis and mixed inflammation (9,10). Osteolysis was apparent in most cases but was not an absolute requirement for the diagnosis.

Prevalence data

For the 19-year interval from 2003 to 2021 inclusive, we determined the total number of canine biopsy submissions, the number specified as skin or subcutaneous masses, and the number specified as digital masses arising from nailbed or bone rather than skin or subcutaneous tissue. We then determined the number of masses subsequently identified via histopathology as subungual squamous cell carcinomas.

Using a list of predisposed breeds adapted from the 5 most relevant previous studies (1,3–6), we determined the prevalence of subungual squamous cell carcinomas in those breeds as contrasted to the prevalence in all dogs within our study population. True breed predilections rather than just overall prevalence were determined by calculating odds ratios (OR) and confidence intervals (CI). We also determined the prevalence in 3 additional breeds not previously documented to be at increased risk to serve as relevant negative controls. Newfoundland dogs were selected as negative controls for “large black dogs” and German shepherd dogs and Bernese Mountain dogs as negative controls for “large black and tan dogs.”

Odds ratios and confidence intervals were calculated for each breed by building contingency tables in the GLIMMIX Procedure in the SAS Software System. Statistical significance was set to P ≤ 0.05.

Assessment of postoperative tumor behavior

We sent 158 surveys to veterinarians who had submitted samples subsequently confirmed as subungual squamous cell carcinomas in giant schnauzers and standard poodles. We limited the surveys to just those 2 breeds because they were the 2 breeds repeatedly mentioned in previous studies as being strongly predisposed and for which we had larger numbers than for any other predisposed breeds. The survey questions were about the color of the dog, the location and timing of the development of any similar tumors on the same or additional digits, the development of metastases, how those metastases were confirmed, and duration of postoperative survival.

The surveys were done at 2 times (in 2012 for cases originally diagnosed for the years 2003 to 2011, and in 2021 for cases diagnosed in the interval 2017 to 2020, inclusive). The questions were identical. In the 2021 survey, we did not attempt to gather information about cases that had been diagnosed more than 4 y before the survey date, and so the survey included only those cases diagnosed during 2017 to 2020. Based on our experience with the 2011 survey, the response rate related to patients last seen > 4 y before the survey date was extremely poor. For both surveys, we followed up with telephone calls to encourage a response from those clinics that had not responded to the written request.

Results

Prevalence data

There were 278 630 canine biopsy submissions over the 19-year study interval. Of these, 1518 (0.5%) were identified as subungual squamous cell carcinomas. The number of submissions that represented skin and subcutaneous masses in general, the number representing digital masses arising from nailbed rather than skin, and those identified specifically as squamous cell carcinomas in those breeds previously identified as being predisposed to subungual squamous cell carcinomas are presented in Tables 2 and 3.

Table 2.

Prevalence of subungual squamous cell carcinomas as a proportion of all skin and digital masses in predisposed dog breeds.

Breed	Skin tumors (including digital tumors)	Number of digital tumors	Number of subungual squamous cell carcinomas	Squamous cell carcinomas as a percentage of all skin masses, including digital
Giant schnauzer	228	114	100	29.2
Standard schnauzera	144	40	37	20.1
Standard poodlea	1036	202	174	14.1
Gordon setter	57	15	13	13.1
Rottweiler	1087	206	133	10.3
Kerry blue terrier	168	24	16	8.3
Briarda	53	9	5	8.1
Scottish terrier	246	37	21	7.4
Bouvier	397	55	33	7.3
Flat-coated retriever	218	24	9	3.7
Dachshunda,b	675	42	23	3.2
Labrador retrievera	9930	431	164	1.7

^aCoat color not specified.

^bCoat type not specified.

Table 3.

Breed predilection for subungual squamous cell carcinomas in dogs.

Breed	Number of overall biopsy submissions	Subungual squamous cell carcinomas Number (%)	P-value	Odds ratio	95% confidence interval
Giant schnauzer	422	100 (23.7%)	< 0.0001	56.7	45.0 to 71.4
Standard schnauzera	370	37 (10%)	< 0.0001	20.3	14.4 to 28.6
Gordon setter	143	13 (9.1%)	< 0.0001	18.3	10.3 to 32.4
Standard poodlea	3032	174 (5.7%)	< 0.0001	11.1	9.5 to 13.1
Kerry blue terrier	327	16 (4.9%)	< 0.0001	9.4	5.7 to 15.6
Rottweiler	3619	133 (3.7%)	< 0.0001	7.0	5.8 to 8.3
Briarda	135	5 (3.7%)	< 0.0001	7.0	2.9 to 17.2
Bouvier	962	33 (3.4%)	< 0.0001	6.5	4.6 to 9.2
Scottish terrier	758	21 (2.8%)	< 0.0001	5.2	3.4 to 8.0
Flat-coated retriever	513	9 (1.8%)	0.0005	3.3	1.7 to 6.3
Dachshunda,b	2372	23 (1.0%)	0.0059	1.8	1.2 to 2.7
Labrador retrievera	26 094	164 (0.6%)	0.0813	1.2	1.0 to 1.4
Newfoundland	692	4 (0.6%)	0.9056	1.1	0.4 to 2.8
Bernese Mountain dog	3059	16 (0.5%)	0.8914	1.0	0.6 to 1.6
German shepherd	7226	20 (0.3%)	0.0026	0.5	0.3 to 0.8
All dogs	278 630	1518 (0.5%)

^aCoat color not specified.

^bCoat type not specified.

Breeds confirmed as having a genuine predilection resemble those in the previously published lists. We specifically compared (Table 4) our results to those published in 1998 by Goldschmidt and Shofer (6) from the diagnostic laboratory at the University of Pennsylvania and in 2021 by Grassinger et al (4) from a diagnostic laboratory in Germany (the only studies that reported odds ratios).

Table 4.

Comparison of dog breed predilection data for subungual squamous cell carcinomas in the 3 largest studies.

Breed Reference	Number of dogs (current study)	(6)	(4)	Probability (current study)	(6)	(4)	Odds ratio (current study)	(6)	(4)	95% CIa (current study)	(6)
Giant schnauzer	100	5	298b	< 0.0001	< 0.00001	≤ 0.001	56.7	14.6	13.6	45.0 to 71.4	5.8 to 36.3
Standard poodle	174	19c	22c,d	< 0.0001	< 0.00001	≤ 0.001	11.1	5.6	4.1	9.5 to 13.1	3.5 to 8.9
Rottweiler	133	7	40	< 0.0001	0.03	≤ 0.001	7.0	2.2	4.7	5.8 to 8.3	0.1 to 4.8
Kerry blue terrier	16	5	3	< 0.0001	< 0.00001	NG	9.4	9.0	NG	5.7 to 15.6	3.6 to 22.2
Gordon setter	13	8	≤ 2	< 0.0001	< 0.00001	NG	18.3	12.5	NG	10.3 to 32.4	6.1 to 25.7
Briardc	5	3	28	< 0.0001	< 0.00001	≤ 0.001	7.0	9.2	5.9	2.9 to 17.2	2.9 to 29.5
Scottish terrier	21	7	≤ 2	< 0.0001	0.01	NG	5.2	2.5	NG	3.4 to 8.0	0.2 to 5.4
Standard schnauzerc	37	10	?	< 0.0001	< 0.00001	NG	20.3	5.3	NG	14.4 to 28.6	2.6 to 10.3
Dachshundc,d,e	23	13	28	0.0059	0.0005	NG	1.8	2.6	3.7	1.2 to 2.7	1.5 to 4.6
Labrador retrieverc	164	28	33	0.0813	0.004	NG	1.2	1.8	1.1	1.0 to 1.4	0.3 to 2.6

^aFor reference (4) 95% CI data were not given.

^bNot specified as giant or standard.

^cCoat color not specified.

^dNot specified as standard, miniature, or toy.

^eCoat type not specified.

NG = Not given.

Survey results

We received responses to 106 of the 158 surveys (67.1%). For giant schnauzers, all dogs were confirmed as being black. In standard poodles, all dogs were black except for 1 described as “coffee-colored.”

In postoperative intervals varying from 12 mo to 5 y, there were no local recurrences following digital amputation. Seventeen dogs had had a previous subungual squamous cell carcinoma removed prior to entry into this study or had more than 1 tumor present at the time of first evaluation. Twenty-two dogs developed squamous cell carcinomas on additional digits within the first year following excision of the first tumor. In 13 additional dogs, there were additional tumors on additional digits first identified between 1 and 2 y after initial surgery. Seven dogs (all standard poodles) developed additional tumors only between 2 and 5 y after removal of the first tumor. Metastatic spread was suspected in 4 dogs (all giant schnauzers; Table 5) based on radiographic evidence of metastatic foci in lung, although none was confirmed by cytology or histology.

Table 5.

Behavior of subungual squamous cell carcinomas in giant schnauzers and standard poodles.

Breed	Number of responses	Number with additional digits affected					Number with suspected metastasis
		At initial presentation	≤ 1 y	1 to 2 y	2 to 5 y	Total (%)
Giant schnauzer	41	7	14	6	0	27 (66%)	4a
Standard poodle	65	10	8	7	7	32 (49%)	0
Total	106	17 (16%)	22 (21%)	13 (12%)	7 (7%)	59 (56%)	4 (4%)

^aIn these 4 dogs, metastasis occurred at 3 mo, 6 mo, 2 y, and 5 y, respectively, after first amputation.

Discussion

This study confirmed, with much larger numbers and therefore with more statistical reliability, virtually all conclusions reported in the most credible of the previous studies on this disease. Those studies reported epidemiologic and clinical information based upon 758, 154, and 79 dogs, respectively (4,5,7). In this study, we had 1518 dogs with histologically confirmed subungual squamous cell carcinoma.

We confirmed the very strong predilection for larger black or black and tan dogs, and yet that predilection was not absolute. Some of the common large breeds that fit this profile such as Newfoundland dogs, Bernese Mountain dogs, and German shepherd dogs have a very low risk of developing subungual squamous cell carcinoma, equal to or less than the risk for dogs in general. Clearly, being moderately large and black is not, by itself, enough to explain the breed predilections.

Data for breed predilections in our study, as well as in all the other published studies, suffer from imprecise information about breed, coat color, and type that can be extracted from submission forms for most diagnostic laboratories. Several of the previous studies, for example, listed data for “schnauzers” without defining whether those data were from giant schnauzers specifically or whether they included data from standard and miniature schnauzers that are separate breeds with distinctive genomes. The same shortcoming applied to data listed only as being derived from “poodles” without specifying whether those dogs were standard, miniature, or toy. In our study, we reported data specifically from giant schnauzers and standard poodles. Even in our study, however, we were unable to determine the coat color for those breeds in which colors other than black are common: Labrador retrievers, briards, standard schnauzers, and dachshunds. Therefore, the relative prevalence of subungual squamous cell carcinomas specifically in black dogs of those breeds is likely to be higher than what we have reported here. This same shortcoming applies to all published reports to date.

We compared (Table 4) our results to those of Goldschmidt and Shofer (6), and Grassinger et al (4). The list of predisposed breeds was similar in all 3 studies, although it was inexplicably shorter in the study from Europe. In 2 of the 3, giant schnauzers had the greatest predilection (consistent with virtually all other published studies, despite not reporting odds ratios). In Grassinger et al (4), schnauzers were not specified as being giant or standard, although these are separate breeds with distinct genomes. In our study, standard poodles had the largest overall number of cases, and were the fourth highest in terms of odds ratio. Based on our survey responses, all affected giant schnauzers and all but 1 of the standard poodles were black.

The most pragmatic and useful information from this study was related to clarification of the risk of seeing additional subungual squamous cell carcinomas on additional digits, how quickly that was likely to occur, and the risk of metastatic spread. These are questions we are asked frequently by primary care practitioners because they are the questions of greatest concern to the owners. Published estimates for the risk of developing tumors on additional digits at between 6% (7/109) and 23% (18/79), and the metastatic risk at anywhere from 0% (0/33) to almost 25% (10/42) are too broad to be useful. Our data, admittedly derived only from the behavior in giant schnauzers and standard poodles, indicated the risk of developing squamous cell carcinomas on additional digits at 56% (59/106) within 2 y after first diagnosis. The metastatic risk was 4% (4/106), with metastases first observed anywhere from 3 mo to 5 y after the first diagnosis. Most dogs destined to develop additional tumors on additional digits did so within 2 y after initial diagnosis (Table 5).

This study, like all previous ones, had limitations intrinsic to retrospective studies based on submissions to diagnostic laboratories. We have already discussed our inability to determine the coat characteristics for several of the breeds on our list of predisposed breeds, which means that the genuine prevalence in black dogs of those breeds is likely to be much higher than what is reported here and elsewhere. Our information about metastatic risk was limited by detection methods used in primary care practice, and there was no clear distinction between radiographic observation of presumed metastases and the presence of actual clinical signs related to metastatic spread. Even in cases in which pulmonary radiographs revealed suspected metastatic spread, in none of these was the tumor confirmed as squamous cell carcinoma by histopathology. CVJ