• Open Access

Postmortem Evaluation of 435 Cases of Intracranial Neoplasia in Dogs and Relationship of Neoplasm with Breed, Age, and Body Weight

Authors


Corresponding author: Rachel B. Song, VMD, The Ohio State University Veterinary Medical Center, 601 Vernon L Tharp, St. Columbus, OH 43210; e-mail: Rachel.baek@gmail.com.

Abstract

Background

Intracranial neoplasia of dogs is frequently encountered in veterinary medicine, but large-scale studies on prevalence are lacking.

Objectives

To determine the prevalence of intracranial neoplasia in a large population of dogs examined postmortem and the relationship between breed, age, and weight with the presence of primary intracranial neoplasms.

Animals

All dogs that underwent postmortem examination from 1986 through 2010 (n = 9,574), including dogs with a histopathologic diagnosis of primary (n = 227) and secondary (n = 208) intracranial neoplasia.

Methods

Retrospective evaluation of medical records from 1986 through 2010.

Results

Overall prevalence of intracranial neoplasia in this study's population of dogs was 4.5%. A statistically significant higher prevalence of primary intracranial neoplasms was found in dogs with increasing age and body weights. Dogs ≥15 kg had an increased risk of meningioma (odds ratio 2.3) when compared to dogs <15 kg. The Boxer, Boston Terrier, Golden Retriever, French Bulldog, and Rat Terrier had a significantly increased risk of primary intracranial neoplasms while the Cocker Spaniel and Doberman Pinscher showed a significantly decreased risk of primary intracranial neoplasms.

Conclusions and Clinical Importance

Intracranial neoplasia in dogs might be more common than previous estimates. The study suggests that primary intracranial neoplasia should be a strong differential in older and larger breed dogs presenting with signs of nontraumatic intracranial disease. Specific breeds have been identified with an increased risk, and others with a decreased risk of primary intracranial neoplasms. The results warrant future investigations into the role of age, size, genetics, and breed on the development of intracranial neoplasms.

Abbreviations
CI

confidence interval

CPN

choroid plexus neoplasm

CT

computed tomography

MRI

magnetic resonance imaging

Intracranial neoplasia in the dog is a frequently encountered disease in veterinary medicine. The most commonly cited estimated incidence of intracranial neoplasms in dogs is 14.5/100,000 based on literature review and necropsy findings across several different species.[1-4] However, another study estimated prevalence can be as high as 3.0%.[2] Meningiomas are the most commonly reported primary intracranial neoplasms in dogs.[1, 5, 6] Glial neoplasms, including astrocytomas, oligodendrogliomas, and oligoastrocytomas, are the 2nd most commonly reported intracranial neoplasms.[2, 7] Certain breeds have been overrepresented in the literature for specific types of neoplasms including meningiomas more commonly in dolichocephalic breeds,[1, 3, 8, 9] and glial neoplasms more commonly in brachycephalic breeds such as Boxers and Boston Terriers.[2, 7, 10, 11] The increased use of computed tomography (CT) and magnetic resonance imaging along with improved biopsy techniques has allowed for improved antemortem diagnosis in dogs. The purpose of this study was to determine the prevalence of primary and secondary intracranial neoplasms, and to determine breed, age, and weight predilections for specific primary intracranial neoplasms in a large population of dogs. We hypothesized that the prevalence of intracranial neoplasms in dogs is higher than it has been previously reported. We also hypothesized that intracranial neoplasms occur more frequently in larger breed dogs than in smaller breed dogs.

Materials and Methods

This study was a retrospective review of the necropsy database of the Matthew J. Ryan Veterinary Hospital at the University of Pennsylvania. The brain and neurocranium are routinely examined grossly in all patients as part of a necropsy. The database was searched for dogs with intracranial neoplasms from January 1, 1986 to December 31, 2010, for which the brain (prosencephalon, mesencephalon, and rhombencephalon), pituitary gland, meninges, and neurocranium were examined histopathologically. Cases with a diagnosis of neoplasia at these sites were included in the study. Necropsy reports were reviewed for information on the type of neoplasm, age, breed, and weight. The same parameters were recorded for all dogs that underwent postmortem examination contemporaneously, regardless of cause of death.

The dogs with intracranial neoplasia were divided into primary and secondary categories. Primary intracranial neoplasms were defined as those originating from cells normally found within the brain and meninges with no extraneural tumor occurrence.[1, 2] Round cell neoplasms such as lymphosarcoma and histiocytic sarcoma were included in this category if no neoplastic round cells were found outside of the central nervous system (brain and spinal cord). Secondary intracranial neoplasms were defined as those with neoplastic foci outside of the brain with corresponding metastasis to the brain or pituitary gland.[1, 12] Additionally, neoplasms originating from adjacent neural and nonneural tissues that affected the brain matter by compression or invasion (eg, pituitary, nasal, and cranial nerve tumors) were included in the secondary tumor category.[1, 12] Dogs were stratified into yearly age categories (eg, 0–1, 1–2, 2–3, … 13–14) except the oldest age category, which encompassed ages 14–24. Dogs were also stratified into weight categories by kilograms into 0–10, 10–15, 15–20, 20–25, 25–30, 30–40, 40–50, and 50–100. The dogs were then analyzed by neoplasm type to determine the relationship among age, weight, and breed. Dogs were classified as either large breed (weight ≥15 kg) or small breed (weight <15 kg).

Single degree-of-freedom tests for linear trend (Cochran–Mantel–Haenszel test)[13, 14] were performed on variables including type of neoplasm, breed, age, and weight. Breeds were rank ordered by their prevalence for each neoplasm type. Breeds with the prevalence of a particular neoplasm type that was greater than the prevalence in the pooled population were defined to be at an increased risk. For each breed and neoplasm type of interest, the statistical significance of the increased risk was assessed by conducting exact binomial tests of the null hypothesis, that breed-specific prevalence was equal to the overall prevalence across all breeds. P-values less than .05 were used to determine which breeds had a statistically significant increased risk for a particular neoplasm. It should be noted that statistical significance depends on both the magnitude of increased risk and on the available sample size. Further analyses compared prevalence of neoplasms between larger and smaller breeds using Chi-square statistics. These comparisons were repeated after excluding specific breeds found to be at higher risk to help isolate breed and size-specific effects.

Dogs less than 1 year of age had much lower risk for intracranial neoplasia (n = 2) despite the fact that the 0–1 year age group was the most commonly represented age group in our study (n = 1,498). Statistical analyses were performed for the general population, and then repeated excluding dogs less than 1 year of age. If significant statistical differences were seen between the analyses, those differences were reported.

Results

There were a total of 9,574 dogs that underwent postmortem examination from 1986 to 2010 of which 7,969 were at least 1 year of age. There were 105 dogs with missing age information and 23 dogs with missing weight information from the necropsy records. These dogs were excluded from the appropriate areas of analyses. A total of 435 dogs with primary (n = 227) and secondary (n = 208) intracranial neoplasms were identified. The prevalence of intracranial neoplasia in dogs was 4.5% (5.5% in dogs >1 year of age) with a prevalence of primary neoplasms at 2.3% of total population (2.8% in dogs >1 year), and secondary neoplasms at 2.2% (2.6% in dogs >1 year). Meningiomas were the most common primary intracranial neoplasms, followed by glial neoplasms, which included astrocytomas, oligodendrogliomas, oligoastrocytomas, glioblastoma and gliomatosis cerebri. One hundred and sixty-two dog breeds were represented in the total population with mixed breed (25%), Golden Retriever (6.3%), Labrador Retriever (6.0%), Rottweiler (5.0%), German Shepherd (4.8%), Boxer (2.2%), Beagle (2.2%), and the Doberman Pinscher (2.2%) as the most frequently represented breeds. All other dog breeds represented less than 2% of the total population.

The mean and median age of all the dogs in the general population were 6.9 and 7 years, respectively (range 0–23). The most commonly represented age group in the general population was 0–1 year (n = 1,498, 15.6% of population). The mean and median weight of all the dogs included in the study were 23.6 kg and 23 kg, respectively (range 0.1–100 kg). The most commonly represented weight group in the general population was between 0 and 10 kg (n = 2,478, 25.9% of population). Among dogs with primary intracranial neoplasms, the most commonly represented age group was 10–11 years (n = 36, 16% of dogs with primary neoplasms). The most common weight group at which primary tumors were observed was between 25 and 30 kg (n = 37, 16.3% of dogs with primary neoplasms). A Cochran–Mantel–Haenszel test for linear trend revealed that there was an increased risk of primary intracranial neoplasms in dogs with higher body weights and older age (P < .0001).

Several breeds had a significantly increased risk of primary intracranial neoplasms when compared with the general canine population's risk of primary intracranial neoplasms. Listed in an order of decreasing significance (number of dogs within the breed with primary intracranial neoplasms/total number of dogs within the breed), breeds at increased risk included the Boxer (P = .0001, 28/212), Boston Terrier (P = .0001, 8/41), Golden Retriever (P = .0002, 30/602), French Bulldog (P = .0059, 3/16), and Rat Terrier (P = .047, 1/2). Among dogs that were >1 year of age, the increased risk in Rat Terriers was no longer significant. There were also breeds that had a lower risk of primary intracranial neoplasia when compared with the general population. No primary intracranial neoplasms were discovered in 117 of 162 examined breeds. Among the dog breeds in which no intracranial neoplasms were identified in the study, the Doberman Pinscher (n = 210) and Cocker Spaniel (n = 166) showed a significantly decreased risk of primary intracranial neoplasia.

Meningioma

One hundred seventeen dogs had meningiomas in the general population, with an overall prevalence rate of 1.2% (1.5% of dogs >1 year), representing 51.5% of primary intracranial neoplasms. The median age and weight for dogs with meningiomas were 11 years and 25.6 kg, respectively. Meningiomas occurred most frequently between ages 12–14 years, representing 3.1% of the general population of dogs within the same age group (Fig 1). No meningiomas were identified in dogs less than 3 years of age. Cochran–Mantel–Haenszel statistics revealed a statistically significant linear relationship between increasing age and increasing prevalence of meningiomas (P < .0001). Dogs with meningiomas most frequently weighed between 20 and 25 kg, representing 2.0% of the general population of dogs within the same weight group (Fig 2). Meningiomas were significantly more likely to occur in large breed dogs compared to small breed dogs with an odds ratio of 2.3 (95% CI 1.4–3.6, P = .0006). Meningiomas were identified in 29 different breeds. Statistical analysis showed Golden Retrievers, mixed breed dogs, Miniature Schnauzers, and Rat Terriers to have a significantly increased risk of meningiomas (P < .05) compared with the risk of the rest of the general canine population (Table 1). When Golden Retrievers (n = 602) were removed from the population, meningiomas were still significantly more likely to occur in large breed dogs compared to small breed dogs, with an odds ratio of 1.8 (95% CI 1.1–3.0, P = .01). When all breeds identified with a significantly increased risk of meningiomas (Golden Retrievers, mixed breed, Miniature Schnauzers, and Rat Terriers) were removed from the population, the odds ratio of the likelihood of meningiomas occurring in large breed dogs compared to small breed dogs was reduced to 1.4 and was no longer statistically significant (95% CI 0.7–2.8, P = .32). There were only two Rat Terriers in our study, one of which had a meningioma.

Figure 1.

Bar graphs depicting (A) number of dogs diagnosed with the most common primary intracranial neoplasms and (B) the % occurrence of the most common primary intracranial neoplasms among all dogs represented within each age group.

Figure 2.

Bar graphs depicting (A) number of dogs diagnosed with the most common primary intracranial neoplasms and (B) the % occurrence of the most common primary intracranial neoplasms among all dogs represented within each weight group.

Table 1. Dog breeds diagnosed with meningiomas on necropsy with total number of dogs within the breed evaluated, number of dogs within the breed diagnosed with a meningioma, percentage of meningiomas from all intracranial neoplasms diagnosed within the breed, percentage of meningiomas diagnosed within the total number of dogs within the breed evaluated, and P values that connote the breed's risk of neoplasia
BreedNumbers of BreedNumbers of Meningioma% Meningioma of Total Intracranial Neoplasms in Breed% Meningioma of Total Numbers of BreedP Value
  1. Breeds are listed in order of decreasing risk of a diagnosis of meningioma. The P values test the null hypothesis that each breed has no elevated risk of developing meningiomas compared with the general populations' risk of meningiomas (1.2%). The smaller the P value, the more evidence that the risk is elevated. Breeds with a P < .05 (italicized) were considered to be at a statistically significant greater risk of meningiomas compared with the rest of the evaluated population.

Golden Retriever 6022273.333.65 .0001
Mixed breed 2,3934468.751.84 .0075
Miniature Schnauzer 10351004.85 .0089
Rat Terrier 2110050 .0243
Malamute3121006.45.0550
Cardigan Welsh Corgi10110010.1157
Norwegian Elkhound1111009.09.1265
Shetland Sheepdog10731002.8.1437
Whippet1811005.56.1985
Bichon Frise7321002.74.2244
Afghan Hound2311004.35.2463
Eskimo dog2711003.7.2825
Rottweiler483337.50.62.3000
Irish Setter3811002.63.3733
Pekingese3811002.63.3733
Airdale Terrier391502.56.3809
Beagle2121250.47.3873
Labrador Retriever577545.450.87.4661
Boxer212414.291.89.5287
English Springer Spaniel9721002.06.6359
Miniature Poodle1012501.98.6391
German Shepherd459583.331.09.8405
Chihuahua971501.031.0000
Chow Chow581501.721.0000
Maltese7511001.331.0000
Samoyed4311002.331.0000
Standard Poodle133266.671.51.0000
Toy Poodle79133.331.271.0000
Weimaraner4211002.381.0000

Glial Neoplasm

There were 83 dogs with glial neoplasms (29 astrocytoma, 51 oligodendroglioma, 1 oligoastrocytoma, 1 glioblastoma, and 1 gliomatosis cerebri) in the total population, giving an overall prevalence rate of 0.9% (1.0% of dogs >1 year) and representing 36.6% of primary intracranial neoplasms. The median age and weight for dogs with glial neoplasms were 8 years and 26.6 kg, respectively. The median age and weight for dogs with astrocytomas were 9 years and 28.9 kg, respectively. The median age and weight for dogs with oligodendrogliomas were 7 years and 26.2 kg, respectively. Following meningiomas, glial neoplasms were the second most common primary intracranial neoplasm. Dogs with glial neoplasms most frequently occurred in the 7–8 year age group, representing 2.02% of all dogs within the same age group (Fig 1). The youngest age at which a glial neoplasm was diagnosed was at 2.4 months (oligodendroglioma in a West Highland White Terrier). Dogs with glial neoplasms most frequently occurred in the 25–30 kg-weight group, representing 1.8% of all dogs within the same weight group (Fig 2). Glial neoplasms occurred more frequently in large breed dogs compared to small breed dogs with an odds ratio of 1.5 (95% CI 0.9–2.5, P = .11). Unlike meningiomas, the peak prevalence of glial neoplasms occurred at a younger age (age 7–8) than the dogs with meningiomas (age 12–14). Like meningiomas, Cochran–Mantel–Haenszel statistics revealed a statistically significant linear relationship between increasing age and increasing prevalence of glial neoplasms (P < .0001). Glial neoplasms were diagnosed in 25 dog breeds. Statistical analysis showed English Toy Spaniels, Boston Terriers, French Bulldogs, Boxers, and English Bulldogs to have a significantly increased risk of glial neoplasms when compared with the risk of the general population (P < .05) (Table 2). When dogs <1 year of age were removed from the general population, Bullmastiffs were also seen to have a significantly increased risk of glial neoplasms (P = .049). The English Toy Spaniel, French Bulldog, Border Terrier, Boston Terrier, Boxer, English Mastiff, and English Bulldog had a significantly increased risk of oligodendrogliomas, whereas only the Boxer and Boston Terrier had a significantly increased risk of astrocytomas (P < .05). In the French Bulldog, all of the glial neoplasms (n = 3) were classified as oligodendrogliomas. In the Boxer, there were twice as many oligodendrogliomas (n = 16) than there were astrocytomas (n = 8).

Table 2. Dog breeds diagnosed with glial neoplasms on necropsy with total number of dogs within the breed evaluated, number of dogs within the breed diagnosed with a glial neoplasm, percentage of glial neoplasms from all intracranial neoplasms diagnosed within the breed, and percentage of glial neoplasms diagnosed within the total number of dogs within the breed evaluated, and P values that connote the breed's risk of neoplasia
BreedNumbers of BreedNumbers of Glial Neoplasms% Glial Neoplasms of Total Intracranial Neoplasms in Breed% Glial Neoplasms of Total Numbers of BreedP Value
  1. A, astrocytoma; O, oligodendroglioma; OA, oligoastrocytoma; GB, glioblastoma; GC, gliomatosis cerebri.

  2. Breeds are listed in order of decreasing risk of a diagnosis of a glial neoplasm. The P values test the null hypothesis that each breed has no elevated risk of developing glial neoplasms compared with the general populations' risk of glial neoplasms (0.9%). The smaller the P value, the more evidence that the risk is elevated. Breeds with a P < .05 (italicized) were considered to be at a statistically significant greater risk of glial neoplasms compared with the rest of the evaluated population.

Boxer 21224 (8 A, 16 O)85.7111.32 .0001
Boston Terrier 418 (2 A, 5 O, 1 GC)10019.51 .0001
French Bulldog 163 (3 O)10018.75 .0003
English Bulldog 1355 (1 A, 3 O, 1 OA)1003.7 .0067
English Toy Spaniel 41 (1 O)10025 .0342
Bullmastiff442 (1 A, 1 O)1004.55.0559
English Mastiff462 (2 O)1004.35.0605
Border Terrier81 (1 O)10012.5.0673
Rottweiler4831 (1 A)12.50.21.1401
Mixed breed2,39314 (5 A, 9 O)21.880.59.1513
German Shepherd4591 (1 A)16.670.22.1998
Miniature Poodle192 (1 A, 1 O)501.98.2184
Old English Sheep Dog331 (1 O)1003.03.2497
West Highland White Terrier561 (1 O)1001.79.3859
Labrador Retriever5773 (1 A, 2 O)27.270.52.3962
Beagle2121 (1 A)250.47.7299
Pit Bull Terrier1931 (1 O)500.52.7360
Yorkshire Terrier1321 (1 O)1000.761.0000
Toy Poodle791 (1 A)33.331.271.0000
Standard Poodle1331 (1 A)33.330.751.0000
Sharpei771 (1 A)1001.31.0000
Pug881 (1 GB)1001.141.0000
Golden Retriever6025 (2 A, 3 O)16.670.831.0000
Chow Chow581 (1 A)501.721.0000
Chihuahua971 (1 A)501.031.0000

Choroid Plexus Neoplasms (CPN)/Ependymoma

There were 12 dogs diagnosed with choroid plexus neoplasms (all were choroid plexus carcinomas) and 1 dog diagnosed with an ependymoma. The overall prevalence of CPN/ependymoma in the general population was 0.1% (0.2% of dogs >1 year), representing 5.7% of primary intracranial neoplasms. The median age and weight of dogs with CPN/ependymoma were 6 years and 23 kg, respectively. No CPN/ependymomas were seen in dogs less than 3 years of age. Age was not significantly associated with tumor prevalence (Fig 1). CPN/ependymoma were significantly more likely to occur in large breed dogs compared to small breed dogs with an odds ratio of 6.5 (95% CI 0.8–50, P = .04) (Fig 2). The Dalmatian and English Setter had a significantly increased risk for developing CPN/ependymomas (P < .05) when compared with the risk of the general canine population (Table 3).

Table 3. Dog breeds diagnosed with choroid plexus neoplasms (CPN)/ependymomas on necropsy with total number of dogs within the breed evaluated, number of dogs within the breed diagnosed with a CPN/ependymoma, percentage of CPN/ependymomas from all intracranial neoplasms diagnosed within the breed, and percentage of meningiomas diagnosed within the total number of dogs within the breed, and P values that connote the breed's risk of neoplasia
BreedNumbers of BreedNumbers of CPN/Ependymoma% CPN/Ependymoma of Total Tumors in Breed% CPN/Ependymoma of Total Numbers of BreedP Value
  1. Breeds are listed in order of decreasing risk of a diagnosis of CPC/ependymoma. The P values test the null hypothesis that each breed has no elevated risk of developing CPC/ependymomas compared with the general populations' risk of CPC/ependymomas (0.1%). The smaller the P value, the more evidence that the risk is elevated. Breeds with a P < .05 (italicized) were considered to be at a statistically significant greater risk of CPC/ependymomas compared with the rest of the evaluated population.

Dalmatian 8321002.41 .0058
English Setter 2611003.85 .0347
Shih Tzu11111000.9.1400
Golden Retriever60226.670.33.1974
Beagle2121 (Ependymoma)250.47.2503
Mixed Breed2,39346.250.17.7780
Rottweiler483112.50.211.0000
Labrador Retriever57719.090.171.0000

Histiocytic Sarcoma

There were 5 cases of histiocytic sarcoma with a prevalence of 0.05% among the total population (0.07% of dogs >1 year) representing 2.2% of primary intracranial neoplasms. The median age and weight of dogs with primary histiocytic sarcoma were 9 years and 33 kg, respectively. Two of the 5 cases of histiocytic sarcoma occurred in Labrador Retrievers.

Lymphoma

There were 4 cases of primary intracranial lymphoma with a prevalence of 0.04% among total population (0.05% of dogs >1 year), representing 1.8% of primary intracranial neoplasms. The median age and weight of dogs with primary lymphoma were 7 years and 48.5 kg, respectively. Rottweiler dogs represented three of the 4 cases and a Golden Retriever represented the other.

Primitive Neuroectodermal Tumor

A secondary search of the necropsy database from 1986 through 2010 performed following statistical analyses revealed identification of six additional dogs diagnosed with a primitive neuroectodermal tumor (PNET). The median age of the dogs diagnosed with PNET was 4.75 years (range 2-7 years), the median weight of the dogs was 29.0 kg (range 6.7- kg). There was one dog of each of the following breeds that were represented including the Boxer, Golden Retriever, Great Dane, Siberian Husky, mixed breed and Pug. These dogs were not included in the statistical analysis. Due to the small number of dogs, the authors do not believe these additional cases impact the overall conclusions of the study.

Secondary Intracranial Neoplasms

There were 208 secondary intracranial neoplasms in the total population of dogs examined, with an overall prevalence of 2.3% of the total population (2.6% of dogs >1 year). In order of decreasing prevalence, the secondary intracranial neoplasms were as follows: hemangiosarcoma (35.1%; n = 73), lymphoma (19.7%; n = 41), metastatic carcinomas including anal sac gland adenocarcinoma, bronchoalveolar adenocarcinoma, squamous cell carcinoma, hepatic carcinoma, mammary adenocarcinoma, exocrine pancreatic adenocarcinoma, neuroendocrine carcinoma, prostatic adenocarcinoma, pulmonary carcinoma, renal carcinoma, thyroid adenocarcinoma and transitional cell carcinoma (19.2%; n = 40), compressive/invasive pituitary neoplasms (11.5%; n = 24), nasal carcinomas (6.25%; n = 13), malignant melanoma (3.4%; n = 7), histiocytic sarcoma (3.4%; n = 7), nerve sheath tumors (0.96%; n = 2), and a poorly differentiated round cell neoplasm (0.48%; n = 1).

Discussion

This retrospective study describes a population of dogs with intracranial neoplasms with postmortem confirmation compared with a general population of dogs examined contemporaneously (1986–2010). We report a higher prevalence of intracranial neoplasia of primary and secondary neoplasms among dogs at 4.5% of the total population (5.5% of dogs >1 year) than has previously been reported at 14.5/100,000.[1-4] A portion of the intracranial neoplasms in this population may represent incidental findings during postmortem examination, as clinical signs associated with the neoplasms were not evaluated in this study. This study also examined the relationship among primary intracranial neoplasms, breed, age, and body weight of the dog. Overall, there was a statistically significant linear relationship between age and weight and the occurrence of primary intracranial neoplasms. Golden Retrievers, mixed breed dogs, Miniature Schnauzers, and Rat Terriers had a significantly higher risk for meningiomas when compared with other breeds. There was a significant linear relationship between age and the occurrence of meningiomas. Large breed dogs also had a significantly increased risk of meningiomas (odds ratio 2.3). Boxers, Boston Terriers, French Bulldogs, English Bulldogs, English Toy Spaniels, and Bullmastiffs were found to have a significantly increased risk for glial neoplasms when compared with other breeds. A significant linear relationship between age and the occurrence of glial neoplasms was identified. Dalmatians and English Setters had a significantly increased risk for CPN/ependymomas when compared with other breeds. Large breed dogs had a significantly increased risk for CPN/ependymomas (odds ratio 6.5). There were 117 breeds in which no primary intracranial neoplasms were observed. More specifically, Doberman Pinschers and Cocker Spaniels were at a significantly decreased risk for primary intracranial neoplasms. This study also showed that Rottweilers were overrepresented for primary intracranial lymphoma.

There have been several publications that have reported overrepresentation of certain breeds for specific intracranial neoplasms. Although overrepresentation for primary intracranial neoplasms in Boxers, Scottish Terriers, Old English Sheepdogs, Doberman Pinschers, and Golden Retrievers has been previously reported,[2, 11, 15] the statistical analyses to determine significance of the overrepresentation have not been previously performed. Contrary to some reports, our study showed Doberman Pinschers at a decreased risk for developing primary intracranial neoplasms. With respect to meningiomas, German Shepherds, Collies, Boxers, Golden Retrievers, Labrador Retrievers, Boxers, and other dolichocephalic breeds have previously been overrepresented.[1, 3, 8, 9, 16-18] Previous reports have also identified some brachycephalic dog breeds to be overrepresented for glial neoplasms.[2, 7, 10, 11] A prior study on choroid plexus neoplasms showed Golden Retrievers to be overrepresented.[19] Our study has identified several other breeds not previously overrepresented for meningiomas and glial neoplasms and our study did not find Golden Retrievers at an increased risk for CPN/ependymomas.

Individual dog breeds represent closed-breeding populations that have been under selection pressure for specific breed standards and traits.[20] The study of breed-specific genotypes and phenotypes can offer investigators the chance to study diseases associated with each breed. The categorization of dog skulls as dolichocephalic (long head), mesaticephalic (medium head), or brachycephalic (short head) is dependent on objective skull morphology such as the cephalic index (ratio between skull width and skull length), the craniofacial angle, or the olfactory bulb angle.[21-23] The Boston Terrier, Boxer, Brussels Griffon, Bull Terrier, Pug, Shih Tzu, Bulldog, English Toy Spaniel, French Bulldog, Japanese Chin, Japanese Pug, Pekingese, Chihuahua, Maltese, and Cavalier King Charles Spaniel have previously been identified to have a brachycephalic skull type.[21] However, there are several other dog breeds with similar shortened facial features that have not been objectively studied that may be considered brachycephalic, such as the Affenpinscher, Lhasa apso, and Rottweiler. Genetic similarities among brachycephalic dogs have been suggested and candidate genes such as THBS2 and SMOC-2 associated with the brachycephalic skull type have been mapped to a region on Chromosome 1.[24] We have identified some brachycephalic dog breeds as having an increased risk for developing glial neoplasms but not all of the previously identified brachycephalic breeds were determined to have an increased risk. In addition, there were several previously identified brachycephalic dog breeds in which no intracranial neoplasms were found, including the Brussels Griffon, Bull Terrier, Japanese Chin, and the Cavalier King Charles Spaniel. There was 1 meningioma and 1 lymphoma identified in the Pekingese, 1 glial neoplasm identified in the Pug, 1 glial neoplasm and 1 meningioma identified in the Chihuahua, and 1 meningioma identified in the Maltese. Theories for an increased prevalence of primary intracranial neoplasms in brachycephalic dog breeds include increased alterations in intracranial pressures, decreased ability of the brain to compensate for increased intracranial pressure causing overt clinical signs that increase clinical index of suspicion, and genetic predisposition.[2, 23-25] Prior studies investigating differences between brachycephalic and dolichocephalic breeds associated with glial neoplasms did not show a statistically significant difference between the two groups.[3, 26]

Inherited tumor suppressor gene p53 germ-line mutations have been linked to increased tumorigenesis in both humans and dogs.[27] Germ-line defects in humans in the mismatch repair genes are the most commonly inherited cancer predispositions (Lynch syndrome).[28] Gene modifications associated with brachycephalic skull types or associated phenotypes may also be related to the increased prevalence of glial neoplasms and brachycephalic dog breeds. However, further genetic studies in breeds with increased risk of specific intracranial neoplasms are still in need.

We also detected a significantly increased risk of primary intracranial neoplasms in large breed dogs when compared to small breed dogs. A previous study determined a single IGF 1 allele as a major determinant for dog size.[20, 29] A common single-nucleotide polymorphism haplotype for the gene IGF 1 was identified in all small breed dogs (<9 kg), but was found to be absent from most giant breed dogs (>30 kg).[20] IGF 1 has not only been identified as a major determinant for dog size, but also has been implicated in regulating longevity in many other organisms as well.[20] Studies and clinical experience show that overall, smaller breed dogs live longer than larger breed dogs, suggesting the possible role of IGF1.[20, 30] In our dog population, there were 2,230 total dogs <9 kg, 1,159 of which were over 1 year of age. Among dogs <9 kg, there were 44 dogs with intracranial neoplasms (25 primary, 19 secondary). The significance of the difference in prevalence of intracranial neoplasia between dogs <9 kg compared with the general population is unknown at this time. Other loci that are highly significant for regulating longevity have also been identified, including CFA 7 and CFA 15.[20] Neoplastic disease is a more frequent cause of death in large breed dogs than in smaller breed dogs,[30] suggesting the possible roles of these genes in longevity. This study warrants further investigation on the possible genetic differences between small and large breed dogs, the potential consequences of the founder effect with inbreeding in pure bred populations, and their possible effects on size, longevity, and intracranial neoplasia development.

The statistically significant linear relationship between age and the presence of intracranial meningiomas and glial neoplasms found in this study is not surprising. However, the peak prevalence of glial neoplasms (7–8 years) was at a younger population when compared with meningiomas (12–14 years). The case of the 2.4-month-old West Highland White Terrier diagnosed with an oligodendroglioma is the youngest reported in the literature. There are several reports of various intracranial glial neoplasms[31-35] occurring in immature dogs, but much fewer reported cases of meningiomas[35, 36] occurring in immature dogs. The seemingly increased reports of glial neoplasms in immature dogs might reflect the trends seen in children.

There are several limitations to this study. As a retrospective study evaluation of the necropsy database, it is possible some cases of intracranial neoplasms may have been missed depending on search criteria. The population of dogs represented in this study may also create a bias because of the demographics of the area the hospital serves. As a tertiary referral hospital, MJR-VHUP serves not only the immediate urban community of Philadelphia but also the suburban and rural communities, including those from the midatlantic and northeastern United States. There were nearly twice as many large breed dogs (≥15 kg, n = 6,197) as smaller breed dogs (<15 kg, n = 3,354) in our sample population. Necropsies are performed on an elective basis, and owners who chose to have a necropsy performed on their dogs may represent a bias that does not accurately represent the general canine population. For example, dogs between 0 and 1 year of age were the most common age group in our general population (n = 1,498). In breeds such as the Australian Cattle Dog, Bull Terrier, Coton de Tulear, Eurasian Dog, Italian Greyhound, Japanese Chin, Portuguese Water Dog, Shiba Inu, and Shiloh Shepherd, more than 50% of the dogs in each breed were in the 0–1 year age group. This bias may represent the increased willingness of an owner to have necropsies performed on puppies to seek information on the cause for death of pets at such a young age. This bias toward young dogs may have caused a bias in weight categories toward lighter dogs as well. The increased risk of intracranial neoplasia in certain breeds may also skew risk comparisons between large and small breed if certain breeds are overrepresented in large numbers. Another limitation of the study is that much of the signalment information, including age and breed, is owner reported, which may represent some inaccuracies. Another limitation of the study is that the body condition of a dog was not taken into consideration. Different body condition scores in a single breed of dog might create a large range of body weights spanning several weight categories established in this study.

This large-scale study provides the most current estimated prevalence of primary and secondary intracranial neoplasms occurring in dogs. The study suggests that primary intracranial neoplasms are much more common among the canine population previously reported. Certain breeds were found to be at a significantly increased risk for specific primary intracranial neoplasms, including Golden Retrievers, mixed breed, Miniature Schnauzers, Rat Terriers for meningiomas; English Toy Spaniels, Boston Terriers, French Bulldogs, Boxers, English Bulldogs and Bullmastiffs for glial neoplasms, Dalmatians and English Setters for choroid plexus neoplasms/ependymomas. Rottweilers were also found to be overrepresented for primary intracranial lymphoma. Based on these findings, further genetic and molecular studies investigating the role of breed, phenotype (eg, brachycephalic skull type) and genetic history are warranted. The study also identified large breed dogs at a higher risk of developing primary intracranial neoplasms. This suggests the clinician's suspicions for a primary intracranial neoplasm in a older, large breed dog presenting with intracranial signs should be higher than for a small breed dog. In contrast, this study suggests that other causes besides primary intracranial neoplasia should be considered more likely if a small breed dog presents with intracranial signs. Based on the data, further studies investigating the role of molecular genetics (eg, role of IGF 1 allele, THBS2 and SMOC 2 genes), size, longevity, and prevalence of neoplasia are warranted. This study has also identified the Doberman Pinscher and Cocker Spaniel to have a significantly decreased risk of primary intracranial neoplasms. Studies comparing dog breeds at increased risk for primary intracranial neoplasms with dog breeds at a decreased risk of primary intracranial neoplasms may offer a better understanding on the role of genetics and phenotype and the occurrence of neoplasia.

Acknowledgments

The authors thank Sharon Hurley and Greg Maislin for their assistance with the statistical analyses of the manuscript.

Grants: None.

Presented: None.

Conflict of Interest: Authors disclose no conflict of interest.

Ancillary