Analyzing early cancer etiology in golden retrievers using Golden Retriever Lifespan Study (GRLS) data
Date
2023
Journal Title
Journal ISSN
Volume Title
Abstract
Background: Although cancer is a burden in both humans and dogs, humans medicine is characterized by established health care organizations, interdisciplinary networks, and databases from which data and research can be complied and shared. No such organization exists in veterinary medicine. Individual registries provide useful data and information on cancer in dogs, but no mechanism exists to summarize data to detect cancer trends, breed-specific measurements of occurrence, and treatment responses. Therefore, there are vast knowledge gaps related to cancers in dogs, especially among early cases. The Golden Retriever Lifetime Study (GRLS) data collected by Morris Animal Foundation is a unique opportunity to evaluate cancer prevalence in a large number of golden retrievers with known pedigree. Data were evaluated for each state and compared to human cancer prevalence provided by the CDC. Differences in cancer prevalence between young and old dogs was evaluated, along with their resident state, sex status, and cancer type. Golden retrievers were recruited from 2012-2015 to participate in the GRLS cohort study and were confirmed to be free of life limiting conditions by a veterinarian. Owners had to have at least a 3-generation pedigree of their dog to be enrolled. Information regarding the dog's health and condition were recorded annually via owner and veterinarian questionnaire, as well as sample collections, and added to the GRLS study data. The GRLS data was refined and cleaned in SAS and R studio evaluate state of diagnosis, age at diagnosis, and sex at diagnosis. The highest prevalence of cancer among GRLS participants was in Louisiana (38.5%) with Arizona as the second highest (17.5%). A cluster of higher prevalence regions were observed in the upper east coast, similarly to the CDC's human data. Although the prevalence was highest in Louisiana and Arizona, neither were found to be statistically significant based on the difference of proportion calculations. A statistically significant difference was found in average age at diagnosis between male neutered and intact cancer dogs, but not when comparing female spayed and intact cancer bearing dogs or when comparing all 4 sex statuses. The average age at diagnosis based on tumor types (mammary, hemangiosarcoma, histiocytoma, lymphoma) was significantly different, most likely due to higher numbers of hemangiosarcoma cases in older dogs and histiocytoma cases observed in younger dogs. Older, male neutered dogs were more susceptible to hemangiosarcoma development (85.5% of cases were old), and younger dogs that had been spayed or neutered were more susceptible to histiocytomas (100% of cases were young). Discussion: One of the interesting findings of this analysis was that there was a statistically significant difference in average age at diagnosis between intact and neutered male dogs, but not between intact and spayed females. Small sample size of cancer dogs could have impacted the power of statistical test results and been a contributor to statistical insignificance seen throughout the analysis. Dogs moving multiple times throughout the duration of the study can affect interpretations and implications from prevalence by state findings. Prevalence was also calculated using only the total GRLS study population the resided in respective states as the denominator, effecting generalizability of the analysis findings.
Description
Rights Access
Embargo expires: 12/29/2024.
Subject
environmental health
human oncology
epidemiology
canine oncology