A Canine Model of Familial Mammary Gland Neoplasia

Intact female Beagles from life-span studies in the Lovelace Respiratory Research Institute colony were examined for mammary tumor incidence. The breeding colony, founded in 1963, produced five generations from 28 founder females. After proportional hazards analysis, two maternal families were shown to have markedly different phenotypes, one susceptible and one resistant to mammary neoplasia, as compared with the entire colony. When tumors were subdivided into benign and malignant based on local invasiveness, familial differences in tumor incidence were preserved for each tumor type. Fifty-seven females in the susceptible family developed 149 benign and 39 malignant tumors, and 95 females in the resistant family developed 70 benign and 20 malignant tumors. The ratio of benign to malignant tumors of about 4:1 for both families was higher than expected. Using Kaplan–Meier and log-rank analyses, the susceptible family had a 50% malignant tumor incidence by age 13.6 years, whereas the resistant family did not have a 50% incidence until 17.0 years (P = 0.0065). Because of marked censoring, Kaplan–Meier analyses could not provide an estimate of the 50% benign tumor incidence; mean incidence age was calculated instead. These estimates for benign tumors for susceptible and resistant families were 10.8 and 13.8 years (P = 0.0001), respectively. Using χ2tests, families had no differences in the occurrence of the types of benign (P = 0.098) or malignant (P = 0.194) tumors or in the ratio of benign to malignant tumors (P = 0.778). Immunohistochemical analysis of malignant tumors from both families did not demonstrate differences in p53 mutation rate or p185erbB-2expression. These results suggest that 1) genetic factors produce familial differences in the age of onset of both benign and malignant mammary tumors; histologic types do not segregate by family; 2) the ratio of benign to malignant tumors is greater than formerly reported; and 3) neither p53 nor p185erbB-2alterations are the basis for the familial predisposition.