Developing and testing prognostic bio‐scoring systems for canine mammary gland carcinomas

Canine mammary carcinomas (CMC) represent a range of histolopathological subtypes with diverse biological behaviours. Several individual factors, including stage, grade, subtypes and presence of invasion, predict outcome. Less is known how these factors interact and impact prognosis. The purpose of this work was to develop and test comprehensive bio-scoring systems in CMCs. Clinical and histopathological data from 127 dogs with MCs treated through two prospective studies were obtained. All dogs underwent standardized pre-surgical staging, treatments and regular follow-up visits. All tumours were evaluated, classified and graded according to published guidelines. Time to primary metastasis was the main endpoint in this study. Two bio-scoring systems were developed: The multivariate scoring (MVS) was based on traditional statistical analysis where only factors significant in the multivariate analysis (tumour size and grade) were kept for the final model. The refined flexible scoring (RFS) system was based on results from subgroup analysis, which guided the development of a flexible system. Progressive worsening prognosis was observed with increasing bio-scores in both systems. MVS: Median primary metastasis-free survival (TTM1 days) was not reached in dogs with bio-scores 0 to 5, 10, 15 and 648, 149, 317, in MVS groups 25, 30, 40, respectively. Similarly, TTM1 was not reached in dogs with RFS 0, 1, 2 and 374, 407 and 149, in dogs with bio-scores 3, 4, 5, respectively. However, a more distinct separation between dogs with high risk vs low risk for metastasis was observed with RFS, suggesting superior overall prognostication regarding the risk for metastasis.