Feline mammary tumors: Comparison of histological grading systems and Nottingham prognostic index in a 3-year survival study (Part II)

Based on the longstanding Nottingham (EE) histological grading system - originally for grading of human breast carcinomas - a mitotic-modified (MMEE) and a revised (REE) grading systems were proposed to better accommodate feline mammary carcinoma (FMC) histopathological features. Additionally, a novel grading system (MGS) has also been proposed. This study aims to directly compare the EE grading with the proposed frameworks using the originally proposed mitotic count thresholds, and to evaluate their association with clinical outcomes. Additionally, we aim to calculate the veterinary Nottingham Prognostic Index (vet-NPI) and assess its prognostic value. Cohen's kappa agreement analysis revealed poor, slight and fair agreement between EE grading and MMEE, REE and MGS grading systems, respectively. Although all histological grading frameworks were associated with overall survival in the univariate analysis, only the EE and REE systems remained significant prognostic factors after adjustment to other covariates. Furthermore, REE and MGS gradings were associated with tumor-specific survival, but the latter lost its significance on multivariate analysis. The MGS system was the only one that was significantly linked with disease-free interval in both univariate and multivariate analyses. While the inclusion of LVI and nuclear form represent good steps toward improving the prognostic value of histological grade, the mitotic count thresholds appeared inadequate for our cohort, raising concerns about its applicability across different settings. Our findings support the inclusion of the vet-NPI in routine diagnostic histopathology reports of FMC to improve TSS prognostic assessment when using the EE and MMEE grading systems.