Statistical partitioning of Ki-67 distribution can refine Molecular Classifica-tion of invasive Breast Carcinoma (Preprint)

Abstract

Breast carcinoma is one of the most common histological types of Breast Cancer, exploring a new approach that allows to do a quantitative description in order to characterize its heterogeneity and refine its classification is one of the main interests for pathologists.

The purpose of our study is to explore further statistically significant subdivisions beyond breast cancer molecular classification that is routinely established in pathology departments.

We conducted a 5-year retrospective study on 1266 invasive breast carcinomas of moroccan pa-tients, collected at the Pathology Department of Ibn-Rochd University Hospital in Casablanca, and followed at King MohammedVI National Centre for the Treatment of Cancers. We elaborated an Estimation-Maximization clustering, based on the main Breast cancer prognosis biomarkers: Ki-67, HER2, oestrogen and progesterone receptors, evaluated by Immunohistochemistry.

Each molecular subgroup could be partitioned into two further subdivisions: Cluster1, with average Ki-67 of 16.26%(±11.9) across all molecular subgroups and higher frequency within luminal sub-groups, and Cluster2, with average Ki-67 of 68.8%(±18) across all molecular subgroups; and higher frequency in HER2 as well as in triple negative subgroups. Overall Survival of the two clusters was significantly different, with 5-year rates of 52 and 37 months for Cluster1 and Cluster2, respectively (p=0.000001). Moreover, patient survival within the same molecular subgroup varied remarkably depending on cluster membership. Three independent datasets (Algerian, TCGA-BRCA and METABRIC) were also analysed to assess the reproducibility of this new “2-clusters partition” through several clustering methods and validation measures. Two different al-gorithms to evaluate the prognostic importance, VSURF and MinimalDepth, confirmed that this new subdivision is able to predict patient survival better than several histoprognostic features.

Our results highlight a new refinement of the breast cancer molecular classification and provide a simple and improved way to classify tumors that could be applied in low to medium income countries. This is the first study of its kind addressed in an African context.