Breast cancer patient stratification using a molecular regularized consensus clustering method

Abstract

Breast cancers are highly heterogeneous with different subtypes that lead to different clinical outcomes including prognosis, response to treatment and chances of recurrence and metastasis. An important task in personalized medicine is to determine the subtype for a breast cancer patient in order to provide the most effective treatment. In order to achieve this goal, integrative genomics approach has been developed recently with multiple modalities of large datasets ranging from genotypes to multiple levels of phenotypes. A major challenge in integrative genomics is how to effectively integrate multiple modalities of data to stratify the breast cancer patients. Consensus clustering algorithms have often been adopted for this purpose. However, existing consensus clustering algorithms are not suitable for the situation of integrating clustering results obtained from a mixture of numerical data and categorical data. In this work, we present a mathematical formulation for integrative clustering of multiple-source data including both numerical and categorical data to resolve the above issue. Specifically, we formulate the problem as a novel consensus clustering method called Molecular Regularized Consensus Patient Stratification (MRCPS) based on an optimization process with regularization. Unlike the traditional consensus clustering methods, MRCPS can automatically and spontaneously cluster both numerical and categorical data with any option of similarity metrics. We apply this new method by applying it on the TCGA breast cancer datasets and evaluate using both statistical criteria and clinical relevance on predicting prognosis. The result demonstrates the superiority of this method in terms of effectiveness of aggregation and differentiating patient outcomes. Our method, while motivated by the breast cancer research, is nevertheless universal for integrative genomics studies.

Keywords: Breast cancer prognosis; Breast cancer subtypes; Cancer patient stratification; Consensus clustering; Integrative genomic.

Figure 1

Integration of molecular expression data with clinically-defined patient stratification. Although sample d_i and sample d_j come from different clinical sub- types (I and II respectively), they come from the same, stable and dense molecular cluster, so they are desired to be combined in the consensus clustering.

Figure 2

Plotted values of NMI of MRCPS and other methods for different values of k.

Figure 3

Prognostic power of different patient stratification methods. Kaplan-Meier survival curves of (a) Histology Type, (b) Tumor Grade, (c) Disease Stage, (d) BCE, (e) HGPA and (f) CSPA listed along with estimated p-values (log-rank test). The numbers of the patients in each stratification are also listed in the parentheses.

Figure 4

Prognostic power of MRCPS. (a) Kaplan-Meier survival curves of MRCPS with its p-value (log-rank test). (b) Disease stage in each subtype; stages earlier than stage III are considered to be early, with the rest considered to be late. (c) Tumor grades for each subtype. Grades less than T3 are considered to be lower grades, with the rest considered to be higher.

Figure 5

A gene network identified from the high-risk group.

Figure 6

Top diseases and functions identified from the subtype-specific genes using IPA.