Integrative cluster classification to predict pathological complete response to neoadjuvant chemotherapy in early breast cancer.

579Background: Integrative Clustering (IntClust) is a breast cancer (BC) classification of 10 different subgroups with distinctive molecular profiles and clinical outcomes (Curtis et al., Nature 2012). We implemented an IntClust classifier based on Copy Number Alterations (CNAs) and explored its prognostic role in a cohort of pre- and post-treatment (ttm) tumors from the neoadjuvant trials GEICAM/2006-03 (NCT00432172) and GEICAM/2006-14 (NCT00841828). Methods: GEICAM/2006-03 HER2-negative pts were selectively treated according to clinical subtypes: triple negative (TN) pts with standard taxane/anthracycline-based chemotherapy (TA-CT) +/- carboplatin, and luminal A patients were randomised to TA-CT vs. hormonotherapy; GEICAM/2006-14 HER2+ pts received TA-CT + anti-HER2 therapy. Shallow-whole genome Illumina sequencing DNA data from 204 paraffin-BC (100 pre- and 104 post-ttm tumors) were used to identify CNAs and grouped samples in the 10–IntClust classification. A functional clustering by grouping IntClust...