Abstract P5-10-01: Clinical relevance of TP53 mutations and genomic instability in node positive breast cancer

Abstract

Abstract Background - Aim: Malignant tumors are currently understood as genetic mosaics. For breast cancer (BC) patient assessment, the challenge remains to use the increasingly available tumor genomic data along with standard clinicopathological parameters. Methods: Histologically reviewed, paraffin tumor tissue DNA samples (N=1771), from patients who had received anthracycline-based adjuvant chemotherapy in the frame of 3 randomized trials by HeCOG (HE10/00, pre-trastuzumab [T]; HE10/05 & HE10/08, post-T era), were investigated with targeted massively parallel sequencing (tMPS) for variants in 58 genes previously implicated in BC (custom panel, Ion Torrent systems). Upon multiple stringent quality filters, pathogenic coding mutations (MUT) and allelic (im)balance (AI, for 5 or more imbalanced SNPs; AI-high: at least 3 AI loci per sample) were evaluable in 1388 cases (78.4%). IHC4 (HER2 ASCO 2013 guidelines, Ki67 20% cutoff) was used for BC subtyping (luminal A, luminal B, luminal HER2, HER2-enriched and TNBC). Results: In total, 1264 tumors carried 1 or more MUT and 788 were positive for AI. No two tumors had the same genetic profile. Top 5 MUT genes (>5% of all cases) were TP53, PIK3CA, CDH1, GATA3 and MLL3; top 5 genes with >50% frequency within the tumor (possible drivers) were TP53, PIK3CA, GATA3, MAP3K1 and PTEN. Top imbalanced (unstable) loci included 5p15.33 (TERT), 7p12 (EGFR), 17p13.1 (TP53), 17q21 (BRCA1), 10p15 (GATA3) and 10q23.3 (PTEN). TP53 MUT were more prominent in HER2-enriched and TNBC (p<0.0001) and were strongly associated with AI (p<0.0001). Tumor genomic characteristics were modelled with 3-year disease-free survival (DFS) as end-point for the evaluation of tumor aggressiveness; patients without infiltrated lymph nodes (LN) were not evaluated in this analysis. In all 3 cohorts, 1-3 positive LN were favourable and >3 LN were unfavourable prognosticators. Among patients with HER2-negative tumors (evaluable N=851), those with >3 positive LN and AI-low tumors fared similarly to patients with 1-3 LN; those with AI-high tumors fared significantly worse (log-rank, p<0.0001). DFS for AI-low TNBC was similar to luminal A and luminal B tumors, while AI-high TNBC fared significantly worse (p=0.0007). Again in HER2-negative patients, TP53 MUT were associated with worse DFS irrespectively of subtype (p<0.0001). Among patients with HER2-positive tumors (evaluable N=246), DFS for T-treated patients (N=100) without TP53 MUT was similar to those who did not receive T, while T-treated patients with TP53 MUT fared significantly better (p=0.011); AI-high was an adverse prognosticator only for patients with >3 positive LN who did not receive T, while this feature did not interfere with outcome in T-treated patients or in non-T-treated patients with 1-3 LN (p<0.0001). Conclusions: This prospective-retrospective translational research study, one of the largest presenting tMPS data from paraffin tissues, confirms the extensive genetic diversity of BC at the individual tumor level. Genomic (in)stability, as assessed with AI and TP53 MUT status, may help in further defining prognosis in BC patients with early aggressive disease. TP53 MUT are indirectly shown to predict for T-specific benefit, which merits validation in larger cohorts. Citation Format: Vassiliki Kotoula, Sotiris Lakis, Zoi Alexopoulou, Elpida Charalambous, Kyriaki Papadopoulou, Aggeliki Lyberopoulou, Eleftheria Tsolaki, Eleni Timotheadou, Dimitrios Pectasides, Flora Zagouri, George Pentheroudakis, Angelos Koutras, Helen Gogas, Konstantine T Kalogeras, Ralph M Wirtz, George Fountzilas. Clinical relevance of TP53 mutations and genomic instability in node positive breast cancer [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P5-10-01.