Intertumor heterogeneity of non-small-cell lung carcinomas revealed by multiplexed mutation profiling and integrative genomics.

Non-small-cell lung cancer (NSCLC) is a heterogeneous disease, with a burden of genomic alterations exceeding most other tumors. The goal of our study was to evaluate the frequencies of co-occurring mutations and copy-number aberrations (CNAs) within the same tumor and to evaluate their potential clinical impact. Mass-spectrometry based mutation profiling using a customized lung cancer panel evaluating 214 mutations across 26 key NSCLC genes was performed on 230 nonsquamous NSCLC and integrated with genome-wide CNAs and clinical variables. Among the 138 cases having at least one mutation, one-third (41, 29.7%) showed two or more mutations, either in the same gene (double mutation) or in different genes (co-mutations). In epidermal growth factor receptor (EGFR) mutant cancers, there was a double mutation in 18% and co-mutations in the following genes: TP53 (10%), PIK3CA (8%), STK11 (6%) and MET (4%). Significant relationships were detected between EGFR mutation and 1p, 7p copy gains (harboring the EGFR gene) as well as 13q copy loss. KRAS mutation was significantly related with 1q gain and 3q loss. For Stage I, tumors harboring at least one mutation or PIK3CA mutation were significantly correlated with poor prognosis (p-value = 0.02). When combining CNAs and mutational status, patients having both KRAS mutation and the highest related CNA (3q22.3 copy loss) showed a significant poorer prognosis (p-value = 0.03). Our study highlights the clinical relevance of studying tumor complexity by integrative genomic analysis and the need for developing assays that broadly screen for both "actionable" mutations and copy-number alterations to improve precision of stratified treatment approaches.