The prognostic effect of single and multiple cancer-related somatic mutations in resected non-small-cell lung cancer

Abstract

IntroductionSomatic mutations are becoming increasingly important biomarkers for treatment selection and outcome in patients with non-small-cell lung cancer (NSCLC). The role of multiple somatic mutations in early-stage NSCLC is unclear. MethodsTissue from 214 patients with resected NSCLC at the Princess Margaret Cancer Centre was analyzed by next-generation sequencing by Mi-SEQ or Sequenom multiplex platforms. Associations between mutation status, baseline patient characteristics and outcomes (disease-free survival (DFS) after surgical resection and overall survival (OS)) were investigated. ResultsSomatic mutations were identified in 184 patients with resected stage I-III NSCLC: None (n = 30), single (n = 101) and multiple (≥2, n = 83). Multiple mutations were significantly associated with younger age (p = 0.0006), female sex (p = 0.012), smoking status (p = 0.002) and adenocarcinoma histology (p = 0.0001).TP53, KRAS and EGFR were the most common mutations. TP53 mutation was the most frequent co-mutation occurring in 72% of patients with multiple mutations. In resected stage I-III patients, multiple mutations were significantly associated with worse DFS (HR = 2.56, p = 0.003) but not OS on univariate analysis. Patients with KRAS and EGFR mutations were also associated with shorter DFS (HR = 2.52, p = 0.016 and HR = 4.37, p = 0.001 respectively) but no OS difference. TP53 mutation was associated with both shorter DFS (HR = 2.21, p = 0.02) and OS (HR = 3.08, p = 0.02). In subgroup univariate analysis, poorer DFS was associated with multiple mutations (p = 0.0015), EGFR (HR = 3.14, p = 0.006), and TP53 (HR = 2.46, p = 0.018) in patients with stage I disease. ConclusionThe presence of known somatic mutations is associated with worse DFS in resected NSCLC. The differences are both statistically significant and clinically relevant. The presence of EGFR, KRAS and TP53 mutations was also associated with adverse outcomes. Larger datasets are required to validate whether mutational status is an independent prognostic factor in early stage NSCLC.