Abstract
INTRODUCTION: Serum proteomics and mutations in the epidermal growth factor receptor (EGFR) and KRAS have been associated with benefit after therapy with EGFR-targeted therapies in non-small cell lung cancer, but all three have not been evaluated in any one study. HYPOTHESIS: Pretreatment serum proteomics predicts survival in Western advanced non-small cell lung cancer patients with wild-type EGFR and independent of KRAS mutation status. METHODS: We analyzed available biospecimens from Eastern Cooperative Oncology Group 3503, a single-arm phase II study of erlotinib in first-line advanced lung cancer, for proteomics signatures in the previously described serum matrix-assisted laser desorption ionization proteomic classifier (VeriStrat) as well as for KRAS and EGFR mutations. RESULTS: Out of 137 enrolled patients, analyzable biologic samples were available on 102. Nine of 41 (22%) demonstrated KRAS mutations and 3 of 41 (7%) harbored EGFR mutations. VeriStrat classification identified 64 of 88 (73%) as predicted to have "good" and 24 of 88 (27%) predicted to have "poor" outcomes. A statistically significant correlation of VeriStrat status (p < 0.001) was found with survival. EGFR mutations, but not KRAS mutations, also correlated with survival. CONCLUSIONS: The previously defined matrix-assisted laser desorption ionization predictor remains a potent and highly clinically significant predictor of survival after first-line treatment with erlotinib in patients with wild-type EGFR and independent of mutations in KRAS.