Ramucirumab plus erlotinib versus placebo plus erlotinib in previously untreated EGFR-mutated metastatic non-small-cell lung cancer (RELAY): exploratory analysis of next-generation sequencing results

BACKGROUND: Ramucirumab plus erlotinib (RAM + ERL) demonstrated superior progression-free survival (PFS) over placebo + ERL (PBO + ERL) in the phase III RELAY study of patients with epidermal growth factor receptor (EGFR)-mutated metastatic non-small-cell lung cancer (EGFR+ mNSCLC; NCT02411448). Next-generation sequencing (NGS) was used to identify clinically relevant alterations in circulating tumor DNA (ctDNA) and explore their impact on treatment outcomes. PATIENTS AND METHODS: Eligible patients with EGFR+ mNSCLC were randomized 1 : 1 to ERL (150 mg/day) plus RAM (10 mg/kg)/PBO every 2 weeks. Liquid biopsies were to be prospectively collected at baseline, cycle 4 (C4), and postdiscontinuation follow-up. EGFR and co-occurring/treatment-emergent (TE) genomic alterations in ctDNA were analyzed using Guardant360 NGS platform. RESULTS: In those with valid baseline samples, detectable activating EGFR alterations in ctDNA (aEGFR+) were associated with shorter PFS [aEGFR+: 12.7 months (n = 255) versus aEGFR−: 22.0 months (n = 131); hazard ratio (HR) = 1.87, 95% confidence interval (CI) 1.42-2.51]. Irrespective of detectable/undetectable baseline aEGFR, RAM + ERL was associated with longer PFS versus PBO + ERL [aEGFR+: median PFS (mPFS) = 15.2 versus 11.1 months, HR = 0.63, 95% CI 0.46-0.85; aEGFR−: mPFS = 22.1 versus 19.2 months, HR = 0.80, 95% CI 0.49-1.30]. Baseline alterations co-occurring with aEGFR were identified in 69 genes, most commonly TP53 (43%), EGFR (other than aEGFR; 25%), and PIK3CA (10%). PFS was longer in RAM + ERL, irrespective of baseline co-occurring alterations. Clearance of baseline aEGFR by C4 was associated with longer PFS (mPFS = 14.1 versus 7.0 months, HR = 0.481, 95% CI 0.33-0.71). RAM + ERL improved PFS outcomes, irrespective of aEGFR mutation clearance. TE gene alterations were most commonly in EGFR [T790M (29%), other (19%)] and TP53 (16%). CONCLUSIONS: Baseline aEGFR alterations in ctDNA were associated with shorter mPFS. RAM + ERL was associated with improved PFS outcomes, irrespective of detectable/undetectable aEGFR, co-occurring baseline alterations, or aEGFR+ clearance by C4. aEGFR+ clearance by C4 was associated with improved PFS outcomes. Monitoring co-occurring alterations and aEGFR+ clearance may provide insights into mechanisms of EGFR tyrosine kinase inhibitor resistance and the patients who may benefit from intensified treatment schedules.