OA14.04 Five-Year Outcomes From the Randomized, Phase 3 Trials CheckMate 017/057: Nivolumab vs Docetaxel in Previously Treated NSCLC

Historically, outcomes for advanced non-small cell lung cancer (NSCLC) have been poor, with 5-year survival rates <5% with conventional chemotherapy. Nivolumab, a programmed death-1 (PD-1) inhibitor, was approved in 2015 for patients with previously treated advanced NSCLC based on two randomized phase 3 trials, CheckMate 017 (NCT01642004; squamous) and CheckMate 057 (NCT01673867; non-squamous), which demonstrated improved overall survival (OS) vs docetaxel. We report 5-year pooled efficacy and safety from these trials, representing the longest survival follow-up for randomized phase 3 trials of an immune checkpoint inhibitor in advanced NSCLC. Patients (N = 854; CheckMate 017/057 pooled) with advanced NSCLC, ECOG performance status (PS) ≤1, and progression during or after first-line platinum-based chemotherapy, were randomized 1:1 to nivolumab 3 mg/kg Q2W or docetaxel 75 mg/m2 Q3W until progression or unacceptable toxicity. After completion of the primary analyses, patients in the docetaxel arm no longer receiving benefit could cross over to receive nivolumab. OS was the primary endpoint for both studies. At 5-year follow-up, 50 nivolumab patients and 9 docetaxel patients were alive. Baseline characteristics of 5-year survivors in both arms were similar to the overall population and patients who survived <1 year, except for a higher percentage of patients with ECOG PS 0 or tumor programmed death ligand-1 (PD-L1) expression ≥1% on nivolumab and ECOG PS 0 and Stage IIIB NSCLC on docetaxel. Nivolumab continued to show long-term OS and progression-free survival (PFS) benefit vs docetaxel with 5-year OS rates 13% vs 3% (HR, 0.68 [95% CI, 0.59–0.78]) and PFS rates 8% vs 0% (0.79 [0.68–0.92]). OS benefit with nivolumab vs docetaxel was observed across subgroups including patients with tumor PD-L1 expression <1%, baseline liver and adrenal metastases, neutrophil-to-lymphocyte ratio < median, lactate dehydrogenase ≥ upper limit of normal or no baseline proton-pump inhibitor use. Among patients with an objective response to nivolumab (20%) or docetaxel (11%), 32% remained in response at 5 years vs none on docetaxel, with a median duration of response of 19.9 vs 5.6 months, respectively. Of the 5-year nivolumab vs docetaxel survivors, 36% vs 0% were on study drug, 20% vs 67% received subsequent immunotherapy (on or off study), and 10% vs 0% were off study drug, progression free, with no subsequent therapy. No new safety signals were observed with longer follow-up. Between 3 and 5 years’ follow-up, 8 of the 31 (26%) nivolumab-treated patients reported a treatment-related adverse event, 1 (3%) grade 3–4. The most common select adverse events (events with a potential immunological cause) were related to skin, in 4 (13%) patients, none of which were grade 3–4. CheckMate 017 and 057 are the first phase 3 trials to report 5-year outcomes for a PD-1 inhibitor in previously treated advanced NSCLC, demonstrating a greater than 4-fold increase in 5-year OS rates with nivolumab (13%) over docetaxel (3%). Nivolumab remained well tolerated with no new safety signals.