Neoadjuvant nivolumab in early-stage non–small cell lung cancer (NSCLC): Five-year outcomes.

Abstract

8537 Background: Neoadjuvant (neoadj) immune checkpoint blockade (ICB) with anti-PD-1 therapy has shown increasing promise for early stage NSCLC, with long-term clinical outcomes still maturing. Our group reported the first phase I/II trial of neoadj nivolumab (nivo) in resectable NSCLC, finding therapy to be safe and feasible. We now present final clinical results from this cohort, representing the longest follow up data for neoadj anti-PD-1 to date. Methods: Two doses of neoadj nivo (3 mg/kg) were given prior to resection in 21 patients (pts) with resectable NSCLC. 5-year (yr) follow-up data, including recurrence-free survival (RFS), overall survival (OS) and association with pathologic response were tabulated. Event time distributions were estimated with the Kaplan-Meier method. All p-values are two-sided with 0.05 significance level. Results: At a median follow up of 63 months, 3-, 4- and 5-yr survival rates were 85, 80, and 80% respectively. RFS rates at 3-, 4- and 5-yrs were 65, 60, and 60% respectively. As previously reported, major pathologic response (MPR: ≤10% viable tumor) was 45%, and pathologic complete response (pCR) rate was 10%. The hazard ratio (HR) for pathologic down-staging was in the direction of improved RFS, without meeting statistical significance (HR 0.36, 95% CI 0.07-1.75, p = 0.2). RFS HR estimates for MPR and an alternative pathologic cut-off of less than 50% residual tumor (RT), were 0.61, (95% CI 0.15-2.44, p = 0.48) and 0.36, (95% CI 0.09-1.51, p = 0.16) respectively. The direction of the effect of pre-treatment PD-L1 positivity (≥1%) was to improve RFS (HR 0.36, 95% CI 0.07-1.85, p = 0.22). At 5-yr follow up, 8 of 9 (89%) pts with MPR were alive and no cancer deaths have occurred. Amongst pts with MPR, 1/9 pts had a cancer recurrence in the mediastinum treated successfully with definitive chemoradiotherapy. Both pts with pCR are alive and without recurrence. Patterns of all recurrences in this cohort are summarized in table 1. No long-term immune-related adverse events have occurred other than one G3 dermatologic event. Conclusions: The 5-yr clinical outcomes for neoadj nivo in resectable NSCLC compare favorably to historical trends. MPR trended toward improved RFS, while definitive conclusions are limited by our cohort size and overall low recurrence rate. Thresholds of %RT beyond pCR and MPR in this setting should be explored in larger prospective studies. PD-L1 expression may play a role in predicting long-term response, but larger prospective studies are needed. Clinical trial information: NCT02259621. [Table: see text]