Neoadjuvant nivolumab (N) or nivolumab plus ipilimumab (NI) for resectable non-small cell lung cancer (NSCLC)

Abstract

Background: Neoadjuvant N induces a 45% major pathologic response (MPR) in resected NSCLCs. We report preliminary results of NEOSTAR (NCT03158129) - a phase 2 trial evaluating neoadjuvant N or NI in resectable NSCLC pts. Methods: Pts with stage I-IIIA (singleN2) resectable NSCLC (AJCC 7th), ECOGPS 0-1, were randomized 1:1 toN(3mg/kg IVQ2 weeks, onD1, 15, 29) or NI (N plus I, 1mg/kg IV on D1) followed by surgery (n=22/arm). Primary endpoint isMPR (≤10% viable tumor cells at surgery), hypothesized to be higher thanMPR to induction chemotherapy historical controls. Immune infiltrates were assessed by flow cytometry on checkpoint inhibitor (CPI)- treated tumors and compared to untreated resected samples (ICON set). Results: As of 9/6/2018, 33 pts were randomized, 17 toN, 16 toNI: mean age 65, 67%males, 21% never smokers, stage I n=10, II n=15, III n=8, 58%adenocarcinomas. 30 pts completed neoadjuvant therapy (2 ongoing, 1 G3 hypoxia [N]) & 26 had surgery (5 unresectable [2N, 3NI], 2pending). OverallMPRratewas 26%(8/31).MPRrate toNandNIwere 25% (4/16) and 27%(4/15), respectively.Median%of viable tumor cells was lower in tumors resected postNI vs.N(28% vs. 65%, p=0.32). There were 5 pathologic CRs (2N, 3NI). RadiographicORR by RECIST v1.1 was 19% (5 PR [N], 1 CR [NI]). 19% of pts had PD (6/ 31, 3 N, 3NI). Surgical complications included 1 bronchopleural fistula (BPF). Treatmentrelated AEs included G5 death due to BPF post steroid-treated pneumonitis (n=1, N);G3 pneumonia (n=1, N), hypoxia (n=1,N);G2 cough (n=3, NI), rash (n=1, N), fatigue (n=1, N). CPIs increased proliferative (Ki67+)& activated(ICOS+) effectorCD8+ & CD4+ TILs in treated vs. untreated tumors (p<0.0001). CD27+CD28+ effector memory CD8+ TILs were higher inNvs. NI (49% vs. 33%, p=0.06). Ki67+CD103+ tissue resident effector CD8+(98% vs. 65%, p=0.1)& CD4+(99% vs. 40%, p=0.03) TILs and Tregs (97% vs. 47%, p=0.06) were higher in NI vs.N. Conclusions: Neoadjuvant CPI is overall safe and induces a 26% MPR rate, with a trend towards less viable tumor after NI. Preliminary results suggest neoadjuvant CPIs induce higher TIL proliferation and activation vs. untreated tumors. NI may induce higher proliferation of different T cell subsets vs. N, which may lead to distinct antitumor immune responses.