Abstract
BACKGROUND: The phase II FRACTION-Lung platform trial (NCT02750514) aimed to evaluate novel immunotherapy combinations in advanced non-small-cell lung cancer (aNSCLC). PATIENTS AND METHODS: Adults with aNSCLC and Eastern Cooperative Oncology Group performance status ≤1 were allocated to tracks 1 to 5 based on programmed cell death protein (ligand) 1 [PD-(L)1] status and/or prior receipt of anti-PD-(L)1 therapy, and received nivolumab (anti-PD-1) as monotherapy or combination therapy [dasatinib (multi-kinase inhibitor), ipilimumab (anti-CTLA-4), relatlimab (anti-LAG-3), or linrodostat (IDO1 inhibitor)]. Patients with progression in one arm could be re-randomized. Primary endpoints were objective response rate (ORR); duration of response (DOR); and 24-week progression-free survival (PFS) rate. Safety was a secondary endpoint; biomarker analyses were exploratory. RESULTS: Overall, 295 patients were treated (191 received no prior immunotherapy, 118 received prior immunotherapy; 14 were re-randomized assigned). Among patients without prior immunotherapy, ORRs ranged from 0% (multiple arms) to 25% (nivolumab plus ipilimumab: 3/12 patients; nivolumab plus dasatinib: 1/4 patients), and among patients with prior immunotherapy from 2.3% (nivolumab plus linrodostat: 1/43 patients) to 5.6% (nivolumab plus ipilimumab: 1/18 patients). Median DOR was not estimated due to limited tumor responses. Twenty-four-week PFS rates ranged from 30.2% [nivolumab-only (n = 40)] to 45.5% [nivolumab plus ipilimumab (n = 12)] among patients without prior immunotherapy and 11.1% [nivolumab plus linrodostat (n = 43)] to 19.1% [nivolumab plus dasatinib (n = 41)] among patients with prior immunotherapy. No new safety signals were observed. Small sample sizes in each treatment group limited analyses, including biomarkers. The study was discontinued early due to an evolving treatment landscape and limited efficacy signals. Given the rolling, adaptive design, this trial was not statistically powered to assess treatment benefits and draw definitive efficacy conclusions, as early termination and the changing standard of care limited accrual. CONCLUSIONS: Despite limited efficacy, the innovative adaptive trial design allowed rapid evaluation of multiple regimens, which may inform future trials.