Atezolizumab Treatment Beyond Progression in Advanced NSCLC: Results From the Randomized, Phase III OAK Study

Introduction: Cancer immunotherapy may alter tumor biology such that treatment effects can extend beyond radiographic progression. In the randomized, phase III OAK study of atezolizumab (anti–programmed death-ligand 1) versus docetaxel in advanced NSCLC, overall survival (OS) benefit with atezolizumab was observed in the overall patient population, without improvement in objective response rate (ORR) or progression-free survival (PFS). We examine the benefit-risk of atezolizumab treatment beyond progression (TBP). Methods: Eight hundred fifty patients included in the OAK primary efficacy analysis were evaluated. Atezolizumab was continued until loss of clinical benefit. Docetaxel was administered until Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1) disease progression (PD)/unacceptable toxicity; no crossover to atezolizumab was allowed. ORR, PFS, post-PD OS, target lesion change, and safety were evaluated. Results: In atezolizumab-arm patients, ORR was 16% versus 14% and median PFS was 4.2 versus 2.8 months per immune-modified RECIST versus RECIST v1.1. The median post-PD OS was 12.7 months (95% confidence interval [CI]: 9.3–14.9) in 168 atezolizumab-arm patients continuing TBP, 8.8 months (95% CI: 6.0–12.1) in 94 patients switching to nonprotocol therapy, and 2.2 months (95% CI: 1.9–3.4) in 70 patients receiving no further therapy. Of the atezolizumab TBP patients, 7% achieved a post-progression response in target lesions and 49% had stable target lesions. Atezolizumab TBP was not associated with increased safety risks. Conclusions: Within the limitations of this retrospective analysis, the post-PD efficacy and safety data from OAK are consistent with a positive benefit-risk profile of atezolizumab TBP in patients performing well clinically at the time of PD.