Early vs late survival patterns of chemoimmunotherapy compared with immune checkpoint inhibitor monotherapy in non–small cell lung cancer.

Abstract

e20603 Background: Immune checkpoint inhibitors (ICIs) have improved long-term outcomes in multiple cancers, and ICI trials often suggest durable benefit in a subset of patients. In first-line non–small cell lung cancer (NSCLC), ICIs are administered either as ICI monotherapy or combined with chemotherapy. Although chemotherapy is generally believed to improve early disease control, its impact on longer-term survival remains unclear. We assessed whether the relative effect of chemoimmunotherapy versus ICI monotherapy differs between early and later time periods. Methods: We reviewed pivotal first-line NSCLC trials of FDA-approved ICIs reporting long-term overall survival (OS) outcomes. Eligible arms were categorized as ICI-only regimens versus ICI plus chemotherapy; six ICI-only and six combination arms were included. Published Kaplan–Meier curves were digitized and individual patient data (IPD) were reconstructed using an IPD-from-KM approach. OS was estimated using Kaplan–Meier methods. To compare early versus later treatment effects, we performed a piecewise Cox analysis with a prespecified cutoff at 12 months, estimating hazard ratios (HRs) for 0–12 months and > 12 months. We also evaluated the PD-L1 ≥50% subgroup where available. Results: A total of 2,054 patients in the ICI-only arms and 1,902 patients in the ICI plus chemotherapy arms were analyzed. In the overall population, the HR for chemoimmunotherapy versus ICI monotherapy was 0.98 (95% CI, 0.89–1.09) during 0–12 months and 1.24 (95% CI, 1.13–1.37) during > 12 months, indicating a neutral early effect with a later disadvantage. In the PD-L1 ≥50% subgroup, the corresponding HRs were 0.84 (95% CI, 0.66–1.06) during 0–12 months and 1.11 (95% CI, 0.91–1.35) during > 12 months, suggesting a more favorable early effect that diminished and trended toward harm in the later period. Conclusions: In this reconstructed-IPD, cross-trial analysis of first-line NSCLC ICI trials, the relative effect of chemoimmunotherapy versus ICI monotherapy appeared time-dependent, with attenuation and potential reversal beyond 12 months. These findings support evaluating early and late survival phases separately when interpreting chemoimmunotherapy outcomes.