Real-world outcomes after second-line treatment in non-small cell lung cancer (NSCLC) patients treated with immunotherapy.

Abstract

e21620 Background: In patients (pts) with advanced non-small cell lung cancer (NSCLC), national guidelines recommend against retrial of immunotherapy (IO) if there is disease progression on IO in the 1st-line (1L). However, optimal 2nd-line (2L) therapy after 1L IO remains unclear and there is significant practice variation. We compared outcomes of 2L approaches after 1L IO or chemoimmunotherapy (chemoIO). Methods: This retrospective cohort study utilized the Flatiron Health EHR-derived de-identified advanced NSCLC database. The study population included pts with disease progression on 1L IO or chemoIO and who subsequently received 2L therapy. Pts with targetable alterations were excluded. We defined the exposure by type of 2L therapy (IO, chemoIO, chemo). Multivariate covariates included age, sex, race, 1L progression-free survival (PFS) and PDL1 level. Median overall survival (mOS) and median real-world PFS (mPFS, based upon abstraction of clinician documentation) times were estimated from Kaplan-Meier curves. A multivariate Cox proportional hazard model computed hazard ratios (HRs) to assess the effectiveness of 2L treatment. Results: 532 NSCLC pts received 1L IO and a 2L therapy, of which 393 (74%) received 1L IO and 139 (26%) received 1L chemoIO. Among 1L IO patients, 2L therapies included chemo (315 (80%)), IO (39(10%), 18/39 (46%) switched IO), and chemoIO (39(10%)). Among 1L chemoIO patients, 2L therapies included: chemo (121 (87%)), IO (8(6%), 6/8 switched IO) and ChemoIO (10 (7%), 8/10 changed chemo used). All pts who received 2L ChemoIO continued the same IO agent. Demographics were well balanced between 2L groups except for higher PDL1 level in the IO-based groups and older age in the IO alone group. There was no statistically significant difference in mPFS or mOS between 2L IO and non-IO containing regimens (Table), nor were there differences among patients switching IO agents in the 2L (multivariate p interaction = 0.2 (PFS), 0.06 (OS)) (Table). Conclusions: Despite national guidelines against this practice, a small proportion of pts in routine care receive 2L IO-based therapies after disease progression on 1L IO or chemoIO. We found similar outcomes between IO and non-IO based 2L therapies after progression on 1L IO or chemoIO. [Table: see text]