Transcriptomic analyses of molecular subsets and correlations with clinical outcomes from the phase 3 IMforte study of lurbinectedin (lurbi) + atezolizumab (atezo) maintenance treatment (Tx) in extensive-stage small-cell lung cancer (ES-SCLC).

8014 Background: Transcriptomic analyses of pre-Tx tumor samples from the Phase 3 IMpower133 study identified 4 molecular subtypes with distinct clinical outcomes to first-line (1L) Tx with atezo and chemotherapy (chemo). Neuroendocrine (NE) tumors with low tumor-associated macrophage (TAM)/high T-effector (T-eff) signal demonstrated longer overall survival (OS) vs non-NE tumors with high TAM/high T-eff, suggesting that TAM contributes to resistance to atezo. Lurbi, an alkylating agent that modifies the tumor microenvironment and synergizes with immune checkpoint inhibitors, could enhance atezo activity. Lurbi + atezo as 1L maintenance Tx significantly improved progression-free survival (PFS) and OS vs atezo in patients (pts) with ES-SCLC in the Phase 3 IMforte study (NCT05091567). We report exploratory biomarker analyses from IMforte. Methods: Adults with Tx-naïve ES-SCLC, without disease progression after induction with atezo + chemo, received maintenance Tx of either 3.2 mg/m² lurbi + 1200 mg atezo or 1200 mg atezo q3w until unacceptable toxicity or disease progression. Transcriptomic analyses were conducted on pre-induction Tx tumor samples to identify subgroups with concordance to previously reported SCLC subtypes (SCLC-A, -N, -I-NE and -I-non-NE), immune gene expression signatures (TAM, T-eff) and SLFN11 . Post-hoc exploratory analyses were conducted for correlation with IRF-assessed PFS and OS. Results: Of 483 randomized pts, 303 had samples for RNA sequencing. Of 303 tissue samples analyzed, 104 (34.3%) were classified as SCLC-A, 89 (29.4%) as SCLC-N, 46 (15.2%) as SCLC-I-NE and 64 (21.1%) as SCLC-I-non-NE. Clinical outcomes in the 303 pts (median [m] PFS: 5.5 vs 2.4 months [mo], hazard ratio [HR] 0.61; mOS: 13.5 vs 11.7 mo, HR 0.72, for lurbi + atezo [n = 150] vs atezo [n = 153], respectively) were consistent with the full analysis set. PFS and OS benefit from lurbi + atezo were comparable between NE (mPFS: 5.5 vs 2.1 mo, HR 0.58; mOS: 13.5 vs 11.7 mo, HR 0.74, for lurbi + atezo [n = 115] vs atezo [n = 124], respectively) and non-NE subtypes (mPFS: 4.6 vs 2.8 mo, HR 0.73; mOS: not reached vs 12.0 mo, HR 0.68, for lurbi + atezo [n = 35] vs atezo [n = 29], respectively). Pts with high TAM/high T-eff in the atezo arm had shorter OS vs those with low TAM/high T-eff. The addition of lurbi improved OS in the high TAM/high T-eff subgroup vs atezo alone (HR 0.56). SLFN11 expression was high at baseline across subtypes and had no predictive value for lurbi + atezo clinical benefit. Conclusions: The prevalence of the 4 SCLC subtypes and the TAM/T-eff pre-Tx data for IMforte are consistent with findings from IMpower133. PFS and OS were longer for lurbi + atezo vs atezo irrespective of molecular subset, though numerical trends suggest that lurbi may overcome TAM-mediated resistance to atezo. Clinical trial information: NCT05091567 .