Comparison of real-world data (RWD) analysis on efficacy and post-progression outcomes with pembrolizumab plus chemo vs chemo alone in metastatic non-squamous non-small cell lung cancer with PD-L1 < 50.

Background: Following the introduction of immunotherapy (IO) in the first-line (1L) treatment in patients with non-small cell lung cancer (NSCLC) without sensitizing EGFR/ALK mutations, increasing real-world data depict how difficult it is to replicate data from clinical trials to clinical practice, with high rates of early treatment failure. In the context of chemo-IO, our study aims to compare platinum-pemetrexed-pembrolizumab combination to platinum-doublet alone in patients with low PD-L1 (<50%). Methods: We retrospectively collected medical records from patients with stage IV non-squamous NSCLC with PD-L1<50%, consecutively treated at our Centre from 2016 to 2021. Patients were grouped according to 1L treatment received: chemo-IO (group A) or platinum-doublet (group B). Survival outcomes were analyzed and compared among the two groups. Results: Overall, 105 patients were included: 49 in group A and 56 in group B. At data cut-off, median follow-up was 12.4 and 34.8 months, with 32/49 and 52/56 events for progression-free survival (PFS) and 21/49 and 29/56 events for overall survival (OS), respectively. No difference in PFS was observed between group B and group A (6.6 versus 8 months, HR 1.12, 95%CI 0.57-1.40). Patients receiving 1L platinum-doublet had significantly longer OS compared to those receiving chemo-IO (median OS 23.8 vs 14.9 months, HR 0.47, 95% CI 1.15- 3.98, p=0.01). 12 month-OS was 58% (95% CI 44-76%) in group A and 78% (95% CI 68-91%) in group B (p=0.040). Subgroup analysis identified KRAS G12C mutation as potentially affecting PFS in patients receiving chemo-IO (HR 0.29, 95% CI 0-10-0.91). The OS benefit of platinum-doublet was consistent across subgroups, with particular benefit in female sex, liver or pleural metastases, PD-L1 negative. Overall, only 46.9% of patients with progression received subsequent treatment in group A (15/32), compared to 86.5% in group B (45/52, all receiving 2L IO), with no difference in PFS to 2L (group A 3.7months, group B 4.1months, p=0.3). Conclusions: Despite small study population and differential follow-up, our study demonstrates that sequential use of 1L platinum-doublet and 2L IO is not inferior to 1L chemo-IO in non-squamous NSCLC with PD-L1<50%. In addition, we identified subgroups who might benefit differentially from the two approaches.