Impact of baseline number of metastases on the efficacy of PD-1 inhibitor monotherapy for non-small cell lung cancer with high PD-L1 expression.

Abstract

e20551 Background: Pembrolizumab monotherapy and pembrolizumab combined with platinum-based chemotherapy are the standard therapies for advanced non-small cell lung cancer (NSCLC) patients with a programmed death ligand 1 (PD-L1) tumour proportion score (TPS) ≥ 50%. Apart from PD-L1 status, predictive factors for the efficacy of pembrolizumab treatment remain unclear. High tumour burden has been shown to be a negative predictor for the efficacy of PD-1 inhibitors in melanoma. Hence, the purpose of this study was to evaluate the association between tumour burden and efficacy of pembrolizumab in NSCLC patients with PD-L1 TPS ≥ 50%. Methods: This single-center retrospective observational study evaluated patients with advanced NSCLC with PD-L1 TPS ≥ 50% who received pembrolizumab monotherapy as first line therapy between March 2016 and November 2018. The primary endpoint was progression-free survival (PFS). Tumour burden was estimated by the baseline number of metastases (BNM) and the baseline sum of the target lesions’ longest diameters (BSLD; measured according to RECIST criteria). Results: Of 52 patients with target lesions, the median age was 69 years; ECOG-PS 0-1 94%; non-squamous histology 79%; and PD-L1 TPS ≥ 75% was 65%. The optimal cut-off values for predicting PFS were 4 for BNM and 115 mm for BSLD, based on the minimum P-value method for PFS. The low BNM group included 20 (38%) patients, and the low BSLD group 33 (63%). Low BNM was significantly associated with longer PFS (median PFS not reached vs 4.0 months, HR = 0.35, P = 0.021) and with higher response rate (60% vs 31%, P = 0.041) compared to high BNM. There was no significant difference in PFS between low BSLD and high BSLD groups (12.4 vs 4.0 months, HR = 0.52, P = 0.098). Multivariate analysis showed that both low BNM (≤ 4 vs. > 4, HR = 0.32, P = 0.02) and PD-L1 TPS (75-100% vs. 50-74%, HR = 0.41, P = 0.032) were significantly associated with longer PFS. There was no association between PD-L1 TPS and BNM (Spearman’s rho = -0.0985, P = 0.487) Conclusions: BNM may be an independent predictive factor of efficacy of pembrolizumab monotherapy. Patients with low BNM can be candidates for pembrolizumab monotherapy.