Abstract
BACKGROUND: It remains unclear whether assessing programmed death-ligand 1 (PD-L1) expression by SP142 plus 22C3 adds value for predicting the response to immunotherapy in non-small cell lung cancer (NSCLC). METHODS: This retrospective multicenter study included patients with advanced NSCLC treated with immune-checkpoint inhibitors. We constructed tissue microarrays (TMAs) and performed immunohistochemical staining with 22C3 and SP142 assays. We denoted the PD-L1 tumor proportion score (TPS) obtained from clinical medical records based on 22C3 staining as "22C3 (C)" and that obtained with 22C3 staining using our TMA as "22C3 (TMA)". SP142 staining was evaluated in both tumor cells and immune cells. We assessed the concordance between each PD-L1 assessment method and analyzed the objective response rate (ORR), progression-free survival (PFS) and overall survival (OS) based on the PD-L1 expression level determined using the 22C3 and SP142 assays. RESULTS: In total, 288 patients were included. Among those with 22C3 (TMA) ≥50%, 60% of patients showed SP142 TC3 or IC3; among patients with 22C3 (C) <1%, 9% and 18% exhibited 22C3 (TMA) ≥1% and SP142 TC1/2/3 or IC1/2/3, respectively. Among patients with 22C3 (C) ≥50% treated with immune-checkpoint inhibitor monotherapy, the SP142 TC1/2/3 or IC1/2/3 group showed significantly better ORR, PFS and OS than the SP142 TC0 and IC0 group (54% vs. 29%, P=0.040, median =11.0 vs. 3.2 months, P=0.002, median =27.9 vs. 12.6 months, P=0.030, respectively). Multivariate analysis revealed that SP142 TC0 and IC0 was an independent unfavorable prognostic factor for PFS and OS in patients with 22C3 (C) ≥50% treated with immune-checkpoint inhibitor monotherapy. For those with 22C3 (C) ≥50% and SP142 TC0 and IC0, immune-checkpoint inhibitor concurrent with chemotherapy tended to result in a longer PFS and OS than immune-checkpoint inhibitor monotherapy (median =13.7 vs. 2.3 months, P=0.054, median = not estimable vs. 12.0 months, P=0.064, respectively). CONCLUSIONS: SP142 evaluation contributes to the prediction of immune-checkpoint inhibitor efficacy in NSCLC with high PD-L1 expression assessed by 22C3.