Clinicopathologic correlates of first-line pembrolizumab effectiveness in patients with advanced NSCLC and a PD-L1 expression of ≥ 50%
- 1 November 2020
- journal article
- research article
- Published by Springer Science and Business Media LLC in Cancer Immunology, Immunotherapy
- Vol. 69 (11), 2209-2221
- https://doi.org/10.1007/s00262-020-02613-9
Abstract
Background Single-agent pembrolizumab represents the standard first-line option for metastatic non-small-cell lung cancer (NSCLC) patients with a PD-L1 (programmed death-ligand 1) expression of >= 50%. Methods We conducted a multicenter retrospective study aimed at evaluating the clinicopathologic correlates of pembrolizumab effectiveness in patients with treatment-naive NSCLC and a PD-L1 expression of >= 50%. Results One thousand and twenty-six consecutive patients were included. The objective response rate (ORR) was 44.5% (95% CI 40.2-49.1), while the median progression free survival (PFS) and overall survival (OS) were 7.9 months (95% CI 6.9-9.5; 599 events) and 17.2 months (95% CI 15.3-22.3; 598 censored patients), respectively. ECOG-PS >= 2 (p < 0.0001) and bone metastases (p = 0.0003) were confirmed to be independent predictors of a worse ORR. Former smokers (p = 0.0002), but not current smokers (p = 0.0532) were confirmed to have a significantly prolonged PFS compared to never smokers at multivariate analysis. ECOG-PS (p < 0.0001), bone metastases (p < 0.0001) and liver metastases (p < 0.0001) were also confirmed to be independent predictors of a worse PFS. Previous palliative RT was significantly related to a shortened OS (p = 0.0104), while previous non-palliative RT was significantly related to a prolonged OS (p = 0.0033). Former smokers (p = 0.0131), but not current smokers (p = 0.3433) were confirmed to have a significantly prolonged OS compared to never smokers. ECOG-PS (p < 0.0001), bone metastases (p < 0.0001) and liver metastases (p < 0.0001) were also confirmed to be independent predictors of a shortened OS. A PD-L1 expression of >= 90%, as assessed by recursive partitioning, was associated with significantly higher ORR (p = 0.0204), and longer and OS (p = 0.0346) at multivariable analysis. Conclusion Pembrolizumab was effective in a large cohort of NSCLC patients treated outside of clinical trials. Questions regarding the effectiveness in clinical subgroups, such as patients with poorer PS and with liver/bone metastases, still remain to be addressed. We confirmed that the absence of tobacco exposure, and the presence of bone and liver metastasis are associated with worse clinical outcomes to pembrolizumab. Increasing levels of PD-L1 expression may help identifying a subset of patients who derive a greater benefit from pembrolizumab monotherapy.This publication has 36 references indexed in Scilit:
- Pembrolizumab versus Chemotherapy for PD-L1–Positive Non–Small-Cell Lung CancerThe New England Journal of Medicine, 2016
- Pembrolizumab for the Treatment of Non–Small-Cell Lung CancerThe New England Journal of Medicine, 2015
- Treatment of Elderly Patients With Non–Small-Cell Lung Cancer: Results of an International Expert Panel Meeting of the Italian Association of Thoracic OncologyClinical Lung Cancer, 2015
- Bi-Force: large-scale bicluster editing and its application to gene expression data biclusteringNucleic Acids Research, 2014
- New response evaluation criteria in solid tumours: Revised RECIST guideline (version 1.1)European Journal of Cancer, 2009
- Power, selection bias and predictive performance of the Population Pharmacokinetic Covariate Model.Journal of Pharmacokinetics and Pharmacodynamics, 2004
- The Effect of Collinearity on Parameter Estimates in Nonlinear Mixed Effect ModelsPharmaceutical Research, 1999
- The meaning and use of the area under a receiver operating characteristic (ROC) curve.Radiology, 1982
- Asymptotically Efficient Rank Invariant Test ProceduresJournal of the Royal Statistical Society. Series A (General), 1972
- Nonparametric Estimation from Incomplete ObservationsJournal of the American Statistical Association, 1958