(72) Trends and Disparities in Real-World Biomarker Testing and Overall Survival (OS) Among US Patients with Advanced Non-Small Cell Lung Cancer (aNSCLC), 2015–2020

Sunday, August 28, 2022

8:00 AM – 1:45 PM CEST

Location: Hall C

Publication Number: 1117

Presenting Author(s)

Wenzhen Ge, PhD
– Regeneron Pharmaceuticals, Greenwich, Connecticut, United States

Co-Author(s)

Scott Antonia
– Center for Cancer Immunotherapy, Duke University School of Medicine, Durham, NC, USA, United States
Ning Wu
– Regeneron Pharmaceuticals, Inc, Tarrytown, NY, USA, United States
Ruben GW Quek
– Regeneron Pharmaceuticals, Inc, Tarrytown, NY, USA, United States
Hilary Dietz
– Center for Cancer Immunotherapy, Duke University School of Medicine, Durham, NC, USA, United States
Jean-Francois Pouliot
– Regeneron Pharmaceuticals, Inc, Tarrytown, NY, USA, United States
Jessica Jalbert, PhD
– Regeneron Pharmaceuticals, Sleepy Hollow, New York, United States
James Harnett
– Regeneron Pharmaceuticals, Inc, Tarrytown, NY, USA, United States

Background: The National Comprehensive Cancer Network and the European Society for Medical Oncology recommend biomarker testing to guide optimal aNSCLC treatment.

Objectives: To describe the latest trends and disparities in real-world biomarker testing, focusing on programmed cell death-ligand 1 (PD-L1), and OS by testing status in first-line (1L) aNSCLC.

Methods: In patients with aNSCLC initiating 1L treatment (index date) between 1 January 2015 and 30 November 2020 in the Flatiron Health electronic health records database, disparities and trends in pre-1L testing of PD-L1 and major actionable oncogenic driver alterations (ALK, EGFR, ROS1, BRAF, KRAS) were assessed. OS was stratified by biomarker testing status and index year; OS benefits of biomarker testing were estimated using a multivariate Cox proportional hazard model.

Results: Among included patients (N=27,398; median age 69 years [Q1–Q3: 62–76]; 47.3% female), proportion of patients with PD-L1 testing increased from 11.3% in 2015 to 74.9% in 2017 and then ranged from 79.3% to 80.1% from 2018 through 2020. Across study period, 22C3, SP142 and SP263 assays were used in 74%, 3%, and 3% of PD-L1 tests, respectively. Pre-1L PD-L1 testing was less likely among older patients, males, non-Caucasians, those treated in community settings, and those with squamous histology or de novo advanced disease. Oncogene driver testing and OS increased over the study period; proportion of patients undergoing oncogene driver (PD-L1, ALK, EGFR, ROS1, KRAS, BRAF) testing (any test/all six tests) increased from 68.3%/4.3% in 2015 to 88.6%/53.8% in 2020. From 2015 through 2019, median OS from 1L (95% confidence interval [CI]) improved from 11.1 (10.6–11.7) to 14.1 (13.4–15.2) months and was 12.4 (11.6–13.0) in 2020 (with median follow-up of < 5 months). Compared to those without biomarker testing, the adjusted hazard ratio (95% CIs) for death was 0.89 (0.83–0.95) for patients tested for PD-L1 only and 0.76 (0.73–0.80) for patients tested for both PD-L1 and ≥1 oncogene.

Conclusions: During 2015–2020, a swift uptake in biomarker testing, especially PD-L1, was observed for 1L-treated patients with aNSCLC. PD-L1 testing gaps remained and significant disparities in PD-L1 testing were identified. PD-L1 testing may be associated with improved OS potentially by enhanced treatment decision making. Future research is warranted to understand underlying reasons for biomarker non-testing and associated disparities.