PD-L1 expression profile and immunotherapy (IO) experience in African American (AA) and Hispanic (H) lung cancer (LC) patients: Addressing disparities at a minority-based academic cancer center.

Abstract

e18073 Background: Immune checkpoint inhibitors have changed the treatment paradigm for metastatic LC. Minority populations are under-represented in large IO clinical trials. Among 989 pts with newly diagnosed LC at Montefiore Medical Center - a community-based academic center from 2014-2015, 330 (33%) were AA and 195 (20%) were H. In line with practice-changing clinical studies, PD-L1 expression testing and IO have been incorporated into LC treatment. Methods: Pts receiving IO and/or had PD-L1 testing between 1/1/14-12/31/16 were identified from records obtained from pathology, pharmacy, oncology clinics and Clinical Looking Glass. Retrospective chart review was conducted. PD-L1 testing was performed using 22C3pharmDx IHC. Results: We identified 111 pts with LC who received IO and/or had PD-L1 testing, with a median age of 66. 55% were female. Based on race, 52 (47%) were AA, 24 (22%) were White, 26 (24%) were Other, and 9 (8%) were race unknown. Based on ethnicity, 30 (27%) were H, 73 (66%) were non-H and 8 (7%) were ethnicity unknown. 82% were former/current smokers. Adenocarcinoma was the dominant histology (60%). The majority were EGFR WT (91%) and ALK neg (98%). PD-L1 testing was performed in 67 (60%), including 32 (29%) AAs and 20 (18%) Hs. Archival tissue was used in 63%. PD-L1 TPS > 50% was found in 30%, 1-49% in 24%, < 1% in 37%. 62 pts received IO, including 26 (42%) AAs and 18 (29%) Hs. Nivolumab was the most commonly used agent (77%). In AAs, 8 (31%) received IO as 1st line, 13 (50%) as 2nd line, 5 (19%) as 3rd line and above. In Hs, 1 (5%) received IO as 1st line, 10 (56%) as 2nd line, 7 (39%) as 3rd line and above. Immune-related adverse events (IRAEs) were reported in 31% of AAs and 39% of Hs. Data analysis on survival is ongoing. Conclusions: Unlike the low numbers of minority pts in large clinical trials, we found no significant difference in PD-L1 testing and IO across racial and ethnic groups treated at our center. Compared to large clinical trials, we observed lower rates of IRAEs in our cohort. Our current and ongoing observations in these populations may have future implications in narrowing health disparities based on race/ethnicity.