Abstract 1242: PD-L1 and PD-L2 mRNA are associated with outcome and high negative predictive value in immunotherapy-treated non-small cell lung cancer
Aileen I. Fernandez,Niki Gavrielatou,Leena McCann,Saba Shafi,Myrto K. Moutafi,Sandra Martinez-Morilla,Ioannis Vathiotis,Thazin Nwe Aung,Vesal Yaghoobi,Yalai Bai,Jodi Weidler,Michael Bates,David L. Rimm
DOI: https://doi.org/10.1158/1538-7445.am2022-1242
IF: 11.2
2022-06-16
Cancer Research
Abstract:Background: Immune checkpoint inhibitors (ICIs) are a class of immunotherapy that enhance a patient's anti-cancer immune response, but only a small percentage of patients who are treated respond. This means patients who do not respond undergo costly and side effect-inducing treatment for no benefit indicating a need for improved selection criteria. Current predictive biomarkers include immunohistochemical (IHC) detection of PD-L1 but are insufficient for determining who will respond or, more importantly in the adjuvant setting, who will not respond. Methods: This study utilizes a research use only (RUO*) quantitative real-time reverse transcription polymerase chain reaction assay, the GeneXpert® (GX) PD-L1 panel prototype assay, to test for the association between 4 target immune genes, CD274 (PD-L1), PDCD1LG2 (PD-L2), CD8A, and IRF1, and response to ICI therapy. Tissues were collected from 122 patients with advanced non-small cell lung cancer prior to ICI therapy in a retrospective cohort, macro-dissected, and analyzed using the PD-L1 prototype assay. Lysates were run on the GX instrument using the PD-L1 prototype assay. Individual transcripts were quantitated for each sample and the association with response was assessed. Median mRNA expression was used as a cutpoint to look at survival, clinical response and positive and negative predictive value (PPV, NPV). Optimal cutpoint was determined using Rstudio to also assess PPV and NPV. This study was approved by Yale Human Investigation IRB protocol ID 9505008219. Results: Both high PD-L1 and PD-L2 mRNA, defined by median, were associated with improved long-term benefit at 24-months (PD-L1, p=0.0416; PD-L2, p=0.0435) and overall survival (PD-L1, p=0.047; PD-L2, p=0.047). Furthermore, low PD-L1 mRNA, defined by optimal cutpoint, showed a negative predictive value of 0.92. Conclusions: High PD-L1 and PD-L2 mRNA, measured by GX, are associated with improved long-term benefit and overall survival. Importantly, low PD-L1 mRNA has a high negative predictive value. Given the simplicity and reproducibility of the GX system, with further validation this assay could be an improved method for selecting patients for treatment in the advanced setting, or more importantly, to determine which low stage patients should not be treated in the adjuvant setting. *For research use only. Not for use in diagnostic procedures. Citation Format: Aileen I. Fernandez, Niki Gavrielatou, Leena McCann, Saba Shafi, Myrto K. Moutafi, Sandra Martinez-Morilla, Ioannis Vathiotis, Thazin Nwe Aung, Vesal Yaghoobi, Yalai Bai, Jodi Weidler, Michael Bates, David L. Rimm. PD-L1 and PD-L2 mRNA are associated with outcome and high negative predictive value in immunotherapy-treated non-small cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl) nr 1242.
oncology