Abstract 1235: Presence of TLS and combined high densities of PD-L1+ macrophages & CD8+ T cells predict long-term overall survival for patients with advanced NSCLC treated with durvalumab

Abstract

Introduction: Predictive biomarkers of anti‒PD-(L)1 therapies have largely focused on the tumor - T cell axis where tumor cell PD-L1 expression has demonstrated its clinical utility in predicting overall survival (OS) in patients with advanced non-small cell lung cancer (NSCLC). Although, other immune cell subsets were shown to be associated with clinical efficacy, their relative impact and combined effect in predicting improved long-term survival warrant further investigation. Using computational image analysis of multiplex immunofluorescence (mIF) and immunohistochemistry (IHC) immune marker panels, we sought to identify single and combined biomarkers of the tumor immune contexture in association with long-term OS in advanced NSCLC patients treated with Durvalumab. Methods: Pre-treatment tumor samples from advanced NSCLC patients (n = 210) enrolled in durvalumab nonrandomized phase 1/2 trial (10 mg/kg Q2W, CP1108/NCT01693562), were stained using IHC and 6-marker mIF panels to detect markers of immune cells, cell functional state and tertiary lymphoid structure (TLS) (PD-L1, CD8, PD-1, Ki67, CD68, CD20, CD1c, NKp46, CD66b, ICOS, FOXP3). Cell marker density (cells/mm2), distribution and proximity were quantified and analyzed in association with overall survival. Results: Computational image analysis of the tumor immune contexture revealed a greater immune inflamed phenotype, both innate (macrophages, dendritic cells) and adaptive (T and B cells), in NSCLC patients with long OS >2 years compared to those with short OS < 1 year (fold change > 2, p < 0.0001). Patient subgroup with high density of individual immune subsets show a median OS (mOS) of 10-20 months (p < 0.01 high vs. low subgroups) while combined markers of innate and adaptive immune cells show an improved mOS > 2 years (p < 0.001). Specifically, among the key findings, combined biomarkers of CD68+ PD-L1+ macrophages and CD8+ T cells predicts for a significant increase in OS for patients with high vs low marker density (HR = 0.21, 95%CI 0.12 - 0.39, p <10-7; mOS 39.5 months [high], 6.5 months [low]). Whereas, in patients with high density of either single biomarkers CD68+ PD-L1+ macrophage or CD8+ T cells mOS is 20.2 or 18.4 months respectively. Moreover, high density of CD20+ B cells, reflective of presence of TLS, associates with improved OS (mOS NR [high], 11 months [low], p = 0.003). In addition, TLS enriched tumors show an increased level of macrophages expressing PD-L1 in synapsis with CD8+ T cells (activated PD1+ or proliferative Ki67+ T cells). Conclusion: These findings demonstrate the importance of both tertiary lymphoid structure and high pre-existing innate-adaptive immunity in driving long-term overall survival of durvalumab-treated patients with NSCLC and highlight the need for the development of multiparametric predictive biomarkers beyond tumor-T cell axis. Citation Format: Lina Meinecke, Jorge Blando, Thomas Herz, Michael Surace, Thomas Padel, Monica Azqueta Gavaldon, Harald Hessel, Farzad Sekhavati, Megha Saraiya, Anmarie Boutrin, Karma Da Costa, Jaime Rodriguez Canales, Ashok Gupta, Carl Barrett, Zachary Aaron Cooper, Ikbel Achour. Presence of TLS and combined high densities of PD-L1+ macrophages & CD8+ T cells predict long-term overall survival for patients with advanced NSCLC treated with durvalumab [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):Abstract nr 1235.

579 Novel digital image approach of multiplex immunofluorescence based PD-L1 expression enables the stratification of advanced NSCLC patients treated with durvalumab

“…Methods Pre-treatment tumor samples from advanced NSCLC patients enrolled in durvalumab nonrandomized phase 1/2 trial (CP1108/NCT01693562) 2 , were stained by mIF panel containing PD-L1. 6 Similarly to IHC PD-L1 QCS, mIF PD-L1 QCS consists of two deep-learning models, first to segment epithelium regions and second to detect membrane, cytoplasm and nuclei of each epithelium cell, transferring for the second model annotations from IHC to mIF domain. 7 The mIF images are normalized based on batch statistics prior to image analysis.…”