Abstract Background: Immune checkpoint inhibitors (ICIs) targeting PD-(L)1 following concurrent chemoradiotherapy (cCRT) have shown clinical activity in locally advanced, unresectable NSCLC, e.g. with the PACIFIC regimen 1. PD-L1 expression is typically assessed by pathologist scoring of immunohistochemically (IHC) stained tissue using the tumor cell (TC) score. In this work, we present PD-L1 Quantitative Continuous Scoring (PD-L1 QCS) to digitally assess PD-L1 expression (Ventana SP263 assay), and analyze its utility for identification of patient subgroups with higher likelihood of treatment response. Method: 368 digitized whole slide images (WSI) of fresh biopsy and archival samples were obtained from the PACIFIC trial (NCT02125461), including 126 patients who received placebo and 242 patients who received Durvalumab post-cCRT 1. WSI analysis included two deep learning models for epithelial TC segmentation followed by optical density (OD) quantification of membrane expression 2. Initially, PD-L1 QCS was calibrated to compute the TC score at 1% cutoff by determining an optimal OD cell positivity threshold. Next, other scoring approaches were investigated, which included various OD thresholds and biomarker cutoffs. Results: PD-L1 QCS-based TC scoring was able to maintain median overall survival (mOS) of 57.4 months in the biomarker positive (BM+) ICI cohort, while increasing prevalence to 191 out of 242 patients (78.9%) compared to 177 (73.1%) with manual TC score at 1% cutoff. Extending the scoring paradigm to the PD-L1+ TC density >= 542.5 cells/mm² lead to a more pronounced increase of prevalence to 200 patients (82.6%) in the ICI population, while mOS was not reached and additionally showed comparable hazard ratios (HR) when evaluating against the BM+ placebo cohort (HR=0.54 0.39, 0.76 resp. HR=0.52 0.37, 0.74). Results further demonstrated that PD-L1 QCS density scoring of the ICI treated population allowed for a clearer separation of patients with less treatment benefit (34.2m mOS, N=42, log rank p=0.0067) compared to manual scoring (47.4m mOS, N=65, log rank p=0.2). Conclusion: We investigated a computational pathology approach for selecting patients for treatment with the PACIFIC regimen and compared it against established manual pathology scoring. Results indicate that PD-L1 QCS TC proportion scoring can increase prevalence, while PD-L1 QCS TC density scoring has the potential to identify an even larger patient subgroup with increased survival benefit. References: 1. Antonia, Scott J., et al. "Durvalumab after chemoradiotherapy in stage III non-small-cell lung cancer.” New England Journal of Medicine 377.20 (2017): 1919-1929. 2. Lesniak, Jan, et al. " Quantitative computational assessment of PD-L1 enables robust patient selection for biomarker-informed anti-PD-L1 treatment of NSCLC patients." J. Immunother. Cancer, Vol. 10., 2022. Citation Format: Jan Martin Lesniak, Ina Bisha, Simon Christ, Lina Meinecke, Thomas Padel, Laura Sebastian Monasor, Pallavi Sontakke, Tze Heng Tan, Dasa Medrikova, Andrea Storti, Anatoliy Shumilov, Hadassah Sade, Günter Schmidt, Ross Stewart, Yashaswi Shrestha. Quantitative continuous scoring of PD-L1 for superior patient selection for anti-PD-L1 treatment of locally advanced or unresectable NSCLC abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7617.