Artificial intelligence (AI) -powered H&E whole-slide image (WSI) analysis of tertiary lymphoid structure (TLS) to predict response to immunotherapy in non-small cell lung cancer (NSCLC).

3135 Background: Recent data indicate that immune cells present in the TLS within the tumor microenvironment (TME) could potentially serve as a favorable predictor for immune checkpoint inhibitor (ICI) therapy. To examine this further, AI model was created to provide an assessment of TLS in H&E WSI, and to determine its correlation with survival outcomes of NSCLC patients treated with ICI. Methods: We collected H&E-stained WSI of primary and metastatic tumors and clinical data in a retrospective study of advanced-stage NSCLC patients treated with ICI (with or without chemotherapy) at Northwestern University from May 2015 to November 2022. When evaluating the TLS status in the biopsy specimen collected from the lymph nodes (LNs), TLSs were only considered if they were admixed with tumor cells and located distant from the residual parenchyma, ensuring the exclusion of innate lymphoid follicles. Following the assessment of the presence of TLS by a pathologist, the identical WSIs were evaluated by the AI-powered H&E WSI analyzer. We used Lunit SCOPE IO, an AI-powered H&E WSI analyzer developed using 21,096 H&E stained WSIs, which can segment various classes of tissue area including TLS within the TME. Survival analysis was performed using Kaplan-Meier curves, and differences in survival outcomes among TLS groups were evaluated using the log-rank test. Results: In this analysis, out of a total of 73 patients, 24 (33%) cases were identified as containing TLS (TLS +) by the AI analyzer. In contrast, the pathologist classified 17 (23%) cases as TLS+. The AI model demonstrated an accuracy of 84.9% (with a sensitivity of 88.2% and specificity of 83.9%) in identifying the presence of TLS compared to the pathologist's interpretation. The median value of the proportion of TLS area within TME (TLSP) in samples containing TLS was 0.418% (Q1: 0.228%, Q3: 1.311%). TLSP was not found to be correlated with survival outcomes. The number of clusters of TLS was not found to be correlated with survival outcomes. Conclusions: AI-powered TLS analysis can be potentially utilized as a predictive biomarker of survival outcomes of NSCLC patients treated with ICI. Table: see text