M. K. Farrugia1, M. Chen1, E. Repasky1, K. Attwood1, K. M. Catalfamo1, S. Yao1, D. Mattson1, A. K. Witkiewicz1, and A. K. Singh2; 1Roswell Park Comprehensive Cancer Center, Buffalo, NY, 2Department of Radiation Medicine, Roswell Park Comprehensive Cancer Center, Buffalo, NY
Purpose/Objective(s): Esophageal cancer (ESC) is an aggressive disease which often presents at an advanced stage. Despite trimodal therapy, 40-50% patients can develop metastatic disease by 18 months. Identification of patients at risk for metastatic spread is challenging and there is a need for improved prognostication. CheckMate 648 and 577 established the efficacy of immunotherapy in the treatment of ESC and immune evasion may play a role in distal failure. Therefore, we investigated whether the immune landscape of pre-treatment tissue was associated with relapse in ESC patients. Materials/
Methods: Between April 2010 and October 2018, we identified 25 patients who had undergone trimodal therapy for ESC and had pre-treatment biopsies suitable for analyses. To assess the immune landscape, we performed high-throughput multiplex immunofluoresence (mIF) staining on formalin-fixed paraffin-embedded biopsy samples. Analysis was performed using Vectra Polaris Quantitative Pathology Imaging System. Analysis of unique populations via immune and exhaustion mIF panels was performed and expression was normalized to total cell counts. Differential expression by relapse was assessed by the Mann Whitney U test. Results: Of the 25 patients analyzed, the median age was 68 years, 23 (92%) were male, 21 (84%) had adenocarcinoma, and most patients had T3 (92%) node-positive (80%) disease. Five patients (20%) had a pathological complete response (pCR) to neoadjuvant chemoradiation. The median follow-up time was 23.9 months, during which 12 (48%) patients suffered a relapse with a median time to progression of 13.1 months. Multiplex mIF staining revealed higher expression of HLA-DR (p=0.019), CD8/LAG3 (p=0.046), CD8/CTLA4 (p=0.027), and CD163+/PDL1+ (p=0.045) among patients without relapse. None of the evaluated immune cell populations were differentially expressed by pCR. To further characterize this observation, time to progression (TTP) and disease-specific survival (DSS) were stratified by median expression of each significant subpopulation and formally tested by the log-rank test. Higher than median expression of HLA-DR (p=0.027), CD8/LAG3 (p=0.039), CD8/CTLA (p=0.039), but not CD163+/PDL1+ (p=0.12) were significantly associated with TTP. Similarly, HLA-DR (p=0.0069) and CD8/CTLA4 (p=0.036) were significantly associated with improved DSS whereas no significant observations were found with CD8/LAG3 (p=0.11) or CD163+/PDL1+ (p=0.2) expression. In Cox proportional regression models, the hazard ratios (HR) associated with high vs. low HLA-DR and CD8/CTLA4 expression were HR=0.16 (p=0.018) and HR=0.27 (p=0.051), respectively. Conclusion: High expression of HLA-DR and CD8/CTLA4 within pre-treatment biopsy ESC samples was associated with significantly reduced rates of relapse. Increased presence of these markers suggests that an improved immune landscape is associated with less aggressive disease and may provide an opportunity for risk-based treatment strategies.
