Tumor-Infiltrating Immune Cell Signature Predicts the Prognosis and Chemosensitivity of Patients With Pancreatic Ductal Adenocarcinoma

Objective: Tumor-infiltrating immune cells might add predictive value for prognostic stratification of patients with pancreatic ductal adenocarcinoma (PDAC) and chemotherapy response. We aimed to develop a prognostic model based on immune cell infiltration signatures to improve the prediction of survival and chemotherapy benefits of patients with PDAC. Methods: The abundance of tumor-infiltrating immune cells for 661 patients with PDAC from 4 different cohorts with survival data was collected in the training cohorts. Cox regression analysis and meta-analysis of immune cells were conducted to generate the tumor immune cell score (TICS) for prognostic stratification. 2 other independent cohorts including 188 patients were then used to validate the model. Those patients who underwent chemotherapy were used to further analyze the value of TICS for predicting the chemotherapy response. Furthermore, the difference in the somatic mutations and immune-related molecules between the TICS subgroups was analyzed. Results: Totally, 6 out of 28 immune cells were found to be significantly associated with PDAC prognosis in training cohorts (all P<0.05). The developed TICS could significantly predict PDAC survival and chemotherapy benefit both in the training and external validation cohorts (log-rank test, P<0.05). Significant differences were found in different TICS subgroups in terms of immune characteristics, checkpoint genes, and tumor mutational burden. Functional and pathway analyses further proved that TICS was significantly related to tumor immunity response in patients with PDAC. Conclusion: TICS could be used to predict PDAC patients with better survival and greater chemotherapy benefit. Keywords: pancreatic ductal adenocarcinoma; immune cell; prognosis; chemotherapy response