A protein-based machine learning approach to the identification of inflammatory subtypes in pancreatic ductal adenocarcinoma.

BACKGROUND/OBJECTIVES: The inherently immunosuppressive tumor microenvironment along with the heterogeneity of pancreatic ductal adenocarcinoma (PDAC) limits the effectiveness of available treatment options and contributes to the disease lethality. Using a machine learning algorithm, we hypothesized that PDAC may be categorized based on its microenvironment inflammatory milieu. METHODS: Fifty-nine tumor samples from patients naïve to treatment were homogenized and probed for 41 unique inflammatory proteins using a multiplex assay. Subtype clustering was determined using t-distributed stochastic neighbor embedding (t-SNE) machine learning analysis of cytokine/chemokine levels. Statistics were performed using Wilcoxon rank sum test and Kaplan-Meier survival analysis. RESULTS: t-SNE analysis of tumor cytokines/chemokines revealed two distinct clusters, immunomodulating and immunostimulating. In pancreatic head tumors, patients in the immunostimulating group (N = 26) were more likely to be diabetic (p = 0.027), but experienced less intraoperative blood loss (p = 0.0008). Though there were no significant differences in survival (p = 0.161), the immunostimulating group trended toward longer median survival by 9.205 months (11.28 vs. 20.48 months). CONCLUSION: A machine learning algorithm identified two distinct subtypes within the PDAC inflammatory milieu, which may influence diabetes status as well as intraoperative blood loss. Opportunity exists to further explore how these inflammatory subtypes may influence treatment response, potentially elucidating targetable mechanisms of PDAC's immunosuppressive tumor microenvironment.

Infiltration of CD3+ and CD8+ lymphocytes in association with inflammation and survival in pancreatic cancer.

BACKGROUND: Pancreatic ductal adenocarcinomas (PDAC) have heterogeneous tumor microenvironments relatively devoid of infiltrating immune cells. We aimed to quantitatively assess infiltrating CD3+ and CD8+ lymphocytes in a treatment-naïve patient cohort and assess associations with overall survival and microenvironment inflammatory proteins. METHODS: Tissue microarrays were immunohistochemically stained for CD3+ and CD8+ lymphocytes and quantitatively assessed using QuPath. Levels of inflammation-associated proteins were quantified by multiplexed, enzyme-linked immunosorbent assay panels on matching tumor and tissue samples. RESULTS: Our findings revealed a significant increase in both CD3+ and CD8+ lymphocytes populations in PDAC compared with non-PDAC tissue, except when comparing CD8+ percentages in PDAC versus intraductal papillary mucinous neoplasms (IPMN) (p = 0.5012). Patients with quantitatively assessed CD3+ low tumors (lower 50%) had shorter survival (median 273 days) compared to CD3+ high tumors (upper 50%) with a median overall survival of 642.5 days (p = 0.2184). Patients with quantitatively assessed CD8+ low tumors had significantly shorter survival (median 240 days) compared to CD8+ high tumors with a median overall survival of 1059 days (p = 0.0003). Of 41 proteins assessed in the inflammation assay, higher levels of IL-1B and IL-2 were significantly associated with decreased CD3+ infiltration (r = -0.3704, p = 0.0187, and r = -0.4275, p = 0.0074, respectively). Higher levels of IL-1B were also significantly associated with decreased CD8+ infiltration (r = -0.4299, p = 0.0045), but not IL-2 (r = -0.0078, p = 0.9616). Principal component analysis of the inflammatory analytes showed diverse inflammatory responses in PDAC. CONCLUSION: In this work, we found a marked heterogeneity in infiltrating CD3+ and CD8+ lymphocytes and individual inflammatory responses in PDAC. Future mechanistic studies should explore personalized therapeutic strategies to target the immune and inflammatory components of the tumor microenvironment.