Introduction: The Know Your Tumor (KYT) pancreas cancer program enables molecular analysis of both tumor DNA and mRNA to characterize tumor phenotypes and guide potential treatment options. The primary objective of the current study was to develop and apply criteria to establish treatment response from real-world clinical data and to characterize the matching molecular and immunologic features of the pancreatic ductal adenocarcinoma (PDAC) tumor biopsies to identify tumor phenotypes that associate with treatment response. Experimental Procedures: Formalin-fixed, paraffin-embedded tumor samples and related clinical data were collected from n=240 patients diagnosed with PDAC and treated at medical centers in the United States. Samples underwent retrospective mutation analysis using the Tempus xT targeted panel of 648 genes and whole transcriptome RNA-sequencing also from Tempus. To identify molecular profiles associated with tumor treatment response, the duration of first therapy and CA19-9 levels taken after the start of first therapy were analyzed to establish consensus responder and non-responder criteria. Mutation associations using Fisher exact tests and differential gene expression and immune signature analysis using Wilcoxon tests were performed on all FOLFIRINOX-treated subjects that meet consensus responder (n = 12) or non-responder (n = 10) definitions to identify molecular underpinnings of FOLFIRINOX resistance. Gene expression analysis initially focused on tumor intrinsic genes (TIGs) because the expression of these genes may specifically reflect PDAC tumor biology. Differentially expressed TIGs with an unadjusted p-value < 0.05 were submitted for gene set enrichment analysis (GSEA) and manual investigation. Results: No association between DNA mutations and response to FOLFIRINOX was observed in this small dataset. In contrast, results from RNA-based gene expression analysis suggested that differences in tumor biology may contribute to response. GSEA and manual interrogation of differentially expressed TIGs revealed non-responder tumors may have high expression of apoptosis-related genes and genes associated with energetics. GSEA of all differentially expressed genes (unadjusted p-value < 0.05) refined this interpretation and corroborated initial findings by suggesting non-responders may have high expression of hypoxia, glycolysis- and apoptosis-related genes and low expression of SMAD4 and its associated target genes. Summary and Conclusions: These results support the testable hypothesis that FOLFIRINOX non-responders evade apoptosis through hypoxia adaptations and autophagic flux. These hypotheses may warrant further investigation with preclinical models such as cell culture, PDX or organoid models. Finally, the analysis plan provides a roadmap for using the KYT cohort and other real-world datasets to generate hypotheses about molecular mechanisms of treatment response.

Citation Format: James M. Davison, Greg Mayhew, Kirk Beebe, Joel R. Eisner, Dennis Ladnier, Eric A. Collisson, Lynn M. Matrisian. Initial retrospective analysis of mechanisms of FOLFIRINOX resistance using clinical and molecular data from the Know Your Tumor (KYT) pancreatic ductal adenocarcinoma (PDAC) cohort [abstract]. In: Proceedings of the AACR Virtual Special Conference on Pancreatic Cancer; 2021 Sep 29-30. Philadelphia (PA): AACR; Cancer Res 2021;81(22 Suppl):Abstract nr PO-002.