DEFINING THE KINOME LANDSCAPE IN PANCREATIC CANCER FOR PRECISION MEDICINE

Effective therapies for pancreatic ductal adenocarcinoma (PDAC) have been largely elusive. Prior proteogenomic studies of PDAC have been limited to bulk samples limiting the ability to define tumor-intrinsic targets and understanding tumor heterogeneity. Here, we perform Multiplexed kinase Inhibitor Beads and Mass Spectrometry (MIB-MS) on 216 patient derived xenografts (PDX) derived from 14 unique primary PDAC to define the tumor-intrinsic kinome landscape. For our baseline kinome studies, our findings uncover three kinome subgroups making up two dominant tumor-intrinsic kinome subtypes that we call kinotypes. The kinotypes show enrichment of different kinase classes and explain the biological differences in previously described molecular subtypes, basal-like and classical. The kinotype characterizing basal-like tumors shows enrichment of receptor tyrosine kinases, whereas the kinotype characterizing classical tumors is enriched in understudied kinases involved in Wnt signaling and immune pathways. We validate our findings in two clinical trials and show that only patients with basal-like kinotype tumors, who are thought to be refractory to chemotherapy, derive significant benefit from EGFR inhibitors. We also define the adaptive kinome landscape in our PDX and uncover MEK2 inhibitor induced classical subtype specific reprogramming. Increased MAP2K6 expression and activation of p38 MAPK signaling is suppressed by dual MEK2 and p38 MAPK inhibition leading to significant response specifically in classical subtype tumors. Our results provide a comprehensive tumor-intrinsic kinome landscape of PDAC that strongly supports actionable kinotype specific baseline and adaptive kinase targets, potential for sensitive quantitation of patient samples, and provides a roadmap for the use of kinase inhibitors for the treatment of PDAC.