The poor correlation of mutational landscapes with phenotypes limits our understanding of pancreatic ductal adenocarcinoma (PDAC) pathogenesis and metastasis. Here we show a critical role of oncogenic dosage-variation in PDAC biology and phenotypic diversification. We found gene-dosage increase of mutant KRASMUT in human PDAC precursors, driving both early tumorigenesis and metastasis, thus rationalizing early PDAC dissemination. To overcome limitations posed to gene-dosage studies by PDAC´s stroma-richness we developed large cell culture resources of metastatic mouse PDAC. Integration of their genomes, transcriptomes and tumor phenotypes with functional studies and human data, revealed additional widespread effects of oncogenic dosage-variation on cell morphology/plasticity, histopathology and clinical outcome, with highest KrasMUT levels underlying aggressive undifferentiated phenotypes. We also identify alternative oncogenic gains (Myc, Yap1 or Nfkb2), which collaborate with heterozygous KrasMUT in driving tumorigenesis, yet with lower metastatic potential. Mechanistically, different oncogenic gains and dosages evolve along distinct evolutionary routes, licensed by defined allelic states and/or combinations of hallmark tumor-suppressor alterations (Cdkn2a, Trp53, Tgfβ-pathway). Thus, evolutionary constraints and contingencies direct oncogenic dosage gain and variation along defined routes to drive early progression and shape downstream PDAC biology. Our study uncovers universal principles in Ras-driven oncogenesis with potential relevance beyond pancreatic cancer.

2-1). The collection and processing of the specimens via PancoBank was supported by Heidelberger 48 Stiftung Chirurgie. I.R. is a recipient of an Odysseus I fellowship of the Fund for Scientific Research 49 Flanders (FWO). E.E. was a recipient of an EMBO long-term fellowship (ALTF 344-2013).

PurposeDevelopment of a supervised machine-learning model capable of predicting clinically relevant molecular subtypes of pancreatic ductal adenocarcinoma (PDAC) from diffusion-weighted-imaging-derived radiomic features.MethodsThe retrospective observational study assessed 55 surgical PDAC patients. Molecular subtypes were defined by immunohistochemical staining of KRT81. Tumors were manually segmented and 1606 radiomic features were extracted with PyRadiomics. A gradient-boosted-tree algorithm was trained on 70% of the patients (N = 28) and tested on 30% (N = 17) to predict KRT81+ vs. KRT81- tumor subtypes. A gradient-boosted survival regression model was fit to the disease-free and overall survival data. Chemotherapy response and survival were assessed stratified by subtype and radiomic signature. Radiomic feature importance was ranked.ResultsThe mean±STDEV sensitivity, specificity and ROC-AUC were 0.90±0.07, 0.92±0.11, and 0.93±0.07, respectively. The mean±STDEV concordance indices between the disease-free and overall survival predicted by the model based on the radiomic parameters and actual patient survival were 0.76±0.05 and 0.71±0.06, respectively. Patients with a KRT81+ subtype experienced significantly diminished median overall survival compared to KRT81- patients (7.0 vs. 22.6 months, HR 4.03, log-rank-test P = <0.001) and a significantly improved response to gemcitabine-based chemotherapy over FOLFIRINOX (10.14 vs. 3.8 months median overall survival, HR 2.33, P = 0.037) compared to KRT81- patients, who responded significantly better to FOLFIRINOX over gemcitabine-based treatment (30.8 vs. 13.4 months median overall survival, HR 2.41, P = 0.027). Entropy was ranked as the most important radiomic feature.ConclusionsThe machine-learning based analysis of radiomic features enables the prediction of subtypes of PDAC, which are highly relevant for disease-free and overall patient survival and response to chemotherapy.