Despite its marked molecular heterogeneity and variable response to gemcitabine-based therapies, pancreatic ductal adenocarcinoma (PDAC) remains poorly understood at the pathway level, particularly across age and treatment contexts. We applied a conversational artificial intelligence framework (AI-HOPE-TP53 and AI-HOPE-PI3K) to analyze clinical and genomic data from 184 PDAC tumors stratified by age and gemcitabine exposure. Pathway-centric analyses were performed and validated using conventional statistical methods. TP53 alterations were more frequent in early-onset compared to late-onset PDAC among gemcitabine-treated patients and showed a similar trend in early-onset untreated cases. In late-onset PDAC without gemcitabine exposure, absence of TP53 alterations was associated with improved overall survival. PI3K alterations were enriched in late-onset gemcitabine-treated tumors. Notably, late-onset patients without PI3K alterations who were not treated with gemcitabine demonstrated significantly improved survival. TP53 and PI3K pathway dependencies in PDAC are context-specific, varying by age and treatment exposure. These findings highlight the value of conversational AI for integrative precision oncology analyses and molecular stratification.
Diaz et al. (Sat,) studied this question.