将他汀类药物重定位用于局部晚期食管腺癌：整合scRNAseq、基于类器官的高通量药物筛选和患者层面的临床数据

Introduction: Overcoming chemoresistance remains a significant challenge in esophageal adenocarcinoma (EAC). Drug repurposing has gained interest for its potential to identify new uses for approved drugs, with advantages including lower costs, established safety, and accelerated timelines. Seeing as EAC tumors demonstrate increased reliance on lipids as an alternative energy source which is associated with therapeutic resistance, we sought to investigate, through an integrated multi-omic approach, the potential of repurposing lipid-lowering drugs as a therapeutic intervention. Materials and Methods: We integrated scRNAseq datasets, a gene network-dependent drug repurposing machine learning algorithm (NeMoCAD) and Hight Throughput Drug screening (HTS) on tumor organoids (PDOs) derived from EAC patients undergoing neoadjuvant chemotherapy to investigate therapy resistance. We applied scRNAseq data in NeMoCAD pipeline to identify agents within the LINCS database (approx. 30,000 drugs), capable of reversing the chemoresistance-specific tumor transcriptional states to states toward chemosensitive or normal phenotypes. Additionally, we leveraged a clinically characterized dataset on 450 EAC patients who underwent perioperative chemotherapy followed by surgical resection. Approximately one-third of which were prescribed statins for cardiovascular reasons. PDOs from non-statin users were selected to evaluate the synergy between hydrophobic statins (atorvastatin, simvastatin) and triplet chemotherapy in vitro . Results: NeMoCAD analysis predicted that hydrophobic statins, which inhibit 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR), can shift EAC tumor gene expression patterns toward those observed in normal gastroesophageal tissues. From our retrospective cohort of patients with pre-operative chemotherapy, statin users had a significantly higher 5-yr overall survival rate (58% vs. 38%; HR 0.55, p=0.018), as well as a 26% increase in major pathological response and significant reductions in recurrence (20%) and distant metastasis (37%). This improvement in survival is despite, statin users being older, higher BMI, more likely to have severe systemic diseases compared to non-statin users. HTS experiments with PDOs demonstrated response with several clinically approved statins, and focused validation drug testing showed pronounced synergistic effect when low doses of hydrophobic statins were combined with standard of care chemotherapeutic regimes. Conclusions: Our findings combining clinical patient outcome, scRNAseq data, drug repurposing machine learning pipelines, and functional validation with PDOs based HTS suggest that statins may represent a valuable adjunct to standard-of-care treatment for EAC.