PURPOSE Combined mitogen-activated protein kinase kinase (MEK) and autophagy inhibition has shown antitumor efficacy in preclinical pancreatic ductal adenocarcinoma (PDAC) models. We evaluated this therapeutic strategy in the autochthonous K-ras LSL.G12D/+ ; Trp53 R172H/+ ; Pdx-1-Cre (KPC) mouse model and in the MEKiAUTO phase I trial of patients with metastatic KRAS -mutant PDAC, with or without anti–PD-L1 antibody atezolizumab. Methods Short-term intervention studies in KPC mice and a phase I clinical trial based on a time-to-event continual reassessment method design were performed. Tumor response was assessed through multiplex immunofluorescence, reverse-phase protein array, and single-nucleus RNA sequencing of serial tumor specimens. RESULTS MEK and autophagy inhibition suppressed tumor growth in KPC mice. In the clinical trial, 14 patients were treated and four experienced a dose-limiting toxicity. The median progression-free survival (PFS) and overall survival were 7.7 weeks (95% CI, 6 to 15.6) and 20.7 weeks (95% CI, 15.6 to 46.1), respectively. snRNA-seq on paired tumor biopsies identified substantial malignant cell heterogeneity, including an epithelial subtype enriched for autophagy and mitogen-activated protein kinase (MAPK) signaling that decreased after treatment. Higher baseline abundance of this subtype correlated with longer PFS but was rare in TCGA PDAC tumors. By contrast, KPC tumors and PDAC cell lines exhibited reduced heterogeneity and uniform enrichment of this MAPK/autophagy-high state, consistent with their stronger preclinical responses. CONCLUSION Combined MEK and autophagy inhibition showed limited tolerability in human PDAC. Divergent efficacy between preclinical and clinical settings likely reflects differences in tumor cell state heterogeneity between models. Integration of diverse, representative preclinical models is critical to guide development of effective therapies in PDAC.
Labadie et al. (Mon,) studied this question.