AMPK Pathway Activation Markers in GBM: Expression Patterns and Clinical Associations in a Real-World Study

Phosphorylated ACC (pACC), a downstream effector of the LKB1/AMPK pathway, may support tumor bioenergetics and exert a tumor-promoting role in glioblastoma (GBM). In the randomized REGOMA trial, pACC expression was associated with improved overall survival (OS) in GBM patients treated with regorafenib, but its validity in large cohorts is unclear. This study aimed to characterize pACC and phosphorylated AMPK (pAMPK) expression in a real-world GBM cohort and to assess their prognostic and predictive value in patients receiving second-line regorafenib or fotemustine/lomustine. In this retrospective, multicenter translational study, pACC and pAMPK were assessed by immunohistochemistry on FFPE tumor sections from 174 IDH-wild-type GBM patients treated with regorafenib (n = 84) or fotemustine/lomustine (n = 90). Staining was performed with a validated anti-pACC monoclonal antibody and quantified by digital pathology. Survival analyses included univariable and multivariable Cox proportional hazards models, with interaction testing. pACC and pAMPK were expressed in 85.6% and 95.4% of samples, respectively, with heterogeneous spatial patterns. Neither marker was significantly associated with OS in the univariable analyses, and neither retained independent prognostic value in multivariable analysis. Regorafenib was associated with significantly longer OS than alkylating agents (10.4 vs. 6.3 months; p = 0.0021). However, the treatment-by-pACC interaction test was not statistically significant (adjusted p = 0.610), providing no formal evidence of a differential treatment effect by pACC status. Although pACC-positive expression was enriched among patients experiencing longer overall survival under regorafenib, formal treatment-by-biomarker interaction was non-significant. These exploratory findings indicate that pACC is not an established predictive biomarker and warrant prospectively powered evaluation.