Metformin may restore PARP inhibitor sensitivity in PTEN-deficient prostate cancer by resetting NADP+ levels

Prostate cancer is a leading cause of cancer illness in men, and PARP inhibitors (PARPi) such as olaparib help only a minority of castration-resistant patients, mainly those with BRCA1/2 or related repair-gene mutations. About half of metastatic castration-resistant cases lose the tumor suppressor PTEN. Biologically that should make them vulnerable to PARPi, yet clinical trials of PARPi alone have not improved survival in this group. This study asks why, and whether a widely used, inexpensive drug can fix it.

The researchers first analyzed a retrospective cohort of 461 men with metastatic castration-resistant prostate cancer and confirmed DNA-damage-repair gene defects. All were treated with olaparib at four Chinese tertiary centers between 2017 and 2024. The primary endpoint was PSA50 response (at least a 50% PSA drop); secondary endpoints were progression-free survival (PFS) and overall survival (OS). They then ran cell experiments in PTEN-deficient (PC3, LNCaP) and PTEN-positive (DU145, 22Rv1) lines, using viability, colony, comet (DNA damage), apoptosis and 3D spheroid assays. They also measured NAD+/NADP+, ROS, intracellular pH and NHE1 activity, ran RNA-seq on PTEN-knockdown cells, and designed mouse xenograft studies with the PARPi mefuparib and metformin.

In the cohort, 63.6% of patients achieved a PSA50 response. Patients with PTEN mutation who also took metformin had a 97.8% response rate (44/45), versus 60.6% in the controls. Metformin did not change response in PTEN-wild-type patients (57.1% vs 61.2%). In the 279 patients with survival follow-up, the PTEN-mutant plus metformin group had longer PFS (12.9 vs 9.5 months) and OS (26.9 vs 24.2 months). Multivariate models identified this combination as an independent favorable predictor. In the lab, PTEN-deficient PC3 cells became about three-fold less sensitive to PARPi in lactate-rich or acidic (pH 6.5) medium, with less DNA damage and less apoptosis. PTEN-positive cells were unaffected. Metformin alone did nothing under normal conditions but abolished the resistance in acidic, lactate-rich conditions in PTEN-deficient cells only.

The abstract reports the proposed mechanism; the supplied full text cuts off before these results. In PTEN-deficient cells, lactate activates the NHE1/PKC/NOX1 axis and sustains elevated NADP+. NADP+ competes with PARPi in forming PARPi-PARP-DNA complexes, which causes resistance. Metformin raises NADP+ further, which the authors say alters PARP's structure through an allosteric effect and restores drug efficacy. The team also built and validated machine-learning models (11 algorithms, 11 predictors, SHAP interpretation) for predicting PSA50 response, deployed as a Streamlit web tool.

If confirmed, this suggests metformin could be a low-cost way to widen PARPi use in PTEN-deficient prostate cancer. Several caveats apply. The data are retrospective, the metformin-plus-PTEN group is small, and the work is mostly in cell lines. The supplied text also has internal inconsistencies in group sizes and statistical wording, and the mechanistic, animal and machine-learning results could not be verified in detail from the available text.