TMZ-induced inflammatory remodeling exposes a ROCK vulnerability in recurrent GBM; fasudil improves survival in preclinical models

In a new preprint, temozolomide (TMZ; alkylating chemotherapy) was evaluated in a window-of-opportunity recurrent glioblastoma (GBM) cohort within NCT05236036: 4 patients received additional pre-op TMZ (150mg/m²/day ×5) before resection vs 3 recurrent comparators without added pre-op TMZ (total n=7 tumors profiled by single-cell/spatial methods and multiplex imaging). Pre-op TMZ exposure was associated with enrichment of inflammatory/chemokine/adhesion/wound-healing programs (GSEA q<0.05), with malignant-cell inflammatory activity covarying with integrin-binding and wound-healing transcriptional programs (median partial ρ=0.45 and 0.42), and spatial vascular CXCL12 plus MYL9/p-MYPT1/nuclear p-STAT3 correlations consistent with ROCK–myosin cytoskeletal activation (e.g., MYL9 vs CXCL12 ρ=0.61 across specimens). Functionally, TMZ increased reactive-state markers and scratch closure in vitro; CXCL8/CXCL12 increased MLC2 phosphorylation; and the ROCK inhibitor fasudil attenuated inflammatory signaling programs and reduced scratch closure/p-MLC2, while fasudil+TMZ improved survival vs TMZ alone in 3 orthotopic models (model-stratified HR 0.21, 95% CI 0.07–0.61; p=0.001)—not yet directly comparable to recurrent-GBM clinical benchmarks where TMZ rechallenge typically yields PFS-6 on the order of ~23–58% (best in MGMT-methylated subsets) and median OS roughly ~5–13 months depending on selection and prior TMZ-free interval.