The role of R-loop aberrations in lower-grade gliomas: prognostic, immune, and metabolic implications from multi-omics and machine learning analysis

Background: The prognostic and therapeutic implications of R-loop regulators in lower-grade glioma (LGG) remain largely unexplored. Methods: We constructed a R-loop prognostic index (RLPI) by benchmarking 9 machine-learning algorithms through nested cross-validation. Associations between RLPI and clinical outcomes, immune microenvironment, and metabolic features were assessed across -11 bulk transcriptomic, 2 single-cell RNA-sequencing, and 1 spatial transcriptomic LGG cohorts. Functional validation was performed for key RLPI genes. Results: Among the 9 algorithms tested, Elastic Net demonstrated the best predictive performance and was selected to construct the RLPI, which comprised 35 R-loop regulators. Elevated RLPI robustly predicted poorer prognosis in 1556 patients across 6 independent LGG cohorts. High-RLPI glioma cells were enriched in angiogenesis and hypoxia pathways and exhibited a mesenchymal-like state with heightened metabolic activity. Notably, RLPI correlated with enhanced antigen presentation and immunosuppressive signaling, alongside increased infiltration of immune cells with pro- or anti-tumor functions. Moreover, high RLPI correlated with resistance to radiotherapy and temozolomide but improved response to immune checkpoint blockade therapy. Functional assays revealed that knockdown of INCENP suppressed glioma cell proliferation, migration, and invasion while promoting apoptosis, whereas NCAPG silencing impaired migration and invasion. Conclusion: This study establishes the RLPI as a promising biomarker for personalized risk stratification and treatment guidance in LGG.