Expression Levels of RAD51 Inversely Correlate with Survival of Glioblastoma Patients

SIMPLE SUMMARY: Identifying prognostic and predictive biomarkers for glioblastoma (GBM), a primary brain tumor, is essential in improving patient survival. We utilized gene expression profiling to investigate a uniform population of GBM patients who had been treated with surgery and adjuvant radiation therapy versus normal brain tissue, and identified high RAD51 expression as a poor prognostic marker that is amenable to therapeutic intervention. This observation was confirmed utilizing a publicly available gene expression dataset in a cohort of GBM patients. ABSTRACT: Treatment failures of glioblastoma (GBM) occur within high-dose radiation fields. We hypothesized that this is due to increased capacity for DNA damage repair in GBM. We identified 24 adult GBM patients treated with maximal safe resection followed by radiation with concurrent and adjuvant temozolomide. The mRNA from patients was quantified using NanoString Technologies’ nCounter platform and compared with 12 non-neoplastic temporal lobe tissue samples as a control. Differential expression analysis identified seven DNA repair genes significantly upregulated in GBM tissues relative to controls (>4-fold difference, adjusted p values < 0.001). Among these seven genes, Cox proportional hazards models identified RAD51 to be associated with an increased risk of death (HR = 3.49; p = 0.03). Kaplan–Meier (KM) analysis showed that patients with high RAD51 expression had significantly shorter OS compared to low levels (median OS of 10.6 mo. vs 20.1 mo.; log-rank p = 0.03). Our findings were validated in a larger external dataset of 162 patients using publicly available gene expression data quantified by the same NanoString technology (median OS of 13.8 mo. vs. 17.4 mo; log-rank p = 0.006). Within this uniformly treated GBM population, RAD51, in the homologous recombination pathway, was overexpressed (vs. normal brain) and inversely correlated with OS. High RAD51 expression may be a prognostic biomarker and a therapeutic target in GBM.