gms | German Medical Science

Objective: Several parameters are known to predict the survival of glioblastoma (GBM) comprising extend of resection and MGMT promotor methylation. Staining for glial fibrillary acidic protein (GFAP) is being frequently assessed during histological work-ups, however its clinical utility for GBM remains unclear. Aim of the present study was to analyse the predictive value of quantitative GFAP measurements for survival of patients with GBM.

Methods: Of institutional database containing patients with primary GBM who underwent surgery between 2011 and 2014, individuals with tissues examined for immunohistochemical staining of GFAP were included to this study. The percentage of observed GFAP staining was measured in 5% steps. Overall survival was assessed as continuous variable. In addition, long-term survival was addressed using two- and three-years survival. Statistical assessments between GFAP values and survival data were performed using univariate analysis and multivariate regression models. Clinical relevant cut-offs for GFAP staining were identified by receiver operating characteristic curves (ROC).

Results: Of 327 patients in the database, quantitative GFAP measurements were available for 272 cases (mean age 62 [±11.1] years, 117 females [43%]). Overall survival of the final cohort was 11.4 months (±8.6). Mean GFAP value was 66.4% (range 5–95%). Clinically relevant cut-off for correlation between GFAP and overall survival was identified at 75% (area under the curve in the ROC: 0.691). Kaplan-Meier survival analysis also showed poorer survival of GBM patients with GFAP≥75% (p=0.021). Finally, multivariate analysis adjusted for patients’ age, extent of resection, preoperative Karnofsky performance index and MGMT methylation status, confirmed independent predictive value of GFAP≥75% for overall survival (p=0.032). Accordingly, patients with GFAP values ≥75% showed significantly lower rates of long-term survival than individuals with GFAP <75%: 5.8% vs 14.7% (p=0.0282) and 0.8% vs 8% (p=0.0076) for two- and three-years survival respectively.

Conclusion: Routine immunohistochemical assessment of GFAP with quantitative measurements might become a novel biomarker for overall and, especially, long-term survival of patients with GBM. Prospective multi-centric validation of GFAP value for GBM survival is mandatory.