Evaluation of surgical decision making and resulting outcome in patients with highly eloquent glioblastoma: Results of a multicenter assessment

- Multicenter retrospective study on resectability in GB patients amenable for a sub-total resection only.
- Assessment of the potential EoR in these patients by volumetric assessment of a blinded central reviewer.
- Patients with a NTR showed the highest benefit from resection. STR and stereotactic biopsy did not differ significantly.
- Reviewer's recommendation differed significantly with clinical decision making in the cohort of subtotal resection.
- We found a high potential to increase EoR in this cohort. Potential EoR was significantly underestimated in our series.

IntroductionTreatment of glioblastoma(GB) patients amenable only for a subtotal resection(STR) is controversial. Since outcome of patients is affected by surgical management, our aim was to assess surgical decision making and resulting outcome in patients with highly eloquent GBs.Patients and methodsWe retrospectively assessed GB patients with intended sub-total resection (STR) or stereotactic biopsy (STX) of 3 neurooncological centers operated between 2008 and 2013. A volumetric assessment of overall extent of resection(oEoR), presence of complications, new permanent neurological deficits(nPNDs) was performed. A central reviewer reassessed all cases blinded and gave recommendation on surgical management and on a potential EoR(pEoR) based on imaging data. We compared outcome data using Mann-Whitney-U-test and Sign-Rank-Test. Survival was assessed based on Kaplan-Meier-estimates.Results97 patients were included. In 17 patients received STX, 70 patients a STR and 10 patients a near total resection (NTR, EoR > 95%). Median OS was significantly different from STX patients only if NTR was reached (16 vs. 7 months, p = 0.042). The central reviewer recommended a more aggressive strategy(NTR or STR resp.) in 41 patients and a less aggressive strategy in 13 patients. Overall, management recommendation was significantly different to clinical treatment (p < 0.001). Mean pEoR was significantly higher than oEoR (85.7% vs. 71.3%, p = 0.001). Regarding the different OR subgroups, no significant differences were found in the NTR group(12/13 ties, p = 1) and in STX group (14/17 ties, p = 0.125). In STR group, a significant difference was found (p = 0.001). In 38/69 patients a NTR and in 13/77 patients a STX was recommended.ConclusionSurgery in GB patients with intended STR requires precise preoperative planning since potential EoR is mainly underestimated. Especially, patients with lesions amenable for a NTR should not be missed.