Glioblastoma surgery with and without intraoperative MRI at 3.0T

Abstract

Background: Gross total or near total resection (GTR/NTR; resection ≥95%) of glioblastoma (GBM) seems correlated with a longer survival. Intraoperative MRI (ioMRI) is one method to evaluate the extent of resection (EOR) in order to improve it during the same anesthesia. We compared GBM resections using a 3.0T ioMRI and then without considering the EOR, safety, survival and discussed the indications for using this expensive modality.

Methods: Between March 2006 and November 2011, 56 GBM resections were performed using an ioMRI, and 38 without (control group). The only criterion in order to have access to the ioMRI was its availability. We compared the variables EOR, Karnofsky Performance Scale scores and survival in both groups.

Results: In the ioMRI group, 15 patients (26.8%) underwent an immediate second resection increasing the GTR rate of 10.7% and the GTR/NTR rate of 8.9%. There was a significant difference between the use of an ioMRI and the control group in reaching a larger EOR (P=0.049, Fisher's exact test). The effect of using the ioMRI or not on the overall survival, with EOR as covariate, was not significant (P=0.147, Likelihood ratio test). However, the EOR alone had a significant effect on survival (P=0.049, Wald test), with a shorter survival for the patients with a partial resection (PR) than a GTR/NTR (Hazard ratio=1.6, 95% CI HR: 1.00-2.69), with a median overall survival of 15.26 months (95% CI: 12.34-19.08) for the GTR/NTR subgroup versus 10.26 months (95% CI: 6.64-15.82) for the PR subgroup. Multivariate regression analysis also identified age, sex and adjuvant chemotherapy as factors significantly associated with overall survival.

Conclusions: A 3.0T ioMRI improved the quality of resection by 17.8% and increased the GTR/NTR rate by 8.9% up to 73.2% without additional morbidity. A GTR/NTR improves survival duration by about 50%. Thus, it remains reasonable to increase the EOR to reach GTR/NTR using an intraoperative control. However, ioMRI should be limited to the cases for which a GTR/NTR seems preoperatively possible.

Keywords: Extent of resection; Glioblastoma; Glioblastome; High-field MRI; IRM à champ élevé; Imagerie par résonance magnétique per-opératoire; Intraoperative magnetic resonance imaging; Survie; Survival; Étendue de résection.