Early post-operative magnetic resonance imaging in glioblastoma: correlation among radiological findings and overall survival in 60 patients.

ABSTRACT

OBJECTIVES: To evaluate early post-operative magnetic resonance (EPMR) as a prognostic tool after resection of glioblastoma. METHODS: Sixty EPMR examinations were evaluated for perioperative infarct, tumour growth between diagnosis and EPMR, contrast enhancement pattern, and extent of resection (EOR). The EOR was approached with the subjective evaluation of radiologists and by quantifying volumes. These parameters were tested as predictors of survival using the Kaplan-Meier method. RESULTS: Contrast enhancement was found in 59 patients (59/60; 98 %). Showing a thin-linear pattern of enhancement was the most favourable finding. Patients with this pattern survived longer than patients with thick-linear (median overall survival (OS) thin-linear=609 days; thick-linear=432 days; P = .023) or nodular (median OS = 318 days; P = .001) enhancements. The subjective evaluation of the EOR performed better than its quantification. Patients survived longer when resection was total (median OS total resection=609 days; subtotal=371 days; P = .001). When resection was subtotal, patients survived longer if it was superior to 95 % (median OS resection superior to 95 %=559 days; inferior to 95 %=256 days; P = .034). CONCLUSIONS: EPMR provides valuable prognostic information after surgical resection of glioblastomas. A thin-linear pattern of contrast enhancement is the most favourable finding. Further prognostic stratification may be obtained by assessing the EOR. KEY POINTS • Some kind of contrast enhancement may be found in most EPMR examinations. • Thin-linear enhancements in the EPMR may be considered benign findings. • The EOR evaluated in the EPMR may stratify prognostic groups of patients. • The subjective evaluation of the EOR performs slightly better than its quantification.