Level of evidence in the literature concerning brain tumor resection

Microsurgical resection is a cornerstone in the treatment of brain tumors. However, the benefit of radical resection still remains controversial. We attempted to analyze the level of scientific evidence (LOE) and methodological aspects of studies concerning the impact of the extent of resection (EOR) on outcome. The LOE classification was Ia: 0%, Ib: 0%, IIa: 0.8%, IIb: 5.8%, IIc: 0%, IIIa: 13.3%, IIIb: 52.5%, IV: 10.8%, V: 16.8%. 72.5% observed a positive effect of total tumor removal. 84.2% did not report the criteria for treatment assignment, 62.5% did not define the terms gross total; radical; partial; or subtotal resection. The average age of the treatment groups was reported in 29.2%, the Karnofsky index in 75.8%. Tumor size was reported in 32.5%, location in 51.7%. Assessment of EOR was based on the surgeon's impression in 75.0%, determined by postoperative CT/MRI scans in 20.8%, quantified by CT/MRI-based volumetry in 3.4%, and assessed by histological analysis in 0.8%. To date, no studies with high LOE are available addressing the benefit of gross total brain tumor removal. Although the majority of the reports found a positive effect of radical resection, the reviewed articles contain methodological limitations which may significantly influence the results. (C) 2004 Elsevier B.V. All rights reserved.