Oncologic benefits of dural resection in spinal meningiomas: a meta-analysis of Simpson grades and recurrence rates

OBJECTIVE While resection of the dural attachment has been shown by Simpson and others to reduce recurrence rates for intracranial meningiomas, the oncological benefit of dural resection for spinal meningiomas is less clear. The authors performed a systematic analysis of the literature, comparing recurrence rates for patients undergoing various Simpson grade resections of spinal meningiomas to better understand the role of dural resection on outcomes after resection of spinal meningiomas. METHODS The PubMed/Medline database was systematically searched to identify studies describing oncological and clinical outcomes after Simpson grade I, II, III, or IV resections of spinal meningiomas. RESULTS Thirty-two studies describing the outcomes of 896 patients were included in the analysis. Simpson grade I, grade II, and grade III/IV resections were performed in 27.5%, 64.6%, and 7.9% of cases, respectively. The risk of procedure-related complications (OR 4.75, 95% CI 1.27-17.8, p = 0.021) and new, unexpected postoperative neurological deficits (OR (infinity), 95% CI NaN-(infinity) p = 0.009) were both significantly greater for patients undergoing Simpson grade I resections when compared with those undergoing Simpson grade II resections. Tumor recurrence was seen in 2.8%, 4.1%, and 39.4% of patients undergoing Simpson grade I, grade II, and grade III/IV resections over a mean radiographic follow-up period of 99.3 +/- 46.4 months, 95.4 +/- 57.1 months, and 82.4 +/- 49.3 months, respectively. No significant difference was detected between the recurrence rates for Simpson grade I versus Simpson grade II resections (OR 1.43, 95% CI 0.61-3.39, p = 0.43). A meta-analysis of 7 studies directly comparing recurrence rates for Simpson grade I and II resections demonstrated a trend toward a decreased likelihood of recurrence after Simpson grade I resection when compared with Simpson grade II resection, although this trend did not reach statistical significance (OR 0.56, 95% CI 0.23-1.36, p = 0.20). CONCLUSIONS The results of this analysis suggest with a low level of confidence that the rates of complications and new, unexpected neurological deficits after Simpson grade I resection of spinal meningiomas are greater than those seen with Simpson grade II resections, and that the recurrence rates for Simpson grade I and grade II resections are equivalent, although additional, long-term studies are needed before reliable conclusions may be drawn.