Surgical risk of CSF leakage following endoscopic transorbital approach for anterior and middle skull base pathologies: a systematic review and meta-analysis.

Abstract

The endoscopic superior eyelid transorbital approach (SETOA) has demonstrated considerable versatility and effectiveness in managing various paramedian anterior and middle skull base pathologies. However, as with any relatively new technique, potential complications remain. We conducted an extensive literature search in MEDLINE and Embase in accordance with PRISMA guidelines including case reports and surgical series reporting cerebrospinal fluid (CSF) leak rate following SETOA for intracranial pathologies. Factors analyzed included lesion location (extra- or intra-axial), reconstruction techniques, and complication management. ROBINS-I tool was employed to assess the risk of bias. Twenty-five studies including 240 cases were eligible. The majority of lesions were intradural extra-axial (68.3%), while trigeminal schwannomas comprised all extradural cases (25.0%). Sixteen patients (6.6%) presented intradural intra-axial tumors. Osteodural reconstruction involved dural substitutes in one third of the cases (32.5%) either alone (14.2%) or combined with fat free graft (18.3%). CSF leak occurred in 6 patients (2.50%), mostly resolving via conservative management (66.6%). The risk of postoperative CSF leak was found to be significantly higher in patients undergoing resection for intra-axial tumors (OR 0.13, 95% CI: 0.04-0.49) compared to those undergoing resection for extra-axial (OR 0.01, 95% CI: 0.00-0.02; I² = 0%; p < 0.001). Key limitations include the retrospective nature and small sample sizes among included studies as well as data heterogeneity and lack of standardized protocols for reconstruction across studies. SETOA appears safe for addressing selected extradural and intradural skull base pathologies with a low postoperative CSF leak rate. The natural repositioning of the orbital content to its original position may be instrumental in preventing its postoperative occurrence. The investigation followed a prespecified protocol registered on PROSPERO (PROSPERO 2024 CRD42024614111).