Objective: Skull base chordomas are complex lesions centered on the clivus with or without multicompartmental extension, and surgically can be approached both from a midline or lateral corridor. The present study aimed to compare the surgical outcomes of skull base chordomas from midline versus lateral approaches by conducting a systematic review of the literature and meta-analysis.
Methods: A systematic review and meta-analysis was conducted using the PRISMA protocol. Electronic database search was performed to identify studies between the years 2000 and 2018. The primary outcome was gross total resection rate (GTR). The secondary outcomes were recurrence rate at last follow-up, rate of cerebrospinal fluid (CSF) leak, and new cranial nerve palsy. Odds Ratios (ORs) were calculated using Mantel–Haenszel random effect model for each outcome using studies describing both approaches. An indirect (proportion) meta-analysis was performed pooling all studies describing either of the approaches. Individual patient data were analyzed to see difference in GTR or recurrence rate with different tumor extensions.
Results: Total 33 and 10 studies comprising a total of 951 and 406 patients were found suitable for indirect and direct meta-analysis, respectively. Various midline approaches included endonasal, transoral, transmaxillary and approaches, and lateral approaches included pterional, orbitozygomatic, retrosigmoid, transpetrosal approaches, etc. The overall odds of having GTR with midline approach as compared with lateral approach was found to be 0.83 (95% CI: −0.43 to 1.62; Fig. 1). The pooled estimates for GTR in midline and lateral approaches were 38 and 34%, respectively (p = 0.84; Fig. 2). The ORs for recurrence, CSF leak, and new cranial nerve palsy were 0.82 (95% CI: 0.26–2.56), 1.49 (95% CI: 0.14–15.5), and 0.14 (95% CI: 0.04–0.56), respectively (Figs. 1 and 3). The pooled estimates with midline versus lateral approaches for recurrence was 27 versus 38% (p = 0.13); for CSF leak was 11% versus 0% (p ≤ 0.001) and for cranial nerve palsy was 5 versus 39% (p = 0.06; Figs. 2 and 4). With individual patient data analysis, the difference in GTR or recurrence rates between the two approaches did not reach to significance with different tumor extensions, although the GTR rate was double with lateral approaches (51 vs. 25%) in patients with significant lateral extension.
Conclusion: The current meta-analysis failed to show any significant difference in overall GTR or recurrence rates between the midline versus lateral approaches for skull base chordomas. Tumor extension to different compartments did not affect the GTR or recurrence rates, although lateral approaches showed a slight nonsignificant better resection rate in tumors with significant lateral extension. In terms of complications, CSF leak rate was higher in midline approaches in contrast to the postoperative cranial nerve palsy which was higher in lateral approaches.
