Surgical Outcomes of Endoscopic Endonasal Versus Transcranial Resections of Adult Craniopharyngioma: A Meta-Analysis

Background The endoscopic endonasal approach (EEA) has been gaining popularity for resection of adult craniopharyngiomas. However, the safety and effectiveness of the procedure in comparison to the traditional transcranial approach (TCA) remains unestablished as previous reviews are outdated. Methods A literature search without language restriction was conducted in PubMed, Cochrane database, and Web of Science from conception to July 9, 2021. Cohort studies and case series that compared EEA with TCA and assessed postoperative complications, recurrence, and 30-day mortality were included. Articles, where data for adult populations could not be extracted or calculated, were excluded. Article selection and data extraction in a predesigned data extraction form were conducted in duplicate. Pooled participant data were included in a random-effects model. Results The search yielded 227 articles, from which eight cohort studies containing 11,395 patients were included (EEA: 6,614 patients, TCA: 4,781 patients). Six studies were good quality and two were fair quality according to the Newcastle Ottawa Scale. There were significantly higher rates of cerebrospinal fluid leak (risk ratio [RR]=0.23, 95% confidence interval [CI] 0.17–0.32, p<0.00001, I²=0%) and lower rates of postoperative hypopituitarism (RR=1.40, 95% CI 1.30–1.51, p<0.00001, I²=0%), hydrocephalus (RR=6.95, 95% CI 5.78–8.36, p<0.00001, I²=0%), visual impairment (RR=1.52, 95% CI 1.34–1.73, p<0.00001, I²=0%), and 30-day mortality (RR=5.63, 95% CI 3.87–8.19, p<0.00001, I²=0%) after EEA. Non-significant lower rates of postoperative diabetes insipidus (RR=1.12, 95% CI 0.78–1.61, p=0.53, I²=85%) and recurrence of tumor (RR=2.69, 95% CI 0.35–20.81, p=0.34, I²=47%) were seen after EEA. Conclusion EEA may be associated with reduced postoperative hypopituitarism, hydrocephalus, visual impairment, and 30-day mortality and higher rates of cerebrospinal fluid leak. These findings do not account for differences in tumor size and extension between the EEA and TCA cohorts. Further research on patients with comparable tumor characteristics is required to fully assess outcomes.