← Back to Previous Page

 

P253: INCIDENCE OF CEREBROSPINAL FLUID LEAK FOLLOWING ENDOSCOPIC SURGICAL MANAGEMENT OF PETROUS APEX LESIONS: A SYSTEMATIC REVIEW AND META-ANALYSIS
Khushal Gupta, MBBS¹; Michael Karsy, MD, PhD, MSc, FAANS, FCNS²; ¹Bharat Ratna Late Shri Atal Bihari Vajpayee Memorial Medical College; ²University of Michigan Health

Introduction: Endoscopic surgery has broadened access to the petrous apex, offering less morbidity than traditional open approaches. Yet, cerebrospinal fluid (CSF) leak remains a frequent and feared complication. Reported rates vary widely, and most studies are small single-institution series. We performed a systematic review and meta-analysis to provide pooled estimates of CSF leak incidence and related outcomes after endoscopic surgery for petrous apex lesions.

Methods: This review followed PRISMA guidelines. PubMed and Embase (inception–July 2025) were searched using terms “petrous apex,” “endoscopic,” and “surgery.” Studies were eligible if they included ≥5 patients, reported endoscopic surgical management of petrous apex lesions, and provided extractable outcome data. Case reports, purely technical notes, and reviews were excluded. Data were extracted into a structured template capturing study design, sample size, pathology, surgical approach, reconstruction method, follow-up, and outcomes.

The primary endpoint was postoperative CSF leak. Secondary endpoints included extent of resection (EOR), recurrence or restenosis, and mortality (procedure-related/≤30-day and all-cause, analyzed separately), as well as complications such as cranial nerve palsy and epistaxis. Event counts and denominators were pooled using a DerSimonian–Laird random-effects model with logit transformation and continuity correction for zero-event studies. Wilson score confidence intervals were used for study-level proportions. Heterogeneity was assessed with Cochran’s Q and I². Analyses were performed in Python 3.11 with routines replicating methods implemented in the metafor package (R Foundation).

Results: Six studies encompassing 82 patients met inclusion criteria. The pooled incidence of postoperative CSF leak was 10.2% (95% CI, 5.2–19.2%) with no heterogeneity (I²=0%). Gross total resection was reported in two studies, yielding a pooled rate of 65.9% (95% CI, 41.6–83.9%), though mixed pathologies limited interpretation.

Recurrence or restenosis was reported in three studies (40 patients), with a pooled rate of 15.5% (95% CI, 7.3–29.7%). Mortality was described in five studies (69 patients). The crude all-cause rate was 2/69 (2.9%). With continuity correction, the pooled all-cause estimate was 7.5% (95% CI, 3.0–17.6%), inflated by small zero-event series and inclusion of deaths labeled “unrelated.” Restricting to procedure-related or ≤30-day deaths, the pooled rate was essentially 0%.

Complications varied. Any new cranial nerve palsy occurred in 15.6% (95% CI, 6.1–34.3%), mainly CN VI palsy at 14.7% (95% CI, 5.1–35.8%). Epistaxis occurred in 8.1% (95% CI, 2.6–22.3%). Single cases of ICA injury, otitis media, and postoperative sinusitis were reported.

Conclusion: This systematic review and meta-analysis demonstrates that endoscopic management of petrous apex lesions carries an approximately 10% risk of CSF leak, with consistent rates across studies. Gross total resection is achievable in selected cases, though data remain limited. Recurrence rates approach 15%, and cranial nerve morbidity is a significant concern. Mortality was low; the pooled all-cause estimate was inflated by small-study effects, while procedure-related mortality was near zero. These pooled benchmarks support surgical counseling and highlight the importance of robust reconstruction and standardized multicenter reporting to refine complication profiles and improve outcomes.

View Poster

 

← Back to Previous Page