Role of lumbar drains in contemporary endonasal skull base surgery: Meta-analysis and systematic review.

Abstract

Historically, lumbar drains (LD) have played a prominent role in endonasal skull base surgery. Over the past few decades, advancements in techniques have augmented our ability to successfully reconstruct complex skull base defects and often obviates the need for cerebrospinal fluid (CSF) diversion. Clarity on the appropriate use of LDs is needed. To examine the literature for the need for LDs in contemporary skull base reconstruction after resection of skull base tumors. A systematic literature review of English language articles by using PubMed and Ovid. Search terms included "lumbar drain," "CSF leak," and "endoscopic endonasal reconstruction." Articles were included when they pertained to adults, used current methods for reconstruction (i.e., multilayered repair or vascularized tissue), and addressed CSF leak rates secondary to endoscopic resection of skull base masses. All the studies discussed CSF leaks that resulted from traumatic-, idiopathic-, or sinus surgery-related iatrogenic causes were excluded. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were followed. A total of five articles met inclusion criteria. These comprised 376 endoscopic skull base tumor resection and reconstruction cases; 5.59% developed postoperative CSF leaks. A meta-analysis was conducted by incorporating three studies that met criteria based on testing for heterogeneity. The average odds ratio for postoperative CSF leak for patients who did not have an LD relative to patients who had an LD was 0.590 (95% confidence interval, 0.214-1.630). Given a p value of 0.30, the results demonstrated a lack of statistically significant improvement between patients who had an LD and patients who did not have an LD. Various details, such as the defect size and type of CSF leak, were not consistently reported among studies. Available evidence for the use of LDs in skull base surgery is of poor quality. Analysis of the literature revealed heterogenous and varied reporting in the primary literature. Further studies that include randomized controlled trials are needed.