Outcomes of Autologous versus Synthetic Inlay Grafts After Skull Base Reconstruction for High‐Flow Defects: A Multicenter Case‐Control Analysis

The most well-known risk related to endoscopic skull base surgery (ESBS) is a postoperative cerebrospinal fluid (CSF) leak, which, if untreated, portends a risk of meningitis. Literature has shown that the risk of postoperative CSF leaks range greatly from 8 to 33% with factors such as BMI, defect location, and type of reconstruction affecting outcomes [1, 2]. As such, skull base reconstruction after ESBS is necessary to create a watertight seal to prevent communication between the sterile subdural CSF space and the contaminated sinonasal tract. For large, high-flow dural defects, multilayer reconstruction with inlay (subdural or epidural) and onlay flaps and grafts remains the gold standard [3, 4]. Specifically for inlay grafts, surgeons may choose to use autologous grafts (those originating from the patient, e.g., fascia lata or fat) or synthetic grafts (commercially available, man-made or processed grafts, e.g., collagen matrix) [5]. A previous systematic review and meta-analysis by our group has shown that CSF leak rates are comparable for reconstructions utilizing either graft type; however, the studies included in the analysis had small sample sizes or were single-institutional, with no head-to-head comparisons [6]. The present multicenter study compares postoperative CSF leak rates following skull base reconstruction utilizing either autologous and synthetic inlay grafts in patients undergoing ESBS with high-flow intraoperative CSF leaks. Although autologous inlay grafts have traditionally used for skull base reconstruction, growing experience suggests comparable efficacy when commercially available synthetic grafts are used for inlay reconstruction. There is no head-to-head comparison between autologous versus synthetic inlay grafts and how they relate to skull base reconstruction outcomes. This multicenter study shows that autologous and synthetic inlay grafts appear to afford comparable rates of reconstructive success, even for large high-flow defects. A multicenter retrospective chart review involving nine institutions (Baylor College of Medicine, Henry Ford Health, Hofstra/Northwell Health, Kaiser Permanente Redwood City, Macquarie University, Rush University, Southern Illinois University, University of California-Irvine, University of Pennsylvania) was conducted from July 1, 2018 to January 31, 2024. Local Institutional Review Board approval was obtained. Demographic and clinical data were collected from patients undergoing ESBS resulting in high-flow CSF leaks with subsequent primary reconstruction. High-flow CSF leak was defined as dural defect >1×1 cm and/or in continuity with a cistern/ventricle [7]. Demographic and treatment data collected include age, sex, pathology, defect site (anterior cranial fossa [ACF], suprasellar [SS], sellar, posterior cranial fossa [PCF]) and size, autologous versus synthetic inlay graft type, and onlay reconstruction (i.e., use of free graft vs. vascularized flap), use of lumbar drain, tissue sealant use, and nasal packing use. The primary outcome was postoperative CSF leak, with meningitis being a secondary outcome. Synthetic grafts included dural substitutes (BioDesign, Cook Medical, Bloomington, IN; AlloDerm, LifeCell, Branchburg, NJ) and collagen matrix (DuraGen, Integra, Princeton, NJ), while autologous grafts include fascia lata and fat grafts harvested intraoperatively. Statistical analyses were performed using R (version 4.2.2; The R Foundation for Statistical Computing) in RStudio (version 2022.12.0). A p value of <0.05 was considered statistically significant. Outcomes related to inlay reconstruction using autologous versus synthetic grafts were compared. Patients were propensity score-matched into autologous and synthetic inlay groups according to their defect site, defect size, pathology, and the use of nasoseptal flap for reconstruction. Propensity scores were calculated using logistic regression, and 1-to-1 matching without replacement was performed using the nearest neighbor method. Logistic regression was performed to identify the effect of graft type on postoperative CSF leak rates. A total of 210 ESBS patients (57.6% female) with an average age of 54.0 ± 15.1 years were included for analysis. The most common pathologies were pituitary adenoma (27.6%), meningioma (25.2%), and craniopharyngioma (19.0%). The average defect size was 296.87 ± 239.15 mm2 (range: 24.00–1289.60 mm2). A total of 110 (60.1%) patients had inlay reconstructions exclusively utilizing synthetic grafts and 187 (89.0%) patients had a NSF used for reconstruction. 108 (51.4%) of patients had a lumbar drain placed. Additionally, the most common dural sealant type was Adherus (82 patients; 40.5%) and the most common packing type was matrix/xeroform/strip gauze (122 patients; 58.1%). A total of 20 (9.6%) patients experienced a postoperative CSF leak and 6 (3.8%) developed meningitis. Demographic and treatment data are tabulated in Table 1. Following PSM and controlling for defect site and size, pathology, and NSF use, there was no difference on Fisher's exact test in postoperative CSF leak rates between autologous (three out of 28; 10.7%) versus synthetic (one out of 28; 3.6%) inlay reconstructions (p = 0.611), which was further confirmed to be noncontributory on multivariate logistic regression (OR: 0.486, 95% CI: 0.075–2.233; Table 2). Herein, it was shown that, for patients with a high-flow intraoperative CSF leak following tumor resection, those that underwent skull base reconstruction with synthetic grafts, when compared with autologous grafts, had no significant differences in postoperative CSF leak rates, even controlling for defect and other risk factors. The results must be interpreted in the context of most reconstructions utilizing NSF, a relatively low matched subcohort (n = 28/group), and an overall low event rate (<10%) [7]. This is the largest multi-institutional study comparing postoperative CSF leak rates between different inlay materials, with adequate granularity of data. Previous systematic review and meta-analysis of 16 studies by our group explored the rates of postoperative CSF leak and meningitis with autologous and synthetic inlay reconstruction, with only one study having a level of evidence of 2 and low sample size (n = 32) [6]. Another systematic review also explored different reconstructive strategies and associated outcomes, but at the same time showcased the heterogeneity of techniques utilized [8]. Ultimately, this study showed that the rate of postoperative CSF leaks was not different among autologous and synthetic inlay grafts, especially with most cases also using a NSF. This highlights the importance of meticulous technique as opposed to specific tissue type, as there were comparable rates of reconstructive success between grafting material. Additionally, surgeons must consider the availability and the financial cost of synthetic materials, which must be balanced with donor site morbidity in autologous grafts [9, 10]. N. D. A. and J. N. P. are associated with Acclarent, Optinose, and 3-D Matrix. E. C. K. is a consultant for Stryker and 3-D Matrix and receives royalties from Springer.