Systematic review and meta-analysis of imaging characteristics in Chiari I malformation: does anything really matter?

Abstract

Previous studies have attempted to evaluate the utility of preoperative magnetic resonance imaging (MRI) parameters in predicting outcomes in Chiari I malformation. We performed a systematic review and meta-analysis to determine what preoperative imaging features (if any) predict (1) presence of preoperative symptoms or associated findings, (2) need for surgical decompression, or (3) improvement after surgical decompression. All publications through June 2018 on PubMed, Embase, and Cochrane Library databases were searched using the keywords "Chiari I malformation" AND "decompression" OR "imaging." One thousand two hundred ten publications were identified, and 20 were included for our systematic review; nine were included in the meta-analysis. Tonsil position, clivus gradient, and scoliotic curve of > 20° were all associated with the presence of preoperative syrinx. Degree of scoliotic curve was associated with length of syrinx. Pre-operative findings of central syrinx morphology, shorter syrinx, and scoliotic curve < 20° were associated with post-operative stability/improvement. Post-operative symptomatic improvement was associated with preoperative pB-C2 line ≥ 3 mm, absence of scoliosis, and presence of syrinx. By meta-analysis, there was no significant difference in post-operative improvement between patients with and without syrinx (OR = 0.89; 95% CI 0.58-1.37). Meta-analysis showed no significant difference in post-operative improvement between patients with and without basilar invagination (OR = 1.31; 95% CI 0.72-2.36). Multiple studies have attempted to identify preoperative imaging parameters to predict post-operative improvement, but no consistently reliable criteria have been defined. This review and meta-analysis highlight the importance of considering each patient's clinical history and physical exam within the context of associated radiographic abnormalities.