Classification and Reliability of Lumbar Facet Cysts Grading Scales

Abstract

The presence of symptomatic lumbar facet cysts has been associated with segmental instability. Given this association, decompression versus decompression with fusion is a frequently debated topic. Multiple grading scales have been devised to identify patients at high risk for development of cyst recurrence; however, there exists no external evaluation of these scales. A retrospective review of 54 patients undergoing initial treatment for lumbar synovial cysts at a single institution over the past 12 years was conducted. Surgical treatment consisted of decompression with cystectomy without fusion. Patients were assessed and classified according to the NeuroSpine Surgery Research Group (NSURG) and Rosenstock Classification systems. Five neurosurgeons reviewed the preoperative magnetic resonance images, and results were classified. Interrater reliability was assessed using both Gwet's AC1 coefficient and Krippendorff's alpha. A 1-way analysis of variance was used to evaluate predictive ability of both classification systems. In total, of the 54 patients who underwent decompression, 7 had cyst recurrence. Overall cyst recurrence was most common in NSURG grade 2 cysts (3/12, 25%) followed by grade 1 cysts (4/27, 14.8%). Of the NSURG grade 3 and 4 patients, none had cyst recurrence. In the Rosenstock grades the most common recurrence was in grade 3 cysts (1/4, 25%) followed by grade 1 cysts (5/26, 19.2%). Interrater reliability demonstrated good reproducibility on Gwet's AC1 and Krippendorff's alpha on both grading scales. Neither score was predictive of cyst recurrence (P > 0.05). The Rosenstock and NeuroSpine scores demonstrate good overall interrater reliability but are inconsistent in their ability to predict recurrence of lumbar facet cysts.