The natural course of bridging osteophyte formation in diffuse idiopathic skeletal hyperostosis: retrospective analysis of consecutive CT examinations over 10 years.

OBJECTIVE The aim of this study was to evaluate the natural progression of bridging osteophyte formation in diffuse idiopathic skeletal hyperostosis (DISH) on CT by a newly proposed scoring system. METHODS CT examinations of the thoracic/lumbar spine of DISH patients (Resnick criteria) obtained at two or more time points within a minimum of 3 years were evaluated. Twenty-six patients (mean age at first CT 57 years, 21 males) fulfilled the entry criteria. A semi-quantitative scoring system for osteophyte progression was evaluated for intra- and interreader reliability on 68 vertebral units (VUs) in five patients. CT sagittal reformates of all 26 study patients were scored by two readers in consensus. RESULTS Scoring intra- and interobserver intraclass correlation coefficient values were high (0.971 and 0.893, respectively). The average time points per patient was 3.6 in 398 VUs analysed for 93 time points. The average time between the first and last scans was 5.6 years (range 3-10). The scores of six patients were unchanged. The scores of 20 patients increased by 3 units in 48 VUs over 5.6 (s.d. 3.1) years. The time for a DISH score to increase by 1 scoring unit was 1.6 (s.d. 0.4) years. Two bridging patterns were observed: osteophyte fusion associated with a calcified anterior longitudinal ligament (ALL, 66%) and osteophyte fusion without apparent ALL calcification (33%). Both patterns were observed concomitantly in 15 patients. CONCLUSION The new scoring system may enable earlier diagnosis and help predict disease progression into its final confluent osteophyte form. The two described patterns may indicate an underlying inflammatory rather than a degenerative pathogenesis.