Non-linear Effect of Preexisting Cranial Adjacent Disc Degeneration on Cumulative 12-year Revision Risk Following Lumbar Fusions

Study Design. Retrospective analysis of prospectively collected data Objective. To evaluate how preexisting adjacent segment degeneration status impacts revision risk for adjacent segment disease (ASD) after lumbar fusions. Summary of Background Data. ASD incurs late reoperations after lumbar fusion surgeries. ASD pathogenesis is multifactorial. Preexisting adjacent segment degeneration measured by Pfirrmann is suggested as one of the predisposing factors. We sought to find deeper insights into this association by using a more granular degeneration measure, the Combined imaging score (CIS). Methods. A total of 197 consecutive lumbar fusions for degenerative pathologies were enrolled in a prospective follow-up (median 12 years). Preoperative cranial adjacent segment degeneration status was determined using Pfirrmann and CIS, which utilizes both radiographs and magnetic resonance imaging. Based on CIS, patients were trichotomized into tertiles (CIS <7, CIS 7–10, and CIS >10). The cumulative ASD revision risk was determined for each tertile. After adjusting for age, sex, body mass index, sacral fixation, and fusion length, hazard ratios (95% confidence intervals, CI) for ASD revisions were determined for each Pfirrmann and CIS score. Results. Patients in the intermediate CIS tertile had a cumulative ASD revision risk of 25.4% (17.0% to 37.0%), while both milder degeneration (CIS <7) [13.2% (6.5% to 25.8%)] and end-stage degeneration (CIS >10) [13.6% (7.0% to 25.5%)] appeared to be protective against surgical ASD. Pfirrmann failed to show a significant association with ASD revision risk. Adjusted analysis of CIS suggested increased ASD revisions after CIS 7, which turned contrariwise after CIS 10. Conclusions. The effect of preexisting adjacent segment degeneration on ASD reoperation risk is not linear. The risk seems to increase with advancing degeneration but diminish with end-stage degeneration. Therefore, end-stage degenerative segments may be considered to be excluded from fusion constructs. Level of Evidence. Therapeutic Level III