Crossing the Bridge from Degeneration to Deformity: When Good Outcomes Necessitate Sagittal Correction in Adult Spinal Deformity Surgery

Background: Patients with less severe adult spinal deformity undergo surgical correction and often achieve good clinical outcomes. However, it is not well understood how much clinical improvement is due to sagittal correction rather than repair of degenerative disc changes.
Hypothesis: There are baseline thresholds in radiographic parameters that, when exceeded, result in dramatic clinical improvement from surgical correction.
Study Design/Setting: Retrospective cohort study of a single-center ASD database

Methods: ASD patients with BL and 2-year(2Y) data included. Parameters assessed: SVA, PI-LL, PT, T1PA, L4-S1 Lordosis, C2-C7 SVA (cSVA), C2-T3, C2 Slope (C2S). Outcomes: Good Outcome (GO): [Meeting either: 1) SCB for ODI by 2Y (change greater than 18.8), or 2) ODI < 15 and SRS-Total >4.5 by 2Y. Binary logistic regression assessed each parameter to determine if correction was more likely needed to achieve GO. Conditional inference tree (CIT) run machine learning analysis generated baseline thresholds for each parameter, above which, correction was necessary to achieve GO.

Results: Included: 447 ASD patients. There were 223 (50%) of patients achieving GO by two years. Binary logistic regression analysis demonstrated correction of all five thoracolumbar parameters was more often needed to achieve GO (all p<.001). Upon CIT, thresholds were derived for each parameter, depicted in Table, with a baseline PI-LL above 10° (74% of patients meeting GO) most likely needing correction to achieve GO (OR: 17, [7.0-41.4]; p<.001). Of patients with baseline T1PA above the threshold, 95% required correction to meet Good Outcome (95% vs. 54%, p<.001). All three cervical parameters were assessed and more often required correction to achieve GO (all p<.001). CIT-generated thresholds were significant for each parameter, with a baseline C2 slope above 15° necessitating correction more often to obtain clinical success (OR: 8.1, [4.1-16.2]; p<.001). Conversely, patients above the C2-T3 threshold only required correction 11% of the time to achieve GO (11% vs. 45%, p<.001).

Conclusion: While degenerative thoracolumbar changes can be debilitating and require surgical intervention, our study highlighted there is a point at which sagittal correction has an exponential influence on clinical improvement, reflecting the line where deformity becomes a significant contributor to disability. These new thresholds delineate patients suitable for sagittal correction, as opposed to degenerative disc repair or replacement only, and may better increase the utility gained from surgical intervention for adult spinal deformity