Subsidence after total lumbar disc replacement is predictable and related to clinical outcome

Purpose: As yet, there are no studies describing a relationship between radiographic subsidence after lumbar total disc replacement (TDR) and patient symptoms. To investigate if subsidence, in terms of penetrated bone volume or angular rotation over time (ΔPBV and ΔAR), is related to clinical outcome. To assess if subsidence can be predicted by position implant asymmetry (IA) or relative size of the TDR, areal undersizing index (AUI) on direct post-operative radiographs. Methods: Retrospective cohort study consists of 209 consecutive patients with lumbar TDR for degenerative disc disease. A three-dimensional graphical representation of the implant in relation to the bony endplates was created on conventional radiographs. Consequently, the PBV, AR, IA and AUI were calculated, direct post-operative (DPO) and at last follow-up (LFU). For clinical evaluation, patients with substantial pain (VAS ≥ 50) and malfunction (ODI ≥ 40) were considered failures. Results: At a mean follow-up of 16.7 years, 152 patients (73%) were available for analysis. In 32 patients, revision by spinal fusion had been performed. Both ΔAR (4.33° vs. 1.83°, p = 0.019) and ΔPBV (1448.4 mm³ vs. 747.3 mm³, p = 0.003) were significantly higher in the failure-compared to the success-group. Using ROC curves, thresholds for symptomatic subsidence were defined as ΔPBV ≥ 829 mm³ or PBV-LFU ≥ 1223 mm³ [area under the curve (AUC) 0.723, p = 0.003 and 0.724, p = 0.005, respectively]. Associations between symptomatic subsidence and AUI-DPO ≥ 0.50 (AUC 0.750, p = 0.002) and AR-DPO ≥ 3.95° (AUC 0.690, p = 0.022) were found. Conclusion: Subsidence of a TDR is associated with a worse clinical outcome. The occurrence of subsidence is higher in case of incorrect placement or shape mismatch.