Patient-specific statistical shape modeling for optimal spinal sagittal alignment in lumbar spinal fusion

PURPOSE The present study compared patients developing ASD after L4/5 spinal fusion with a control group using a patient-specific statistical shape model (SSM) to find alignment-differences between the groups. METHODS This study included patients who had undergone spinal fusion at L4/5 and either remained asymptomatic (control group; n = 25, follow-up of > 4 years) or required revision surgery for epifusional ASD (n = 22). Landmarks on preoperative and postoperative lateral radiographs were annotated, and the optimal spinal sagittal alignment was calculated for each patient. The two-dimensional distance from the SSM-calculated optimum to the actual positions before and after fusion surgery was compared. RESULTS Postoperatively, the additive mean distance from the SSM-calculated optimum was 86.8 mm in the ASD group and 67.7 mm in the control group (p = 0.119). Greater differences were observed between the groups with a larger distance to the ideal in patients with ASD at more cranial levels. Significant difference between the groups was seen postoperatively in the vertical distance of the operated segment L4. The patients with ASD (5.69 ± 3.0 mm) had a significant greater distance from the SSM as the control group (3.58 ± 3.5 mm, p = 0.034). CONCLUSION Patients with ASD requiring revision after lumbar spinal fusion have greater differences from the optimal spinal sagittal alignment as an asymptomatic control group calculated by patient-specific statistical shape modeling. Further research might help to understand the value of SSM, in conjunction with already established indexes, for preoperative planning with the aim of reducing the risk of ASD. LEVEL OF EVIDENCE I Diagnostic: individual cross-sectional studies with consistently applied reference standard and blinding.