<h3>BACKGROUND CONTEXT</h3> The durability of adult spinal deformity (ASD) surgery is key for cost-effective treatment. Malalignment is one of the main reasons for revision surgery. <h3>PURPOSE</h3> To investigate risk factors for loss of correction within the instrumented lumbar spine following ASD surgery. <h3>STUDY DESIGN/SETTING</h3> Retrospective cohort study of a prospective database. <h3>PATIENT SAMPLE</h3> This study included 321 ASD patients with minimum 3 -year follow-up. <h3>OUTCOME MEASURES</h3> L1-S1, L4-S1 and L1-L4 lumbar lordosis. <h3>METHODS</h3> A total of 321 patients who underwent fusion of the lumbar spine (=5 levels, LIV S1/ilium) with a revision-free follow-up =3 years were identified. Patients were stratified by the change in PI-LL from 6 weeks to 3 years postop as Maintained vs Loss > 5°. Those with a loss due to instrumentation failure (broken rod, screw pullout, etc.) were excluded before comparisons. Changes in regional and focal lordosis over time were investigated with repeated measures ANOVA and factor comparison. <h3>RESULTS</h3> Mean age 64 yrs, BMI 28 kg/m<sup>2</sup>, 80% female. The baseline alignment (PI-LL=21±19°, T1PA=26±12°) was corrected to PI-LL=3±13°, and T1PA=18±10° at 3 yrs (mean 3Y follow-up: 45±11m, with 44±11 mo between early and late follow-up). Eighty-two patients (25.5%) lost >5° of PI-LL correction (mean loss 10±5°). After exclusion of patients with instrumentation failure, 52 patients (Loss) with a mean loss of correction of 8.6±2.9 were compared to 222 controls with maintained LL correction. Demographics were similar between groups (age: 63 vs 61 p=0.15; Sex: 78.4% F vs 88.5% p=0.10). There were no significant differences in use of osteotomy, 3CO, and IBF between Loss and Maintained LL (all p >0.1). There was no significant difference in number of levels fused (11 vs 12 p=0.39), rod material (CoCr 59.4% vs 58.8% p=0.34), and BMP use (88.7% vs 84.6% p=0.41) but Loss had less supplemental rod use (5.8% vs 23.4% p=0.004). Comparison between Maintained and Loss revealed a similar PI-LL mismatch at pre-op (16.7±18.7 vs 20.9±18.3 p=0.14) and final postop (1.8±12.5 vs 5.1±14.5 p=0.11) but a significantly smaller PI-LL for Loss at early postop (0.6±12.8 vs -3.5±13.7 p=0.41). Distally, Maintained had a significant improvement in L4-S1 lordosis from preop to early postop (p=0.013), with no significant difference from early to final follow-up. In contrast, patients in the Loss group had no difference in L4-S1 from preop to early postop (p=0.14), but showed a significant loss of correction at 3 years (p <0.001). Proximally, significant increase between pre and early postop in L1-L4 (all p <0.001), with Loss and Maintained reaching similar early postop value (p=0.34). However, from early postop to final follow up, Loss showed a significant decrease in proximal lordosis (p <0.001) while Maintained did not (p=0.08), corresponding with a smaller absolute L1-L4 in the Loss cohort at 3 yrs (p=0.002). Screw orientation showed a significant decrease from early to late follow-up between the L1 and S1 screws for Loss (1.3±4.1 p=0.031), without any changes across the L4-S1 segment (-0.1±2.9 p=0.97). <h3>CONCLUSIONS</h3> Approximately a quarter of patients lose an average of 10° of their 6-week correction by 3 years. Lordosis is lost proximally through the instrumentation (ie, tulip/shank angle shifts and/or rod bending), but lost distally through bone "settling" through the instrumentation itself. The use of supplemental rods and avoiding sagittal overcorrection may help mitigate this loss. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.
Read full abstract