Instability and instrumentation failures after a PSO: a finite element analysis.

STUDY DESIGN: Finite element analysis. BACKGROUND DATA: Pedicle subtraction osteotomy (PSO) is associated with a high rate of mechanical complications and implant failures. The biomechanical reasons for these failures are unclear. OBJECTIVES: Using finite element analysis (FEA): to analyze the biomechanical instability after a PSO, to compare the effect of constructs with different rod contours and analyze the mechanical forces acting on these constructs to explain the mechanisms of failure. METHODS: A 3D validated FE model of the spine from L1 to the sacrum was used. The model was modified to simulate a PSO of L4 in different situations: healthy, high dehydrated and completely degenerated discs. Loads were applied and range of motion (ROM) was measured. Pedicle screw constructs from L2 to S1 with different rod contours were added to the most instable scenario. Bending, torsion, shear moments and stress were measured. RESULTS: PSO alone had a moderate impact on the ROM of basic movements (flexion, extension and lateral bending). Secondary motion (torsion) in lateral bending increased 200 %. Greatest increase in ROM was observed with the PSO and degenerated discs. Secondary motion (torsion) in lateral bending increased +625 %. The instability after a PSO is rotational. Mean reduction of ROM was 95 % for all constructs tested. Rod contour affected the location of bending moments and stress. Sharp angle bend showed maximum bending moments (2,208 Nmm) and stress at the PSO level. Smooth contour of the rod showed maximum bending moments (1,940 Nmm) and stress at the sacral connection. Anterior support below the PSO reduced bending moments along the rod (-26 %). CONCLUSION: The instability observed after a PSO is mainly rotational and increases with disc degeneration. Shape of rod contour affects the location of maximum stress in the constructs. These findings may explain different instrumentation failures.

Complications and risk factors of primary adult scoliosis surgery: a multicenter study of 306 patients.

STUDY DESIGN: A multicentric retrospective study on primary adult scoliosis patients operated on between 2002 and 2007. A 3-step statistical analysis was performed to describe the incidence of complications, the risk factors, and the reoperation risk with survival curves for the entire cohort. OBJECTIVE: To describe complication rate and risk factors as well as survival curves associated with adult primary scoliosis surgery in patients aged 50 years or older. SUMMARY OF BACKGROUND DATA: Adult deformity surgery is classically associated with a high rate of complications. The identification of risk factors for developing such complications is consequently of major interest as well as survival curves that can provide useful information on reoperation risks. Although many reports exist in the literature, the cohorts analyzed are often heterogeneous and the actual prevalence of complications varies widely. This study represents to our knowledge the largest series on adult patients aged 50 years or older operated for the first time for lumbar or thoracolumbar scoliosis and excluding every other possible diagnosis. METHODS: A retrospective review of prospectively collected data from 6 centers in France. A total of 306 primary lumbar adult or degenerative scoliosis patients older than 50 years undergoing surgery between 2002 and 2007 were included. Demographics, comorbidities, x-ray parameters, surgical data, and complications were analyzed. Statistical analysis was performed to obtain correlations and risk factors for developing complications. Reoperation risk was calculated with Kaplan-Meier survival curves. RESULTS: A total of 306 patients aged 63 years (range, 50-83), with 83% women. Mean follow-up was 54 months. Mean Cumulative Illness Rating Scale score was 5 (range, 0-26). Main curve was 50° (range, 4-96) with apex between T12 and L2. Ten percent of patients had anterior surgery only, 18% had double anteroposterior approach, and 72% had posterior surgery only. Seventy-four percent (226 patients) had long fusions of 3 or more levels and 44% (134 patients) were fused to the sacrum. Forty percent (122 patients) had a decompression performed and 18% had an osteotomy. There were 175 complications for 119 patients (39%). No cases of death or blindness were reported. General complication rate was 13.7%, early infection occurred in 4% (12 patients), and late infection occurred in 1.2%. Neurological complications were present in 7% with 2 cases (0.6%) of late cord-level deficits and 12 reoperations (4%). Prevalence of mechanical complications was 24% (73 patients), with 58 patients (19%) needing a reoperation. Risk factors for mechanical or neurological complications were number of instrumented vertebra (P ≤ 0.01) fusion to the sacrum (P ≤ 0.001), pedicle subtraction osteotomy (PSO) (P = 0.01), and a high preoperative pelvic tilt of 26° or more (P ≤ 0.05). Kaplan-Meier survival curves showed reoperation risk of 44% at 70 months. Long fusion risk was 40% at 50 months and fusions to the sacrum reoperation risk was 48% at 49 months. CONCLUSION: Overall complication rate was 39%, and 26% of the patients were reoperated for mechanical or neurological complications. Risk factors include number of instrumented vertebra, fusion to the sacrum, PSO, and preoperative pelvic tilt of 26° or more. There is a 44% risk of a new operation in the 6-year-period after the primary procedure.