Risk factors for L5 pedicle fractures after single-level posterior spinal fusion.

ABSTRACT

BACKGROUND CONTEXT Pedicle fractures are a rare but potentially devastating complication of posterior instrumented spinal fusion (PSF). Preoperative awareness of the possible risk factors may help prevent these fractures by modifying the surgical plan. However, the risk factors have not yet been identified. PURPOSE: To determine the preoperative parameters associated with postoperative L5 pedicle fracture after L4/5 PSF. STUDY DESIGN: Case control study. PATIENT SAMPLE Patients undergoing L4/5 PSF at a single academic institution between 2014 and 2020. OUTCOME MEASURES: Occurrence of postoperative L5 pedicle fracture. METHODS: Of 253 patients (femalemale, 145108) undergoing L4/5 PSF from 2014 to 2020, patients with postoperative L5 pedicle fractures were identified retrospectively as "cases" (n = 8, all female, age 70 ± 10.7 years). As a control group all remaining patients with a follow-up of more than 12 months were allocated (n = 184, 104 females, age 64.27 ± 13.00 years). In all but 16 cases, anterior support with transforaminal or posterior interbody fusion was performed. Demographic and clinical data (body mass index (BMI)), surgical factors, and comorbidities) were compared. Radiological assessment of spinopelvic parameters was performed using pre- and postoperative standing lateral radiographs. RESULTS: The overall incidence of L5 pedicle fractures after L4/5 spinal fusion was 3.16%, with a median time from index surgery to diagnosis of 25 days (range, 6-199 days) (75% within the first 32 days postoperatively). Patients with L5 pedicle fractures had higher pelvic incidence (PI) (71° ± 9° vs. 56° ± 11°; p=.001), sacral slope (SS) (45° ± 7° vs. 35° ± 8°; p=.002), L5 slope (30° ± 11° vs. 15° ± 10°, p=.001), L5 incidence (42° ± 14° vs. 26° ± 11°; p= .003), L1-S1 lumbar lordosis (LL) postop (57° ± 10° vs. 45° ± 11°; p=.006), and L4 -S1 LL postop (33° ± 7° vs. 28° ± 7°; p=.049) compared with the control group. Pelvic tilt and PI- LL mismatch were not significantly different. Female gender was a significant risk factor for L5 pedicle fractures (p=.015). BMI (kg/m2) was statistically equal in patients with or without pedicle fractures (28.37 ± 5.96 vs. 28.53 ± 16.32; p=.857). There was no significant difference between the groups for approximative bone mineral density assessment (Hounsfield units; 113 ± 60 vs. 120 ± 43; p=.396) using the L3 trabecular region of interest (ROI) measurement. The correlation analysis demonstrated that most of the identified risk factors except for the postoperative L4-S1 lordosis show significant positive associations among each other. All eight patients in the fracture group underwent revision surgery, and the instrumented fusion was extended to the sacrum, with the addition of sacral-alar-iliac or iliac screws, in six cases. CONCLUSIONS: L5 pedicle fractures occurred in 3% of the patients after single level L4/5 PSF. Risk factors are female gender, higher PI, SS, L5 slope, L5 incidence, and LL postop but not high BMI. These findings can be used for surgical planning and decision of fusion levels.