Abstract
Mechanical failure of spine posterior fixation in the lumbar region Is suspected to occur more frequently when the sagittal balance is not properly restored. While failures at the proximal extremity have been studied in the literature, the lumbar distal junctional pathology has received less attention. The aim of this work was to investigate if the spinopelvic parameters, which characterize the sagittal balance, could predict the mechanical failure of the posterior fixation in the distal lumbar region.
All the spine surgeries performed in 2017-2019 at Rizzoli Institute were retrospectively analysed to extract all cases of lumbar distal junctional pathology. All the revision surgeries performed due to the pedicle screws pull-out, or the breakage of rods or screws, or the vertebral fracture, or the degenerative disc disease, in the distal extremity, were included in the junctional (JUNCT) group. A total of 83 cases were identified as JUNCT group. All the 241 fixation surgeries which to date have not failed were included in the control (CONTROL) group. Clinical data were extracted from both groups, and the main spinopelvic parameters were assessed from sagittal standing preoperative (pre-op) and postoperative (post-op) radiographs with the software Surgimap (Nemaris). In particular, pelvic incidence (PI), sagittal vertical axis (SVA), pelvic tilt (PT), T1 pelvic angle (TPA), sacral slope (SS) and lumbar lordosis (LL) have been measured.
In JUNCT, the main failure cause was the screws pull-out (45%). Spine fixation with 7 or more levels were the most common in JUNCT (52%) in contrast to CONTROL (14%). In CONTROL, PT, TPA, SS and PI-LL were inside the recommended ranges of good sagittal balance. For these parameters, statistically significant differences were observed between pre-op and post-op (p<0.0001, p=0.01, p<0.0001, p=0.004, respectively, Wilcoxon test). In JUNCT, the spinopelvic parameters were out of the ranges of the good sagittal balance and the worsening of the balance was confirmed by the increase in PT, TPA, SVA, PI-LL and by the decrease of LL (p=0.002, p=0.003, p<0.0001, p=0.001, p=0.001, respectively, paired t-test) before the revision surgery. TPA (p=0.003, Kolmogorov-Smirnov test) and SS (p=0.03, unpaired t-test) differed significantly in pre-op between JUNCT and CONTROL. In post-op, PI-LL was significantly different between JUNCT and CONTROL (p=0.04, unpaired t-test). The regression model of PT vs PI was significantly different between JUNCT and CONTROL in pre-op (p=0.01, Z-test).
These results showed that failure is most common in long fused segments, likely due to long lever arms leading to implant failure. If the sagittal balance is not properly restored, after the surgery the balance is expected to worsen, eventually leading to failure: this effect was confirmed by the worsening of all the spinopelvic parameters before the revision surgery in JUNCT. Conversely, a good sagittal balance seems to avoid a revision surgery, as it is visible is CONTROL. The mismatch PI-LL after the fixation seems to confirm a good sagittal balance and predict a good correction. The linear regression of PT vs PI suggests that the spine deformity and pelvic conformation could be a predictor for the failure after a fixation.
