The Relationship Between Intraoperative Lumbar Lordosis Following Lateral Decubitus Position Spinal Fusion Surgery and Postoperative Standing Lumbar Lordosis

Introduction: Lumbosacral alignment and spinal balance are important in long-term patient success following spinal fusion surgeries. Pre- and intraoperative spinal alignment planning solutions have been developed to predict postoperative alignment. However, there is question of how well intraoperative measurements, taken in the lateral decubitus (LD) position, correlate with true standing measurements. This study aims to characterize the relationship between intraoperative and standing postoperative lumbar lordosis measurements.

Methods: A post-hoc review of a prospectively collected database at a single institute was undertaken. A total of 462 consecutive patients in the LD position treated with XLIF and/or ALIF were included. Patient demographic, radiographic (alignment measures of pelvic incidence (PI), pelvic tilt (PT), and lumbar lordosis (LL)), and treatment variables were collected and assessed for their relationship between intra- and postoperative LL.

Results: Mean patient age was 64.7 years (±11.3 years), mean body mass index was 29.3, and 232 (50.2%) patients were female. Most common diagnoses were spondylolisthesis, 276 patients (59.7%), and degenerative disc disease, 105 patients (22.7%). Majority of patients were treated as a single level (62.1%) construct, with two, three, and four level surgeries performed in 125, 29, and 21 patients, respectively. Preoperative PI, PT, and LL were 56.5°, 19.8°, and 51.9°, respectively. Intraoperative LL was 55.1° (±10.4°) and postoperative LL was 53.9° (±10.5°). Average change between intra- and postoperative LL was 1.2° (±7.3°). Mean postoperative PT was 18.7° (±7.4°).
There was moderate correlation between intraoperative LL and postoperative LL, r2=0.57, p< 0.001. Factoring for sex, BMI, age, diagnosis, number of levels treated, preoperative PI, and PT did not materially change the correlation between intra- and postoperative LL.

Conclusion : While the change from intra- to postoperative LL was modest (1.2°), there was a high degree of individual variability (7.3° standard deviation), with only a moderate correlation between intra- and postoperative LL. This suggests less than 60% of the variability between intra- and postoperative LL is accounted for by intraoperative LL. The model did not improve further with other variables. Thus, intraoperative LL was only a moderate predictor of postoperative standing LL, and further study is necessary to better predict postoperative LL from an intraoperative measurement.