Study design:

Retrospective cohort.

Objective:

To determine if intraoperative on-table lumbar lordosis and segmental lordosis coincide with perioperative change in lordosis.

Summary of background data:

Improvements in sagittal alignment are believed to correlate with improvements in clinical outcomes. Thus, it is important to establish whether intraoperative radiographs predict postoperative improvements in lumbar lordosis or segmental lordosis.

Methods:

Electronic medical records were reviewed for patients ≥18 years old who underwent single- and two-level anterior lumbar interbody fusion (ALIF) with posterior instrumentation between 2016 and 2020. Lumbar lordosis, segmental lordosis, and the lordosis distribution index (LDI) were compared between pre-, intra-, and post-operative radiographs using paired t-tests. A linear regression determined the effect of subsidence on segmental lordosis and lumbar lordosis.

Results:

A total of 118 patients met inclusion criteria. Of those, 75 patients had one-level fusions and 43 had a two-level fusion. Lumbar lordosis significantly increased following on-table positioning (delta (Δ): 5.7°, P<0.001). However, lumbar lordosis significantly decreased between the intraoperative to postoperative radiographs at 2-6 weeks (Δ: -3.4°, P=0.001), while no change was identified between the intraoperative and >3 month postoperative radiographs (Δ: -1.6°, P=0.143). Segmental lordosis was found to significantly increase from the preoperative to intraoperative radiographs (Δ: 10.9°, P<0.001), but it subsequently decreased at the 2-6 week follow up (Δ: -2.7, P<0.001) and at the final follow up (Δ: -4.1, P<0.001). On linear regression, cage subsidence/allograft resorption was predictive of the Δ segmental lordosis (β=0.55; 95% CI, 0.16-0.94; P=0.006), but not lumbar lordosis (β=0.10; 95% CI, -0.44-0.65; P=0.708).

Conclusion:

Early post-operative radiographs may not accurately reflect the improvement in lumbar lordosis seen on intraoperative radiographic imaging, but they are predictive of long-term lumbar sagittal alignment. Each millimeter of cage subsidence or allograft resorption reduces segmental lordosis by 0.55°, but subsidence does not significantly affect lumbar lordosis.

Levels of evidence:

4.