Introduction: Symptomatic lumbar spinal stenosis can be treated with decompression surgery. A recent review reported that, after decompression surgery, 1.6-32.0% of patients develop postoperative symptomatic spondylolisthesis, and may subsequently be indicated for lumbar fusion surgery. The latter can be more challenging due to the altered anatomy and scar tissue formation. It remains unclear why some patients develop postoperative symptomatic spondylolisthesis, though some risk factors have been suggested. This study evaluates the association between key demographic, biological and radiological factors and postoperative symptomatic spondylolisthesis after lumbar decompression.  

Methods: This retrospective cohort study included patients who had undergone lumbar spinal decompression surgery between January 2014 and December 2016 at one of two specialist Spine Centres in the Netherlands or Switzerland. Patient characteristics and details of the surgical procedure were extracted from patient charts. Preoperative MRI-scans and X-rays of the lumbar spine were evaluated for multiple morphological and pathological characteristics. Patients with and without postoperative spondylolisthesis, confirmed with postoperative MRI-scans, were compared. Univariate and multivariable logistic regression analyses were used to identify statistical predictors of postoperative spondylolisthesis for single and multilevel surgeries.

Results: 1094 surgical levels in 741 patients, were included in the analyses. Preoperative X-rays were available for 488 patients, while MRI was available for all. ICCs for intraobserver and interobserver reliability of measurement of X-ray and MRI variables were all >0.60 (range intra 0.81-0.99, range inter 0.67-0.97). In the group of single level decompression surgery, 52/420 surgical levels (12.4%) developed postoperative symptomatic spondylolisthesis. In the group of multilevel decompression surgery, 48/685 surgical levels (7.0%) in 39/295 patients developed postoperative symptomatic spondylolisthesis. Multivariable logistic regression identified six significant independent risk factors in patients with single level decompression surgery: female sex (odds ratio (OR) 3.639, 95%CI 1.559-8.492), lower BMI (OR 0.868, 95%CI 0.789-0.956), Rheumatoid Arthritis (OR 3.601, 95%CI 1.075-12.067), preoperative spondylolisthesis (OR 7.124, 95%CI 3.194-15.888), and increased cross-sectional area of M. psoas (OR 68.155, 95%CI 4.505-1031.160). Facet degeneration grade 2 resulted in a lower risk compared to grade 3 (OR 0.419, 95%CI 0.182-0.963). Multivariable logistic regression identified one significant independent risk factor in patients with multilevel decompression surgery: preoperative spondylolisthesis (OR 38.857, 95%CI 16.499-91.515). The Hossner and Lemeshow test (P=0.937 in single level surgeries) and Nagelkerke R-square (P=0.315 in single level surgeries and P=0.396 in multilevel surgeries) showed a good model fit.
Discussion: Being female and having lower BMI, Rheumatoid Arthritis, higher facet degeneration grade, preoperative spondylolisthesis and an increased cross-sectional area of M. psoas were associated with a higher risk of having postoperative symptomatic spondylolisthesis after single level decompression surgery. Preoperative spondylolisthesis was associated with a higher risk of having postoperative symptomatic spondylolisthesis after multilevel decompression surgery. These associations can be used for shared-decision making when deciding for decompression surgery versus lumbar fusion surgery in patients with lumbar spinal stenosis.