Factors associated with transforaminal epidural steroid injection efficacy and prognosis for symptomatic treatment of lumbar stenosis associated with degenerative scoliosis.

Background: Although transforaminal epidural steroid injections (TFESIs) have been very commonly used for symptomatic treatment in patients with lumbar spinal stenosis, there is a paucity of literature on their use in patients with stenosis associated with degenerative lumbar scoliosis. There have been various published studies on parameters affecting the outcome of TFESIs, however there are no studies focusing on the effects of anatomical parameters associated with lumbar scoliosis on the outcome of these injections. The primary goal of this study is to assess anatomical and structural factors affecting efficacy of TFESI in the treatment of patients with lumbar degenerative scoliosis. A secondary goal of this study is to assess whether positive response with TFESI could predict better outcomes of subsequent lumbar surgery. Methods: This is a retrospective cohort study involving patients who were treated at an academic medical center. The patient sample is made of 149 patients with degenerative lumbar scoliosis who underwent TFESI during the time period of 2018 to 2023 for symptoms of axial and radicular pain. 19 of these patients underwent subsequent lumbar surgery. Scoliosis was defined as having a Cobb's angle of at least 20 degrees. The primary outcome measure is TFESI efficacy, assessed by the difference in pain on the numerical rating scale (NRS, score 0-10) for patients between score before TFESI and score at minimum 2 weeks after the injection. A minimal clinically important difference (MCID) was defined as net change in NRS ≥2.0 to identify the patients with a clinically significant reduction in pain level. A secondary outcome measure is subjective clinical improvement or lack thereof for patients after lumbar surgery. Variables analyzed to assess TFESI efficacy include laterality of injections with respect to the primary lumbar curve, distance of an injection to the apex of the primary lumbar curve, severity of Cobb's angle, severity of stenosis, and degree of spinal lateral listhesis. Lumbar surgery success was compared between patients who had either previously achieved or had not achieved MCID with TFESI. Statistical analyses included Chi-Square tests, Fisher exact tests, and a linear regression. Results: We found an improvement in TFESI efficacy for injections completed on the convex-side of the primary scoliosis curve over concave-side when patients had radicular pain at the L5-S1 spinal level (likelihood ratio with convex-side injection 0.21, likelihood with concave-side injection 0.80; p = .037). Trends were also observed for convex-side injections overall and for lower levels of lateral listhesis over higher levels to be associated with improved efficacy, and there was a 100% positive predictive value for prior successful TFESI to predict lumbar surgery success. No significant impact of severity of Cobb's angle nor severity of stenosis were noted on the outcomes with the injections. Conclusions: This study, the first of its kind, suggests that for symptomatic treatment of axial and radicular pain in the treatment of degenerative lumbar scoliosis associated with stenosis, TFESIs performed on the convex side of the primary lumbar scoliosis curve provide more symptomatic relief than concave side injections in treating radicular pain in general including at the L5-S1 level and that a history of symptom relief from TFESI is a positive predictor of outcomes for future lumbar surgery.