Epidural Steroid Injections for Management of Degenerative Spondylolisthesis: Little Effect on Clinical Outcomes in Operatively and Nonoperatively Treated Patients

Abstract

Background: Although epidural steroid injection (ESI) may provide pain relief for patients with degenerative spondylolisthesis in treatment regimens of up to 4 months, it remains unclear whether ESI affects crossover from nonoperative to operative management.

Methods: This retrospective cohort study analyzed 2 groups of surgical candidates with degenerative spondylolisthesis: those who received ESI within 3 months after enrollment (ESI group) and those who did not (no-ESI group). Annual outcomes following enrollment were assessed within operative and nonoperative groups (patients who initially chose or were assigned to surgery or nonoperative treatment) by using longitudinal mixed-effect models with a random subject intercept term accounting for correlations between repeated measurements. Treatment comparisons were performed at follow-up intervals. Area-under-the-curve analysis for all time points assessed the global significance of treatment.

Results: The study included 192 patients in the no-ESI group and 74 in the ESI group. The no-ESI group had greater baseline Short Form-36 (SF-36) Bodily Pain scores (median, 35 versus 32) and self-reported preference for surgery (38% versus 11%). There were no differences in surgical rates within 4 years after enrollment between the no-ESI and ESI groups (61% versus 62%). The surgical ESI and no-ESI groups also showed no differences in changes in patient-reported outcomes at any follow-up interval or in the 4-year average. Compared with the nonoperative ESI group, the nonoperative no-ESI group showed greater improvements in SF-36 scores for Bodily Pain (p = 0.004) and Physical Function (p = 0.005) at 4 years, Bodily Pain at 1 year (p = 0.002) and 3 years (p = 0.005), and Physical Function at 1 year (p = 0.030) and 2 years (p = 0.002). Of the patients who were initially treated nonsurgically, those who received ESI and those who did not receive ESI did not differ with regard to surgical crossover rates. The rates of crossover to nonoperative treatment by patients who initially chose or were assigned to surgery also did not differ between the ESI and no-ESI groups.

Conclusions: There was no relationship between ESI and improved clinical outcomes over a 4-year study period for patients who chose or were assigned to receive surgery for degenerative spondylolisthesis. In the nonsurgical group, ESI was associated with inferior pain reduction through 3 years, although this was confounded by greater baseline pain. ESI showed little relationship with surgical crossover.

Level of evidence: Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.

Fig. 1

Design of the study of the degenerative spondylolisthesis SPORT subgroup. Randomized and observational SPORT cohorts were combined for an “as treated” analysis. The present study only includes patients who did not receive ESI prior to enrollment in SPORT.

Fig. 2

Comparison of patient-reported outcomes (adjusted mean values with standard deviation) over time between patients who received ESI within 3 months after SPORT enrollment and those who did not. The “Surgery p-value” is for the difference between the ESI and non-ESI groups within the surgical cohort, while the “Non-operative p-value” is for that difference within the nonoperative cohort. The “Interaction p-value” is derived from the comparison of treatment (surgery versus nonoperative) between the no-ESI and ESI groups. All p values reflect 4-year area-under-the-curve averages. DS = degenerative spondylolisthesis.