Epidural Steroid Injections Ineffective for Lumbar Spinal Stenosis

S. Andrew Josephson, M.D.


AccessMedicine from McGraw-Hill 

Lumbar spinal stenosis is characterized by degenerative changes leading to narrowing of the spinal canal as well as compression of multiple nerve roots. The progressive pain and weakness of the legs that can result remains a common cause of disability, especially in the elderly. While surgical treatment has been shown to be effective, many clinicians begin with more conservative measures including epidural corticosteroid injections. These epidural injections have increased in frequency in the United States, and it is now estimated that over 2 million are performed annually in Medicare patients alone. Friedly and colleagues (2014) aimed to examine the efficacy of these injections in a randomized trial.

The authors conducted a double-blind, randomized controlled trial at 16 sites in the United States. Eligible patients were 50 years of age or older, with evidence of central lumbar spinal stenosis on imaging along with a modest or severe pain rating. Those with a history of previous lumbar spine surgery or those who had received epidural steroid injections within the last 6 months were excluded. Patients were randomized to injections with lidocaine plus corticosteroid or lidocaine alone. The primary outcomes examined at 6 weeks were scores on a standardized disability questionnaire and ratings of average pain over the past week.

A total of 400 patients were enrolled in the trial with similar baseline characteristics in the two intervention groups except that the lidocaine alone group had a significantly shorter duration of pain prior to study entry. At 6 weeks, both groups improved their standard disability scores but there was no significant difference between the two (adjusted difference, –1.0 points; 95% confidence interval, –2.1 to 0.1; p = .07). Similarly, there were no significant differences found in reported pain between the two groups at 6 weeks. At the 3-week time period, there was indeed a significant difference in disability favoring the corticosteroid group, but this was of a magnitude not thought to be clinically important.

A prespecified secondary analysis looking at different technical approaches to the injection (interlaminar vs transforaminal) also demonstrated no differences. Other scales of physical and mental function that were examined were generally not significantly different at 6 weeks between the groups.

Total adverse effects were not found to be significantly different between the two groups (21.5% in the glucocorticoid group vs 15.5% in the lidocaine alone group), although the adverse event rate on average per person was significantly greater in the corticosteroid group (p = .02). At both 3 and 6 weeks, the authors found that a significantly higher proportion of patients in the corticosteroid group had low morning serum cortisol levels, suggesting systemic absorption of the drug leading to adrenal suppression.

This study was performed without the use of a sham injection group, but since lidocaine alone is not thought to have lasting effects (e.g., weeks), the authors felt that this was a suitable placebo. The failure to demonstrate efficacy of epidural steroid injections for lumbar spinal stenosis suggests that this extremely common practice does not stand on firm evidence. Furthermore, some insurers currently insist that a patient must “fail” epidural steroid injections prior to authorizing a surgical treatment; this policy likely should stop. Unless additional data become available to the contrary, clinicians should not recommend epidural steroid injections to their patients with lumbar spinal stenosis.