Epidural steroid injections resulted in nearly twice the incidence of serious neurologic complications as epidural injections without steroids, according to a retrospective study, which also found soluble steroid injections to be just as safe as nonsteroid injections.
Since 2012, when the FDA issued a letter of warning, “steroids have not been approved for epidural injections,” said principal investigator Byungkwan Hwang, MD, a second-year resident in neurology at Texas Tech University Health Sciences Center, in Lubbock. “However, there was no epidemiologic evidence for this regulation.”
Dr. Hwang said the warning has confused many pain practitioners because most of them use steroids for epidural injections. “This warning has also raised liability and medicolegal issues,” said Dr. Hwang, noting that the same dilemma exists in his home country of South Korea, where there is no longer approval for epidural steroid injections.
The study, presented at the 2017 annual meeting of the American Academy of Pain Medicine (abstract 194), culled data from the Health Insurance Record Review and Assessment Service (HIRA) database for 2009 to 2013. During this period, 1,945,822 patients visited a clinic or hospital for at least one epidural injection. The total number of epidural injections was 9,980,627. The investigators created a retrospective cohort from the insurance claim data set and matched the two groups—steroid versus nonsteroid and particulate steroid versus soluble steroid—using a propensity score to adjust for measured confounders (Charlson Comorbidity Index, age and sex). Patients were excluded if they had neurologic complications before injection or had undergone a spine operation prior to 2010.
The investigators also identified three serious complications—stroke, spinal cord infarction and death—in admitted hospital patients within 24 hours after receiving an epidural injection at an outpatient clinic.
The study found that the complication rate of epidural steroid injections was higher than epidural injections without steroid: 1.38 compared with 0.75 per 100,000 person-days, respectively. “However, this is mainly due to a particulate steroid, methylprednisolone or triamcinolone, rather than a soluble steroid, dexamethasone or betamethasone,” Dr. Hwang said.
The incidence of complications from a particulate steroid injection was 1.52 compared with 0.84 per 100,000 person-days for a nonparticulate steroid injection.
In addition, cervicothoracic injection had a higher risk than lumbosacral injection.
Overall, epidural steroid injections represented nearly double the risk for serious complications (adjusted hazard ratio, 1.82; 95% CI, 1.58-2.22) as epidural injections without steroid. The same was true for epidural particulate steroid injections (adjusted hazard ratio, 1.83; 95% CI, 1.53-2.42) compared with nonparticulate steroid injections.
“Our study confirms the previous result of animal studies, which show that particulate steroids can cause neurovascular complications by obstruction of arterioles,” Dr. Hwang said.
Because the study found particulate steroid to have a higher risk than soluble steroid for epidural injection, “to avoid complication, soluble steroid is recommended for epidural injection,” Dr. Hwang said. “Also, pain physicians should be more careful when planning a cervicothoracic injection than a lumbosacral injection, as the former is riskier.”
To the best of Dr. Hwang’s knowledge, this is the first and only study that has tried to estimate true complication rates for epidural injections. However, a limitation of the study is that the investigators had to define both epidural injections and nerve root blocks as epidural injections because it is difficult to differentiate the two using insurance claim data. “Also, because this study used insurance claim data, we could not differentiate an interlaminar injection from a transforaminal injection, since insurance claim data does not contain specific information about the mode of injection,” Dr. Hwang said.
Still, based on the study’s findings, “I believe the FDA should approve epidural steroid injections because soluble steroids did not increase risk of serious neurological complications,” Dr. Hwang said. “I would like the FDA warning letter to focus more on particulate steroids, rather than all kinds of steroids, when it comes to neurological complications from epidural injections.”