FDA Panel Warns Against Certain Back Pain Injections

Pauline Anderson

November 28, 2014

Epidural steroid injections (ESIs) should be contraindicated under certain circumstances in the management of back pain, the US Food and Drug Administration's (FDA's) Anesthetic and Analgesic Drug Products Advisory Committee has concluded.

After 2 days of presentations and discussions, committee members agreed that injecting a particulate steroid in the cervical region, using the transforaminal approach, increases the risk for sometimes serious and irreversible neurological adverse events, including stroke, paraplegia, spinal cord infarction, and even death.

However, although most members wanted this concern captured in a contraindication, some felt a warning in the product labeling is sufficient.

"We have heard a lot of evidence that particulate steroids using a transforaminal approach in the cervical region has significant risk associated with it," said panel member James Eisenach, MD, professor, anesthesiology and physiology and pharmacology, Wake Forest School of Medicine, Winston-Salem, North Carolina. "We have heard that from a theoretical point of view, we have heard it from animal data, and we have seen it in very dramatic reports of patient catastrophes.

"The additional benefit, if any, of a particulate vs a nonparticulate in this setting is very small, and the risk, although in raw terms and in uncertain terms is very small, it is catastrophic," added Dr Eisenach.

However, many committee members commented on the challenges of assessing sparse, but somewhat complex and sometimes contradictory, data in this field.

New Question

Originally, the question put to the panel was, "Do you recommend FDA add a contraindication to the labeling of injectable corticosteroids for the use of these products in epidural administration?" However, the question the panel eventually voted on was slightly altered: "Do you think there are any clinical situations for which a contraindication is warranted?"

At the end of the meeting, committee members voted 15 to 7 that there are clinical situations for which a contraindication to ESIs are warranted (there was a single abstention).

During further discussions, members elaborated on their vote. John Farrar, MD, associate professor of epidemiology, neurology and anesthesia, University of Pennsylvania, Pittsburgh, said he voted "yes," but added that he "feels quite strongly" that there should not be a contraindication for the soluble or nonparticulate form of steroids.

For Brian Bateman, MD, assistant professor, anesthesia, Brigham and Women's Hospital, Massachusetts General Hospital, and Harvard Medical School, Boston, Massachusetts, not only should transforaminal injections of a particulate steroid be contraindicated in the cervical region but there also should also be a warning against injecting particulate steroids into the thoracic region.

Some members who voted against a contraindication felt the statement presented to them was overly broad, but most did favor some kind of warning or limitation of the use of suspension or particulate steroids in the cervical region, using the transforaminal approach.

For S. Raymond Golish, MD, PhD, spinal surgeon, medical director of research, Jupiter Medical Center, Pam Beach, Florida, however, who voted against a contraindication, "the most worrisome practices are being eliminated in the clinical practice of medicine by increased awareness and leadership."

Some committee members wanted the current warning to include a cross-reference to any contraindication.

Committee chair Randall Flick, MD, director, associate professor, anesthesiology, Mayo Clinic Children's Center, Rochester, Minnesota, noted that preclinical studies that can help clarify the role of injected steroids in the pathogenesis of neurologic injuries should be carried out.

"It's important to note that these studies are relatively simple and could be funded with very little research investment," said Dr Flick. "Other clinical studies could better define both the benefit and potential risk of various applications of injected corticosteroids in the setting of epidural pain procedures."

Common Disorder

According to FDA background materials, back pain is the most common of all chronic pain disorders, with lifetime prevalence reported to be 54% to 80%. Epidural injections of corticosteroids to reduce inflammation is a common procedure in the management of spinal pain.

It is estimated that there are at least 9 million steroid injections a year in the United States alone, and that number is increasing as the population ages.

The FDA's Center for Drug Evaluation and Research (CDER) has never approved an injectable corticosteroid product administered via epidural injection, so this use, although common, is considered "off-label." The only available efficacy information, therefore, comes from published literature, which has limited descriptions of the protocols and may lack original data sets.

The FDA does, however, regulate medications that are injected during this procedure. FDA-approved injectable corticosteroids include betamethasone acetate and betamethasone sodium phosphate (Celestone and Soluspan, Merck), dexamethasone sodium phosphate (generic only), hydrocortisone sodium succinate (Solu-Cortef, Pharmacia and Upjohn Company), methylprednisolone acetate (Depo-Medrol, Pharmacia and Upjohn Company), methylprednisolone sodium succinate (Solu-Medrol, Pharmacia and Upjohn Company), triamcinolone acetonide (Kenalog-10, Kenalog-40, E.R. Squibb & Sons, LLC), and triamcinolone hexacetonide (Aristospan, Sandoz Inc).

Epidural steroid injections are performed in the office setting, hospitals, and surgery centers by a variety of health providers, including anesthesiologists, nurse anesthetists, radiologists, neurologists, physiatrists, and surgeons. These experts often use fluoroscopy to guide the needle insertion and confirm needle location.

Safety Concerns

The FDA has been evaluating the issue of serious neurological events with ESIs since 2009. Its Safe Use Initiative group convened an expert panel to discuss the topic in 2011. In July 2014, the FDA required a class warning to be added to product labels of injectable corticosteroids. The warning says serious neurologic events have been reported with and without use of fluoroscopy.

The FDA is now taking the regulatory approach in asking the committee for advice on further labeling options, which could include a black box warning, changes to the general warning, or a contraindication.

The agency has been criticized for including all corticosteroids in its general warning, when reports in the literature of an increased risk for serious neurologic events pertain to particulate corticosteroids compared with nonparticulate formulations of corticosteroids. The FDA does not use these terms, instead preferring "suspension" and "solution."

The committee heard animal evidence showing that particulate forms of therapy may clump together and block arteries.

"Why would anyone want to have sludge injected into their artery that goes to their brain?" wondered panel member Jeffrey Kirsch, MD, professor and chair, Department of Anesthesiology and Perioperative Medicine, Oregon Health and Science University, Portland.

Injection Approaches

Steroid injection approaches can be interlaminar, caudal, or transforaminal. The choice depends on the type of pain being treated, the anatomic area of steroid deposition, and the level of comfort of the treating clinician.

In interlaminar injections, clinicians insert the needle between (midline) or adjacent and parallel to (paramedian) the spinous processes of two vertebrae and traversing the ligamentum flavum to the epidural space. A caudal injection is placed through the sacral hiatus, with the needle advancing into the caudal canal to a point 1 to 2 cm beyond and posterior to the ventral place of the sacrum.

In the case of a transforaminal injection, the clinician inserts the needle, advances it toward the lower edge of the transverse process, near its junction with the superior articular process, and then directs it toward the superior, lateral, anterior aspect of the neural foramen. The goal is to position the needle in the triangle composed of the pedicle, the nerve root, and the vertebral body.

Committee members heard that although somewhat scant, the research seems to indicate that these injections do have an effect, at least in the short-term, in some patients. However, according to presenter Steven Cohen, MD, professor, anesthesiology and critical care medicine, Johns Hopkins School of Medicine, Baltimore, Maryland, the evidence for long-term benefit is conflicting.

"It may provide long-term relief in some patients and prevent surgery by permitting the body time to heal itself," he said, adding that multiple injections may provide long-term relief.

Dr Farrar agreed the evidence suggests the effect is "on the small side, and short-lived in the sense that if you look 6 months or 12 months down the line, the benefit may not exceed patients who don't get the injection."

Some attendees noted that these injections speed resolution of radicular pain, which typically gets better anyway.

"It's a pain that has a favorable natural history," said presenter James Rathmell, MD, chief, Division of Pain Medicine, Massachusetts General Hospital, and professor, anesthesia, Harvard Medical School. "Yes, we can speed resolution, yes they will get better, but if patients do nothing, they are very likely to get better as well, and that's the rough part about weighing this."

As for adverse neurological outcomes, the evidence suggests these are rare, although the exact incidence is unknown; one expert presenter put the risk for accidental injury at less than 1 in 10,000. Some members stressed, however, that the number of adverse events may be underreported.

According to Sally Seymour, MD, deputy director for safety, Division of Pulmonary, Allergy, and Rheumatology Products, Office of Drug Evaluation, CDER, it is difficult to separate the risk of the procedure from the risk of the injected drug. She noted that most steroids should not be used for intrathecal injections.

Although rare, the neurological outcomes that do occur can be catastrophic and irreversible, and they can occur in patients who were relatively healthy to begin with, the committee heard.

Members agreed that there is no single pain condition that carries higher neurological risks, but also that the evidence suggests that injections into the cervical region, as opposed to the lumbar area, are relatively risky. They heard that the risk for accidental injury in the arterial system is greater in this location.

"The human body is variable, and that means that occasional needles are going to go in the wrong place for a variety of reasons," Dr Farrar said.

At least one member noted that the skill level of clinicians doing these injections varies, and that this may account for a substantial part of the risk associated with the procedure.

Some attendees pointed out that there are also risks to alternative modes of pain management, including nonsteroidal anti-inflammatory drugs and opiates.

As for injection method, the committee heard evidence pointing to the transforaminal approach being less safe than the other approaches.

Some members emphasized the importance of patient selection. Edward Michna, MD, director, Pain Trials Center, Brigham & Women's Hospital, and assistant professor, Harvard Medical School, pointed out that patients come for injections once or twice a year for many years, and that this allows them to continue to function.

"The problem is that people are getting three epidurals, whether they get pain relief or not. You have to evaluate whether this patient is a good responder," said Dr Michna.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.