A widely used method of treating a common cause of back and leg pain – steroid injections for spinal stenosis – may provide little benefit for many patients, according to a new study that experts said should make doctors and patients think twice about the treatment.
Hundreds of thousands of injections are given for stenosis each year in the United States, experts say, costing hundreds of millions of dollars.
But the study, the largest randomized trial evaluating the treatment, found that patients receiving a standard stenosis injection — which combine a steroid and local anesthetic — had no less pain and virtually no greater function after six weeks than patients injected with anesthetic alone. The research, involving 400 patients at 16 medical centers, was published Wednesday in The New England Journal of Medicine.
“Certainly there are more injections than actually should happen,” said Dr. Gunnar Andersson, chairman emeritus of orthopedic surgery at Rush University Medical Center in Chicago, who was not involved in the research. “It’s sort of become the thing you do. You see this abnormality on the M.R.I. and the patient complains, and immediately you send the patient for an epidural injection.”
Some people can still benefit from injections, Dr. Andersson said, but now physicians “will be more cautious” and patients should ask, “Should I really do this?’”
Mostly, steroid injections are safe, carrying small risks of infection, headaches and sleeplessness. But in April, the Food and Drug Administration warned they may, in rare cases, cause blindness, stroke, paralysis or death, noting injections have not been F.D.A.-approved for back pain and their effectiveness has “not been established.”
In an extreme case in 2012, 14,000 people received contaminated steroid injections, causing 751 cases of fungal meningitis and other infections,including 64 deaths. Some of the exposed patients were being treated for spinal stenosis, said Dr. Tom Chiller, a deputy chief at the Centers for Disease Control and Prevention.
Often caused by wear and tear, spinal stenosis occurs when spaces within the spine narrow, putting pressure on nerves and causing pain or numbness in the back and legs. More than a third of people over 60 have some narrowing of the spinal canal, research suggests.
Steroid injections, which reduce inflammation, are often tried when physical therapy or anti-inflammatory medication fails, with the aim of avoiding expensive surgery, which itself is inappropriate for some patients. . Some insurance companies require injections before approving surgery.
The new study provides evidence to tell some patients, “This probably isn’t going to work very well for you,” said Dr. Ray Baker, past president of the North American Spine Society and the International Spine Intervention Society, who was not involved in the study. And because some participants received two injections without greater benefit, he added, “it strongly speaks against the practice of performing multiple injections.”
But the research also leaves the options for some patients unclear.
“We don’t have a lot of good things in our toolbox for spinal stenosis,” Dr. Baker added. “We’re really stuck with a problem, especially with an aging population.”
Spinal injections are considered effective for other conditions, including herniated discs. But of the 2.2 million given annually to people on Medicare, more than half a million are for spinal stenosis, said Dr. Janna Friedly, professor of rehabilitation medicine at University of Washington and lead author of the new study. She said injections cost $500 to $2,000 each.
The study helps answer questions raised by the Spine Society and the Cochrane Collaboration, a group of medical experts. Both issued reviews last year finding insufficient evidence to recommend injections for some types of stenosis.
“If the benefit really isn’t there and you do the procedure more and more, then all you’re doing is compounding the risk,” said Dr. Christopher Standaert, a co-author of the study and a professor at University of Washington.
Dr. Scott Kreiner, co-chairman of the Spine Society’s evidence-based guidelines committee, said the new research should “influence future guidelines and recommendations and everything else.”
A rehabilitation medicine specialist in Phoenix who was not involved in the study, Dr. Kreiner says he will give fewer second injections and may refer some patients to surgery sooner. “This is probably a step toward eliminating or minimizing the use of epidural steroid injections for this problem,” he said.
Still, the research, funded by the federal Agency for Healthcare Research and Quality, leaves important questions unanswered.
Since every patient received injections, and both groups reported similar improvement six weeks later, researchers cannot tell if patients would do as well without any injections at all. Also unclear is whether the anesthetic, lidocaine, did anything helpful when injected alone. Some experts said the benefits patients reported seemed larger than typical placebo effects.
“To me it’s unlikely that the lidocaine has a long-term effect,” Dr. Friedly said, “but there are people who think that it could.”
The study also did not represent all types of stenosis, involving patients with central stenosis, not stenosis on one side, which Dr. Andersson said was more localized and therefore possibly more treatable by injection.
“Stenosis, it’s tricky,” said Dr. Standaert. “It can be caused by all sorts of things. I certainly have patients who have had steroids for years and done well,” but “if the pain isn’t primarily due to inflammation, then maybe it’s not going to help.”
Still, two patients recently tried lidocaine-only injections and “didn’t get better,” he said. He may now emphasize more physical therapy or posture exercises to try avoiding surgery.
Some results suggest the issue is complex. Three weeks after injection, patients receiving steroids reported slightly greater function and less pain than the lidocaine-only group. But by six weeks, the difference evaporated. When results were adjusted for how long patients had suffered from stenosis, the steroid group had tiny advantages in function, but too minimal to be relevant, Dr. Friedly said.
Afterward, patients and their doctors were told which injection they received, and then offered another injection, the same or different. William Johnson, 58, of Plano, Tex., a former Air Force and Postal Service employee now attending college, initially received the steroid. He said his pain “immediately went away,” he did not need another injection and, 18 months later, remains pain-free.
His doctor, Dr. Thiru Annaswamy of the Dallas Veterans Medical Center, said by email: “There are patients who clearly respond to steroid injection. However, it is unclear why some do, and others don’t.” And it was unclear, he said, if Mr. Johnson “may have responded to the lidocaine-only injection too.”
Another participant, Bonnie Merenstein, 73, a retired teacher in Denver, received lidocaine-only injections and requested another. Years ago, she said, steroid injections provided minimal improvement, and before the study, “I really could not walk for more than eight minutes without my legs going numb.”
Afterward, numbness lessened, grocery shopping and museum-going became easier, and she recently biked and canoed with her granddaughter.
“I believe that the lidocaine may have been as effective as a steroid,” she said.