An expensive and common procedure treating back pain from osteoporosis turns out to be ineffective, no better than a fake surgery, two major studies have found.
Called vertebroplasty, the procedure involves injecting surgical cement into the vertebrae of patients to relieve pain from spinal fractures due to osteoporosis. Some 750,000 Americans get these fractures each year. Vertebroplasty and a related procedure called kyphoplasty have soared in popularity in recent years after doctors reported apparently miraculous results after trying it on small numbers of patients.
But early studies reporting positive results for the vertebroplasty procedure did not have control groups, making it unclear whether the purported benefits were simply due to the placebo effect or the natural tendency of chronic pain to subside over time.
The new studies, published in the New England Journal of Medicinetoday, tested vertebroplasty in a far more rigorous way. They randomly assigned patients to get either vertebroplasty or a fake procedure in which patients were injected with local anesthetic but no cement. Researchers followed patients for a few months to see if there was any difference in pain or functioning.
The results? “People who got a fake procedure did just as well as people who got the real thing,” says back pain researcher Richard Deyo of Oregon Health and Science University, who was not involved in either study. Both groups improved, but the placebo patients improved just as much as those who got the real thing.
“It is absolutely shocking,” says Mayo Clinic interventional neuroradiologist David Kallmes, who led one of the studies in conjunction with researchers at the University of Washington. “Vertebroplasty as currently practiced in this country and around the world doesn’t seem to work.”
Kallmes did the study because he always though the results reported for vertebroplasty seemed too good to be true–everyone got good results no matter how much cement was injected or what technique was used. And the mechanism through which vertebroplasty provided pain relief was a bit of a mystery. He figured if vertebroplasty was as good as promised, it would be easy to prove. He never expected results as stone cold negative as they were.
Kallmes says it is too early “to kill the procedure” but that doctors need to stop doing it routinely outside research trials. Some researchers went further.
“I don’t think they should be done” any more outside of research studies, saidMonash University rheumatologist Rachelle Buchbinder, who led the second study, done in Australia. “There are some risks, and we did not demonstrate any benefits.”
Side effects of vertebroplasty can include cement leakage and the possibility that it could increase the risk of future fractures.
Johnson & Johnsonand Stryker are among the companies that make materials using vertebroplasty.
Each year Medicare pays for roughly 40,000 vertebroplasty procedures at a cost of $3,000 or more, says Dartmouth-Hitchcock Medical Centerorthopedic surgeon James Weinstein. If the new studies’ findings are correct, much of that money has been wasted. He estimates that at least that many more expensive kyphoplasty procedures are done.
They findings are likely to provide ammunition to legislators and doctors pushing for more “comparative effectiveness” studies to be done as part of health care reform. The goal is to get better data on which medical procedures and devices provide value for the money. “The results are a very big deal for all of health care,” says Dartmouth’s Weinstein. This is what the president is asking us to do–find the answer and pay for value. If things work lets pay for it, if it doesn’t, lets not.”
Researchers said the negative vertebroplasty findings raise questions about the effectiveness of a related procedure called kyphoplasty, in which the cement is injected after a small balloon is inflated to restore the shape of the broken vertebra. In 2007, device maker Medtronicspent $3.9 billion to buy the company that makes kits used to do the procedure.
“There is no plausible reason why it would work where vertebroplasty doesn’t work,” said Buchbinder, who calls for blinded trials looking at the benefits of kyphoplasty.
Adds University of Washington neuroradiologist Jeffrey Jarvik, who worked on the American trial: “It absolutely calls into question the benefit of kyphoplasty as well There is no strong evidence” that kyphoplasty is better, even though it costs more.
Medtronic says kyphoplasty is different from vertebroplasty; among other things, it aims to not just relieve pain but to restore lost height. It points to a 300-patient trial published in the Lancet this year that found kyphoplasty was better than nonsurgical care for relieving spine fracture pain and improving function.
But that trial was not blinded–patients knew which treatment group they were in. This is an issue because the mere knowledge you are getting the latest in high-tech procedures can make you feel better. Studies have found that fake devices produce bigger placebo effects than placebo pills.
Medtronic argues that a sham surgery trial for kyphoplasty is not needed because the Lancet trials followed patients for a full year, and that a placebo effect would be unlikely to persist for this long. “Placebo effects have never been shown to last as long as one year,” the company said in a statement.
Vertebroplasty was first pioneered in France and has been done in the U.S. since researchers at the University of Virginia reported good results in the 1990s. “It is a very simple procedure to do,” says Kallmes. “Everyone picked up on it and stared doing it. Kyphoplasty, which adds in the balloon, became popular more recently.
But the new studies testing vertebroplasty’s efficacy took years to perform because doctors were reluctant to enroll patients. They were convinced that vertebroplasty worked. The Mayo Clinic/University of Washington trial had to expand to locations worldwide to attract enough patients. It eventually got 131 patients to agree. The Australia trial only enrolled 78 participants, far less than originally planned.
Some doctors even said the trials were unethical. One 2005 editorial in theJournal of Neuroradiology predicted that a randomized trial of vertebroplasty “would almost assuredly merely confirm the data that we already have” and thus “does not seem necessary or ethically justifiable.”
Patients may still ask for vertebroplasty no matter what the results show. Sister Rogene Fox, an 81-year-old nun at Assisi Heights convent in Minnesota, was in the placebo group of the Mayo Clinic study after getting a painful vertebra fracture in 2007. The pain went away 10 weeks after getting the placebo surgery, she says. This year, when she had a second vertebral fracture, she went back to get the real thing. Her pain went away within 10 days. “For me it was better, because within 10 days to two weeks my pain was considerably relieved. I would go back for another one,” she says.