This article is more than 5 years old. Information may no longer be current.
Debate concerning viability of vertebroplasty fueled by AAOS guidelines
Click Here to Manage Email Alerts
In late September, the American Academy of Orthopaedic Surgeons released new clinical practice guidelines that recommended against the use of Vertebroplasty — the use of bone cement to fill fractured vertebrae — in the treatment of osteoporotic vertebral compression fractures.
The guidelines, which specifically focused on those patients who were neurologically intact and presented with spinal compression fractures on imaging with correlated clinical signs and symptoms, were based primarily upon the evidence presented in two level 1 studies published in the New England Journal of Medicine (NEJM) that found vertebroplasty to be no more effective than sham procedures or placebo.
Controversy followed, however, as results of another clinical trial — Vertos II — indicated that vertebroplasty was, in fact, safe and effective when compared to conservative treatment in dealing with acute osteoporotic vertebral compression fractures. Criticisms of the NEJM studies, the Vertos II study and the American Academy of Orthopaedic Surgeons (AAOS) guidelines have since been leveled, with vertebroplasty’s viability as a treatment for osteoporotic vertebral compression fractures the primary point of debate.
“I think the most interesting question is: Is there a proper amount of clinical evidence to support the use of vertebroplasty? The answer is, undoubtedly, no,” said Stephen I. Esses, MD, chairman of the evidence-based research team that formulated the guidelines.
Esses cited the NEJM studies as being the primary support for the AAOS guidelines, but noted there was more to the decision.
|
Image: Brett Chisholm Photography |
“Let us also recognize there are other level 2 studies that did not use a sham procedure as a control — and reported similar results,” Esses told Orthopedics Today. “That is, after 6 weeks, the effect of vertebroplasty was not statistically or clinically significant. That observation goes out for 2 years.”
“When people say there is a lot of conflicting evidence … the only two level 1 studies and the three level 2 studies all have pretty consistent results,” he added.
The problem with the evidence that has already been presented, according to Scott D. Boden, MD, is that studies have not been focusing on narrow enough groups to justify such far-reaching conclusions.
“The study that I would want to do, if I were going to try and prove the efficacy of vertebroplasty in the way that I use it, would be to take patients who have had pain for 10 days and are not seeing any improvement in their pain level and/or are bedridden or not very mobile, and I would randomize those patients to vertebroplasty and non-vertebroplasty,” he said. “Not people who are 6 months out or 4 months out or 3 months out and up and around.”
Boden is not the only physician who believes the studies cited by the AAOS have not fully proven the case against vertebroplasty.
“I am quite familiar with the NEJM articles,” John G. Heller, MD, said in an interview with Orthopedics Today, “I believe that the study designs were inappropriate, and thus led to flawed results and conclusions.”
Heller cited operator skill and inclusion or exclusion criteria as being central to the differences in findings among various studies.
“I believe there is ample clinical evidence that cement augmentation procedures are both safe and effective for the treatment of painful [vertebral compression fractures],” he said.
Patient selection crucial
Inclusion and exclusion criteria could be critical, as there are those who insist the efficacy of vertebroplasty depends entirely upon the patient population. Criticisms of both the NEJM studies and the Vertos II trial have brought up the issue of patient selection. According to Boden, it cannot be ignored in judging the validity of vertebroplasty.
“I think anyone who is closely involved with the procedure knows that there is a subset of patients in whom vertebroplasty is an important and effective tool,” Boden said. “You only have to see one elderly lady go from being bedridden to walking off the table after a cement injection procedure to understand that, in certain circumstances, it is a very viable treatment. The key is to figure out where the risk/benefit ratio makes sense.”
Esses — who has stopped using vertebroplasty and taken to conservatively treating patients with osteoporotic vertebral compression fractures — acknowledged that patient selection has been a basis for some of the criticism against the studies, but refuted the point by noting that there has been no clinical evidence to prove the point.
“One of the criticisms was that vertebroplasty somehow works better for a certain fracture type than another, but there is absolutely not one iota of evidence that that’s the case,” he said.
|
Boden noted the lack of clinical evidence, but added that the AAOS guidelines pointed out a lack of clinical evidence and did not specifically say there was no situation in which vertebroplasty could be efficacious.
“They say the clinical evidence in the randomized trials does not really support the use of vertebroplasty, and that is a true statement,” he said. “That is not the same as saying that it should never be used or it does not work in certain subpopulations of patients.”
Heller, who does not believe the AAOS was justified in their assessment of vertebroplasty’s viability, said patient selection is crucial. “Inappropriate candidates will not benefit from either procedure,” he said. “Safety hinges on proper execution of the procedure with an awareness of the technical pitfalls.”
The NEJM articles
Eric D. Nabors, MD, told Orthopedics Today “the results of any procedure are dependent on patient selection.” He added that patient selection was at the heart of the different conclusions in the NEJM studies when compared to the Vertos II trial.
“The two New England Journal of Medicine articles included patients with symptoms of up to 1 year’s duration, and the study by Kallmes mandated 4 weeks of medical treatment — as opposed to Vertos II, which included only patients with pain duration of less than 6 weeks,” he said. “The NEJM studies were largely treating healed fractures. The Kallmes study mandated 4 weeks of medical treatment and excluded inpatients — the very patients most likely to benefit from vertebroplasty.”
Nabors noted that 32% and 44% of patients in the Buchbinder and Kallmes studies, respectively, were treated within 6 weeks while 100% of patients treated in the Vertos II trial were treated within 6 weeks. He also pointed out the high crossover rate within the two NEJM articles, as well as small patient population — 78 in the Buchbinder study, as compared to 202 in Vertos II.
Patient bias was another large problem with the NEJM articles, according to Nabors.
“By offering sham surgery as a treatment option, the NEJM studies were selecting away from potential treatable patients,” he said. “What patient with an acute, painful fracture would wish to enter a study with a 50% chance of undergoing a sham procedure?”
Esses described the NEJM articles as “very well-constructed level 1 studies — the type that had not yet been done, presented or published,” and said he was surprised at the degree of criticism directed toward the studies.
“The criticism obviously came from entities that have supported the use of vertebroplasty, and rather than come up with or show studies that were contrary to the results, all they did was criticize the methodology — and the criticisms, in my opinion, were not valid,” he said.
Esses noted that the AAOS guidelines specified that both NEJM studies did, in fact, have sufficient power to detect what was referred to as the “minimally clinically important difference in pain.” He also noted that while the crossover in the Kallmes study was criticized heavily, the Buchbinder study did not allow crossover but still came to the same conclusions.
“There was also criticism about low participation rates of eligible patients,” he added. “That is an issue of external validity and does not compromise the internal validity and conclusions.”
Vertos II
Esses stressed that while the findings of Vertos II have been pushed as convincing evidence in favor of using vertebroplasty, there are flaws within the study that need to be examined.
“Let us keep in mind that Vertos II was a level 2 study because there was no blinding or concealment of allocation,” he said. “Although that study was not used in the AAOS guidelines because it was published after our guideline, even if we used or included that study, it would not change the strength of our recommendation because this study would not be the best available evidence.”
According to Nabors, however, there is reason to assert that Vertos II offers more compelling evidence for vertebroplasty’s use than the NEJM articles offer against.
“Vertos II is a rigorously performed, randomized, controlled clinical trial comparing vertebroplasty to nonoperative therapy, and it showed a statistically significant improvement in pain for vertebroplasty — both initially and at 1-year follow-up,” he said. “Vertos II is a much more scientifically rigorous study than either of the NEJM studies … There is adequate evidence to support [vertebroplasty’s] use in clinical practice.”
Heller said he thinks it is important to find “optimal candidates” — such as those patients with acute or subacute fractures who are in severe pain and require assistance with daily activities — when considering the use of vertebroplasty.
“First and foremost, most patients with painful vertebral compression fractures will do just fine with supportive medical care and tincture of time,” he said. “The fractures will heal. Residual pain following successful fracture healing will not be improved by these procedures. The entry criteria for the NEJM studies tended toward patients who were unlikely to experience meaningful benefit from the intervention.”
Boden — who felt the inherent problem in each study was a lack of properly specific patient selection — said he does not plan on stopping the use of vertebroplasty, as he feels there are patients who can benefit.
“The question remains, I think: Are there subpopulations and scenarios where vertebroplasty should be the procedure or treatment of choice? I happen to believe that is the case,” he said. “Nobody has proven that the way I use cement injection procedures is ineffective. The academy’s statement does not prove it, the studies in the New England Journal of Medicine do not prove it. It has not been evaluated in the specific scenario in which I use it. Therefore, those studies have not changed my impression of it.”
Esses concluded by noting that even though the controversy surrounding vertebroplasty has brought it again to the forefront, it is important to remember the underlying issues that led the patient to seek treatment in the first place.
“All too often, when a patient presents with an osteoporotic vertebral compression fracture, treatment is directed solely toward the compression fracture — and sometimes no thought is given to treating the osteoporosis itself, or assessing what treatment the patient has had for osteoporosis,” he said. – by Robert Press
References:
- Buchbinder R, et al. A randomized trial of vertebroplasty for painful osteoporotic vertebral fractures. N Engl J Med. 2009;361:557-568.
- Kallmes DF, et al. A randomized trial of vertebroplasty for osteoporotic spinal fractures. N Engl J Med. 2009;361:569-579.
- Klazen CA, et al. Vertebroplasty versus conservative treatment in vertebral compression fractures (Vertos II): An open randomized trial. Lancet. 2010:25(376):1085-1092.
- Scott D. Boden, MD, can be reached at 59 Executive Park South, The Emory Spine Center, Ste. 3000, Atlanta, GA 30329; 404-778-7143; e-mail: sboden@emory.edu. He receives royalties and is a consultant for Medtronic Sofamor Danek and Osteotech, and receives institutional support from Synthes, Medtronic Sofamor Danek and DePuy.
- Stephen I. Esses, MD, can be reached at Scurlock Tower, 6560 Fannin St., Suite 1016, Houston, TX 77030; 713-333-4410; e-mail: sesses@mysurgeon.com. He has no direct financial interest in any products or companies mentioned in this article.
- John G. Heller, MD, can be reached at 59 Executive Park South NE, Ste. 3000, Atlanta, GA 30329-2208; 404-778-7112; e-mail: jhell02@emory.edu. He receives royalties from Medtronic; is on the speakers bureau for Abbott and Medtronic; receives institutional support from Abbott, Medtronic and Synthes; and owns stock in Medtronic.
- Eric D. Nabors, MD, can be reached at 363 Vanadium Road, Scott Township, PA 15243; 412-429-0880; e-mail: ednabors@yahoo.com. He is a consultant for Stryker.
Is there enough clinical evidence to warrant the use of vertebroplasty?
Change of tide
I think many of the supposed criticisms of the [NEJM] trials over the past year have been flawed, and I still think that our two trials carried a low risk of bias with no major methodological flaws. I think that the tide has shifted over the past year, and there have been more and more editorials and commentaries agreeing with the results of the trials. That has correlated with the new guidelines that the AAOS put out. I think that [the AAOS workgroup] has done a great job in recognizing that these two trials are really the only ones that are not biased, and I think in retrospect when you go back and look at the open studies — including the Vertos II trial — they are all consistent.
In the Vertos II trial there were two major difficulties: One was the trial was unblinded, and the second is that it was compared to usual care and not placebo. These effects would account for about the 25% benefit difference seen between the two groups. I think in that sense the Vertos II trial is completely consistent with the results of our study. There has been good work to show that lack of blinding results in an average overestimate of the relative treatment benefit of about 25%, and that largely explains their results.
Blinding is of major importance when the outcome is subjective … because the major endpoint is whether the pain has been relieved, then knowing which treatment has been received has a big impact.
The second problem is they did not use a placebo — and this is tied in with the blinding. They compared it to usual care. If you do not have a placebo group and patients are not blinded to which treatment they are getting, the treatment benefit is always going to be overestimated. There are many examples of this in other interventions.
The other important thing to note from the Vertos II trial is that this condition — the painful fracture in the spine — is self-limiting. You can see in the Vertos trial that a large proportion of the people they screened were not eligible because by the time they were ready for the treatment, their pain had improved substantially. This is another reason to have a blinded placebo control group.
Rachelle Buchbiner, PhD, is the lead author of
A randomized trial of vertebroplasty for painful osteoporotic vertebral
fractures, which appeared in the New England Journal of Medicine.
Enough clinical evidence exists
In my opinion, the Vertos II study provides enough evidence to support the use of vertebroplasty.
During the past years, we have empirically learned that proper patient selection for vertebroplasty is necessary to warrant clinical success. Patients with osteoporosis should at least meet the following three criteria:
- painful VCF;
- bone edema on MRI; and
- focal tenderness at fracture level, as assessed by a clinician on physical examination.
In the sham studies, bone edema in the fractured vertebral body on MRI was not used as a consistent inclusion criterion and patients were not physically examined before inclusion. Thus, according to the criteria of Vertos II, a proportion of patients in the sham studies had no indication for vertebroplasty. This difference in patient selection largely explains the very limited decrease in pain after vertebroplasty found in these sham studies.
Unlike in Vertos II, there was no control group without intervention [in the NEJM studies]. In the sham intervention, the use of a local anesthetic may have resolved secondary soft tissue and facet joint pain unlikely to be affected by vertebroplasty — especially in poorly selected patients in which we do not know whether the VCF is the cause of pain.
Vertos II provides a head-to-head comparison of the two actual treatment strategies — conservative therapy and vertebroplasty — and thus provides the clinician with directly applicable information on how to best treat the patient.
Caroline A.H. Klazen, MD, is coauthor of Vertebroplasty
versus conservative treatment in acute osteoporotic vertebral compression
fractures (Vertos II): an open-label randomised trial, which appeared in
the Lancet.