Vertebroplasty and the Randomized Study: Where Science and Ethics Collide

Vertebroplasty, which was first performed in the early 1990s, has become widely available in the United States, primarily for the treatment of osteoporotic and malignant compression fractures. Although initially described by Deramond, the first major study describing the effectiveness of vertebroplasty in osteoporotic vertebral compression fractures (OVCFs) was by Jensen et al (1) in 1997. They reported on 29 patients with painful vertebral fractures, with a 90% success rate, and the flag was dropped. With an estimated 700,000 vertebral compression fractures (VCFs) a year, resulting in 150,000 hospital admissions and five million restricted-activity days, the only previous means of treatment (ie, bracing and bed rest) accelerated the disease process. Vertebroplasty was therefore seen as a potential major medical breakthrough, and numerous other studies soon followed. Barr et al (2) found 95% of patients in their retrospective study of 47 had at least moderate relief of pain, and other retrospective studies on groups ranging up to 245 patients have been reported, all of which describe high percentages of success.

The number of prospective studies has been few, and most deal with a small number of patients. In this issue of the AJNR, Huy Do et al report on the largest group of vertebroplasty patients yet to be followed prospectively. Although they have run into the difficulties of following a patient population that typically is cognitive–and long-term survival–challenged, their data are strong evidence that further supports the effectiveness of vertebroplasty in the treatment of medically refractive OCVFs. The limitation of this study, which the authors freely admit, is the lack of a control group to compare their results. In fact, when reading most major articles on vertebroplasty, the lack of a double-blind randomized prospective study is lamented as a major obstacle that needs to be overcome before vertebroplasty can be “proved” to be effective. At this point, however, the question should not be “how could such a study be done?” but “should we do it?”—two questions that may sound similar but approach the problem from entirely different points of view.

Purists, of course, would disdain the idea that a randomized controlled study is not necessary when considering a medical treatment. From a pragmatic and ethical standpoint, however, such a study, though appealing in an intellectual sense, would almost assuredly merely confirm the data that we already have. To begin with, OVCF is a disease that, until vertebroplasty came along, had only one means of treatment—namely, conservative therapy, consisting of bed rest, which can accelerate bone loss, bracing, and narcotic anesthesia, In the past, when conservative management failed, the only option was . . . more conservative management. In most practices and studies, and by using Medicare guidelines, vertebroplasty is not typically considered until the patient has already failed conservative management. In other words, one could argue that the need for a control study has been reduced or eliminated, because these patients have already failed the only other treatment option. In light of this fact, as well as the extremely high success rates reported in the numerous studies of vertebroplasty, including the article in this issue of the AJNR, is it ethically defensible to deprive half the patients in a study of realistically their only chance for pain relief, especially when a side effect of the alternative therapy is to make the disease process accelerate? Some will argue that it is, for the sake of science and future patients, but others, including myself, would argue that what is already is in the literature is so uniformly positive that there is justification for treating medically refractory VCF with vertebroplasty, despite the absence of a randomized double-blind prospective study.

Where a study is needed, however, is in patients with acute osteoporotic compression fractures, particularly those patients who are hospitalized or bedridden from the pain of such fractures. This subgroup of patients has the highest risk of additional bone loss, hospital-associated morbidity and mortality, and costly hospital stays, merely for pain control. Minimizing or eliminating hospital stays and bed rest time by performing vertebroplasty in the acute setting might have significant benefits to such patients, although there is no large study on this population. This type of study would greatly benefit from a control group, because these patients would not have had a chance to try conservative management otherwise, and the data without a control group would therefore be much weaker. A randomized double-blind study would be very useful in determining how aggressive we should be in treating these patients.

Ultimately, however, on the basis of the existing data on vertebroplasty in medically refractory OVCF, a true, long-term, randomized double-blind study does not seem necessary or ethically justifiable.

• Copyright © American Society of Neuroradiology