Effectiveness of Vertebroplasty: A Recent Controversy

Article Outline

• So Why Is This Controversy of Significance to Radiologists in General?
• References
• Copyright

The most important service rendered by the press and magazines is that of educating people to approach printed matter with distrust.

Samuel Butler, novelist (1835–1902)

Many of you are already well acquainted with the procedure known as vertebroplasty, which has become a significant contributor to the management of both osteoporotic and malignant compression fractures. Under imaging guidance, a needle is placed within the collapsed vertebrae and acrylic orthopaedic bone cement is injected. The vast majority of patients with osteoporotic fractures (>90%) report significant and durable pain relief at the treated levels. The procedure was first devised in France in the mid 1980s and by the late 1990s had become widely used not only throughout Western Europe but North America and other parts of the world. Hundreds of thousands of these procedures have been performed, and interventional radiologists who perform vertebroplasty acknowledge that it is one of the most effective and dramatically satisfying procedures they perform. Because of the improvement in patient pain and its excellent safety profile, this procedure has become widely adopted.

However, it has been recognized that the accrued experience with vertebroplasty has not included a double-blinded randomized control trial to document efficacy. In the current evidence-based environment of modern medicine, this deficit must be addressed. It is noted that several practical barriers exist to performing a randomized trial of this procedure. The most problematic barrier is a historical one, secondary to the rapid acceptance of this procedure by not only interventional radiologists, referring clinicians, and patients, but by many third-party payers who saw it as a way of quickly resolving an often difficult medical problem. In spite of these issues, 2 double-blind randomized trials were recently published in the New England Journal of Medicine [1], [2]. Consistent with this rigorous study design, both trials compared vertebroplasty in the treatment arm with a sham procedure that mimicked vertebroplasty on the nontreatment arm, with random allocation to either arm. Consent for the trial was performed before allocation, with the result that prospective volunteers have only a 50% chance of receiving vertebroplasty. As expected, patient accrual in both trials was extremely limited, taking 5 and 4 years, respectively, in large part, because many patients were unwilling to undergo randomization. The published conclusions of these trials indicate that patients in the vertebroplasty arm did no better than those in the sham treatment arm and that vertebroplasty “does not work.”

Several problems exist with these articles. First, these 2 trials sample fewer than 300 patients in both treatment and nontreatment arms. The small size of the sample is in large part because of the difficulty in having patients accept randomization. Patients in severe pain are often unwilling to undergo randomization if they perceive that a safe, fast, and purportedly highly effective treatment exists. The effect of this preselection is to remove those patients in greatest pain, which serves to decrease the measured treatment effect and measured efficacy of vertebroplasty. In spite of this limitation, if we look at the data carefully, in both trials, there was a statistical trend (that did not, however, reach statistical significance) that suggested that vertebroplasty may be more effective than the sham treatment. It is conceivable that, with a larger sample, vertebroplasty may be shown to be effective even in this less symptomatic cohort.

The implications of basing management decisions on these trials are significant. This would mean rejection of what has come to be thought of as one of the most effective and safe treatments for severely painful compression fractures in osteoporosis, a widespread condition that is becoming increasingly significant as the population ages. Because of this, it is very important that these 2 trials be examined critically to determine whether they indeed should be accepted at face value or whether further investigation is truly merited. The investigators admit that they had originally planned to recruit more patients; however, this was changed midstream through the trial when it was decided to accept a smaller number. The fact that a smaller number of patients was studied also raises another difficulty in the analysis of these trials. Different types of compression fractures exist, and it is quite possible that these different categories of compression fracture will respond to vertebroplasty in different ways. Much of the data suggests that patients with fracture planes (so called clefts) are more likely to respond to vertebroplasty. The number of patients in the trial makes it impossible to analyse for fracture subtypes. Because of these limitations, many interventional radiologists believed that these conclusions are premature and indeed are rather unlikely to be substantiated by detailed and larger trials. For this reason, vertebroplasty should continue to be performed while accruing additional data to ensure that many patients who are in significant pain are not denied a potentially highly effective and safe treatment.

Back to Article Outline

So Why Is This Controversy of Significance to Radiologists in General? 

These studies outline the importance of study design and careful interpretation of the results. The investigators of both of these trials had an extremely challenging, very difficult question to answer, with many barriers put in front of them. They did a remarkable job in trying to overcome these difficulties and have provided thought-provoking and useful data. However, because of the limitations outlined above, we do not believe that these 2 trials can be viewed as the final word on this subject. We believe that further research on the clinical efficacy of vertebroplasty is indicated to provide stronger evidence.

Back to Article Outline

References 

1. Buchbinder R, Osborne RH, Ebeling PR, et al. A randomized trial of vertebroplasty for painful osteoporotic vertebral fractures. N Engl J Med. 2009;361:557–568
   • View In Article
   • CrossRef
2. Kallmes DF, Comstock BA, Heagerty PJ, et al. A randomized trial of vertebroplasty for osteoporotic spinal fractures. N Engl J Med. 2009;361:569–579
   • View In Article
   • CrossRef