For vertebral compression fractures, largely in the elderly due to osteoporosis, several treatments (eg, bed rest, bracing, and pain medications) have traditionally fostered modest pain relief and increased activity. Relatively recently, percutaneous vertebroplasty — the injection of cement (polymethylmethacrylate) directly into vertebral fractures to provide immediate stability and pain relief has emerged as a popular treatment option. Despite several studies suggesting positive effects for vertebroplasty as compared with control treatments, there have been no blinded and/or placebo-controlled, randomized trials. Two trials recently reported in the New England Journal of Medicine sought to fill the research gap.
In the first trial — called INVEST (Investigational Vertebroplasty Safety and Efficacy Trial) — researchers in the U.S., UK, and Australia randomly assigned 131 patients with osteoporotic vertebral compression fractures to receive either actual vertebroplasty or a placebo "sham" surgery [Kallmes et al. 2009]. Those in the placebo arm went through the motions of surgery, including being brought into the procedure room and having novocaine injected under the skin and over the bone, but no cement was injected within vertebrae. A second trial, done in Australia, used essentially the same methodology with 78 participants [Buchbinder et al. 2009]. All patients were “blinded”; not told which group they were assigned to.
At one month, the real and sham surgery groups in both trials reported clinically meaningful improvements in pain and disability compared with baseline measures prior to the procedures; however, the extent of improvements in the two groups were statistically equivalent, casting doubt on the effectiveness of actual vertebroplasty. The Australian trial found similar results, and also reported equivalent outcomes at 6 months when 46% of patients receiving surgery and 42% in the sham group perceived their pain as better than prior to the procedures; although, 20% of surgical patients and 14% of sham-group patients said their pain was worse.
These two studies are interesting and important for a number of reasons. First, having two separate research teams conduct essentially the same trials in different, multinational populations adds significant rigor and reliability to the results. Second, the researchers acknowledge longstanding controversies surrounding the use of sham surgeries in unsuspecting (blinded) patients [Buchbinder et al. 2009; also, Macklin 1999]; however, these trials exemplify that without such an approach the extent to which the surgical procedure itself influences outcomes would be unknown. We have recently noted similar situations involving surgeries for migraine headaches [click for blogpost] and real versus sham acupuncture procedures [click for PDF]. Furthermore, several questions and concerns are raised by these studies that impact clinical decisions regarding whether or not to recommend vertebroplasty for patients:
- What might account for both real and sham groups experiencing improvements? As one possibility, the researchers suggest a strong placebo effect in the sham-procedure group; although, one of the trials reported that within 2 weeks 63% of patients in the sham group correctly guessed they had not received actual vertebroplasty [Kallmes et al. 2009]. Another possibility is lasting effects of local anesthesia, which both groups received, or the natural history of vertebral compression factures, which can improve with time, rest, and medications.
- There were no adjustments taking into account other medical treatments that patients in either study group may have received concurrently, which could have been confounding factors. Along those lines, it would have been better if a third group — receiving usual medical treatment — had been included for comparison with real and sham vertebroplasty. Also, the researchers did not evaluate a procedure similar to vertebroplasty, called kyphoplasty, whereby small balloons are first inflated within collapsed vertebra to create caverns for cement infusions. Granted, adding a 3rd and 4th group might have made these trials impractical from a subject-recruitment standpoint.
- The followup time of 6 months, in only one of the trials, may have been insufficient to detect long-term benefits of real vertebroplasty. The Australian research team planned to do a 2-year followup, but this will not be completed until 2010 and there may be insufficient patients by that time to have adequate statistical power for valid assessments.
- Another concern is that, based on these studies in about 200 patients total, the 750,000 Americans with painful vertebral compression fractures each year might be denied insurance coverage for vertebroplasty, which can cost more than $2,000. And, this also may raise questions about other interventional procedures, such as kyphoplasty or nerve blocks, for which evidence from randomized, well-controlled trials may be lacking.
> Buchbinder R, Osborne RH, Ebeling PR, Wark JD, et al. A randomized trial of vertebroplasty for painful osteoporotic vertebral fractures. N Engl J Med. 2009;361(6):557-568.
> Kallmes DF, Comstock BA, Heagerty PJ, Turner JA, et al. A randomized trial of vertebroplasty for osteoporotic spinal fractures. N Engl J Med. 2009;361(6):569-579.