Thursday, April 9, 2009
RCT of Kyphoplasty
Lancet has the results of a multi-national randomized controlled trial of balloon kyphoplasty for vertebral compression fractures. This was an industry-funded trial of ~300 patients with 1-3 acute or subacute (less than 3 months old - mean was 6 weeks old), painful (4/10 or greater) VCFs who were randomized to KP or medical management/routine supportive care (meds, braces, therapy, etc.). They followed patients for 12 months, although the primary outcome was changes in the SF-36 quality of life scale at 30 days. Intention to treat analysis was used; as far as I can tell no one (not even the researchers administering the follow up surveys) was blinded.
Health-related quality of life was improved at 1 month (statistically and likely clinically significant) in the KP group; this had mostly attenuated by a year (although remained statistically significant). Pain improved markedly in the KP group in the short term: from (median) 7/10 before the procedure to ~3.5/10 immediately afterwards. By one month the difference was about 2/10 points between groups (3.5/10 vs. 5.5/10); this gap slowly closed over the year and was about 1/10 points at 12 months (3/10 vs. 4/10). At one month 65% of control group patients were taking opioids vs. 45% in the KP group; this difference was gone by 1 year (both ~30%).
There was one serious procedure related adverse event (hematoma); 27% had cement extravasation (all aysmptomatic). Overall, rates of adverse events were similar. Cancer/myeloma-related VCF patients were included in the study although only 1% of the subjects had these so one cannot draw any conclusions specifically about cancer-related VCFs.
These is the best study done on KP and the results are supportive of KP: rapid and marked pain relief that endures for a year although over months, as presumably the non-KP patients' fractures heal, the benefits attenuate. The benefits in overall health-related QOL are less pronounced, but they're there at least for a few months. It is also reassuring that patients well over a month out from their fractures appear to get good benefit, although they did not present any analysis of whether time-from-fracture was related to the magnitude of the benefit. The biggest concern for me was the lack of blinding; not even in those administering the survey instruments. Patient-blinding is difficult with interventions like these (although it has been done, e.g. at Mayo in the use of sham-procedures in their controlled trial of celiac plexus blocks) of course, but blinding of those doing the analyses and data collection is ideal. In addition, this does not answer the question of whether simple vertebroplasty (which doesn't use balloons) is as efficacious as the more-expensive KP.
Wardlaw, D., Cummings, S., Van Meirhaeghe, J., Bastian, L., Tillman, J., Ranstam, J., Eastell, R., Shabe, P., Talmadge, K., & Boonen, S. (2009). Efficacy and safety of balloon kyphoplasty compared with non-surgical care for vertebral compression fracture (FREE): a randomised controlled trial The Lancet, 373 (9668), 1016-1024 DOI: 10.1016/S0140-6736(09)60010-6
2 Responses to “RCT of Kyphoplasty”
April 12, 2009
Interesting study. The lack of blinding (when possible) is a weakness in the methodology, as you say. The results of the study are not surprising, though. A few comments:
1) This study helps us understand the “natural history” of pain related to compression fractures a little better. Perhaps there are other studies that lend evidence to this. I suspect that many practitioners might have stories about patients who received KP with tremendous results (for instance, I can remember one patient that was on dilaudid 10 mg/hr and afterwards was able to go back to Morphine 60 mg orally in a day). However, patients and practitioners must brace for the fact that there will likely be some chronic pain that results from the fracture, with or without KP (3/10 median vs. 4/10 respectively at one year). This rings true based on the few patients that I have followed longterm after a compression fracture.
2) One might assume that patients with a malignant compression fracture have a worse prognosis than the population in this study- I’m not aware of studies that address the prognostic significance of a compression fracture, and it likely depends on the primary malignancy. Based on this study which shows potentially more benefit on the front end, one might hypothesize that patients with a poor prognosis in the setting of cancer might get more bang for their buck (since many may die before the differences in pain even out).
3) I was intrigued to see that most patients underwent general anesthesia for this procedure. As far as I know, this is usually done under conscious sedation at my institution.
4) Fractures were 5.6 weeks old on average by the time of randomization. This seems like a long time to me. Most patients with compression fracture present acutely with severe pain, and in my limited experience, will get KP relatively quickly (within a few weeks) if they are a candidate.
April 13, 2009
Hi Lyle - agree with #1 - the findings were 'confirmatory' of general opinion, what we all see, etc. (a substantial number of patients feel a lot better very quickly) - that makes the simple natural history findings particularly interesting - just what happens w/r/t pain and function in patients who don't receive any intervention. And yes, it does kinda make one wonder if this potentially means that patients with poor prognoses (say, less than 3 months) are particularly good candidates for this, insofar as most of the net benefit (over medical management) - pain, function, hrqol, etc. occurs in those first few months....a counterintuitive example to argue that poor prognosis actually is an argument to do an invasive procedure insofar as one could expect this to benefit someone for 100% of their lifespan (assuming it's <6mo) as opposed to ~~5% for someone with 10 y to live? That's a somewhat perversely compelling argument i have to say.
Post a Comment