— A new meta-analysis examines hyaluronic acid for treating knee OA.
by Hilda Bastian February 3, 2015
Is hyaluronic acid (HA) just an expensive placebo treatment with adverse effects? Or can HA injections into knees damaged by osteoarthritis improve lubrication and shock absorption?
Whatever your position has been — for, against, or on the fence — you could dig out a meta-analysis to support it. You could choose between professional society opinions, too. A new network meta-analysis puts a strong new spin on the trial evidence, though.
Before this, there had been conventional systematic reviews and meta-analyses, including one based on limited patient-level data. Those can combine similar enough data from multiple studies.
A network, or mixed treatment meta-analysis, is different. It uses modeling more extensively. Say treatments A and B have been compared with each other well enough, and so have B and C. Then it’s possible to reckon how A and C compare with each other, even if they haven’t been tested head-to-head.
Raveendhara Bannuru and colleagues from Tufts University found 137 suitable trials involving 33,243 people with osteoarthritis of the knee, setting up a network that included oral placebo, intra-articular placebo, intra-articular HA, intra-articular corticosteroids, acetaminophen, and key NSAIDs. It means that something new could be teased out: intra-articular placebo could be compared with oral placebo.
They standardized the outcome measurements into the most common scale, enabling almost all of these patients to be compared on pain as an outcome. (For function, stiffness, and adverse effects, there was decidedly less available data.) The minimum level of effect for clinically significant improvement was set for a widely accepted standard, too: a change of at least 20 points on a scale of 0 to 100. And then they ran the numbers.
Intra-articular HA came out on top — more effective than any of the other treatments. Everything was better than oral placebo. But acetaminophen was the least effective.
How effective were the injections? Key analyses are in the article’s supplement (figure 1 and table 7). Whereas acetaminophen didn’t reach the 20 improvement points needed for clinical relevance, intra-articular HA was almost 30.* If you take only the best studies though — those with more than 50 or 100 participants in each group — the injections lost up to a third of their effectiveness. Diclofenac pulled ahead in the largest studies. (More data is needed on safety, but both kinds of injection fared better than the medicines.)
Why did the HA injections come out so strongly in this analysis? The authors point to the effect of the placebo injection here. You couldn’t see this in a traditional meta-analysis. It would show a marginal improvement of the active injection versus the placebo injection. But that doesn’t show the effect of the “placebo” injection versus doing nothing.
It’s not yet clear what’s happening when something is injected into the knee — or when aspiration accompanies it. It’s not clear what HA does either, or what it might do in the long term. There’s a compelling narrative about it working by being a lubricant and shock absorber, but that might not be exactly what’s happening. Everything from some harm, to no effect, or even a disease-modifying effect, is conceivable. Some other substance may outperform it.
And there’s still the issue of differing interpretations of the same data. With sets of evidence as complicated and large as this one, multiple explorations with varying choices along the way can seem to muddy the waters. But multiple perspectives can ultimately help get a clearer picture.
The American College of Rheumatology (ACR) has already issued a statement responding to the new study. Their 2012 clinical practice guidelines conditionally recommended HA injections when there was an inadequate response to initial therapy. The American Association of Orthopedic Surgeons, on the other hand, recommended against HA injections in 2013, based on a conventional meta-analysis with a smaller pool of studies.
Utilization data from Washington state show a rapid increase in use of HA injections by Medicaid patients from 2006, apparently leveling off in 2011 and 2012. The growing number of baby boomers who will have painful knees is sure to keep the controversy over knee osteoarthritis treatments heating up. The competition to treat them has been fierce, with inevitable widespread financial conflicts of interest.
This network meta-analysis was funded by the Agency for Healthcare Research Quality (AHRQ). ACR reports that it was commissioned by CMS. When it comes to the crunch, costs will be critical to people’s options — especially if evidence emerges for benefit from longer courses. As it is, a course can cost up to $1,000. If knee injections could reduce the demand for knee replacements, that would be a bargain. It would be pretty steep for a placebo, though!
* The exact effect size of absolute treatment effects on the WOMAC 0-100 scale was 29.44 (credible interval 24.17-34.93). Credible intervals are the Bayesian counterpart to confidence intervals.
Hilda Bastian is a senior clinical research scientist. She works at the National Institutes of Health as editor for the clinical effectiveness resource PubMed Health and as editor for PubMed Commons, PubMed’s scientific publication commenting system. She is an academic editor at PLOS Medicine, and blogs for PLOS (Absolutely Maybe) as well as on a personal cartoon clinical epidemiology blog (Statistically Funny). The thoughts Hilda Bastian expresses here at Third Opinion are personal, and do not necessarily reflect the views of the National Institutes of Health or the U.S. Department of Health and Human Services.