Thursday, February 23, 2006

Knee osteoarthritis- glucosamine and chondroitin- we still don't know

Seldom are the results of a randomized controlled trial (RCT) so definitive that the issue is settled and and more questions are not raised. The GAIT trial is no exception.

Published in the February 23, 2006 issue of the NEJM this multi-center,placebo and celecoxib controlled trial found that glucosamine hydrocholride and chondroitin sulfate alone or in combination did not reduce pain effectively in the overall group..but the combo "may be effective" in the subgroup with moderate to severe knee pain.

"Hydrocholride" is in bold print because as the editorialist suggests, perhaps the"wrong" type of glucosamine was used. While one recent meta-analysis of 8 RCTs showed no difference from placebo when glucosamine hydrochloride was used, another meta-analysis of 7 RCTs which used a particular brand (Rotterpharm brand) of glucosamine sulfate showed significant symptom improvement and functional improvement versus placebo. On the face it appears to be another example of dueling meta-analysis but not if we consider different types of glucosamine. It is possible that the sulfate works better than the hydrocholride. Further, there are two trials that purport to show slowed radiographic progression from the sulfate preparation.

RCTs almost always are paired with an editorial that is replete with many "howevers", "Yes, buts" and "on the other hands" and typically a plea for more data collection. Yet, when the desired data is published there is usually another round of "howevers". There is a reason for that;RCTs often do not settle the issue.

Often , the best we can squeeze out of a RCT is a tentative general "Game plan", subject to future modification or even refutation. Here as the editorial suggested it may be:

If the patient wants to try a nutrional supplement, he should go with glucosamine sulfate plus chondroitin sulfate for about 3 months and if no better discontinue them.

Clinical trials are not rocket science. At its best rocket science deals with deterministic, differential equations that tell us with impressive precision where the rocket will hit absent the ambiguity, uncertainty, seemingly uncontrollable biological variability and the problems that arise from placing too much confidence in our statistical techniques to control for or "quantitate" the uncertainty that are intrinsic to clinical trials.

No comments: