Evaluating clinical trials: details matter.

Many of us are trying to figure out which herbs and nutritional supplements we should take. It's easy to clip newspaper stories about the latest research and find study abstracts on the Internet. However, even when we have a stack of clippings, it can be quite difficult to summarize the key points or pick out the examples that apply most directly to our own situation.

[ILLUSTRATION OMITTED]

Mary L. Hardy, MD, medical director of the Simms/Mann-UCLA Center for Integrative Oncology in Los Angeles, has a special interest in herbal medicines and nutritional supplements. Each year, she reviews the latest studies and notes the points that don't always end up in the popular press. "When you're designing a clinical study, the first issue is, what question do you want to answer?" she says. "If you don't know your question, you can't design the right trial. Are you asking whether a product is safe, or whether it works better than placebo, or whether it's as good or better than something that's on the market now? Those are three different questions."

The questions researchers ask, and the expected effect of the substance they're testing, will determine how many patients they need to enroll in the study. "Suppose they expect to see 50% improvement as an average response," Hardy says. "That's a pretty big response, so it doesn't take many subjects to statistically demonstrate it. If the response is smaller or less consistent, you need more people enrolled in the trial in order to observe the response and separate it out from random events."

Some studies compare one therapy to another treatment that's already in use. "In that sort of study you want to demonstrate that what you're testing works as well as the accepted therapy, or perhaps better," Hardy says. "You usually need more people enrolled in that sort of trial. You're discriminating between two active therapies, so the size of the effect is usually smaller."

The classic way to set up a clinical trial is to compare a substance to a placebo, a "dummy pill." "In that case, you're supposedly comparing it to a pill that doesn't have any effect. However, that's just not true," Hardy says. "We know that under certain conditions, the placebo effect can be as large as the effect from your intervention."

[ILLUSTRATION OMITTED]

Glucosamine and Chondroitin: Don't Stop at the Headlines

Many of these issues are exemplified in a recent study of glucosamine and chondroitin, known as the Glucosamine/Chondroitin Arthritis Intervention Trial (GAIT) study. It was funded by the National Center for Complementary and Alternative Medicine (NCCAM) and the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) and cost just over $12.5 million.

In this large, double-blind, multi-center trial, reported in the February 23, 2006 issue of the New England Journal of Medicine, about 1600 patients with painful knee arthritis were randomly assigned to one of five different treatments. They included the following:

  1500 mg glucosamine daily;
  1200 mg chondroitin sulfate daily;
  Both glucosamine and chondroitin sulfate;
  200 mg of celecoxib daily; or placebo.

This was a sophisticated study design, Hardy points out. "They had both a positive and a negative control. By positive control, we mean that they compared glucosamine and chondroitin to something that was known to work, in this case, a Cox-2 inhibitor. They also compared them to something that should not work, the placebo, but oddly enough, in this case, it did work."

The study used appropriate dosage levels. However, some experts point out that it used glucosamine hydrochloride, while most people nowadays take glucosamine sulfate, which is said to be more bio-available.

The researchers compared the treatments in patients with mild knee pain and in those with moderate-to-severe knee pain. They looked at whether the treatment decreased knee pain by 20% over the 24-week study period, and they reported that "overall, glucosamine and chondroitin sulfate were not significantly better than placebo in reducing knee pain by 20%."

That's the summary, and that's what made newspaper headlines. However, when we look at the results in detail, we get a somewhat different impression. The first key pointto notice is that the placebo response was 60.1%. The authors write, "as compared with the rate of response to placebo (60.1%), the rate of response to glucosamine was 3.9 percentage points higher (P = 0.30), the rate of response to chondroitin sulfate was 5.3 percentage points higher (P = 0.17), and the rate of response to combined treatment was 6.5 percentage points higher (P = 0.09)."

In most studies, the placebo response rate is around 30%. "The placebo effect is the huge, undiscussed elephant in the room," Hardy says. "It turns out that if you tell someone they have a reason to get better, the body's pretty good at making you better. Why aren't we investigating this? But that's another subject." In this study, the placebo effect was a dramatic 60%. "We should be really clear that 64% of the glucosamine patients had a clinically positive response, and they're calling that a failure," Hardy says.

The second key point is that for patients who started out with moderate-to-severe pain, the study found that glucosamine and chondroitin were clearly more effective than placebo. For this group, "the rate of response was significantly higher with combined therapy than with placebo (79.2% vs. 54.3%, P = 0.002)," the researchers write. That seems like an important finding! But it didn't make the headlines. "People with more severe pain, in whom you would expect the placebo response to be less effective, showed a clear benefit," Hardy observes. "Yet this was reported in the press as a completely negative trial."

A third key point from the GAIT trial: the safety record on glucosamine and chondroitin was good, with no significant adverse events in a study that looked at almost 1600 people for six months. "This was reported in the research paper, but none of the big media outlets emphasized it in their coverage," Hardy notes.

The GAIT study includes a second part that will assess whether additional glucosamine and chondroitin treatment can reduce or halt arthritis progression. Researchers will compare X-rays taken at the start of the study and after one and two years of treatment to observe changes in the knee joints. Results are expected in spring 2008.

On the whole, Hardy believes GAIT was a well-designed study, and the paper in the New England Journal of Medicine offered a fair, balanced summary of the results. She's curious about the extremely high placebo response rate and wonders whether some members of that group might have been taking glucosamine and chondroitin on their own. "I don't think we're going to see another big trial on this subject. We might possibly see a smaller trial using glucosamine sulfate," Hardy says. "The point is, this was not a slam-dunk negative trial, but it ended up getting reported that way."

Find Help to Sort Through Available Information

Because so much information is available now about herbal medicines and nutritional supplements, it's best to find a health care provider who's familiar with the literature to help you identify safe, effective treatments. "When I first started focusing on these areas, I didn't know that much," Hardy recalls. "I said to my patients, 'Give me what you've got, and I'll see what else I can find out.' You need a healthcare practitioner who's willing to be open-minded, look through the literature, and try to understand how something works and whether it's worth the money."

More and more allopathic physicians are learning about these subjects, Hardy says. Naturopathic physicians are generally trained in these areas as part of their core training. "Many pharmacists are taking the lead these days and getting additional training, so they could be a valuable resource for patients or for a doctor who is just starting out."

However, in many situations, after reviewing all the available information, there may be no clear-cut answer; the evidence is inconclusive. In the end, Hardy says, "The bottom line for any individual person is, how do you feel? Has your health improved? If you're better because of the placebo response, or if you're better because of glucosamine, that may not matter so much to you as an individual. On the whole, you just want to feel better."

The research perspective is very different from the individual's perspective, she notes. "The researcher wants to separate out what's true from not true. At a policy level, we need research to determine what we should pay for. But your perspective as a patient, or as someone caring for an individual patient, is incredibly pragmatic. Is this going to make my patient feel better and be healthier? That's really all I care about. Most of my patients who take glucosamine find it's helpful. I'm happy, because that means they'll exercise more, which will benefit their health in a thousand ways."

Elaine Zablocki is the former editor of CHRF News Files.

Resources

For a clear, detailed summary of the GAIT study, see the six-page, recently updated backgrounder from NCCAM at http://nccam.nih.gov/research/results/gait/D310_BKG.PDF.

For an abstract of the GAIT study, "Glucosamine, Chondroitin Sulfate, and the Two in Combination for Painful Knee Osteoarthritis," go to http://www.ncbi.nlm.nih.gov/pubmed/16495392.