Features March 2005 Issue

Clinical Trials: Proof Or Spoof?

Gauging the impact—now or future—of today’s drug studies.

It’s hard to go a week without seeing a newspaper or TV report about some new research study related to arthritis. Occasionally, when a story is important enough, as in the recall of Vioxx, the media goes to great lengths to put the study in context, helping you figure out just what it means.

But in most cases, research stories don’t come with handy explanation kits. Instead, you’re left on your own to determine if the latest story seems credible or overly dramatic, whether its implications may be a long way off or are something you need to discuss with your doctor next week.

The arthritis story that makes tomorrow’s headlines could come from anywhere on the research spectrum, from basic laboratory science all the way to final-stage clinical testing. One part of making sense of any research study is determining its place on this continuum. Another is paying careful attention to the key phrases used by investigators when describing the study.

Here we try to give you a sense of the types of research used to develop new drugs, devices, and ideas about arthritis, as well as explain why some trial results may only be suggestive, while others can yield solid conclusions that can change the practice of medicine.

Variety of players
The cast of arthritis research is a broad one that includes thousands of specialists, from Ph.D. scientists, to rheumatologists, orthopaedic surgeons, immunologists, physical therapists, exercise physiologists, and others.

Where the research takes place is just as diverse: it may occur in a variety of health-care settings—hospitals, outpatient centers, private laboratories, pharmaceutical companies, or in federally funded labs.

The source of the funding is a bit more concentrated, coming from the federal government, private foundations, or private industry. The federal government (through the National Institutes of Health or the National Institute for Arthritis and Musculoskeletal and Skin Diseases) spends about $127 million on arthritis research every year. The Arthritis Foundation provides another $21 million. And the big spenders are the drug companies which, collectively, spend some $32 billion a year on R&D, though what fraction of that is arthritis-related is unknown.

Research categories
All this money is used to support two major categories of arthritis research: basic and clinical.

Basic research usually occurs in a test tube or in animals, and attempts to study arthritis at the molecular level, looking to understand how a disease or treatment may be related to changes in proteins, cells, or genes.

Clinical, or applied, research takes a promising new hypothesis or treatment that emerges from years of basic research and tests it among a group of people. If it’s a hypothesis about a new risk factor for arthritis, investigators want to study various populations with and without the factor to test its predictive power. If it’s a new drug or device, researchers want to see how safe and effective it may be compared to the standard drug or device currently being used.

Brewing on the bench
Though basic research takes place on many fronts, some of the most active involves the hunt for genetic causes of arthritis, obtaining a more detailed understanding of the immune response, identifying better materials for use in artificial joints, and identifying biomarkers that can lead to earlier detection and treatment.

“A better understanding of the genetics of arthritis opens up the possibility of novel treatments and causes,” says Elaine Husni, M.D., staff physician in the departments of Rheumatology and Orthopaedic Surgery at The Cleveland Clinic. Such a treatment might correct for a mutant gene that leads to abnormal cartilage formation.

When talking about basic research, scientists may describe it as pre-clinical, investigational, or in vitro (test tube) studies. Look for these words as a way to know you’re reading about basic science that is likely many years away from standard use in patients.

Choosing the type
Though clinical trials come in many types—randomized clinical trials, cohort studies, case-controlled studies, systematic reviews of previous studies, and various types of population or epidemiological studies—the type chosen for a particular line of research depends upon the question being asked, as well as the time and funds available.

“With any research, the idea is to put together a study that will yield the strongest possible statistical support for your findings,” says Dr. Husni, who conducts clinical studies of risk factors for systemic rheumatic disease.

Randomized trials
Randomized clinical trials are the gold standard since their design leaves the least room for statistical bias. Rather than select who gets the experimental treatment and who gets the placebo, investigators randomly assign participants to one of the two options. Such studies carry maximum statistical power when neither participants nor researchers know who is getting what until after the study is complete (also known as double-blind).

Cohort studies
Randomized trials are not always feasible. “When all participants have a serious medical problem, it may not be ethical to randomize half to a placebo,” says Joseph Iannotti, M.D., co-director of the Orthopaedic Research Center and chairman of Orthopaedic Surgery at The Cleveland Clinic.

In such instances, the researcher might do a cohort study instead, in which half get the new treatment and half get standard therapy. The participants are then followed over time. A cohort study is less statistically rigorous than a randomized trial since bias can creep in as both investigators and participants know upfront who is getting what treatment.

One rung lower on the statistical ladder are case-controlled studies. Here, researchers follow two groups, one with disease, and one without, backward in time (retrospectively), trying to identify risk factors that may have led to the onset of disease. “Case-controlled studies are used with diseases whose symptoms take many years to develop,” says Dr. Husni. The statistical power of such studies improves if there are only a few risk factors that could have played a role in the difference between the groups.

“The increasing power of computers and availability of hospital databases are making case-controlled studies easier to do,” says Dr Iannotti. “They are much faster and less expensive that a randomized trial or cohort study since you don’t need to recruit patients, hire staff, or wait years for results.”

Combining studies
Because many arthritis studies involve only a few dozen patients, they are too small to pack statistical power. To overcome this, researchers sometimes conduct a systematic review, or meta-analysis, in which they combine results from several small studies in order to reach larger conclusions that carry more weight. “The challenge is being sure to compare apples to apples,” says Dr. Iannotti. “Researchers must be careful to explain how the studies were chosen and analyzed.”

Knowing the phases
If you read about promising new arthritis drugs, then you probably know that drugs undergo at least three phases of clinical testing. In Phase I, researchers try to determine if a drug is safe, and give it to small number of healthy volunteers to determine the maximum safe dosage.

Phase II trials involve a larger group who have a particular type and stage of arthritis. Researchers want to know if a drug is effective and to monitor side effects. The results of Phase II trials are preliminary at best.

In Phase III, investigators want to find out if a new drug is at least as effective as other current drugs. These studies involve still larger groups of patients, take place over a year or two, and are usually randomized clinical trials. Results may be strongly suggestive, but are not conclusive. Widespread use of the drug or treatment (after FDA approval) may be required for final proof of safety and effectiveness or the lack thereof (as in the case of Vioxx).

As a promising new drug or idea about a risk factor moves from basic research to clinical testing, a number of studies and stories will likely be published. It’s best to treat such stories as mere facts in an ongoing debate that is far from settled. Remember that no one study wins the battle, and that change in medicine comes slowly, requiring a build-up of compelling evidence over time.