Clinical trials: Where the rubber hits the road for science-based medicine

Until recently, most of my research was laboratory-based. It included cell culture, biochemical assays, molecular biology, and experiments using mouse tumor models. However, one of the reasons that I got both an MD and a PhD was so that I could ultimately test ideas for new treatments on patients and, if I’m good enough and lucky enough, ultimately improve the care for cancer patients. If there’s one thing, though, that I’ve learned in my nearly nine years in academic surgery, it’s that far fewer patients end up enrolled on clinical trials than are eligible to participate in them. Every meeting regarding clinical trial accrual I’ve ever attended, or so it seems, concentrates on how disappointing accrual has been or how accrual has been flat or even declining. That was especially true at my first job post-training, although it is less true at my current job, mainly because the clinical trial office there is superior to that at my old job.

So it was with great interest that I saw this article in the Los Angeles Times that acknowledged a problem that those of us in academic medicine have known about for years:

A year ago, U.S. researchers launched what they deemed a high-priority study to determine if women with an often-fatal type of breast cancer could live longer by taking a specific combination of drugs. If the study found that to be true, the average rate of survival — four years — could be significantly extended.

A worthy question to address? It would seem so. But the answer may be a long time coming.

The U.S. arm of the international trial got underway nine months behind those in other countries. And researchers now expect to enroll only 350 U.S. patients of the original 3,500 sought.

Before drugs and therapies to save lives or reduce suffering can reach the market, they must be tested to ensure they’re safe and effective. But the reluctance of Americans to participate in clinical trials has been a serious drag on medical research.


The irony of this problem is that we are in an unprecedented era in which the fruits of science are producing potential therapies at a rate not seen before: biological therapies, gene therapies, new drugs, and even new genetic tests that can predict disease or response to therapy better than ever before:

“It’s sort of ironic that there are more compounds available to study than ever before, and most are based on some new scientific principles, yet we are still having trouble recruiting participants,” says Diane Colaizzi, executive advisor of the nonprofit Coalition of Cancer Cooperative Groups, founded by research groups to promote study participation.

Actually, it’s more than “sort of” ironic. Only about 5% of patients who are potentially eligible for clinical trials actually participate in them–and that’s just patients seen at academic medical centers, where the bulk of clinical trials are up and running. If you count patients being seen in the community by private doctors who don’t enroll patients on clinical trials, the actual percentage of patients who participate becomes very tiny indeed.

I can’t understate how much of a problem this is. Although from my position it’s tempting to blame increasingly onerous federal regulations, I rather suspect that that is a relatively small contributor to the difficulties. Moreover, although there is a fine balance between too little regulation such that patient safety is compromised and so much regulation that clinical trials become so expensive and onerous that more and more physicians don’t want to put themselves through the pain of doing them, patient safety must never be compromised. True, no system designed by humans is perfect and there will always be some degree of risk in a clinical trial. After all, if we knew how an experimental drug will work and what its risk profile would be it would no longer be an experimental drug. However, we must do what we can to minimize that risk; so whatever we do to try to increase clinical trial accrual, it must not involve doing anything that might jeopardize the human subjects protections built up over so many years. If we do, we risk more problems with clinical trials that result in patient harm, each of which does more damage to efforts improve clinical trial accrual than almost anything else.

One of the reasons for this problem, unfortunately, is that patients don’t understand what modern clinical trials are about. There seems to be a persistent belief that participants in clinical trials will be randomized to groups that receive a placebo. For most treatment trials that just isn’t true anymore. In cancer, for instance, most clinical trials test the standard of care versus the standard of care plus the new drug. Medical ethics for human subjects research, as set forth in the Helsinki Declaration and the Belmont Report, demand no less. Indeed, if I may digress a minute, this fundamental misunderstanding seems to be what leads the antivaccine movement to be so obtuse when it comes to their insistence on a “vaccinated versus unvaccinated” trial. Either they are too ignorant to understand that it is unethical to give anything less than the standard of care to persons participating in clinical trials, or they have a superficial understanding of how clinical trials are done that leads them to believe that only a double-blind, randomized placebo-controlled clinical trial is the way to answer a question. In a world of pure science, that might be true, but ethics limit how purely we can apply the scientific method to human subjects participating in clinical trials. Unfortunately, as similar misunderstanding seems to lead all too many potential subjects believing that by signing up for a clinical trial they might risk receiving placebo only.

Of course, that this is true is clearly a failing on our part as clinicians and scientists. It’s also a failing among us that, as the article points out, some centers are simply trying to treat patients more like human beings as a strategy to improve clinical trial enrollment rates.

We have a saying in oncology that clinical trials represent the best patient care. Although there is some risk in the experimental arm, overall the benefits of clinical trials outweigh it. Indeed, a well-respected oncologist and former cancer center director who used to be one of my bosses liked to say it all the time, and there is evidence to suggest it:

The one area of medicine in which clinical-trial enrollment is relatively common has positive results to show for it. In pediatric cancer centers, more than 60% of children receive experimental therapies. That decades-long practice, which began out of desperation to treat children who would otherwise die, is now credited with vastly improved rates of survival, such as the astonishingly high cure rate associated with childhood leukemia.

“I think the real issue here is that in adult medicine, the clinical trial process is not an integral component of the entire treatment paradigm,” Comis says. “What we’re trying to do is make it more mainstream.”

It may be difficult to translate the success in pediatric cancers to adult cancer and adult diseases, however. The reason is that pediatric cancers are relatively rare, and there are relatively few centers that have the expertise available to treat them. It’s therefore much easier, given the small number of patients and the horror of pediatric cancers, to entice parents to enroll their children in such trials.

However, one example doesn’t demonstrate that clinical trials provide in general the best care; so let me explain why we in academic medicine generally think that the participation in well-designed clinical trials represents the best possible care. One reason is, as has been explained, that no group gets only a placebo, except in the case of prevention trials or, far less commonly, in treatment trials for which no effective treatment exists. Medical ethics has evolved to the point where a placebo control is considered neither necessary nor ethical in most cases; new modalities must be tested against or with the standard of care. Another reason is that patients in clinical trials are followed far more closely by far more people than most patients undergoing standard therapy. The protocol must be followed, and there are people there who make sure that the doctors and nurses actually do follow the protocol or explain why they don’t think it’s appropriate for an individual patient. In other words, because of the weight of human subjects protections, arguably the chances of medical error tend to be lower.

Other strategies being tried are databases that patients have access to, including ClinicalTrials.gov, TrialCheck, and BreastCancerTrials.org. Future efforts will lead to linking up electronic medical records to such databases so that potential subjects can be identified:

The databases are only a first step. As electronic medical records become the norm, researchers hope to tie the records into an automated clinical trial matching system that will alert doctors and patients alike. One company, Impac Medical Systems of Sunnyvale, Calif., offers this service to oncology practices that use its health records software. After a patient’s health record has been entered into a computer, the software searches TrialCheck to see if that person may be eligible for clinical trials.

Comis describes the ultimate goal for patients: “They will have access to their own health records and they can say to their doctor, ‘It says here I’m eligible for three studies. Why didn’t you talk to me about that?’ “

However, perhaps the most important reform that could help boost the number of patients willing to sign up for clinical trials is the following:

Researchers and patient advocates in various states are also fighting for federal legislation that would require insurance companies to pay for the patient care involved in clinical trials, following 23 states — including California — that already have such laws. The federal bill was developed specifically to address enrollment in cancer trials, which make up a majority of clinical research. Called the Access to Cancer Clinical Trials Act, it was introduced by Sen. Sherrod Brown (D-Ohio) last year and is before a House committee.

Though the research sponsor, such as a drug company, usually provides the drug or therapy free of charge, some insurance policies don’t pay for care, such as blood tests, scans or examinations, linked to the treatment, says Dr. James Thomas, director of clinical trials at the Ohio State University Comprehensive Cancer Center.

This is not requiring insurance companies to pay for experimental treatments. Rather, it is requiring them to cover accepted medical expenses that flow from clinical trials. Given the potential benefits in terms of better treatments, in the long run this could end up saving insurance companies money, although it would admittedly cost more up front and any cost savings possible would be in the long term. It would greatly streamline the process of clinical trials by eliminating the need to try to parse what blood tests (for example) are being done for the trial and what tests would be done for routine medical care.

Clinical trials are, as I have said before, where the rubber hits the road for science- and evidence-based medicine. Without well-conducted clinical trials, all that elegant science, all that hard-won understanding culled from biochemistry, molecular biology, cell culture experiments, animal experiments, and early measurements in humans will be for naught. All that science has to be tested in humans to see if it actually works. If we don’t find a way to persuade more people that it is in their best interest to participate in clinical trials, we cede that ground to Europe, whose human subject protections are every bit the equal of ours, but whose academic medical centers appear better able to persuade patients to enroll in clinical trials, an advantage that might have something to do with nationalized health care systems and centralized databases. Worse, we encourage drug companies to look elsewhere for subjects, and they are doing exactly that–in China and Third World Countries, where the human subjects protections are–shall we say?–a bit more lax than they are here. If we cannot do the clinical trials, we risk being unable to capitalize on the promise of science- and evidence-based medicine. We’ll in essence, have all these neat new treatments, but no way to determine if they’re actually better than our old treatments.