Children in research trials: Have we gone too far?

IRBs also should consider appropriate levels of risk

Throughout the ’90s, the push was on to include more children in medical research. The U.S. Food and Drug Administration (FDA) encouraged pharmaceutical companies to begin testing new drugs on children as well as adults. The National Institutes of Health (NIH) required applicants for clinical research funding to improve measures for including children in federally funded medical studies.

But with the 1999 death of 18-year-old Jesse Gelsinger during a research study at the University of Pennsylvania and subsequent disclosures of violations of accepted patient-protection protocols at many well-known institutions, medical ethicists are taking a new look at the protections afforded to human subjects of medical research. And some ethicists and pediatric specialists also are questioning whether the drive to include children has progressed too far, too rapidly. Are we moving beyond inclusion to exploitation? How can we protect such a vulnerable population, yet at the same time ensure that new drugs and medical treatments are adequately studied before being administered to kids?

The fear that children will be exploited in clinical research is well grounded, says Lainie Friedman Ross, MD, PhD, assistant director of the MacLean Center for Medical Ethics and assistant professor of pediatrics at the University of Chicago. "Historically, children were very abused. And if you want to talk about the history of children in research, for many years, there were greater protections for animals than there were for children in research," she explains.

Friedman Ross and other experts in child health and medical ethics gathered in May at the University of Maryland for a conference discussing the history of research protections for children and whether current protections are sufficient — and sufficiently enforced. In the 19th and early 20th centuries, Friedman Ross explains, children often were the subjects of research projects that exposed them to significant risk and harm without their knowledge, she explains. Institutionalized children, particularly those in orphanages or who were wards of the state, were vulnerable to research abuse.

Following the adoption of the Nuremburg Code in the 1940s, it became commonly accepted that subjects in research projects should give informed consent — agreeing to participate only after being informed of the risks and benefits involved. It wasn’t until the late 1960s and early ’70s that society began questioning whether there were some people — children and the mentally disabled — who were unable to understand the risks involved and unable to give informed consent, Friedman Ross says.

In 1978, the National Commission for Biological and Behavioral Research drafted the first national guidelines on protections for humans involved in research projects. These guidelines became known as the Common Rule because they were adopted by all federal agencies conducting or sponsoring research: the Department of Health and Human Services, the FDA, the Department of Defense, and others.

"The Common Rule starts getting written in the 1970s, but the pediatric version isn’t finalized until 1983," she adds. "The regs are very specific. First, it says basically that research should be done first on animals, whenever possible, then on adults, then on older children before younger children."

Federal regulations for kids

Because children are a vulnerable population, they required special protections beyond those given to adult subjects. The federal regulations for children focus on three areas: the level of risk involved, the potential (or lack of potential) for therapeutic benefit to the child, and the issue of consent. "For example, they have said that you need not only the parent’s permission and the child’s assent, but also they are specific about whether you need one or both parents," Friedman Ross explains. "For research involving no risk or minimal risk, the consent of only one parent is required."

The regulations also are specific about the need to allow the child to opt out or dissent if he or she wants to, she adds. However, there is a section in the regulations about when an institutional review board can waive the need to get the child’s assent and rely only on parental informed consent. "There are a lot of situations where you do not have to get the child’s assent, although it is supposed to be taken into consideration," Friedman Ross says.

Requiring assent of the child

Although some ethicists feel that children should always be able to opt out, there are valid situations in which the assent of the child should not be required, Friedman Ross argues. "It’s very easy to sit here and say we should listen if the kids say no," she explains. "But kids say no for good reasons that we would expect from any patient in that situation, and they say no because they just don’t like being in the doctor’s office or don’t feel like it or whatever."

There are valid reasons why researchers might not need the child’s assent in every situation, she says. But if they are not going to respect it, researchers should not ask for it. "My biggest problem with researchers being able to waive assent is that if you ask for the kid’s assent, you should respect it," she says. "My issue is that there are a lot of situations in which I just would not ask for the child’s assent. I would inform the child. But I get very nervous when you have different parts of the protocol where the kid needs to sign, and if the child says no, you’re still going to do it, anyway."

In England, researchers are required to get assent from any child who is age 7 or older, notes Paul J. Edelson, MD, a researcher at the Center for the Study of Society and Medicine in the Columbia College of Physicians and Surgeons in New York City. He is conducting a three-year study of the history of children as research subjects. "They have taken a nuanced approach that says while that person is still too young to judge independently whether he or she can consent, you must get the assent," he says.

Children are equally entitled to personal respect, just as adults are, Edelson notes. But the respect may be played out in a different way — seeking assent in certain situations, not in others, making policies to respect assent when it is sought, etc. "I think it should be the first step in encouraging a longer discussion with the child and family about participation," he says.

Protection vs. inclusion

During the 1980s, with the memory of past abuses of children still strong, very little medical research included children as subjects, Friedman Ross says. "Some people argue that because children are not competent to give informed consent themselves, they should be excluded from research," she notes.

But with the advent of the AIDS epidemic, researchers began questioning the ethics of excluding children from trials of potentially lifesaving or life-prolonging medications. "All of a sudden, there is a realization that research can be very therapeutic," she says. "All of the drugs that were approved for AIDS didn’t get approved for kids for like five or six years because of the way the studies had to be done."

In the 1990s, medical research moved from being protectionist to emphasizing access, Friedman Ross says. The FDA moved to give companies that conducted research trials in pediatric populations an extra six months of exclusivity of marketing their product. And in 1998, the NIH developed guidance for researchers that stated they must include children in research protocols or be able to state why they did not. Programs that were unable to justify why not received less money, she adds.

"There have been enormous tidal changes in what it means to be a research subject," says Edelson. "I think the pendulum has swung to the point where you feel you may be unjustly disenfranchised somehow if you are not invited to be a subject." In the 1990s, the emphasis shifted from people willing to participate in research because of some inducement to people demanding to participate in order to get access to care or to new medications, he adds.

With children, it is important that they are included in medical research, particularly drug trials, because treatments and medications have different effects in the pediatric population. "What happened for a long time, because children were not seen as a significant market for many high-tech drugs, companies would do studies on adults, get the adult indications, then they were done," Edelson explains. "They received permission from the government to market the drug for a specific use. But once available, a physician has the power to use it any way he or she decides is appropriate. You had all of these drugs that were tested only in adults, but for which there was no alternative but to use them in pediatrics because there was no pediatric equivalent."

That led regulators to question the safety of using drug dosages in pediatric patients that had never been tested. That led to the FDA’s passage of the Best Pharmaceuticals for Children Act that encourages drug trials for children. "There are legitimate reasons for children’s study to be encouraged and promoted, and they need to be," Edelson says. "On the other hand, there are many forces at work."

The drive for access, many now fear, is in danger of going too far. "In certain situations, like AIDS, a [then] nontreatable, deathly illness, you might want to include children more quickly in certain research," Fried-man Ross says. "But there are many reasons why we still need to be protecting kids differently."

Me-too’ trials

An area of key concern is placebo-controlled pediatric drug trials, Friedman Ross says. For every 100 drugs that get to market in the United States, almost half are taken back off the market due to health complications that are discovered once use of the medication is expanded, she states. For every 100 drugs that research is performed on, only one drug actually makes it to market approval. "There is good reason to want to go slow with kids, especially with the so-called me-too’ drugs, drugs that are essentially just like a drug on the market that has already proven effective," she says.

FDA regulations require placebo-controlled trials of all new drugs coming on the market. So children randomized into the control arm of a trial of a me-too drug will get a placebo — not the drug already on the market that has proven effective and safe treatment for the same condition, she says. "In these situations, we know there is a drug available that is good for kids, but we are not going to give it to them. We are going to test them with a new drug, that is probably good, but some of these kids will be on placebo when we know they need medicine," she explains.

Unacceptable levels of risk?

In addition to discussing ways to improve informed consent for children involved in research, it is also important for IRBs to examine whether there are some levels of risk that are inappropriate to expose children to, whether the child assents and the parents consent or not, Friedman Ross says. "If you are going to talk about consent, you have to talk about whether parents have the right, in a sense, to expose their children to risk for no benefit to the child," she says. "Some experts have argued that children should not participate in nontherapeutic research even if it involves no risk."

But parents place their children at risk all the time, even when there is no benefit to the child; so it is pointless to argue whether they have the right or not, Friedman Ross says. "If I take my children to the store to buy a bottle of wine for Mommy and Daddy in the rain, I am putting them at risk of getting into a car accident or something, and that is not for their benefit," she illustrates. "Parents do things like this all of the time."

Parents obviously have an obligation not to abuse or neglect their children and obligations to help develop and protect the child, she says. But parents need a wide latitude of freedom about things that are clearly not harmful or neglectful, that may not promote an individual child’s best interests, yet may be promoting the family’s interest or the interests of other people. "I am comfortable saying that parents should be able to enroll their children," she says. "Yet, I like the idea of an IRB limiting the amount of risk that they are willing to expose children to, because children do not belong to their parents."

IRBs sometimes forget they have the function of protecting subjects, not just ensuring that the subjects give informed consent to whatever level of risk the researchers are proposing, she adds. "We focus a lot on consent," she says. "But consent is the second step. We need to focus on research that is as minimally risky as possible, especially for children. The 1990s regulations emphasized equitable access to research benefits. But don’t forget that for every benefit, there are hundreds of risks."