Paediatric drugs are finally being tested on children. But, as pharmaceutical companies vie to capture part of the market, Lainie Friedman Ross fears that children aren't always benefiting.
The phrase "women and children first" refers to the rescue policy on a sinking ship. But in medical research, the traditional policy has been "women and children last".
The justification was paternalistic: women and children were vulnerable and needed additional protection. However, in the past decade, policy and guidelines in the US and the UK have shifted away from the women and children last philosophy. While the shift for women is long overdue, that for children is less clear cut. Those who favour it point out the negative repercussions caused by the additional protections: health issues unique to children are understudied and underfunded.
These concerns were expressed as early as the 1960s when Harry Shirkey, a US paediatrician, said children would become "therapeutic orphans" if denied access to new drugs because government and industry did not support paediatric drug investigations. In 1995, his fears were confirmed when the American Academy of Pediatrics stated that more than 80 per cent of drugs prescribed to children had not been tested on them.
By the late 1990s, the pendulum in the US had swung away from the children-last philosophy as researchers emphasised the need for children to have access to research participation. The justifications given were the need to improve scientific understanding of paediatric health and to ensure that medical advances would accrue to children.
Policies were promulgated by the National Institutes of Health, the US Food and Drug Administration and the US Congress that encouraged the participation of children earlier in the research process. In the UK, policy changes over the past two decades have also increased the participation of children in research. Whereas in 1962-63, the Medical Research Council argued that it was illegal for children younger than 12 years of age to participate in non-therapeutic research, in 1980, the British Paediatric Association contended that "non-therapeutic research is not necessarily either unethical or illegal". In 2000, the BPA, now the Royal College of Paediatrics and Child Health, reaffirmed a preference for using adults in research, but stated that "a research procedure that is not intended directly to benefit the child subject is not necessarily either unethical or illegal".
The evolving policy led Priscilla Alderson, a British sociologist, to write an article in the Bulletin of the Institute of Medical Ethics titled, "Did children change or the guidelines?" However, although the policy changes have increased the participation of children, it is not clear whether researchers have achieved their goals to advance pediatric knowledge and promote paediatric health.
Let me illustrate my concern with US clinical asthma trials. In the US paediatric population alone, asthma affects about 5 million children, although there is evidence to suggest that a significant number of children remain undiagnosed. Asthma accounts for 166,000 hospitalisations and 200 deaths a year among US children under 15 years old.
In 1998, the cost of asthma in the US was estimated to be $12.7 billion (Pounds 7.36 billion). Prescriptions accounted for about 25 per cent of the cost. The two main classes of asthma drug therapy are anti-inflammatory medication and smooth muscle airway relaxants. Anti-inflammatory medications are recommended as daily prophylaxis for all children with asthma except those with the most modest intensity. Since the 1980s, inhaled corticosteroids (ICS) have become the anti-inflammatory medication of choice. Not surprisingly, a number of pharmaceutical companies want to capture part of the market and have developed their own ICS.
Between January 1, 1998, and December 30, 2001, my colleagues and I identified 70 clinical asthma trials performed in the US that included at least some paediatric subjects. Sixty-six of these studies documented partial or complete pharmaceutical funding, suggesting that most were sponsored by pharmaceutical companies to obtain FDA approval. Forty-five of the 70 studies were placebo-controlled trials, meaning that the experimental inhaled corticosteroid was compared with a "dummy pill", in contrast to clinical trials that compared one anti-inflammatory medication with another. Placebo-controlled trials are considered the gold standard by the FDA because it wants the drug to be shown to be superior (better than nothing) rather than equivalent (that the new compound is as good as an already marketed anti-inflammatory medication). While placebo-controlled trials may be scientifically neater, they expose children with asthma to the risk of a potentially life-threatening attack.
Our analysis found that subjects in placebo-controlled trials were more likely to have an acute exacerbation of asthma than subjects in equivalency trials. We also found that subjects in the placebo arm did worse than peers receiving the experimental drug. This was not a surprise to the researcher who commented: "Asthma symptoms would be expected to worsen in the placebo group during the treatment period because these patients were dependent on inhaled steroids but were not allowed treatment with inhaled steroids while in the study."
New US policies expose children to additional risks because they have led to an increase in the number of children enrolled in adult trials. In fact, 52 of the 70 US asthma trials enrolled children and adult subjects, and only 18 enrolled exclusively children. Expanding the research to include children should enhance the wellbeing of the individuals or promote knowledge about children as a class. One would therefore assume that any clinical drug trial that enrolled children would have as one of its goals an assessment of the safety and efficacy of the drug on children.
Unfortunately, only eight of these studies differentiated between adult and paediatric subjects at baseline, and only one of the 52 studies that included children and adults did subset analyses to determine if treatment responses in children were different to responses in adults or to see whether adverse events and withdrawals occurred more frequently in paediatric or adult subjects. That is, children helped researchers achieve accrual goals, but the research results did not advance paediatric knowledge.
In response to two asthma studies that included subjects older than 12 years and were published in 2001 in the Journal of the American Medical Association , I wrote a letter asking researchers why they included adolescents if not enough children were to be enrolled to make useful subset analyses. One principal investigator agreed that "studies involving children must balance the generalisability of results against the risk of participation", but another PI responded that "neither trial was designed specifically to evaluate if the response of children or adolescents differed from adults; rather patient selection was based on criteria that would permit the results to be generalised to the patient populations for which these medications were approved by the US Food and Drug Administration."
That is, the second author was following the letter but not the spirit of the new policies that require the inclusion of children. I would have liked to ask the second principal investigator to provide moral justification for enrolling children if there were no plans to analyse paediatric data separately. While it helped researchers achieve enrollment criteria more quickly, using children in these studies was unethical because it placed some at risk without plans to benefit individuals or children generally.
Children are a vulnerable population in research and need additional protection, and we should maintain a children last philosophy with two caveats. First, "children last" does not mean that children should be ignored. I support paediatric research when the research is promising and can be done only in children.
Second, "children last" should be a presumption, but exceptions should be permitted. For example, exceptions are desirable if the condition is life-threatening and no therapy exists. If children are enrolled, there must be a plan to enroll enough children to perform subset analyses in order to benefit children as a class. This is clearly stated in the guidelines of the Royal College of Paediatrics and Child Health; it is not so clearly stated in any US guidelines.
But in both countries, the loosening of restrictions could present problems. The current policies to promote paediatric participation in clinical trials are being interpreted to permit too much focus on access, exposing children to significant and unnecessary risks. The pendulum has swung too far.
Lainie Friedman Ross is associate professor of paediatrics and associate director of the MacLean Centre for Clinical Medical Ethics, University of Chicago. Her book Children in Medical Research is published by Oxford University Press.