Drug Regulations and the Paediatric Promise
The introduction of the sulphonamide antibiotics to clinical practice also indirectly had a major impact on drug regulation. While drugs have been regulated in the UK since 1540 – initially under the Apothecaries Wares, Drugs and Stuffs Act – the primary focus of concern, as was the case drug regulatory agencies in most countries up to the early 20th century, related to identifying commercial preparations for the purpose of securingpatent and proprietary rights. Early regulatory frameworks paid little attention to the efficacy or safety of approved medications, notably with respect to use in children. A classic case study of this disregard is the approval of a product known as Dr. James Soothing Syrup Cordial, advertised as a cure for cholera infantum or infectious diarrhoea and suggested for use for colic and teething pain (11). The active ingredient in Dr. James Soothing Syrup Cordial was heroin – which the company claimed was not habit forming. Other examples include cocaine drops marketed for teething pain and marijuana cigarettes, listed on the US Pharmacopeia for headache.
Following the introduction of the first sulphonamide preparations on the market, it became clear that the very large commercially available tablets were difficult for children to take, especially children under the age of six. To address this problem, a chemist at S.E. Massengill dissolved sulfanilamide in diethylene glycol, an effective solvent and, unfortunately, a potent nephrotoxin. With the addition of pleasantly flavored raspberry syrup, the liquid sulphonamidewas then marketed as Elixir of Sulfanilamide in September of 1937. Although the toxicity of diethylene glycol was known –and had been documented – since 1930, this had not been appreciated by the development team at Massengill. In addition, prior to commercializing the product or obtaining approval to market, no animal testing was performed and no rigorous safety appraisal was completed, as neither were required for such testing at the time– and Sadly, very shortly after Elixir of Sulfanilamide became available, a number of patients – mostly children – sustained significant renal injury and more than 100 died (11, 12). The public outrage following this event lead to the development and passage of the 1938 Food, Drug and Cosmetic Act, the legislation that was fundamentally responsible for the structure of the US Food and Drug Administration (“FDA”) as it exists today. This legislative framework– which established that Federal drug regulatory agencies have a fiduciary duty to ensure that marketed drugs are safe – has been adopted by national drug regulatory agencies worldwide.
Changes in drug policy and regulation driven by tragedies in childhood unfortunately did not end with the Elixir of Sulfonamide disaster. Despite the preliminary protections provided by the early versions of the Food, Drug and Cosmetic Act, paediatric patients remained a uniquely vulnerable population with respect to marketed drug products. As new therapies for a variety of conditions became available, paediatricians and other child health care providers rapidly adopted them into practice. This use was primarily based on literature from adult populations, as the need for paediatric testing was not required, and was rarely available. . For example, after the discovery of the antimicrobial efficacy of chloramphenicol in 1947, the drug was widely used, including for the therapy of presumptive sepsis in neonates. When used for this indication – without efficacy or safety data in children - a number of babies developed cardiorespiratory collapse, termed the Chloramphenicol Grey Baby Syndrome (13, 14). It has subsequently been recognized that this was due to the developmental immaturity of drug metabolizing systems, specifically glucuronidation, paired with a developmentally driven reduced renal clearance of the parent drug in the neonate resulting in accumulation of toxic concentrations of chloramphenicol (15). This, in turn, produced mitochondrial dysfunction and cellular toxicity, impairing cardiac function resulting in the Chloramphenicol Grey Baby Syndrome. This tragedy was key in driving foundational research in developmental pharmacology, manifesting the truth of Abraham Jacobi’s quote 1889 that “pediatrics does not deal with miniatures men and women, with reduced doses and the same class of diseases in smaller bodies, but it has its own independent range and horizon and gives as much to general medicine as it receives from it” (16).
A further tragedy – and one with equally monumental consequences for pharmacotherapy for children – occurred when the German pharmaceutical company Chemie Grünenthal marketed thalidomide as a sedative-hypnotic in 1957 (17). The product launch was very successful; it was marketed in a number of other European countries including the UK, and was used by more than a million women in Germany alone (18). Four years after the product launch, William McBride, an Australian obstetrician, described an association between in utero exposure to thalidomide and congenital limb malformations, more specifically phocomelia, preferential shortening of the upper limb segments which has a natural incidence of between 1 and 4 per 100,000 live births but was seen in up to 20% of women taking thalidomide over the time of embryogenesis (19). This was subsequently confirmed by Lenz, with approximately 8,000 children effected worldwide, mostly in West Germany, the UK and Japan (1, 20). While there were some cases in North America, there were relatively few, as thalidomide was only approved in Canada in April 1961, and subsequently withdrawn in 1962. Thalidomide was never approved in the United States due to the concerns of Dr. Francis Kelsey, a reviewer at the FDA who was troubled by reports of embryopathy from Australia and Europe. Of note, Dr. Kelsey’s graduate work at the University of Chicago had included studies with Gelling exploring the toxicity of sulfanilamide.
Despite the fact that thalidomide had never been approved in the United States, knowing that thousands of children worldwide had been harmed by drug therapy during pregnancy lent urgency to critically important drug-related legislative changes proposed by Estes Kefauver, a Senator from Tennessee, and Oren Harris, a Representative from Arkansas. This important legislation – known as the Kefauver-Harris Amendment – was passed in October 1962, and served to strengthen the Food, Drug and Cosmetic Act in the United States. These changes made the approval of the Food and Drug Administration mandatory prior to the release of a drug to market, and granted powers to the FDA enabling them to require manufactures to demonstrate the effectiveness of their products and report serious adverse events following market authorization. These changes also transformed quality of clinical trials to be used for the purpose of drug approval, set controls for the marketing of generic products and established good manufacturing practices for industry clearly delineated the requirement that drugs must be shown to be both safe and effective in the populations for which they were indicated before they could be approved for marketing and regulated the sale and marketing of generic drugs as well as the conduct of clinical trials (21). This amendments had an immediate and profound effect on the drug approval process notably on the quality of new drug submissions, and are credited with transforming the landscape of drug development and approval.
Unfortunately, this landmark Amendment also had an equally profound and quite unintended effect on pediatric pharmacotherapy. With an aim to secure the safety of marketed pharmaceuticals, the Act stipulated that there must be substantial evidence of efficacy pursuant to the conditions of use as outlined in the product monograph. Adverse effects of the drug also needed to be studied during drug development and had to be stated in detail in the product monograph.. The unique circumstances of conducting research in children at the time created a number of pragmatic problems that made this difficult including a relative lack of clinical trial data in children, small population sizes in most centres mandating multi-centre and network studies, lack of validated end-points for clinical trials, the ethics of appropriate informed consent and cost of conducting studies. In many cases, it was determined that if the product indication did not include children there was no legislative or regulatory requirement to conduct studies in children. This frequently led to strategic decisions during drug development to not include children in clinical studies of new therapeutic agents. Therefore, as most new drugs were not studied in children prior to market authorization, to ensure compliance with the regulations, product monographs often contained exclusionary language that recommended against use of the drug in children.
This paradoxical situation where legislation intended to enhance safety testing in children produced a decrease in drug research in children, resulting in the majority of drugs used in the care of children being used “off label”, that is to say without guidance from the product monograph – and often without evidence to guide clinicians as to dose or safety, a problem for children most eloquently described by Harry Shirkey in 1968 as children being “therapeutic orphans” (22).
Thus, an important amendment intended to protect all citizens from harm by strengthening the authorities of the FDA, and enforcing regulations designed to increase the safety and efficacy of prescribed medications, resulted in children being excluded from the new regulatory protections. For many years following the passage of the Kefauver-Harris Drug Control Act, there was neither incentive nor regulatory imperative to include paediatric data in new drug submissions unless the drug was specifically indicated for children.
This issue persisted for decades, despite calls by professional societies and pharmacologists for change (23-26), and despite clear and compelling evidence that there was “off label” use was becoming an unacceptable norm for paediatric practice (23, 40, 41). Coupled with the barriers identified with respect to the conduct of clinical trials in children, there has been a persistent and incorrect belief that “children rarely require prescription medication”. This, however, is simply not the case. Children are frequently prescribed medications, and for a wide range of indications (27-39). Having studied a cohort of nearly a million children in Canada for a year, researchers found that, on average, children received 4 prescriptions per year (27, 28). American studies have demonstrated similar numbers; in one study, 27% of all children studied had received a prescription in the past 30 days (33, 36). For inpatient care, a multinational study of drug utilization demonstrated that 96% of children admitted to hospital in the UK received at least one prescription medication (34).
In addition to a large number of medications being prescribed, there are large range of medication classes used (27-36). While antimicrobials were found to be the most commonly prescribed drugs for children, up to 1200 other drugs from a wide range of classes were used (27-36). Although the use of drugs for children is to some extent variable from country to country, probably reflecting national differences in child care practices and drug availability, one trend that has emerged as well was that over time fewer antimicrobials were used and more psychoactive drugs are being prescribed (34, 36). To add complexity, over the past decade, the number of biological agents prescribed to children has increased almost logarithmically (37-39).
An additional consideration is that the use of pharmacotherapy for children is not uniformly distributed across the population. For many children, pharmacotherapy is indeed infrequent and largely confined to antimicrobials and respiratory medications. However for children with serious and/or chronic disease, therapy is not only frequent but also frequently involved polypharmacy (27, 28). As an example, it has been demonstrated that for children with epilepsy they were treated with up to seven different drugs over the first year after diagnosis (28).