FormalPara Key Points

The “therapeutic orphans” dogma has had negative as well as positive effects.

Some paediatric studies required to market drugs are impractical, scientifically unjustified, or ethically questionable.

The methods designed and used to test new drugs in adults may not be appropriate or even needed in all minors in order to enable the safe and effective treatment of children and adolescents.

1 Introduction

Since 2007, when the European Union (EU) paediatric legislation came into force [1], new drugs need a paediatric investigation plan (PIP) accepted by the European Medicines Agency’s (EMA’s) Paediatric Committee (PDCO) for registration [2, 3]. PIPs must discuss the adult disease, its counterparts in children, and “propose” child-friendly formulations (e.g., syrup), juvenile animal studies, clinical studies, and more [1,2,3]. The aim of this article is to critically evaluate how well the EU paediatric legislation is achieving the stated goal of the PIP system: “to improve the health of the children of Europe” [3]. We analyzed the intellectual roots of the US paediatric legislation [4, 5], which preceded the EU legislation, especially key publications by Shirkey [6] and statements by the American Academy of Pediatrics (AAP) [7, 8], then analyzed the EU paediatric legislation (EUPL), PIP decisions and decision patterns for their effects on paediatric healthcare.

2 Children as “Therapeutic Orphans”

Since the 1962 amendments to the US pharmaceutical law, drugs have needed proof of safety and efficacy (S&E) from clinical studies [4, 5]. These amendments were introduced in response to the thalidomide catastrophe. Gradually, most countries introduced comparable legislation. Regulatory authorities became a third pillar of the healthcare system, alongside the medical profession and industry. To avoid liability lawsuits, manufacturers emphasized whenever their respective drug had not been investigated in children. Shirkey in 1963 concluded that this denied children the use of new medications and made them “therapeutic orphans” [6]. The AAP took up Shirkey’s concerns. It stated in 1977 that use of drugs not labeled for children created a dilemma for physicians—avoid them and deprive children of potential benefits, or prescribe them despite the lack of Food and Drug Administration (FDA) certification for children—and that it was “unethical to adhere to a system which forces physicians to use therapeutic agents in an uncontrolled experimental situation virtually every time they prescribe for children” [7]. In 1995 AAP claimed, “There is a moral imperative to formally study drugs in children so that they can enjoy equal access to existing as well as new therapeutic agents” [8], that “When drugs have been administered to children without sufficient pharmacology studies to identify the optimal therapeutic approach, children have occasionally suffered severe toxic effects, including death” [8], and “Growth, differentiation, and maturation can alter the kinetics, end organ responses, and toxicities of drugs in the newborn, infant, child, or adolescent as compared to the adult” [8].

In fact the 1962 US pharmaceutical law did not distinguish between children and adults; it addressed “human beings” [4, 5]. Paediatric disclaimers had limited legal weight. They protected manufacturers in lawsuits and were sometimes used by insurance companies as a reason to not pay, but neither labels nor the FDA could forbid the drugs’ use in children [9]. Drugs with paediatric disclaimers could not be advertised for children because after 1962 the FDA controlled drug advertising [10]. Shirkey’s claim that children were denied the use of many new drugs was misleading: the 1962 amendments did not specifically restrict drug use in children [9] and the AAP’s 1977 warning about toxic effects and death (repeated in 2010 [11]) referenced only two publications on antibiotics in premature babies [8]. These publications did not prove there were different “kinetics, end organ responses, and toxicities of drugs” in all age groups.

In fact, AAP documents on off-label use emphasized repeatedly physicians’ right to prescribe what helped. The AAP supported knowledge-based, off-label treatment of children: “Evidence, not label indication, remains the gold standard from which practitioners should draw when making therapeutic decisions for their patients” [12]. Yet the AAP also saw paediatric labels as the path towards optimal clinical care for children.

Additionally, both Shirkey and the AAP used “children” in its contemporary sense, i.e., persons below the age of majority. There was a lack of appreciation that a legal rather than a physiological definition for “children” was used. Endorsed and amplified by AAP statements that toxicity and death are risked by treatment of those who are legally minors, the “therapeutic orphans” concept resulted in the AAP and FDA concluding that minors needed separate proof of S&E studies. This has contributed to the development of paediatric clinical pharmacology, to US paediatric pharmaceutical legislation, and other positive outcomes, but has also become a dogma that needs critical reassessment.

3 Progress in Child Healthcare and Labels

Many diseases with high under-five mortality were gradually prevented by vaccination, better housing, sanitation, nutrition and other improvements; others became treatable. Paediatric oncology evolved with systematic use and investigation of adult cytotoxics in children [13, 14]. It did not focus on labels. Many paediatric subspecialties successfully used new drugs when they became available, including neonatology [15] and paediatric cardiology [16]. Children were not “therapeutic orphans”. The call for them to benefit more from pharmaceutical progress is justified and well-intended, but real-world observations show that their healthcare is not necessarily or exclusively linked to paediatric labels.

4 Paediatric Clinical Pharmacology, US Paediatric Legislation, and the FDA

Absorption, distribution, metabolism and excretion (ADME) are clearly different in newborns and infants [17]. Since 1997 US paediatric legislation has offered patent extension in exchange for doing FDA-requested (usually industry-suggested) paediatric studies. This resulted in many regulatory paediatric studies done by industry [2, 18]. However, the ADME differences between adults and children are most pronounced in newborns and infants; differences become less with increasing age [17]. The upper age limit for “children” for the FDA is 16 years. In 2014 the FDA explained its current view of evidence needed for paediatric labeling [19]: (1) paediatric clinical studies in an indication different from the adult one; (2) paediatric studies in the same indication as the adult one; and (3) adult studies plus additional information in the specific paediatric population. Point (3) is a departure from the therapeutic orphans dogma. Indeed, in 2016 the FDA expanded the indication for partial onset seizures for anti-epileptic drugs down to 4 years [20], in 2017 it approved avelumab down to 12 years, based on only additional population pharmacokinetic (PK) data [21], and recently approved ivacaftor for further cystic fibrosis mutations, based only on in vitro studies [22].

5 EU Views of Off-Label Use

Some European publications on paediatric off-label use were pragmatic [23], some conceded that adverse drug reactions might be balanced by increased therapeutic efficacy [24], and some were accusatory, e.g., “Children and infants deserve the same right to treatment as adults” [25]. The latter was not supported by data, but was based on a moral argument, similar to the AAP’s 1995 statement [8].

5.1 EU Debate and Paediatric Legislation

A 1998 EMA report warned that a “lack of appropriate dosage recommendations for children” resulted in frequent off-label use in children [26]. But paediatric dosing recommendations did exist. Paediatricians and physicians were trained to use them. Children were treated based on their doctors’ best knowledge and opinion. Adjustment of practice in the light of new findings is a normal process in medicine. But instead regulatory studies in children of all age groups were required [27], without differentiating between subpopulations or considering whether studies mirroring those for adult approval are necessary for all minors.

A 2004 EMA paper characterized off-label use in children as dangerous without mentioning paediatric oncology, neonatology, other paediatric subspecialties, or differentiating between beneficial and potentially dangerous off-label use [28]. The EU arguments were and are formal, often biased and ignore many pathways of knowledge transfer. Physicians first study and receive clinical training, then gain experience through patient care and continuing medical education. The clinical impact of drug labels depends on physician and setting. Paediatric oncology and neonatology are practiced in specialized centers where label changes will change little. In contrast, label changes can be expected to impact general practitioners more.

5.2 The EU Law

Recital 3 and the definition of the “paediatric population” in article 2(1) of the EUPL are shown in Table 1. These reflect AAP statements discussed above, and do not differentiate between paediatric subpopulations. Not all minors below their 18th birthday need separate adult-style S&E studies. The 18th birthday is a legal, not a physiological border.

Table 1 EUPL recital 3 and article 2(1)
Full size table

5.3 Positive Impact of PIPs, and PIP Shortcomings

US and EU paediatric legislation have both improved understanding of and support for paediatric studies in academia, industry, regulatory bodies, governments and the public. This has increased the number of investigators, industry experts, regulators, legislators, parents and patients who are interested in as well as capable of making meaningful contributions to our understanding of what studies need to be done and how better to do them.

Some examples of problematic outcomes of the PIP system follow.

5.3.1 PIP-Required Clinical Studies

PIPs are required at the end of phase 1, when no efficacy data exist yet. PIPs must address the targeted adult disease, corresponding childhood diseases, and “propose” juvenile animal studies, formulation development (e.g., syrup), clinical studies, and more [3]. PIPs do not prioritize even when multiple drugs are being developed for the same adult disease that has a vanishingly low incidence in children. The feasibility of any single drug’s PIP is based on the assumption that the entire paediatric population with a given disease would be available to recruit for that drug. Studies of other drugs being studied or even already approved in other countries for the same disease or indication are not considered. The EMA PDCO decides about PIPs even though final authority about drug registration rests with another EMA committee. If the PDCO judges “proposals” to be insufficient, the PIP is rejected and adult approval is blocked. Thus, sponsors can be pressured to make study “proposals” that they feel are impractical, unnecessary or even undoable. The system rewards “proposing” such studies with permission to proceed with drug development, while discouraging reasoned dissent. PIP decisions are published on the EMA website and can be googled using the PIP number. The PIP system requires studies in the “paediatric population” (Table 1, [1]) that mirror adult registration studies.

5.3.2 Questionable Studies in Adolescents

PIPs routinely require S&E studies either in adolescents alone or in both adolescents and children.

Several PIPs request studies in “children” between the body’s maturity and 18th birthday, maturity being defined by sexual puberty or by closure of the growth plates:

  • EMEA-001264-PIP01-12 (chondrocytes for joint repair) requires a study in children between closure of growth plates and the 18th birthday.

  • EMEA-000250-PIP01-08-M02 and EMEA-000658-PIP01-09 (contraceptives) require PK comparisons between young women after menarche and before their 18th birthday versus women 18–50 years old.

  • EMEA-001492-PIP01-13 (trifarotene, anti-acne) requires one study of different trifarotene concentrations versus placebo in “children” versus adults, and one of trifarotene versus vehicle, again in “children” versus adults; “children” are defined as “from puberty to less than 18 years of age”.

Clinical studies in older adolescents or young adults can make sense, such as studies of some psychiatric drugs, but they need not be done exclusively in adolescents, or they could be done using opportunistic study designs. However, the examples above and studies that include doses already shown to be sub-therapeutic in young adults (see Sect. 5.3.3) are unlikely to produce therapeutically relevant findings and could be considered unethical [29].

5.3.3 Unfeasible Studies

In 2007 the EMA established a list of PIP-exempted diseases (“class waivers”). These were diseases where paediatric studies were judged to be either unfeasible or unnecessary. Some waivers were soon revoked [30], including for melanoma in children >12 years of age, resulting in the initiation of unfeasible studies. The EMA claimed that enough adolescents with melanoma exist for clinical trials. Twelve melanoma PIP trials were approved [31]. Six company-sponsored international studies were initiated either for minors only with melanoma or with one of a number of malignant tumors including melanoma [31,32,33,34,35,36,37]. Two were terminated in 2016 due to slow recruitment [31,32,33]; four are ongoing [31, 34,35,36,37]. Six adolescents with metastasized melanoma were enrolled in a vemurafenib “dose escalation” trial that included treatment below the FDA-approved dose [31, 32].

PIPs routinely require paediatric studies for all new compounds that target predominantly adult diseases; even when the diseases only very rarely occur in minors. In addition to melanoma [31, 38, 39], the same approach has been used for leukemia subtypes [14], multiple sclerosis [40] and psoriasis [41]. In 2015 the EMA again revised its class waivers, making PIPs mandatory for diseases even more rare in children, including liver cancer [42].

5.3.4 Questionable Trials in a Frequent Disease

Based on a standard allergen PIP developed by the EMA and the Paul-Ehrlich-Institute, 58 double-blind, 5-year, placebo-controlled studies of specific immunotherapy (SIT) should enroll many thousands of children and adolescents with allergic rhinitis into PIP-required studies scheduled until the end of 2031. These PIP-required studies would prevent effective treatment of allergic rhinitis during years of placebo treatment, exposing some minors to progression of disease to asthma as a result of the “atopic march” [43]. Switzerland, however, simply allowed retroactive approval of SIT allergens [44].

5.3.5 The EMA’s 10-Year Paediatric Reported “Successes”

The “successes” mentioned in this report [45] are misleading at best. The “increasing numbers of medicines becoming available to children” (page 4 of the report) refers to the number of drugs with new paediatric labels rather than to drugs made available by the EUPL. The higher percentage of trials that include children is also given as evidence of success (Figure 2 on page 6 of the report), irrespective of what percentage of trials failed to produce meaningful data or whether or not the trials made scientific or clinical sense. Some of the successes listed are of limited clinical importance. For example, two oncology “successes” are reported: a new asparaginase now “available” for acute lymphoblastic leukemia (ALL), and dinutuximab for neuroblastoma. ALL has been successfully treated for decades using asparaginase-containing regimens, with survival rates approaching 90%. Asparaginase was available for adults and children without a PIP. Additionally, neuroblastoma occurs only in paediatric patients. Dinutuximab did not need a PIP to be developed. In the USA, for example, studies of new oncology drugs are approved only by the Oncology Division of the FDA. Even in the EU, oncology marketing authorization decisions are made by the EMA, not by the PDCO. The PIP system is redundant and unnecessary for drugs used exclusively in a paediatric population.

6 Discussion

Shirkey, the AAP and FDA noted that more information on drugs and their interaction with the child’s body (and vice versa) were needed for improved pharmaceutical treatment, specifically in small children. The AAP and FDA concluded that the “therapeutic orphan” should be fixed by more paediatric labels based on separate S&E trials in all age groups, almost as if they were another species. Since then, paediatric clinical pharmacology has expanded, attitudes about clinical studies in children have changed, paediatric research infrastructure and training has improved dramatically, and industry has performed many paediatric regulatory studies. Today, our understanding of the developing body is much greater than half a century ago. But the term “therapeutic orphans” also ignored the difference between small children and adolescents and the positive effects of some “off-label” treatments. This was amplified when the AAP extrapolated dangers, including death, from use of antibiotics in premature newborns to the entire paediatric population and claimed that ADME are different from adults in children of all age groups [7, 8, 11]. This dogma also initially resulted in the FDA requiring only adult-style S&E studies for paediatric labeling. Recent FDA decisions such as extrapolation of efficacy for antiepileptic drugs [20] and the anticancer compound avelumab [21] suggest a rethinking. The EMA has not shown comparable willingness to change its approach.

The EU paediatric legislation has certainly advanced awareness that clinical research in children is needed. But defining the “paediatric population” as everyone under age 18 (Table 1) and routinely requiring S&E studies that mirror the adult testing paradigm is not science-based and can harm patients enrolled in questionable studies. The adolescents enrolled in the vemurafenib “dose escalation” trial [31, 32] should have been treated as adults. Instead, they became a commodity required for drug approval. This PIP-requested trial lacked common sense and exposed patients to unjustified risks/harm [31, 38, 39]. The more compounds are developed for liver cancer [42], the more rare underage patients with liver cancer will be searched for to enroll in PIP-required studies that cannot enroll sufficient numbers for meaningful results. These rare diseases patients deserve treatment driven by therapeutic beneficence [29], not by dogma.

With the exception of psychiatric drugs and perhaps a few others, adult doses can be used in adolescents. In school-age and younger children, dose estimation using modeling and simulation is possible with confirmation by small PK/pharmacodynamic (PD) studies or with opportunistic PK/PD sampling of adolescents and children treated off-label [31, 46,47,48].

There are a number of limitations to the reach of paediatric regulatory authority in the US paediatric legislation that do not exist in the EU paediatric legislation. Thus, there are no mandatory paediatric studies for orphan diseases, mandatory studies are required only in the same disease as adults, and there are no additional requirements for drugs developed solely for paediatric diseases. Minors do not need just more studies, but rather reasonable approaches to obtain the information needed for effective and safe treatment.

Doing unnecessary or unfeasible global studies that are terminated because they fail to recruit represents considerable, unjustified costs. So far, industry has accepted this as bearable and necessary for adult development and registration. Such studies, however, have negative impacts on both the cost of medicines and on enrolled patients.

7 Conclusions and Alternatives

The FDA, AAP and EMA have claimed that the best way for children to benefit from pharmaceutical progress is more paediatric labeling. Labels provide important information, but responsible physicians will always use what they consider as the best available treatment, including off-label. There are also ways to develop useful paediatric dosing information other than adult-style S&E studies. Examples include the use of registry data and opportunistic use of PK/PD data collected in children receiving off-label treatments. Additional improvements to the EU paediatric legislation are beyond the scope of this manuscript, but would include exempting drugs being developed for purely paediatric, rare or orphan diseases from the PIP process; allowing for prioritization when multiple drugs are being developed for the same rare disease or condition; and delaying the PIP until at least some adult efficacy and toxicity data are available. Other concrete suggestions for change include improved external oversight/review of PIP decisions, as well as regularly scheduled, very clearly described, more clinically meaningful outcomes for the PIP system. Examples include the number of studies that resulted in statistically valid results or meaningful changes in practice or dosing rather than just the number of PIP studies done.

US and EU paediatric legislation has treated all paediatric subpopulations the same. With few exceptions, adolescents with an adult body should receive adult doses. If reimbursement institutions refuse to pay, this is a legal, not a medical problem. Legal challenges should not be fixed with medical actions. New pharmaceutical legislation should explicitly allow physiologically adult adolescents to receive appropriate drugs and doses based on the treating physician’s educated opinion.

Today’s industrial and scientific revolution offers an increasing number of new medications for both adult and paediatric diseases. For adult diseases that rarely also affect children, there are better ways to make treatment available than blindly enforcing adult-type regulatory trials in all paediatric subpopulations, especially when multiple drug studies compete for the same very small number of rare paediatric patients; this is being done by four ongoing studies recruiting minors with melanoma or other solid tumors [31, 34,35,36,37].

The potential benefits of mandatory consideration of children in drug development should neither be ignored nor overestimated. Legislation can offer a useful framework for drug development, but cannot replace scientific progress. Asking for child-friendly formulations is reasonable; juvenile animal studies routinely requested by the EMA/PDCO while of questionable value do not harm children. Indiscriminately requiring undoable, unjustified or poorly designed paediatric studies can.

Companies who resist doing such studies risk being criticized for being against “better medicines for children”. In the short-term it can be easier to accept the requirement to conduct a study even when management feels the study makes no sense or is likely to fail to recruit. But companies have both a moral and fiduciary responsibility to protect children against unrealistic or irrational EMA/PDCO requirements. They can seek help in doing so from institutional review boards (IRBs) and ethics committees (ECs). When they consider EMA/PDCO requests to be irrational, unethical or unfeasible, they should say so in writing, even if later during the PIP procedure, based on the written EMA feedback, they must “propose” such studies. Otherwise EMA/PDCO can refuse their PIP and block approval. If this happens, companies should then share their concerns with IRBs/ECs who are asked to approve the studies and to document when this occurs.

It is hoped that this paper will provide the clinicians and members of IRBs/ECs who are involved in paediatric trials a better understanding of the PIP system so that questionable PIP-required studies can be more easily identified. Additionally, researchers should continue to develop and test improved, alternative ways to collect dosing and efficacy data in all age groups and lobby for their use.

A statement on off-label use from the European Academy of Paediatrics (EAP) comparable to the AAP’s [12] could also be helpful. Bodies representing European paediatricians and other physicians who treat children should also support indicated revisions of the PIP system.