THE IMPACT OF THE FDA MODERNIZATION ACT
ON THE RECRUITMENT OF CHILDREN FOR RESEARCH
Vera Hassner Sharav
ETHICAL HUMAN SCIENCES & SERVICES
Summer 2003, vol. 5 pp. 83-108
This paper argues that contrary to the claims made by the research stakeholders in industry, academia and government, the shift in public policy since the enactment of the FDA Modernization Act (FDAMA) of 1997 and its financial incentives to industry to test drugs on children, has had a deleterious impact on children’s dignity, health and welfare. Those lucrative incentives offered an opportunity to accelerate the pace of FDA approval for pediatric drug marketing. FDAMA resulted in a radical shift in federal policy to accommodate an expansion of pediatric trials. Children who are precluded from exercising a human adult’s right to informed consent to research are increasingly sought as test subjects even when the trials offer no potential benefit for them. Prior to FDAMA children were protected under federal regulations that prohibited their recruitment for experiments that were not in their best interest. But as the cases described will demonstrate children have been subjected to experiments that exposed them to pain, discomfort, and serious risks of harm – including suicide. Babies have died testing a lethal heartburn drug; children have been subjected to “forced dose titration” in antidepressant drug trials that resulted in several suicide attempts. Toddlers are currently being subjected to Ritalin dose tolerance tests without evidence of any pathological condition. Healthy teenagers are being exposed to antipsychotic drugs known to induce severe pathological side effects in speculative “schizophrenia prevention” experiments.
The ubiquitous influence of the pharmaceutical industry on healthcare, academia and government oversight agencies has transformed the culture of medical research (Angell, 2000). Children, who cannot exercise free will, have been particularly affected adversely. This paper examines the impact of corporate business ethics on research as well as the culture of research in which children have been subjected to pain and exposed to risks of potential long-term harm.
A public policy shift has accelerated the pace of drug testing and expanded the use of children as research subjects. The Prescription Drug User Fee Act (PDUFA, see U.S. Public Law 102-571, 21 USC 301) of 1992 speeded up the drug approval process and officially ushered drug company influence on this process at the United States Food and Drug Administration (FDA) by instituting company user fees. These fees created a conflict of interest at the FDA. Congress passed the PDUFA under intense pressure from AIDS activists seeking immediate approval of potentially life-saving but experimental drugs. The demands of these activists inadvertently presented the pharmaceutical industry and its allies with an opportunity to “loosen regulatory brakes” (Sharav, 2002; Willman, 2000a) by demanding the same “lenient” and “fast-track” standards for the review of drugs not aimed at treating “life-threatening illness” (Lurie & Wolfe, 1998; Public Citizen, 2001b; Willman, 1998).
In 1997, the FDA Modernization Act (FDAMA, see U.S. Public Law 105-115, 21 USC 301) expanded the PDUFA’s policy shifts by codifying a number of practices that had become common at the FDA. According to Public Citizen (2001b), “FDAMA expanded the use of ‘accelerated approval’ mechanisms for drugs for life-threatening conditions and using surrogate endpoints in clinical trials. It also included a number of mechanisms for speeding FDA review and changed the legal standard for new drug approval to a single clinical trial (instead of two).” The law also offered pharmaceutical companies financial incentives to enroll children as research subjects. FDAMA extended the length of a manufacturer’s market exclusivity by six months for any patented drug or one under development, if the drug was tested on children in controlled clinical trials. For individual top selling drugs, such an extension can mean revenues of $284 million to $975 million (Zimmerman, 2001)
Prior to FDAMA, children were protected under Federal regulations adopted in 1983 (United States Code of Federal Regulations, 1983), which restricted the recruitment of children to research that offers them a potential direct benefit. The restriction aimed to protect children, who cannot volunteer and are vulnerable to exploitation, from undue risks of harm even if research offers a prospect of direct benefit to the child: “Research involving greater than minimal risk but presenting the prospect of direct benefit to the individual subjects” is permissible only if “the relation of the anticipated benefit to the risk is at least as favorable to the subjects as that presented by available alternative approaches.” If the research involves even a “minor increase” over minimal risk but does not have a prospect of direct benefit to the child, the regulations restrict such research to that which “is likely to yield” knowledge that is of “vital importance” for understanding or amelioration of the child’s disorder or condition (45 CFR 46.406 )
Unfortunately, regulations do not have the force of law. Unlike laboratory animals (protected by law since 1966), children subjected to research are not protected by an independent oversight system. No one maintains a record of the number or nature of clinical trials or of the subjects’ disposition following research (Institute of Medicine, 2002). Children are now at the mercy of local institutional review boards that, as will be shown in this paper, may approve experiments that put children’s health and welfare at risk of harm. Furthermore, researchers who have violated existing regulations have not been held accountable.
FDAMA was enacted following intense lobbying by biomedical industry stakeholders – including industry-supported children’s “advocacy” organizations (see Public Citizen, 2001b) who claimed that financial incentives were necessary to prod drug companies to conduct pediatric trials that would provide data to legitimize the use of drugs in children. It was argued that clinical trials are needed to provide physicians with safety and efficacy information, and that the absence of FDA-approved, pediatric dose information put children at risk of adverse drug reactions from wrong doses (Greeley, 2002) Congress accepted these claims at face value, although a reliable, widely used guide to inform pediatricians about pediatric dosing, the Harriet Lane Handbook (Siberry, Iannone, & Childs, 1996), has long existed. Furthermore, FDAMA failed to provide safeguards against child exploitation or ensure that patent exclusivity incentives would lead to studies of vital medical importance for children. Instead, the (brand name) pharmaceutical industry and its lobbying arm, the Pharmaceutical Research and Manufacturers
Association of America (PhRMA) have been accused of distorting the public health goal of FDAMA by demanding patent exclusivity “for virtually any study in pediatric subjects, no matter how trivial the study design, and regardless of the medical significance or utility of the data obtained” (Teva Pharmaceuticals, 2000). Congress also failed to conduct an impact assessment of the new law. According to Dembner (2001a), the number of young subjects enrolled in clinical trials in the United States swelled from 16,000 in 1997 to 45,000 in 2001.
FDAMA offered the research enterprise opportunities for increased profits and expansion. In 1999, the National Institute of Child Health and Human Development expanded the network of pediatric pharmacology research units at key academic institutions from 7 to 13. Steinbrook (2002) reported that the network conducted many of the studies related to pediatric exclusivity, and that additional funding is planned for the units. These lucrative partnerships between industry and academia facilitate a utilitarian goal – the rapid development and marketing of new drugs – but, arguably, have precipitated morally dubious policy changes. Public health officials set in motion a strategy for loosening federal restrictions that had been adopted to protect children from exploitation. FDA (1998) adopted the “Pediatric Rule” requiring that “new drugs and biologic products be studied in children at the same time, or soon after approval [for use in adults]. The final rule establishes a presumption that all new drugs and biologics will be studied in pediatric patients” (p. 66634). The National Institutes of Health (1998) issued new policy guidelines for grant applications stating that children “must be included in all human subjects research, conducted or supported by the NIH unless there are scientific and ethical reasons not to include them.” In sum, whereas the goal of the 1983 regulations had been to minimize the exposure of children to the risks of experimental research, the goals of recent changes have been to remove those barriers to child exploitation.
The FDA (1999) acknowledged that prior to FDAMA the use of children as subjects in Phase I drug studies “had been primarily limited to life threatening diseases and children who had the disease” in question. FDA and the Department of Health and Human Services (DHHS) convened advisory panels to facilitate the recruitment of children for drug trials. The panels obliged by “reinterpreting” federal prohibitions protecting children from experiments involving greater than minimal risk when no potential direct benefit to them or their condition exists. These advisory committees sanctioned the recruitment of healthy children to be used as “risk-bearing normal control subjects” (DHHS, 2001; Sharav, 2001). The rationale given was that even children free of a particular medical condition may be “at risk” of suffering from the condition and may therefore be recruited and subjected to research risks to stave off the possible future condition. FDA acknowledged that the post-FDAMA policy change “led to an increasing number of proposals for studies of safety and pharmacokinetics, including those in children who do not have the condition for which the drug is intended” (Murphy, 2001). Since the adoption of the “Pediatric Rule,” the FDA received 229 proposals from drug manufacturers and issued 191 written requests for pediatric studies (Murphy, 2001). According to Dr. Rosemary Roberts (2002) FDA’s Deputy Director of Pediatric Drug Evaluation, there were only six “significant” label changes for dosing or safety. In 2002 Congress extended the financial incentive provision of FDAMA for five years, and enacted measures “to stimulate pediatric studies of drugs’ under development and for drugs no longer patented. (Biotechnology Industry Organization, 2003) Zimmerman (2001) estimated expected additional revenues to drug companies resulting from the marketing of approved drugs for children to total $6 billion annually.
The financial incentive may in part explain how FDAMA accelerated the pace of bringing new drugs into the pediatric market even before their safety (i.e., the rate of adverse effects) in adults was known. In effect, the new policy targeting children and infants as test subjects in clinical trials was advanced with little debate by those who see nothing wrong with exposing a few young human beings to risks of harm for the claimed greater good of society. The 1983 regulations were adopted after a vigorous debate regarding society’s moral obligation toward children who cannot give a mature and informed consent. Ethicist Ramsey argued that this moral obligation overarches the obligation of scientific exploration: “Children should not be made the subjects of medical experimentation unless, other remedies having failed to relieve their grave illness, it is reasonable to believe that the administration of a drug as yet untested or insufficiently tested on human beings · may further the patient’s own recovery” (cited in Glantz, 1998, p. 234, italics in original). Rather than confront that moral question, some have claimed that it is safer to medicate a child under controlled research conditions than under a physician’s care. However, in research studies individual needs typically give way to standardized protocols, including fixed doses. Since doses often need to be individualized, clinical observation may be superior in this regard than controlled trials (Healy, 2002a). An examination of the new policy’s impact suggests that children have been actually put at increased risks of receiving inappropriate or harmful drugs both in clinical practice and research. And it contradicts the advice of FDA’s own commissioner, who cautioned against taking a new drug until its adverse effects are known, at least one year after it has been on the market (Neergaard, 2000a).
To facilitate pediatric research of greater than minimal risk and no direct benefit to subjects constituted a major policy change, which was effected without an open regulatory review process. Instead, the National Institutes of Health (1998) and FDA (1998, 1999) set about to “reinterpret” federal restrictions on the use of children by convening advisory committees, issuing guidelines, and publishing articles arguing to broaden the interpretation of regulatory language. They argued against being held to a consistent standard for the assessment of “minimal risk” and “minor increase over minimal risk.” Vitiello, Jensen and Hoagwood (1999), leading child psychiatrists from the National Institute of Mental Health (NIMH), argued for a relativist ethics approach, stating that “the interpretation [of minimal risk] varies across clinical context, institutional settings, and IRBs” (p. 1046). They proposed that in nontherapeutic research – where no potential benefit for the child subject exists – the determination of risk level should be the ratio between risk to the subject and “the scientific value of the project” (p. 1048).
ocating consensus statement by FDA’s pediatric Ethics Working Group (FDA, 2000a) opened the gate for easier access to children. The advisory panel first recognized that, “In general, pediatric studies should be conducted in subjects who may benefit from participation in the trial. Usually this implies the subject has or is susceptible to the disease under study.” It then indicated that it chose to utilize “a broad definition of potential benefit,” citing the example of the common ear infection. The panel argued that since most children will at some time get an ear infection, then every child is “at risk,” and, therefore, every child may potentially derive a benefit from testing a new treatment for ear infection. This reasoning allows any child to be used as a subject by speculating about a possible future benefit. Vitiello et al. (1999) wrote that “a particular category of potential research subjects is that of children who do not present with disorders or psychopathology but are considered at risk for mental illness” (p. 1046). At FDA’s pediatric subcommittee deliberations (November 5, 1999), it was suggested that the regulatory definition of “minimal risk” (“probability and magnitude of harm encountered in everyday life”) does not exclude death: “a risk of everyday life includes death.” It would appear, therefore, that at least some members on FDA’s Pediatric Advisory Subcommittee consider death “minimal risk” for research purposes.
The stunning shift in federal policy following FDAMA demonstrates how existing regulations can be bent. This utilitarian approach absolves the research community from the time-honored medical injunction – “First, do no harm” – and casts aside the “best interest of the child” standard, as well as the need to demonstrate a favorable risk/benefit ratio before permitting the use of children in painful experiments. Additionally, Vitiello et al.’s own acknowledgement casts doubt about whether psychiatry has even met the “scientific value” standard: “thus far, research on the biological substrates of mental illness has yielded relatively little specific information on the pathogenesis of psychiatric disorders” (p. 1045).
Research stakeholders, government officials, politicians, and the editors of the New York Times (“The Need to Test,” 2002) have claimed that by conducting clinical trials on children, the benefits to be gained for all children outweigh the risks to the subjects. However, those making the claim may not be free from financial conflicts of interest. Pharmaceutical industry influence is not marginal or isolated, it is ubiquitous, affecting all facets of biomedical research. Academia, government, professional journals, and advocacy organizations are all under its influence (Angell, 2000; Public Citizen, 2001a), so too, are bioethicists and the media (Elliot 2001; Sharav, 2001). Those who make claims that promote recruitment of children for experiments involving risks of harm and pain for the benefit of all children do not volunteer their own children or grandchildren. When the argument was made in a court of law in a case involving research on lead poisoning, it was unequivocally rejected by Maryland’s highest court (Grimes v. Kennedy Krieger, 2001, see discussion ahead).
As will become apparent from the case descriptions that follow, following FDAMA many children have been subjected to experiments that did not serve their best interest (Stolberg, 2001). Children have been subjected to a lethal heartburn drug and a plethora of psychoactive drugs leading some to serious self-destructive behavior (King et al, 1991) Young children have even been subjected to painful spinal taps and the attendant risks involved, without therapeutic intent (e.g., Mittleman et al, 1997; Castellanos et al, 2002).
Case 1: Testing a Heartburn Drug in Babies
In 1999, nine-month old Gage Stevens was recruited to test Propulsid (cisapride), a heartburn drug approved by the FDA’s advisory panel in 1993 despite clinical trial evidence showing that Propulsid prolonged the QT interval (a potentially fatal heart rate irregularity). FDA’s advisory panel had recommended approval of Propulsid without consulting with the agency’s own cardiology division (Willman, 2000b). In 1996 the FDA had informed the drug’s manufacturer that Propulsid was “not approvable” for children (Neergaard, 2000b), and in 1997 the director of the FDA’s gastrointestinal drugs division acknowledged that “at least” three children had died after being given the drug (Willman, 2000b). Yet, in 1998, FDA approved the recruitment of 100 babies for clinical trials of Propulsid and apparently allowed these trials to continue despite knowledge of the drug’s deadly effects (Mazo, 2000). The babies had been born with gastroesophageal reflux, a minor condition that resolves before one year of age.
In the year that Gage was recruited, Propulsid sales reached $950 million. “As of December 31, 1999,” according to the FDA (2000c), “there were 341 cases of cardiac arrythmia and 80 deaths reported.” Gage died of cardiac arrhythmia – as had six-month-old Chase Brown the year before. By the time Propulsid was taken off the market in 2000, 19 children had died. In an editorial, the Pittsburgh Post-Gazette asked whether – to facilitate recruitment of children – the research consent forms had falsely claimed that the FDA had approved the drug for use in children (Full disclosure, 2000). After his death, Gage’s parents were quoted as saying, “Little did we know that Gage was basically a guinea pig· They never told us · there had been deaths” (Mazo, 2000). The final blow was delivered when the parents learned from the autopsy report that their baby’s esophagus “failed to show signs of significant inflammation or other hallmarks of gastroesophageal reflux” (Spice, 2000). In other words, the baby did not appear to have the condition for which he was entered as a subject into a clinical trial.
A Boston Globe investigative series by Alice Dembner (2001a, 2001b, 2001c) has revealed how ethical standards have been violated and children have suffered and died in clinical trials. For example, some children were induced with Toys ‘R Us gift certificates to assume risks in drug trials not in their own best interests, while some parents in need of money have been offered as much as $1,000 to enroll their children in drug experiments (see also Jetter, 2000). Physicians violated the 1998 Code of Ethics of the American Medical Association by accepting $5,000 referral fees for recruiting the children. Dembner (2001a) reported that between 1994 and 2001, at least eight children died in medical experiments and hundreds suffered harmful effects. She suggested that, given the tendency of researchers not to report adverse events in clinical trials and given the threefold increase in the number of child research subjects in four years, “there is strong reason to believe that deaths and injuries in research involving children are more widespread” than available statistics would indic
ate. (Dembner, 2001a; Sabo, 2001) In their review of 561 pediatric research studies published in five medical journals, Bauchner and Sharfstein (2001) found that 40% failed to report whether the procedures “were in accordance with the ethical standards of the responsible committee on human experimentation.” The authors concluded: “Unless we enhance our system of safeguards, an unethical study could be published. While the primary responsibility for assuring ethical conduct of research rests with investigators, peer review journals should be more active in protecting human subjects” (p. 319).
Psychoactive Drugs: Are the Risks for Children Justified?
Seven of eight cases and controversies discussed in this paper pertain to psychoactive drug tests because of the inherent ethical and diagnostic problems involved and their explosive rise in pediatric use in clinical practice and research. (Diller, 2000; Grinfeld, 1998; Rushton, 2000; Vitiello, 2001; Voelker, 1999; Zito, Safer, dosReis, Gardner, Boles, & Lynch, 2000). Serious safety concerns have been raised about psychotropic drugs and the trials testing them. A body of evidence reveals that psychiatric drugs may have significant adverse neurological and physiological effects, including brain damage with long-term use (critical reviews by Glenmulen, 2000; Healy, 1997; Whitaker, 2002). These concerns are magnified when children are experimental subjects.
Before a physician can legitimately prescribe a drug to treat or prevent a medical condition, four methodological issues must be addressed:
1. whether the diagnosis is valid and the instruments used are objective;
2. whether minor or transient symptoms predict the risk of future full-blown pathology;
3. whether drugs used to treat a manifest condition, such as “schizophrenia,” are most appropriate to prevent the condition; and
4. whether the risk/ benefit ratio favors the drug treatment.
Several problems exist with prescribing (or testing) perception- and behavior-altering drugs for children. There is the absence of objective criteria for diagnosing children with pathological behavioral problems. Psychotropic drugs are prescribed for adults and children who have been led to believe, without proof, that they have a “chemical imbalance” in their brain (e.g., Breggin & Cohen, 1999; Kaiser 1996; Moore 1998; Valenstein, 1998, p. 96). Demonstrable evidence exists linking some psychotropics to severe adverse effects. Researchers know, but treating clinicians and the public often do not, that some psychotropics, such as neuroleptics, also produce profound, long-lasting cognitive and functional alterations in the brain, and that the drugs may have in fact worsened patients’ prognosis for recovery (Bowers & Swigar, 1988; Hegarty, Baldessarini, Tohen, Waternaux, & Oepen, 1994; VanPutten & Marder, 1987; Whitaker, 2002). Whether used illegally or by prescription, the mechanisms by which all psychoactive drugs work are identical (Konradi, Leveque, & Hyman, 1996). In a much cited article, Hyman and Nestler (1996) stated that “chronic administration of psychoactive drugs creates perturbations in neurotransmitter function” which “cause molecular and cellular changes in neural function” (p. 151). They explained that repeated “perturbations” (i.e., chronic use of psychoactive drugs) “usurp normal homeostatic mechanisms within neurons” (i.e., interfere with normal brain function) “thereby producing adaptations that lead to substantial and long-lasting alterations in neural function” (p. 153). Hyman and Nestler noted that these neural adaptations might not be beneficial to the organism, citing the possibility that neural adaptation to stimulant drugs results in addiction.
Case 2: The Attention-Deficit/ Hyperactivity Disorder/ Methylphenidate Controversy
In 1998, a panel of experts convened by the National Institutes of Health failed to reach a consensus about either the diagnostic criteria or the best treatment of ADHD. Even mainstream critics are alarmed about evidence of misuse and overuse of psychoactive drugs in children (Coyle, 2000; Diller, 2000). Critics point to the absence of any objective diagnosis to justify the risks (Baughman, 2000; Carey, 2000, 2002) and are concerned about indicators pointing to long-term harm related to chronic use of psychoactive drugs in children (Breggin & Breggin, 1994; Diller, 1998; Coyle, 2000; Caser Reports, 2002). It is unclear why NIMH has initiated few, if any, retrospective outcome surveys to determine whether psychoactive drugs have done more harm than good to the generation of children to whom they have been widely prescribed.
Proponents of psychostimulant drug treatment for ADHD – among them, members of NIMH and its Research Unit on Pediatric Psychopharmacology (RUPP), the American Psychiatric Association (APA), and the American Academy of Child and Adolescent Psychiatry (AACAP) – seem undeterred by a growing body of evidence suggesting the possibility of long-term harm. Hyman and Nestler’s (1996) laboratory findings about the psychostimulants’ link to addiction in animals are backed up by evidence gathered by the Drug Enforcement Agency (DEA, 1995; 2000) and a 26-year follow-up study of 492 children in the San Francisco Bay area, half of whom had been prescribed Ritalin as children, half did not (Lambert, 1998;; Lambert & Hartsough, 1998). Those who had been exposed to Ritalin as children were higher users of both tobacco (40% compared to 19%) and cocaine (21% compared to 10%). Furthermore, Volkow et al. (1995) found that methylphenidate acts like cocaine but remains in the brain longer. Vastag (2001) reported that Volkow’s team was startled by a new finding: “·instead of being a less potent transport inhibitor than cocaine, methylphenidate was more potent·the data clearly show that the notion that Ritalin is a weak stimulant is completely incorrect” (online) If Ritalin is more potent than cocaine, and the ADHD diagnosis has not been validated, one must ask what is the justification for exposing millions of children to a drug has the potential to result in addiction, and whose long-term effects are uncharted.
Influential psychopharmacologists such as Biederman and Spencer (2000) claim that “ADHD is one of the best researched disorders in medicine; in fact, the overall data on its validity are far more compelling than for many other medical conditions” (p. 77). However, the American Academy of Pediatrics’ ADHD clinical practice guideline (AAP, 2000), published at the height of controversy about the overprescribing psychoactive drugs for behavioral problems, shows that the claim is unfounded. The guideline attempted to legitimize the ADHD diagnosis while acknowledging the absence of any objective verifiable criteria to support it. AAP claimed that a “consensus” exists about the DSM-IV criteria, but on the same page acknowledged that these criteria “remain a consensus without clear empirical data supporting the number of items required for the diagnosis” and “remain subjective and may be interpreted differently
by different observers” (p. 1163). The guideline recognized that “the questions on which [ADHD] rating scales are based are subjective and subject to bias” (p. 1164). AAP also acknowledged that current ADHD criteria “do not take into account gender differences or developmental variations in behavior,” (p.1163) and that “symptoms may not be apparent in a structured clinical setting,” and may be bound up with “the demands and distraction of the home and school” (p.1160). Finally, the AAP guideline acknowledged that “brain imaging studies and electroencephalography do not show reliable differences between children with ADHD and controls” (p.1167). Nevertheless, AAP endorsed an ADHD diagnosis in children aged 6 to 12 years on the admittedly subjective basis of the DSM-IV criteria and reports (from parents and teachers) of “inattention,” “hyperactivity,” “academic underachievement,” or “behavior problems.”
In addition, the Agency for Healthcare Research and Quality (1999) reviewed 78 randomized controlled, peer reviewed, published studies of the drug treatment of ADHD. It found that in 87% of the studies, the number of dropouts and withdrawals and the reasons for such withdrawals were not described, neither were the long-term effects and severity of adverse effects: “overall, numerous deficiencies in the reporting of available randomized controlled trials limit the assessment of their validity, relevance, precision, and, therefore, their clinical application. Most studies did not clearly describe clinically important information such as the primary outcomes of interest” (pp. 4-5). The reviewers could not conduct a comparative analysis of the findings because of “the low quality of reporting and the large number and heterogeneity of outcome measures and tests used in the studies.” They reported that these scientifically flawed studies “show a trend to general improvement over time regardless of treatment·Ritalin appears to reduce behavior problems in ADHD children as long as it is taken·” (pp. 4-5).
Case 3: The Multimodal Treatment of ADHD (MTA) Study
The MTA Cooperative Group (1999) study has been hailed by those who promote psychotropic drugs for children as “a landmark in the history of treatment research in child psychopathology” (Barkley, 2000, p. 595). But others are unconvinced about the justification of using psychostimulants for children whose behavior may be within the range of normalcy (Carey, 2002) and criticize the study for bias and lack of scientific rigor (Breggin, 2000; Leo, 2002).
The NIMH-sponsored study was conducted over a period of 14 months at 10 research centers on 579 children, aged 7 to 10 years, selected from 4,541 children who were screened for participation in the experiment. The study included four treatment regimens to which children were randomly assigned: (1) medication alone, (2) combined medication and behavioral therapy, (3) behavioral therapy alone, or (4) standard community care which included psychoactive drugs. The purpose of the study was to find the most effective long-term treatment for ADHD. However, no long-term follow-up was planned. Parents and teachers rated the children’s behavior as improved on groups 1 and 2, but blinded classroom observers found no difference among the four intervention groups (MTA, 1999a, table 5, pp. 1082-1083). As in previous studies, the children’s academic performance did not improve (pp 1080-1081). Parents reported side effects of varied severity in 63.7% of the children (most were “mild,” p. 1073), but no trained professionals observed or evaluated either the children’s symptoms or adverse drug reactions.
The MTA study “findings” are a matter of dispute even among the investigators, who have focused their reporting mostly on whether drugs alone or a combination of drugs and behavioral therapy improved children’s behavior. Oddly, one of the key findings of the study (not accentuated by the original study researchers or ever cited by other psychopharmacologists who have referred to the original study) was that “more than three fourths [of the] subjects receiving behavioral treatment were successfully maintained without medication throughout the study” (MTA, 1999a, p. 1083). Instead, advocates for drug intervention claim that the study demonstrated drug efficacy (Swanson et al., 2001).
Several factors complicate the debate surrounding psychotropic drug experiments such as the MTA. Investigators who seek to test use of psychoactive drugs in children do not address the fundamental ethical and scientific questions posed by Baughman (2000), Breggin (2000), Carey (2002), and Coyle (2000):
- Are the children and adolescents sick, requiring medical intervention?
- Are their conditions sufficiently serious and disabling to necessitate pharmacologic intervention?
- Are the risks associated with the drugs justified?
- Do the researchers have financial ties to drug companies, and are their evaluations unbiased?
Glantz (1998) pointed to another (unacknowledged) factor that interferes with impartial selection and diagnosis of test subjects: “Obviously, if it’s the researchers doing the diagnosis, it is in their interest to diagnose kids with ADHD because they need them for the study” (cited in Montero, 2001).
A survey by Angold, Erkanli, Egger, and Costello (2000) of 1,422 school children found that 7.3% had been prescribed Ritalin, although only 3.4% had been diagnosed with ADHD. Even more alarming, Zito and colleagues (2000), examined data on more than 20,000 children aged 2 to 4 and found that between 1991 and 1995 there was a substantial increase in prescriptions for children of this age group of psychostimulants, antidepressants and clonodine (a blood pressure drug with sedative effects). A highly critical editorial by Coyle (2000) accompanied the Zito article, in which he raised concerns about the long-term impact of these drugs on children’s developing brains. These articles generated a national dialogue. (Then) First Lady Hilary Clinton convened a White House conference in March 2000 to address public concerns about inappropriate prescription of psychoactive drugs for children. Hyman asked, “How can we tolerate a situation in which drugs are prescribed to an increasing number of preschoolers without safety and efficacy data?” (cited in Huget, 2001). The chairman of the Judiciary Committee of the House of Representatives called for a federal investigation (Martin, 2000), and several class action lawsuits were filed. However, the White House conference was dominated by proponents of psychoactive drugs for children. The conference generated no call for retrospective follow-up studies that would have involved no added risks for children and might have provided much-needed details for rational decision-making. Nor did conference participants call for any curbs on prescribing psychoactive drugs for children. Instead what followed was an aggressive campaign for new initiatives to test
psychotropic drugs on ever-younger children. The deliberations of other high level conferences, and the reports that followed, were “informed” by powerful psychoactive drug advocates whose work is supported by pharmaceutical companies. Among those who control the NIMH research agenda in pediatric psychiatry are its Research Unit on Pediatric Psychopharmacology (RUPP) centers established in 1996-1997 to test the safety and efficacy of drugs. Influential pharmacologic stakeholders promote the idea that a “mental health crisis” exists, claiming that 14 million (20%) of American children aged 9 to 17 have a diagnosable mental illness (American Psychiatric Association, 2002).
By October 2000, NIMH and FDA representatives met to discuss the need to develop psychopharmaceuticals for preschoolers. Vitiello (2001), NIMH’s director of Child and Adolescent Treatment and Preventive Interventions Research Branch, acknowledged to the assembled that “the ability to formulate a valid and reproducible diagnosis of disorders and syndromes is a prerequisite for clinical trials” (p. 985). He recognized that this pre-requisite was lacking by acknowledging (repeatedly) “the diagnostic uncertainty surrounding most manifestations of psychopathology in early childhood” (p. 983). He admitted that “very little research has been done to demonstrate replicability across raters and external validity of these diagnoses in preschoolers” (p. 986). Therefore, “clinical trials of these agents in preschoolers do not seem possible given the current uncertainties about diagnostic validity of mood disorders in children <6 years old" (p. 985). Vitiello acknowledged the impact of FDAMA, affirming that "pediatric psychopharmacology has recently seen an unprecedented expansion·. NIMH-funded research for clinical trials in youths has more than doubled in the last few years" (p. 987). We are arguing here that the sharp rise of psychopharmacologic testing in pediatric research is a response to two contributing factors: widespread illegitimate prescribing of psychiatric drugs to children, and the financial incentives of FDAMA.
Case 4: Testing Ritalin in Preschool Children
Despite the explicit recognition by a prominent NIMH researcher that the essential ethical and scientific pre-requisite to clinical investigation of psychotropics in pre-schoolers was lacking, NIMH embarked on a radical trial exposing three-year-old children to psychoactive drugs. In November 2000 NIMH initiated the Preschool ADHD Treatment Study (PATS), a $5 million, multi-center, pediatric research initiative that will expose 312 children three years of age to psychostimulant drugs to test the drugs’ safety and long-term effects. The experiment is designed to test the children’s tolerance level for methylphenidate at increased doses – from 2.5 mg once a day, to 15 mg three times a day. The experiment has proceeded despite a lack of evidence that the study participants either have any validated medical condition or a demonstration that benefits of using the study drugs outweigh risks to the participants. Parents will receive $645 if their child completes the 43 study visits and teachers will receive $340 to fill out rating forms.
Even the study’s principal investigator acknowledged that ADHD is “not a well-defined psychiatric disorder in this age group” (Marshall, 2000a, p. 1280). Given that permanent harm may be caused to the central nervous systems of young children recruited to test the drugs, the experiment would seem to be highly questionable, if not altogether unethical and repugnant. Those initiating the trials have argued that it is safer to expose children to these drugs under controlled clinical conditions than to rely on pediatricians and primary care physicians to prescribe these drugs (essentially) in uncontrolled clinical trials. However, if the children’s behavior is within the range of normalcy (Carey, 2002), the experiments cannot be classified as therapeutic, and the risks are not justifiable. Indeed, nontherapeutic experiments that put children at greater than minimal risk are unapprovable under federal regulatory standards – yet, the preschool methylphenidate experiment is sponsored by the government.
Hyman has stated: “Without good clinical data, every child who receives this medication represents an uncontrolled experiment – that is entirely unacceptable” (cited in Marshall, 2000a, p. 1281). But is it morally acceptable to ask little children to assume risks by testing the safety of psychoactive drugs they may not need, in order to accumulate “good clinical data?” If that rationale were endorsed, would our society allow some children to be exposed to (potentially) deadly products in controlled clinical trials to obtain “scientific” information about how to protect other children? This very issue was the focus of a recent lawsuit involving a lead abatement experiment in which healthy children were exposed to the risks of lead paint so that researchers could determine the effectiveness of varying degrees of lead paint abatement. The academic research establishment – including Johns Hopkins University, the University of Maryland, the Association of American Medical Colleges (AAMC), and the Association of American Universities – attempted to intervene, arguing that the research would help all other children (Brief of amici curiae, 2001). The Maryland Court of Appeals unequivocally rejected their arguments in a strongly worded, landmark decision that severely criticized researchers and the institution involved for exposing children to risks for the advancement of science (Grimes, et al, and Higgins v. Kennedy Krieger Institute, 2001) The Court cited the Nuremberg Code and the Declaration of Helsinki, proclaiming the primacy of individual rights:
- whatever the interests of the general public in fostering research that might, according to a researcher’s hypothesis, be for the good of all children, this court’s concern for the particular child
- over-arches all other interests. It is, simply, and we hope, succinctly put, not in the best interest of any healthy child to be intentionally put in a nontherapeutic situation where his or her health may be impaired, in order to test methods that may ultimately benefit all children. (p. 80)
Case 5: Ignoring Risks of Antidepressants
The 1988 launching of Prozac, the first selective serotonin reuptake inhibitor (SSRI) antidepressant was a watershed in psychotropic drug marketing. It set the stage for countless upbeat news reports, such as a 1990 cover story in Newsweek – “Prozac: A Breakthrough Drug for Depression.” Prozac and the SSRIs that followed were promoted as safer, more effective, with significantly milder side effects than the earlier tricyclic antidepressants (TCA), which had been shown to cause serious, for some, life-threatening cardiac effects, especially in children (Wilens & et al., 1996 Mezzacapp
a et al., 1998), some of whom died of “sudden death” (Riddle, Geller, Ryan, 1993; Kutcher, 1997; Swanson, Jones, Krasselt, Denmark, Ratti, 1997; Varley & McClellan, 1997). The academic literature reflects the unreserved promotional claims made about SSRIs as a breakthrough, safe treatment for depression and the prevention of suicide, but also for a wide range of “conditions” such as grief, shyness, social dissatisfactions and problems in living. The wide application of SSRIs adds to the perception that they are “magic bullets.” Blind faith seems to have led doctors to prescribe SSRIs without hesitation – even to infants. Grinfeld (1998) reported that FDA data compiled by an industry research firm indicated that that although not approved for children, Prozac “was prescribed 349,000 times to pediatric patients under 16, including 3,000 times to infants under 1 year of age.”
However, SSRIs have been linked to severe adverse behavioral and neurological side effects. (Healy, 1997) In children, the most common of these effects include increased restlessness (akathisia), insomnia, excitability, hyperkinetic behavior and impulsiveness (King et al., 1991; Riddle, Hardin, King, Scahill, & Woolston, 1990). Other serious concerns involve severe withdrawal syndrome (Black, Shea, Dursun, & Kutcher, 2000) and the possible risk of brain damage – as found in animals (Kalia, O’Callaghan, Miller, & Kramer, 2000; Rustad, 2000). Indeed, Glenmullen (2000) and Healy (2000) an others have compared the damage of SSRIs to that caused by antipsychotic medications, concluding that future generations may look back on the use of antidepressants and other psychiatric drugs as “a frightening human experiment” (see also Goode, 2000).
Recent analyses of reports submitted to the FDA reveal that the resoundingly positive claims made in published reports are not supported by scientific evidence. In clinical trials, SSRIs were found to be no more effective than either TCAs or placebo (Kirsch, Moore, Scoboria, & Nicholls, 2002; Moore, 1997). Yet, children continue to be recruited to test antidepressants despite the risks. The FDA (2000b) acknowledged “the preponderance of negative studies of antidepressants in pediatric populations.” Indeed, 13 studies comparing the efficacy of TCAs with placebo in children found no difference in efficacy (Birmaher et al., 1998; Klein et al., 1998; Kye et al., 1996). Given the negative findings in pediatric trials testing TCAs, it is reasonable to suspect that SSRIs, which were shown to lack efficacy in adults, will prove no more effective in children. Indeed, Ambrosini et al (1999) acknowledged that “[r]esults of early controlled trials of [SSRIs] in pediatric subjects were also discouraging” *
Teicher, Cole and Glod (1990) released a fly into the ointment when they first reported the emergence of “intense, violent and suicidal preoccupation” in six patients after two to seven weeks of fluoxetine (Prozac) therapy. The authors estimated the risk of developing violent suicidal preoccupation on Prozac to be 3.5%. But these findings were anathema to the psychiatric establishment as much as to the pharmaceutical industry: the professional literature is weighted with reports claiming to disprove Teicher’s findings. In a follow up article, Teicher et al (1993) reported: “Although antidepressants diminish suicidal ideation in many recipients, about as many patients experience worsening suicidal ideation on active medication as they do on placebo. Furthermore, at least as many patients attempted suicide on fluoxetine and tricyclic antidepressants as on placebo·” (p. 186) Wirshing, van Putten, et al (1991) reported that their patients with no prior history of suicidal behavior became suicidal during treatment with fluoxetine, “all described their distress as an intense and novel somatic-emotional state; all reported an urge to pace that paralleled the intensity of the distress; all experienced suicidal thoughts at the peak of their restless agitation; and all experienced a remission of their agitation, restlessness, pacing urge, and suicidality after the fluoxetine was discontinued· [l]ike the akathisia in the neuroleptic-treated schizophrenic population, ‘fluoxetine akathisia’ can apparently be associated with suicidal ideation, sometimes of ruminative intensity.” Wirshing 1992, p. 580- 581)
Case 6: FDA Data Corroborates a Significant Risk of Suicide
Clinical trial reports submitted by drug manufacturers to the FDA for licensure purposes became publicly accessible, ironically, through a provision of FDAMA. These documents reveal the scope and severity of adverse side effects experienced by the subjects in those trials, and they reveal a significant risk of suicide in psychiatric drug trials. The incidence of suicidal preoccupation, attempts, and actual suicide in patients testing the new psychotropic drugs occurs with disturbing frequency, far exceeding the risk for patients on placebo (Kahn, Warner, & Brown, 2000; Moore, 1997; Whitaker & Kong, 1998; see also Sherman, 2002).
Using antidepressant clinical trial data submitted by manufacturers to the FDA for drug approval purposes, Khan and colleagues (2000) sought to assess the safety and efficacy of placebo. These authors found that patients on placebo were actually far less likely to commit suicide than those testing an investigational drug. Among the 19,639 patients testing 7 SSRI antidepressants in clinical trials, 34 patients had committed suicide and 130 had attempted suicide. However, only 2 patients who committed suicide and 15 of the suicide attempters were in the placebo arms of the trials. Khan, Khan, Leventhal, Krishnan, and Gorman (2002) presented an expanded analysis of the data to include recently approved antipsychotics as well as SSRI antidepressants. Between for 1985 to 2000, more than 71,604 patients participated in clinical trials testing all psychotropic drugs (Sherman, 2002). Khan et al. (2002) found that despite efforts to exclude suicidal patients from clinical trials, the suicide rates were exceedingly high in clinical trials of short duration. The suicide rate within a year of testing an atypical antipsychotic was 752 per 100,000 persons and for those testing an SSRI the suicide rate was 718 per 100,000.
These FDA data essentially corroborate the Teicher et al. (1990, 1993) findings and those reported by Glenmullen (2000) and Healy (2000). Healy (and others) believe that akathisia – drug induced severe restlessness, anxiety and agitation – is the catalyst that triggers suicidal or even homicidal behavior in some patients. However, FDA has failed to require disclosure of the evidence to those most in need of accurate information – clinicians and the public. They have been mostly kept in the dark about the true risks of psychotropic drugs, except when the press reports about lawsuits that have brought to light documents that revealed evidence linking SSRIs to violence and suicidal acts (e.g., Boseley, 1999; Garnett, 2000; Zuckoff, 2000).
Children, more than adults, have endured severe adverse effects in clinical t
rials. Although King and colleagues (1991) noted the need to study the incidence of medication-related agitation, self-injury and emerging suicidal obsession in children taking SSRIs, neither the NIMH nor the FDA has initiated such study. Neither the drugs’ serious risks nor their lack of clinical efficacy has dissuaded psychopharmacologists from subjecting ever more children to clinical trials that expose them to these drugs. Moreover, existing publications of pediatric trials may present only a partial and biased portrait of the available evidence gathered by investigators. Klassen and associates (2002) found that more than 40% of medical studies conducted on children are never published. These authors traced unpublished data that had been presented at the Society for Pediatric Research and found that invariably such studies had produced negative findings that raised questions about the safety and/or effectiveness of new drugs. Since many of these drugs were subsequently brought to market, one can argue that parents and clinicians were misled about the drugs’ actual safety or effectiveness. Some have suggested that this skewing of the scientific literature is a form of “scientific misconduct” (Picard, 2002). However, as the following case reveals, even when studies are actually published, authors may not report all the relevant data, especially concerning serious adverse effects.
Case 7: Forced Dose Titration Tests in Children
Pfizer (1996, 1997) submitted a two-part expert report to the FDA in response to an FDA memorandum that raised concerns about the “emergence of suicidal ideation, gestures, and attempts in association with sertraline [Zoloft] use” as a treatment for obsessive-compulsive disorder (OCD) in children The Pfizer report refers to two completed pediatric trials in the United States and two extension trials in children aged 6 to 12 years and adolescents aged 13 to 17 years, diagnosed with either depression or OCD. One study, a 12-week, multi-center, double blind, placebo controlled trial, recruited 92 children to sertraline and 95 to placebo. Of the latter group, 67 subsequently went onto sertraline in an extension trial. A second open label study lasting 51 days and aiming to explore the pharmacokinetics of and tolerance to sertraline after single and multiple doses, recruited 61 children, of whom 44 were diagnosed with depressive disorders and 17 with OCD.
Pfizer (1997) reported that during the first four weeks of the second study the children’s dose was increased to 200mg, higher than in adult trials: “the mean maximum daily dose of sertraline was considerably higher in the paediatric studies (185mg) than in the adult OCD studies (148mg).” Pfizer (1997) provided no clear rationale for testing a higher doe in children, simply stating: “This higher mean maximum daily dose is due to the design of the paediatric studies” (p. 31) It is also unclear why the FDA approved, what Alderman, Wolkow, Chung, and Johnston, (1998) call the “forced titration” study design, which surely put children under increased stress. Indeed, in the completed studies, six children on sertraline attempted suicide and a number of other children became suicidal. Within the group of 44 depressed 4 children (9%) attempted suicide. Suicide attempts in the main occurred within a few days of dose escalation. One of the children who became suicidal was an eight-year-old boy who had been in the sertraline dose tolerance study for 36 days. Pfizer (1996) states: “Patient was hospitalized for a suicide gesture, and dropped from the study. The patient mutilated himself by cutting his feet with a razor blade and tying a tie around his neck. There was no previous history of self-mutilation or suicidality, although family history was significant for affective disorder (mother, maternal uncle) and suicide (maternal uncle). The event was attributed to study drug by the investigator.”(p. 23, Table 1) In the same study a 14-year old who had been receiving 200mg/day sertraline was hospitalized on the 35th day of the study for “a moderate suicide gesture” in which he ingested “400 mg of sertraline·10 mg of lorazepam and unknown amount of organophospate insecticide. The suicidal ideation was thought to have resolved within one day and the patient was not discontinued. The following day the patient ingested 8 g of chloral hydrate. Nevertheless, the investigator continued the patient in the study five days later, without apparent further sequelae “(Table 1) Pfizer’s report further notes that this eight year old boy, “had been treated with methylphenidate 40mg qd for over five years, and this was discontinued immediately prior o entering the study.” This would indicate that the boy was continuously prescribed psychoactive drugs since he was nine years old. To what effect, one might ask?
Pfizer (1997) reported 21 “serious adverse events” (SAE) in 16 patients on sertraline (7.5%) in the OCD studies (p. 28). In the non-OCD studies, there were 11 SAE in 7 patients on sertraline (18.4%). “These included three reports of suicidality, and single reports of malaise, intentional drug overdose, medical/surgical·aggressive reaction, aggravated depression·” (p. 29) According to Pfizer, “The adverse events most frequently associated with discontinuation were psychiatric symptoms, most commonly agitation” (p. 21). Children (6 to 12 years) on sertraline were especially prone to agitation: 15.1% (compared to 1.9% on placebo). Among adolescents (13 to 17 years), 10.3% became agitated (compared to 2.4% on placebo). Pfizer’s report acknowledges: “there were no serious adverse events reported in the 95 patients in the placebo group” (p. 28). Of significance is the fact that these suicide-related events were reported “spontaneously” – the child-subjects in these trials were not monitored for suicidal preoccupation. The Pfizer (1996) report acknowledges that “there is minimal information in most reports about the individual’s history or course of treatment” (p. 13) and “it is well known that spontaneous reports are not a good indication of the true frequency of events.” (p. 17)
The academic publications stemming from these studies (Alderman et al., 1998; March et al., 1998) validate Klassen and colleagues’ (2002) findings of skewed reporting, since the publications fail to report the scope and nature of the SAE suffered by participants in the trials– only one suicidal event is mentio