Re-Examining the Treatment of Lead-Exposed Children (TLC) Trial

In May 2001, the New England Journal of Medicine published the TLC trial’s primary results, “The Effect of Chelation Therapy with Succimer on Neuropsychological Development in Children Exposed to Lead”, which found no significant difference in IQ between the succimer and placebo groups at 36 months. Within a year, the finding was absorbed into the policy apparatus that had been watching the trial since its inception.

NIEHS Director Kenneth Olden, Ph.D. said:

“For more than twenty years, NIEHS has sponsored much of the research showing that lead at these levels was harmful to children’s brain function, and that succimer lowered blood lead. We had hopes that the treatment would prevent or reduce lead-induced damage in these children, who are mostly poor, African-American, and living in deteriorated housing in big cities. The results of the trial show clearly that treatment after the fact does not undo the damage among 5 year olds. We must prevent these children from being exposed in the first place.” (NIEHS Newsroom — May 9, 2001)

The CDC’s Advisory Committee on Childhood Lead Poisoning Prevention (ACCLPP) published “Managing Elevated Blood Lead Levels Among Young Children”—a 200-page case management guide edited by former AAP president Birt Harvey. It recommended against chelation therapy for children with blood lead levels below 45 µg/dL, citing TLC as the basis. This was the first formal guideline to close the door on medical treatment for the 20–44 µg/dL range that TLC had studied.

Walter Rogan—TLC’s project officer and an ex officio member of ACCLPP for 16 years—presented the trial’s results directly to ACCLPP. He told the committee that TLC “did not produce evidence to demonstrate that succimer is beneficial to children” and that “the findings do not support conducting another trial.” He described succimer as “an expensive drug” that “resulted in symptoms in children who were previously asymptomatic.” The committee received this as settled science.

The American Academy of Pediatrics published “Lead Exposure in Children: Prevention, Detection, and Management.” It cited TLC and recommended against chelation below 45 µg/dL. Rogan, who served as NIEHS Liaison to the AAP’s Committee on Environmental Health for 36 years and was “primary author on several such statements,” was primary author on this guidance statement.

The AAP reaffirmed the recommendation in “Prevention of Childhood Lead Toxicity.” The 45 µg/dL chelation threshold held.

Dr. Rogan’s NIEHS oral history described the TLC trial as a success:

“We went to where you might imagine we went. Where was the other centers? Philly, Newark, Baltimore, and Cincinnati. We followed 780 kids, half of whom had gotten Succimer, half of whom got placebo… We lowered their blood leads pretty dramatically, and we changed their IQs not at all.”

“That study ended drug treatment, which had been being promoted as something that you ought to do to these kids. It also stopped the idea of what we call secondary prevention… and moved the attention back to primary prevention, not letting them get exposed to lead in the first place.”

“Nothing I’ve done would have been possible outside of a government agency, either NIH or CDC or EPA… My career has been very much a government scientist career.”

—Walter J. Rogan, MD, NIEHS Oral History, 2016

The U.S. Preventive Services Task Force published its recommendation on screening for elevated blood lead levels in children and pregnant women. The accompanying evidence review identified TLC as the only randomized controlled trial of chelation at lower blood lead levels and found “insufficient evidence” for treatment. The USPSTF gave screening an “I” statement—meaning it could not determine whether the benefits outweigh the harms—for children without obvious risk factors. TLC was the evidentiary centerpiece.

The result is a treatment gap. The CDC’s current reference value is 3.5 µg/dL—the level at which a child is considered to have “elevated” blood lead. The chelation threshold is 45 µg/dL. Between those two numbers, there is no recommended medical intervention. A child with a blood lead level of 40 µg/dL—more than ten times the reference value—receives the same pharmacological treatment as a child with a level of 4: none.

This policy framework rests on a single trial—a trial that returned children to lead-contaminated homes where lead dust levels rebounded to precleaning values within three to six months (Campbell 2003), failed to sustain the blood lead separation it was powered to detect—the gap between treatment and placebo groups largely closed by week 30 (TLC Group 2000, Figure 1), never analyzed iron status as an effect modifier, withheld electronic adherence data, and never tested for treatment-by-site interactions despite massive differences in environmental intervention across its four clinical centers.

628 of 780 enrolled children were exposed to lead-contaminated multivitamin supplements distributed by the trial itself (Rogan 1999). 83% of succimer-treated children required retreatment after the first course (TLC Group 2000).

Each of the trial’s documented limitations—ongoing environmental exposure, unmonitored iron status, inadequate treatment endpoint, unmonitored adherence, contaminated supplements—would, if present, reduce the probability of detecting a treatment effect. Whether any operated in the opposite direction is an open question.

The TLC trial evaluated one specific implementation of chelation therapy—chelating to a target of 15 µg/dL, without environmental remediation, with unmonitored adherence, unreported iron status, and undisclosed protocol changes.

Meanwhile, the problem TLC was supposed to address has not gone away. According to the most recent national survey data, approximately 24 million U.S. housing units still contain significant lead-based paint hazards. The children living in them are disproportionately Black, low-income, and urban—the same demographic profile as the 780 children TLC enrolled thirty years ago. These children have been told, on the basis of one trial, that there is nothing medicine can do for them between 3.5 and 45 µg/dL. The question is whether that trial earned that conclusion.

Clinical guidance regarding the management of lead-exposed children was largely based on guidance from the CDC (1985) and the American Academy of Pediatrics (1987). These statements recommended initiating chelation therapy in children with BLLs >55 µg/dL and clinical discretion between 25 and 55 µg/dL—a range which directly reflected the CDC’s definition of an elevated BLL (≥25 µg/dL). However, by 1990, new research showed IQ loss occurred as low as 10 µg/dL, raising concerns among clinicians as to whether the current recommended thresholds for intervention were sufficiently protective.

Another controversial recommendation was that prior to chelating children with BLLs below 55 µg/dL, clinicians should use EDTA provocation tests—in which a single intravenous (IV) dose of chelator is followed by urinary lead measurement to estimate ‘chelatable lead’. Despite their widespread use, these tests had limited diagnostic value, and animal studies suggested single EDTA doses might paradoxically increase brain lead concentrations, intensifying safety concerns and confusion among clinicians.

The three primary pharmacological agents in use at the time included dimercaprol (BAL)—administered via a painful intramuscular injection in peanut oil solution; calcium disodium EDTA–administered by IV; and D-penicillamine—an orally administered off-label drug approved for Wilson Disease but found to be somewhat effective at increasing urinary lead excretion. Each of the drugs in use carried unique risks, which were gradually coming to light through emerging studies and clinical experience. The 1985 CDC statement identified succimer (DMSA; dimercaptosuccinic acid) as a promising new, but unapproved, oral chelator.

In light of the disarray and lack of consensus among clinicians, in 1990, the Health Resources and Services Administration funded a national survey to determine current chelation practices being used at U.S. academic hospitals. The survey revealed widespread variability in treatment thresholds at levels as low as 20 µg/dL. Succimer, the oral compound mentioned in the 1985 CDC statement, was utilized by only one clinic participating in a special project as it awaited FDA approval.

Bock Pharma had submitted succimer for orphan drug designation in 1984, under the recently passed Orphan Drug Act (1983), and by 1990 preliminary studies suggested succimer was safer, more effective, and easier to administer than the other chelators in clinical use.

Starting January 1st, 1991, the Medicaid Drug Rebate Program (MDRP) came into effect. This meant that if a manufacturer wanted Medicaid to cover its outpatient drugs, the manufacturer would later “true up” the price by sending Medicaid a rebate check tied to how much Medicaid used the drug. Under the new law, state Medicaid programs were legally required to cover any FDA-approved drug if a manufacturer signed a rebate agreement. Most importantly, the law mandated coverage for all “medically accepted indications,” which included off-label uses supported by major medical compendia.

On January 30th, 1991, the FDA granted final marketing approval of succimer as an orphan drug for treatment of lead poisoning in children, specifically children with BLLs >45 µg/dL. McNeil Pharmaceuticals, who would market the drug, was given seven years of market exclusivity. Orphan drugs carried numerous incentives and benefits for pharmaceutical companies, and succimer’s development was largely subsidized by federal grants. To qualify as an orphan, the drug must be indicated for less than 200,000 patients. It remains unclear how the 45 µg/dL cutoff was selected. Clinical trials submitted to the FDA included children with BLLs as low as 30 µg/dL. However, ensuring the treated population included no more than 200,000 patients would have been a requirement for final FDA approval as an orphan drug. Regardless, succimer was expected to be used off-label to chelate children at much lower BLLs.

Later that year, in October 1991, the CDC released their long overdue guidance update: “Preventing Lead Poisoning in Young Children.” The statement included three conflicting recommendations:

1. Lowered the hard line for chelation from 55 to 45 µg/dL, aligning clinical guidance with the FDA-approved indication for succimer. Critically, the guideline allowed clinicians discretion on chelating patients with BLLs 20–45 µg/dL.
2. Lowered the BLL of concern to 10 µg/dL (a level shared by an estimated 1.7 million children), aligning it with evidence of neurotoxicity and creating new cases by definition.
3. Recommended universal lead screening for every 1- and 2-year-old child, substantially increasing the detection of these newly defined cases.

By late 1991, the CDC had mandated screening to find a massive new population of poisoned children, most of whom were low-income Medicaid patients. The MDRP had created a legal mandate for Medicaid to pay for their treatment, even if off-label. And McNeil Pharmaceuticals held a monopoly on the only safe, oral drug that made mass treatment feasible.