|
|
| |||||||||||||||||||||||||||||||||||||||||||||
Background Thousands of children, especially poor children living in deteriorated urban housing, are exposed to enough lead to produce cognitive impairment. It is not known whether treatment to reduce blood lead levels prevents or reduces such impairment.
Methods We enrolled 780 children with blood lead levels of 20 to 44 µg per deciliter (1.0 to 2.1 µmol per liter) in a randomized, placebo-controlled, double-blind trial of up to three 26-day courses of treatment with succimer, a lead chelator that is administered orally. The children lived in deteriorating inner-city housing and were 12 to 33 months of age at enrollment; 77 percent were black, and 5 percent were Hispanic. Follow-up included tests of cognitive, motor, behavioral, and neuropsychological function over a period of 36 months.
Results During the first six months of the trial, the mean blood lead level in the children given succimer was 4.5 µg per deciliter (0.2 µmol per liter) lower than the mean level in the children given placebo (95 percent confidence interval, 3.7 to 5.3 µg per deciliter [0.2 to 0.3 µmol per liter]). At 36 months of follow-up, the mean IQ score of children given succimer was 1 point lower than that of children given placebo, and the behavior of children given succimer was slightly worse as rated by a parent. However, the children given succimer scored slightly better on the Developmental Neuropsychological Assessment, a battery of tests designed to measure neuropsychological deficits thought to interfere with learning. All these differences were small, and none were statistically significant.
Conclusions Treatment with succimer lowered blood lead levels but did not improve scores on tests of cognition, behavior, or neuropsychological function in children with blood lead levels below 45 µg per deciliter. Since succimer is as effective as any lead chelator currently available, chelation therapy is not indicated for children with these blood lead levels.
In 1991, the Food and Drug Administration licensed succimer (dimercaptosuccinic acid), the first approved oral lead chelator, for use in children with blood lead levels of at least 45 µg per deciliter (2.2 µmol per liter).7 Succimer reduced blood lead levels at least as well as parenteral treatment with edetate calcium disodium in children with levels of 30 µg per deciliter (1.4 µmol per liter) or higher.8 Also in 1991, universal screening of children for elevated blood lead levels was recommended by the Centers for Disease Control (CDC),9 and the threshold of concern was lowered from 25 µg per deciliter (1.2 µmol per liter) to 15 µg per deciliter (0.7 µmol per liter) — a level associated with cognitive impairment but not symptoms of lead poisoning. However, the CDC made no specific recommendation about chelation therapy in children with blood lead levels of 20 to 44 µg per deciliter (1.0 to 2.1 µmol per liter).9 Because of the increase in screening, the wide availability of an oral chelator, and the lack of data on lead chelation for the prevention of cognitive impairment, we conducted a multicenter, randomized, placebo-controlled clinical trial. Our study was designed to test the hypothesis that children with moderate blood lead levels who were given succimer would have better scores than children given placebo on a range of tests measuring cognition, neuropsychological function, and behavior at 36 months of follow-up.
Methods
Referral and Prerandomization Activities
We accepted referrals of children 12 to 33 months of age (a range that includes the age at which lead levels peak) who had blood lead levels of 20 to 44 µg per deciliter, had no more than two main residences, and could be tested in English (or Spanish, at one site). Children with blood lead levels greater than 44 µg per deciliter were referred to local clinics for treatment. We measured lead levels in venous blood, serum ferritin levels, blood counts, renal function, and serum enzyme levels. We provided a vitamin and mineral supplement but later recalled one batch because of lead contamination; however, the amounts were too low to affect the children's blood lead levels.10 We inspected the children's homes to determine whether cleaning and minor repairs could be expected to reduce exposure to lead dust.
Children with confirmed blood lead levels of 20 to 44 µg per deciliter who lived in housing that could be cleaned or who could move to lead-safe housing returned for a second visit. A child was enrolled if his or her blood lead level at the time of the second visit was also 20 to 44 µg per deciliter. About half the children referred to us were enrolled; most disqualifications were due to a blood lead level below 20 µg per deciliter.
We vacuumed the residences of the enrolled children (using a high-efficiency particle-arrestor vacuum cleaner), mopped floors and wiped walls and surfaces with a trisodium phosphate solution, made minor repairs, and performed paint stabilization (by scraping loose paint and doing minor carpentry) at about the time of randomization. The homes of 16 of the 396 children given succimer and 5 of the 384 children given placebo were never cleaned; among these, 3 of the children given succimer and 2 of the children given placebo were moved with their families to lead-safe housing.
Using supplies known to be lead-free, we collected venous blood for measurement of blood lead levels twice before randomization and then on days 7, 28, and 42 after the beginning of each course of treatment. The blood lead levels were measured at the Nutritional Biochemistry Branch of the CDC.11 After treatment was stopped, blood lead levels were measured every three to four months. A total of 780 children were enrolled; we randomly assigned 396 to succimer and 384 to placebo. Treatment assignments were stratified according to clinical center, body-surface area, blood lead level, and language (because of the tests administered in Spanish). The study was approved by the institutional review boards at the clinical centers, the Harvard School of Public Health, the CDC, and the National Institute of Environmental Health Sciences. The parents of all the children provided written informed consent at enrollment for prerandomization activities and at the initiation of treatment for treatment and follow-up activities.
Succimer and Placebo
We administered succimer (Chemet) or placebo in 100-mg capsules of identical appearance (both provided by McNeil Consumer Products, Fort Washington, Pa.). We used 26-day courses of therapy and aimed to provide 1050 mg per square meter of body-surface area per day for the first seven days and 700 mg per square meter per day thereafter.12 Children could receive up to three courses of treatment; those who had blood lead levels of 15 µg per deciliter or higher two weeks after the completion of a first or second course of succimer were given another course. Eighty-three percent of the children assigned to succimer required a second course, and 83 percent of those receiving a second course required a third. Children in the placebo group were assigned to retreatment to match the frequency of retreatment in the succimer group, within the strata used for initial randomization. If a child had a confirmed blood lead level above 44 µg per deciliter, the study treatment was stopped, and the child was referred for treatment according to the usual standards of the clinical center.13
Developmental, Neuropsychological, and Behavioral Tests
Before treatment began, we administered to the children the Bayley Scales of Infant Development II (BSID-II),14 the current edition of the most widely used scales of infant development. At 18 months of follow-up, if the child was still younger than 42 months of age, we administered the BSID-II again; if the child was 42 months of age or older, we administered the Wechsler Preschool and Primary Scales of Intelligence–Revised (WPPSI-R).15 At 36 months of follow-up, we administered the WPPSI-R and the Developmental Neuropsychological Assessment (NEPSY),16 a battery of tests designed to identify neuropsychological deficits that interfere with learning. The NEPSY evaluates the domains of attentional and executive, sensorimotor, visuospatial, language, and memory function. At 36 months of follow-up, we also administered to the parent the short form of the Conners' Parent Rating Scale– Revised (CPRS-R).17 The CPRS-R is a 27-item scale that provides four behavioral indexes, of which we used the three that are applicable to younger children: Oppositional Index, Hyperactivity Index, and Attention-Deficit–Hyperactivity Disorder Index. We then averaged the CPRS-R index scores and called the average the Behavioral Index.
The children were tested with a parent or guardian present between the hours of 9 a.m. and 4 p.m. in a quiet room. The tests were rescheduled if the child was acutely ill, and the child was fed if he or she had not recently eaten. The testing psychologists did not know whether the children had been given succimer or placebo. The IQ of the caregiver in attendance (the mother for 88 percent of the children, the father for 4 percent, and another caregiver for 8 percent) was assessed during one of the follow-up visits with the short form of the Wechsler Adult Intelligence Scale–Revised.18
Statistical Analysis
Our study was designed to have 82 percent power to detect a 3-point difference between groups in the children's mean IQ at 36 months of follow-up at a two-sided significance level of 0.05. The actual power of the study was 96 percent, because the number of children with data at 36 months of follow-up was higher than expected and the correlation between base-line and follow-up psychometric tests was better than expected.
Nine children in the placebo group and five in the succimer group attended the 36-month follow-up appointment but were unable to complete the WPPSI-R test for developmental reasons. So that these children could be included in the analysis, their scores were imputed on the basis of the sum of scaled scores for the completed subtests or by assigning the score to the value below the lowest possible score for the corresponding domain. The data from three children who could not be tested were treated as missing.
In addition, 10 children in the placebo group and 8 in the succimer group completed the WPPSI-R at 18 months of follow-up but did not attend the appointment at 36 months. For these children, we substituted the WPPSI-R score at 18 months of follow-up for the score at 36 months, since the correlation between the two test scores was 0.83 for full-scale IQ.
We used the two-sample t-test to compare unadjusted mean scores in the two treatment groups. We also used multiple linear regression analysis to adjust the mean differences for a set of base-line covariates chosen in advance. These covariates included the variables specifically balanced by the strata used for randomization (clinical center, body-surface area, blood lead level, and language [Spanish or English]), the caregiver's IQ, the child's base-line score on the Mental Development Index from the BSID-II, and a term for the interaction between the Mental Development Index score and age. For the analyses of NEPSY subscale scores, we included an indicator variable for the version of the test administered (age of 3 to 4 years vs. age of 5 to 12 years) and a term for the interaction between the version and the age at testing. All analyses were performed according to the intention-to-treat principle. Since drug therapy was completed long before follow-up testing was performed, there was no need for stopping rules or interim analyses.
Results
Randomization, Adherence, and Retention
The two treatment groups were balanced with respect to base-line characteristics (Table 1), so the estimates of the effect of treatment are similar for the adjusted and unadjusted scores. The level of precision, however, is much higher for the adjusted estimates. According to the parents' reports, over 90 percent of the assigned doses of study drug were given. When the pills were counted, about 76 percent of the capsules had been removed from the bottles. Forty percent of the families whose children were given succimer and 26 percent of the families of children given placebo reported difficulty administering the drug. Interruptions in the administration of the drug occurred at similar rates in the succimer group (30 percent) and the placebo group (27 percent). Of the children in whom administration of the drug was interrupted, 39 percent of those receiving succimer and 45 percent of those receiving placebo resumed taking the study medication.13
|
|
We observed the largest estimated mean difference between groups in blood lead levels, 11 µg per deciliter (0.53 µmol per liter), at one week after the beginning of treatment (Figure 1). A rebound in blood lead levels in the succimer group, presumably due to lead stored in calcified tissue, began at one week and continued. At 49 days after the beginning of treatment, the mean blood lead level in children given succimer was 72 percent of the base-line mean, as compared with 88 percent in the children given placebo. Blood lead levels dropped again in the children given second and third courses of succimer, and in each case the levels then rebounded, whereas the mean blood levels in children given placebo declined steadily but slowly. One year after treatment began, the difference in mean blood lead levels between the two groups had largely disappeared. The mean blood lead level of the children treated with succimer was lower by 4.5 µg per deciliter (95 percent confidence interval, 3.7 to 5.3 µg per deciliter [0.2 µmol per liter; 95 percent confidence interval, 0.2 to 0.3 µmol per liter]) than the mean level of the children given placebo over the 6 months after the initiation of treatment, and lower by 2.7 µg per deciliter (95 percent confidence interval, 1.9 to 3.5 µg per deciliter [0.1 µmol per liter; 95 percent confidence interval, 0.1 to 0.2 µmol per liter]) over the 12 months after the initiation of treatment (Figure 1).13
|
Of 10 children who were receiving succimer when their blood lead levels exceeded 44 µg per deciliter, 9 were hospitalized for treatment; of 7 children who were receiving placebo when their blood lead levels exceeded 44 µg per deciliter, 4 were hospitalized. Five of the children given succimer and none of those given placebo were hospitalized for trauma, with no common pattern or site of injury. A history of trauma or evidence of trauma on physical examination was noted for 15 percent of the children given succimer and 10 percent of those given placebo. There was no significant excess of any other category of symptoms, individual symptom, or laboratory abnormality in either group.13 (Full data are available on the study Web site at http://dir.niehs.nih.gov/direb/tlc1/home.htm). We monitored the children's growth and found that those given succimer had grown 0.25 cm (95 percent confidence interval, 0.05 to 0.45) less than those given placebo over 12 months of follow-up and 0.35 cm (95 percent confidence interval, 0.05 to 0.72) less over 34 months of follow-up.
Intelligence, Neuropsychological Development, and Behavior
Unadjusted mean scores on the WPPSI-R, NEPSY, and CPRS-R were similar in the two treatment groups (Table 2). After adjustment for the variables listed in the Statistical Analysis section above, the mean full-scale IQ score on the WPPSI-R for children given succimer was 1.1 points (95 percent confidence interval for the difference, –2.6 to 0.5) lower than that for children given placebo (Figure 2). The children given succimer scored slightly higher on four of the five domains of the NEPSY; the differences did not vary according to age at testing. For the CPRS-R, the behavioral index was 1.2 points (95 percent confidence interval for the difference, –0.5 to 2.8) higher (i.e., worse) in children given succimer. None of these differences approached statistical significance. Adjustment had little effect on the estimates of the differences in means, as would be expected given the similarity of the groups at base line (Table 2 and Figure 2). When we excluded the 32 children for whom we imputed the WPPSI-R scores or substituted the scores obtained at 18 months, the estimates were very similar; for adjusted full-scale IQ scores, the difference (succimer scores minus placebo scores) with imputation was –1.1 (95 percent confidence interval for the difference, –2.6 to 0.5), and without imputation it was –1.3 (95 percent confidence interval for the difference, –2.8 to 0.1).
|
Our study was a randomized trial of chelation therapy in children exposed to lead that was designed to examine developmental end points. Treatment with succimer did not lead to better scores on cognitive, neuropsychological, or behavioral tests than placebo.
Observational data suggest that the effect on IQ of an increase in the blood lead level from 10 to 20 µg per deciliter is a decrease of about 2 to 3 points.19 The difference in blood lead levels between the children given succimer and those given placebo exceeded 10 µg per deciliter only briefly, and the mean difference was 4.5 µg per deciliter (0.2 µmol per liter) during the six months after the initiation of treatment. Thus, it could be that the failure of our study to demonstrate a difference in test scores is due to the small difference in blood lead levels that we observed. However, succimer is as effective as any chelating agent currently available, and we used it for 26 days per course rather than the usual 19 days. We also used the (higher) loading dose for the first seven days, rather than the first five days, of each course, with the dose calculated according to body-surface area rather than weight.20 When the pills were counted, 76 percent were gone, so adherence to therapy was as high as has been reported among children in trials lasting longer than a week.21,22 We believe it is unlikely that another chelation regimen would have been more effective.
Ruff et al.23 treated children with blood lead levels of 25 to 55 µg per deciliter (1.2 to 2.6 µmol per liter) using parenteral edetate calcium disodium as a chelation agent. Chelation was not randomly assigned but was used when clinically indicated and had no relation to blood lead level or IQ at follow-up. However, the children whose blood lead levels fell the most had the greatest improvement in IQ. The nonrandomized design of the study by Ruff et al. made it difficult to control for environmental and parenting differences that might lead to both larger drops in blood lead levels and increased IQ, and the six-month follow-up period was relatively short. However, Australian children whose blood lead levels fell more quickly as toddlers had higher IQ scores at seven years of age.24 The children in our study completed their 36 months of follow-up at about five years of age and are now being evaluated at seven years of age.
Unless a more positive outcome becomes apparent at seven years of age, these results suggest that drug therapy should be used with caution in young children with blood lead levels below 45 µg per deciliter. The treatment in our study did not reduce the number of children whose blood lead levels exceeded 45 µg per deciliter and did not improve the cognitive, behavioral, or neuropsychological outcome 36 months later. The regimen is expensive and a significant burden on the families. In addition, the slight slowing of linear growth and the evidence of more frequent trauma in children receiving succimer are not reassuring. Since lead poisoning and its sequelae are entirely preventable, our inability to demonstrate effective treatment lends further impetus to efforts to protect children from exposure to lead in the first place.
Supported by contracts and intra-agency agreements with the National Institute of Environmental Health Sciences in cooperation with the Office of Research on Minority Health at the National Institutes of Health and the Centers for Disease Control and Prevention.
We are indebted to T. Shaffer, Program Support Center, Department of Health and Human Services, Perry Point, Md., and to E. Helzner, McNeil Consumer Products, Fort Washington, Pa.
Source Information
From the National Institute of Environmental Health Sciences, Research Triangle Park, N.C. (W.J.R., N.B.R.); the Department of Environmental Health, University of Cincinnati, Cincinnati (K.N.D.); the Harvard School of Public Health, Boston (J.H.W., D.W.D., M.S.); the Department of Psychology, Children's Hospital of Philadelphia, Philadelphia (J.R.); the Nutritional Biochemistry Branch, Centers for Disease Control and Prevention, Atlanta (R.L.J.); the Kennedy–Krieger Institute, Baltimore (J.J.C.); and the Environmental and Occupational Health Sciences Institute, University of Medicine and Dentistry of New Jersey, Piscataway (G.G.R.). Presented in part at the National Lead Grantee Conference of the Centers for Disease Control and Prevention, Atlanta, December 11, 2000, and at the Society of Toxicology meeting, San Francisco, March 27, 2001.
Address reprint requests to Dr. Rogan at the Epidemiology Branch, National Institute of Environmental Health Sciences, A3-05, P.O. Box 12233, Research Triangle Park, NC 27709, or at rogan{at}niehs.nih.gov.
References
The following persons also participated in the study: J.R. Serwint, Johns Hopkins Hospital, Baltimore; M. Brophy, C.T. Davoli, M.R. Farfel, and G.W. Goldstein, Kennedy Krieger Institute, Baltimore; J. Rubin, University of Maryland, Baltimore; O. Berger, R.L. Bornschein, C. Wesolowski, and S. Wilkins, University of Cincinnati Medical Center, Cincinnati; G. Maynard-Wentzel and M.E. Mortensen (to 1994), Children's Hospital of Columbus, Columbus, Ohio; S. Adubato, M. Elsafty, M. Heenehan, A. Sheffet, A. Ty (to 1997), and R.P. Wedeen, University of Medicine and Dentistry of New Jersey, New Jersey Medical School, Newark; C. Campbell, F.M. Gill (to 1996), J. Guinn, F. Henretig, D. Knight, and D.F. Schwartz, Children's Hospital of Philadelphia, Philadelphia; B.B. Bowman (to 1996), E. Gunter, D. Huff, D.T. Miller (to 1995), and D.C. Paschal, Nutritional Biochemistry Branch, Centers for Disease Control and Prevention, Atlanta; and A.J. Bernstein, A.I. Damokosh, M.E. Fay, and T.V. Kotlov, Harvard School of Public Health, Boston; Data and Safety Monitoring Committee — C.R. Angle, J. Faison, S.H. Gehlbach (chair), B. Gray-Little, S.A. James, L.A. Moyé, and H.L. Needleman.
| |||||||||||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
