|
|
| |||||||||||||||||||||||||||||||||||||||||||||
Background Secretin is a peptide hormone that stimulates pancreatic secretion. After recent publicity about a child with autism whose condition markedly improved after a single dose of secretin, thousands of children with autistic disorders may have received secretin injections.
Methods We conducted a double-blind, placebo-controlled trial of a single intravenous dose of synthetic human secretin in 60 children (age, 3 to 14 years) with autism or pervasive developmental disorder. The children were randomly assigned to treatment with an intravenous infusion of synthetic human secretin (0.4 µg per kilogram of body weight) or saline placebo. We used standardized behavioral measures of the primary and secondary features of autism, including the Autism Behavior Checklist, to assess the degree of impairment at base line and over the course of a four-week period after treatment.
Results Of the 60 children, 4 could not be evaluated — 2 received secretin outside the study, and 2 did not return for follow-up. Thus, 56 children (28 in each group) completed the study. As compared with placebo, secretin treatment was not associated with significant improvements in any of the outcome measures. Among the children in the secretin group, the mean total score on the Autism Behavior Checklist at base line was 59.0 (range of possible values, 0 to 158, with a larger value corresponding to greater impairment), and among those in the placebo group it was 63.2. The mean decreases in scores over the four-week period were 8.9 in the secretin group and 17.8 in the placebo group (mean difference, –8.9; 95 percent confidence interval, –19.4 to 1.6; P=0.11). None of the children had treatment-limiting adverse effects. After they were told the results, 69 percent of the parents of the children in this study said they remained interested in secretin as a treatment for their children.
Conclusions A single dose of synthetic human secretin is not an effective treatment for autism or pervasive developmental disorder.
In the absence of proven drug treatment for the core symptoms of autism, unproven treatments are widely used.6 Some of these treatments are based on theories of abnormal intestinal permeability, according to which excessive absorption of peptides derived from the gut may affect central serotonergic, opioid, and perhaps other neurotransmitter systems.7 Indeed, there are many parents who report that their children with autism have a variety of gastrointestinal symptoms (e.g., chronic diarrhea).
Recently, a three-year-old child with autism and chronic diarrhea underwent an endoscopic procedure that included the intravenous administration of secretin to assess pancreatic exocrine function. Within a week after the procedure, the child's parents noticed dramatic improvements in his behavior and language skills.88 Media attention focusing on anecdotal reports of secretin-related effects in children with autism has escalated into reports that secretin may represent a "cure" for autism. There are reports of the sale of sham secretin, price gouging, and other forms of profiteering.9 It is estimated that at least 2500 children with autism have received secretin injections.10 We are aware, however, of the published results of only one open-label study, in which it was used in three boys with autism.11
Secretin is a peptide hormone composed of 27 amino acids, which is released from S cells of the duodenum in response to the presence of acidic contents in the stomach. Secretin increases the volume and bicarbonate content of secretions from pancreatic acinar cells.12 Secretin is derived from the duodenum of pigs, and purified porcine secretin (Ferring Laboratories, Suffern, N.Y.) was approved by the Food and Drug Administration in 1981 for single-dose use in the diagnosis of gastrointestinal disorders. No data are available on the safety of repeated doses over time, and there are no data regarding its safety and efficacy in children.
Pure human secretin was synthesized and manufactured in 1998 by ChiRhoClin (Silver Spring, Md.). The human and porcine forms of secretin have equivalent kinetics and pharmacologic effects on the exocrine pancreas in humans.13 We conducted a randomized, double-blind, placebo-controlled trial of synthetic human secretin in children with autistic disorders.
Methods
Subjects
We identified 60 children from a sample of 94 children who were referred by the Treatment and Evaluation of Autism and Communication Handicaps program of the Department of Psychiatry at the University of North Carolina or whose parents responded to notices about this study placed in the newsletter of an autism-support group. The inclusion criteria for the study were a previous diagnosis of autism; an age between 3 and 14 years; no previous secretin treatment; no diagnosis of pancreatitis, inflammatory bowel disease, or gastrinoma; and written, voluntary informed consent for participation by a parent or legal guardian.
Children who met all inclusion criteria were screened for autism. The diagnosis of autism was made according to the criteria of the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV),14 the Childhood Autism Rating Scale,15 and the Autism Behavior Checklist.16 A diagnosis of autism was given to children who met the criteria for autism of the DSM-IV on the basis of observation of the child and a structured diagnostic interview of a parent, who had a score on the Childhood Autism Rating Scale of 30 or more (range of possible scores, 0 to 60), and who had a score on the Autism Behavior Checklist of 60 or more (range of possible scores, 0 to 158). For both scores, a larger value corresponds to more severe impairment. A diagnosis of pervasive developmental disorder, not otherwise specified, was given to children who had severe impairments in either socialization or communication, or who engaged in repetitive behavior, as documented by observation and the diagnostic interview, but who did not meet the DSM-IV criteria for autism. This evaluation represented consensus ratings from the study clinician (on the basis of observation and the DSM-IV diagnostic interview), the referring clinician (on the basis of the Childhood Autism Rating Scale), and the parent (on the basis of the Autism Behavior Checklist).
Of the 60 children who entered the study, 40 met the criteria for autism (mean score on the Childhood Autism Rating Scale, 38.3; mean score on the Autism Behavior Checklist, 74.6), and 20 met the criteria for pervasive developmental disorder, not otherwise specified (mean score on the Childhood Autism Rating Scale, 27.9; mean score on the Autism Behavior Checklist, 41.2). A comparison of the two diagnostic subgroups revealed significantly higher scores on the Childhood Autism Rating Scale (P<0.001) and the Autism Behavior Checklist (P=0.05) in the subgroup with autism than in the subgroup with pervasive developmental disorder.
None of the children had serious medical problems according to the history and medical examination. The human subjects committee of the Olson Huff Center for Child Development approved the study protocol.
Treatment
After a two-week base-line period, the 60 children were randomly assigned to receive either a single infusion of secretin or a single infusion of a saline placebo. A stratified randomization scheme based on a "median split" of the final subject sample according to the median age (eight years) was used.
The vials of synthetic human secretin contained a sterile, lyophilized powder consisting of 16 µg of synthetic human secretin, 1.5 mg of cysteine hydrochloride, and 20 mg of mannitol. The secretin was reconstituted with 8 ml of normal saline per vial. The children were weighed, and a single dose of 0.4 µg per kilogram of body weight was prepared (a volume of 0.2 ml per kilogram). For the children in the placebo group, an identical-appearing syringe containing 0.2 ml of normal saline per kilogram was prepared. The doses were administered by slow intravenous injection over a period of one minute. Secretin and placebo were administered with the use of a double-blind procedure.
Outcome Measures
Detailed assessments of each child's behavioral symptoms were conducted by the study clinicians, by the child's parents, and by the child's school or preschool teacher. Behavioral assessments were conducted at the end of each of the two base-line weeks, on the first and second days after infusion, and at the end of the first, second, and fourth weeks after infusion.
Communication skills were measured with the communication subscale of the Vineland Adaptive Behavior Scales.17 On this subscale, the standard score has a mean of 100 and a standard deviation of 15, and the range of possible scores is less than 20 to greater than 160, where higher scores mean less impairment. The subscale is a 63-item structured clinical interview that measures age-appropriate communication skills and provides a standardized score for communication ability. Separate assessments were conducted of both the parent's perception of the child's behavior in the home and the teacher's perception of the child's behavior in school.
The severity of autistic symptoms was measured with the Autism Behavior Checklist,16 a 57-item standardized clinical-rating scale that provides an overall score for severity (total score, 0 to 158, with higher values corresponding to greater impairment) and five subscores (relating to sensory function, social relatedness, body and object use, language, and socialization). The Clinical Global Impression Scale18 was used to measure eight separate features associated with autism (response to social interaction, social initiation, use of speech, types of repetitive behavior, behavior problems, activity level, sleep problems, and digestive problems) on a standardized, seven-point Likert scale, with 7 indicating "very much worse," 4 "no change," and 1 "very much improved."
Adverse effects of treatment were measured with the Treatment Emergent Symptoms Scale,19 a standardized rating scale that measures the occurrence or nonoccurrence of 24 potential side effects of medication in six categories (gastrointestinal, urinary, respiratory, skin, neurologic, and psychological). Each side effect is indexed in terms of observable behavior and is scored on a three-point Likert scale, with 2 denoting severe, 1 mild, and 0 absent. The Vineland Adaptive Behavior Scales and Treatment Emergent Symptoms Scale were completed by study clinicians, and the Clinical Global Impression Scale and Autism Behavior Checklist were completed by the child's parents. All persons who performed the behavioral assessments were unaware of each child's treatment status.
Statistical Analysis
For each outcome measure, two-by-six analysis of variance with repeated measures was used to assess the effects of the treatment group (secretin or placebo), the time of assessment (at base line and day 1, day 2, week 1, week 2, or week 4 after infusion), and the interaction between treatment group and time of assessment.20 For each outcome measure, the changes from base-line scores were computed for all assessment periods after treatment, and the resulting changes in scores were analyzed with the use of t-tests and estimates of 95 percent confidence intervals.20 Chi-square tests with Yates' correction20 were used to compare the rates of response in treatment groups and to determine whether response was related to diagnostic variables (autism vs. pervasive developmental disorder), demographic variables (sex and age), or clinical variables (use of adjunctive medication). To examine the response to treatment at the level of the individual subject, a criterion for clinically significant positive response was established before analysis. Significant positive response was defined as a score on the Clinical Global Impression Scale of either 2 (much improved) or 1 (very much improved) for any of the eight features assessed at two or more of the five follow-up evaluations (day 1, day 2, week 1, week 2, or week 4). P values for all statistical tests were two-tailed. All data were collected before the codes for treatment group were broken.
Results
Subjects
Fifty-six of the 60 children completed the study (28 per group). Four children discontinued the study after the administration of the infusion. Two children (both assigned to secretin treatment) were withdrawn because of their parents' wish that their children begin receiving secretin treatment outside the study protocol, and the families of two children (both assigned to placebo treatment) could not be reached for collection of follow-up data.
The clinical and demographic characteristics of the two treatment groups are shown in Table 1. There were no significant differences between the treatment groups with respect to any of the clinical or demographic variables measured at base line.
|
Each of the outcome measures was examined for changes in the severity of symptoms that were related to treatment (placebo or secretin), time of assessment (base line, day 1, day 2, week 1, week 2, or week 4), or the interaction of these factors. Two patterns of results were found. First, secretin was not associated with significantly greater improvements on any of the 16 outcome measures than was placebo (P values ranged between 0.20 and 0.88). Scores were not significantly better with secretin treatment than with placebo for the overall severity of autistic symptoms (P=0.33), communication skills (P=0.21), specific diagnostic features of autism (P=0.73), or nonspecific features of autism (P=0.88). An example of this pattern is shown in Figure 1.
|
|
|
|
Nine of the 27 children in the secretin group for whom there were complete data (33 percent) and 7 of the 25 children in the placebo group for whom there were complete data (28 percent) were considered to have had a response to treatment. There was no significant difference between the treatment groups in the proportions of children who met the criteria for a positive response (P=0.68). Within the secretin group there were no significant associations between response status and diagnosis, age, IQ, presence or absence of digestive symptoms, or use or nonuse of adjunctive medication (P>0.35 for all comparisons).
Adverse Events
None of the children had treatment-limiting side effects. Comparisons of the secretin and the placebo groups showed marginally higher severity scores on the Treatment Emergent Symptoms Scale for the children in the secretin group than for those in the placebo group with respect to decreased activity level (P=0.15).
Parent Interviews
Parents of all of the children who completed the study were interviewed at its conclusion to determine their intentions with respect to future treatments for their children. Before the interview, the parents of each child were made aware of their child's treatment assignment, the specific study outcomes of their child, and the overall results of the study with the use of a standardized script. Sixty-nine percent of the parents remained interested in secretin as a form of treatment for their children (63 percent of the parents of children in the secretin group and 76 percent of the parents of children in the placebo group).
Discussion
We used a set of standard outcome measures to compare children with autism or pervasive developmental disorder who received either a single dose of synthetic human secretin or placebo under double-blind conditions. We found no significant effects of secretin treatment. This lack of effect was seen overall, for the subgroup with autism, and for individual children. Furthermore, decreases in the severity of symptoms with respect to some of the outcome variables were reported over time for both treatment groups. The 95 percent confidence intervals for the differences from base-line scores provide estimates of the plausible changes in the severity of symptoms that could be obtained with the measures we used. The 95 percent confidence intervals for the differences between the groups were narrow, indicating that the differences provide a reasonable estimate of the plausible effects of secretin treatment. Furthermore, the bounds of the 95 percent confidence intervals indicated that secretin-related improvements in the severity of symptoms amounted to a change of no more than 15 percent from base-line levels for any of the measures we used. This finding suggests that although we could not rule out small improvements related to secretin treatment, improvements of this magnitude would not be clinically meaningful for children with autism or related disorders.
Our study had several weaknesses. First, this was a short-term study, and it is unlikely that significant changes in behavior can occur in a brain-based disorder within days or weeks. Second, this was a single-dose study, and multiple doses may prove to be more efficacious. We conducted a short-term evaluation of a single dose of secretin to provide a replication of the circumstances of the anecdotal reports of improvement observed after secretin treatment.8,9,10,11 Third, recent studies have used other diagnostic schedules to identify autistic disorders,21 and specific subtypes of autism quantified according to these schedules may differ in the potential for response to treatment with secretin. Finally, we used synthetic secretin even though the majority of anecdotal reports of the effects of secretin have been based on the use of the biologic (porcine) product. Given that synthetic human secretin and porcine secretin have equivalent pharmacologic effects in humans13 and that porcine secretin carries the additional risk of immunogenicity in humans, we chose to examine the effects of the synthetic form of human secretin.
The lack of positive effects of secretin treatment with respect to any of the features of autism stands in contrast to the anecdotal reports of positive and even dramatic effects that have appeared recently on television, in print, and on the Internet. The widespread circulation of anecdotal reports of the benefits of secretin in the treatment of autism may have raised expectations among parents and care providers and biased them toward perceiving improvement.
Suboptimal or ineffective treatments for autism are often promoted and widely accepted in the absence of empirically validated treatment. In this respect, secretin may become one of the several proposed alternative treatments for autism (others include special diets, vitamin therapy, facilitated communication, and sensory integration) that continue to be used despite a lack of empirical validation.6,22 We found that most parents remained interested in secretin as a form of treatment for their child's autism even after being told that we found no evidence of benefit. It is important to note that other treatments for autism have been empirically validated. These include direct behavioral instruction to improve social interaction and communication skills23,24 and behavioral25 and medical3,5,26,27,28,29 interventions for managing the aberrant types of behavior associated with autism. In summary, we found that a single dose of synthetic human secretin was similar to placebo for the treatment of children with autistic disorders.
Supported by the Thoms Health Services Foundation and by a Public Health Service grant (30615) from the National Institutes of Child Health and Human Development.
We are indebted to Drs. Olson Huff, Don Stedman, Jim Favell, and Iverson Riddle for their encouragement in the conduct of this trial; to Jim Barnett for assistance with infusions; to Caroline Freeze, Vicki Harper, and Dawn Parker for assistance with data collection; and to ChiRhoClin for providing human secretin.
Source Information
From the Olson Huff Center for Child Development, Thoms Rehabilitation Hospital, Asheville, N.C. (A.D.S., J.D., M.A.G., V.S.); the Human Development Research and Training Institute, Western Carolina Center, Morganton, N.C. (K.A.S., J.W.B.); and the Department of Psychiatry, University of North Carolina at Chapel Hill, Chapel Hill (J.W.B.).
Address reprint requests to Dr. Sandler at the Olson Huff Center for Child Development, Thoms Rehabilitation Hospital, 68 Sweeten Creek Rd., Asheville, NC 28803, or at adsandler{at}pol.net.
References
| |||||||||||||||||||||||||||||||||||||||||||||
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
