In children with autism, is intravenous secretin more effective than placebo in improving social skills, communication, behaviour or global functioning?

Abstract

This kind of question arises frequently in an autism follow-up clinic or other developmental disability setting. The parents of children with autism are bombarded with information about their child’s condition: “you should try…” or “you should have avoided…”, and often ask their physician for support in the form of either information, services or advocacy. Physicians rarely have good information on which to base their decision to support such a family. The secretin story illustrates a number of important clinical issues. It is often difficult to evaluate the efficacy of new therapies, especially in a chronic behaviour-based condition like autism. Because secretin is administered by injection, either once or by infrequent repeats, it is relatively easy to administer in a ‘blinded’ fashion. Other interventions are not as easily studied. Outcomes are difficult to measure in autism, even relatively concrete ones like ‘gastrointestinal symptoms’. It is rare that a new therapy will be subjected to this kind of research evaluation (ie, multiple randomized controlled trials) so quickly, so we are usually left to make decisions or recommendations without good evidence. Interventions should be assessed on their theoretic basis (does it make any sense given what we know about the condition?), their potential risks and costs (including ‘opportunity costs’), and on the evidence, whether anecdotal or more scientifically validated. It is wonderful to have even one treatment for autism where there is scientifically valid information. The secretin experience is an important lesson in the politics of alternative therapies. There has been argument about the type of secretin to use (porcine versus human recombinant), about the number of injections to give (and if multiple injections were safe!), and about the possibility that there is a subgroup of responders hidden within the group data. Some physicians felt comfortable giving this injection, while others did not. Some families were able to find supplies, while others couldn’t. One of the research teams commented on the ease with which they recruited families for study, given the high interest and lack of clinical access. The company manufacturing human recombinant secretin (RepliGen, Waltham, USA) recently discontinued their Phase III study of children with autism for lack of evidence of efficacy (1), but they continue with trials for other indications. A very important outcome of multiple secretin studies was the frequency of ‘placebo response’ that was identified. Approximately 30% of children in the various studies had ‘improvement’ in their symptoms of autism during the study period, although this turned out to be unrelated to secretin. This implies that in other clinical situations there is a similar chance of improvement associated with interventions in which one believes (whether of proven or of unproven benefit). Chronic disabling conditions like autism require ongoing management, and children with autism often receive therapies of no proven benefit (whether accepted or ‘alternative’). It is very easy to interpret improvement in some children during treatment as ‘proof’ of benefit, which would appear to be the basis for the many anecdotally ‘successful’ therapies in autism or other chronic disabling conditions.