Place of Pilot Trials in Drug Development for ASD

Friday, May 16, 2014: 3:30 PM
Imperial A (Marriott Marquis Atlanta)
L. Scahill, Pediatrics, Marcus Institute, Emory University, Atlanta, GA
Background:  Large-scale, NIMH-funded, multisite trials in ASD take time (three to five years) and money (5 to 10 million dollars depending on the sample size and length of trial).  If we are to use resources wisely, the selection of compounds for large-scale study warrants careful consideration.

Objectives:  The purpose of this presentation is to describe the role of pilot studies in drug development. The discussion will also highlight the research questions that pilot studies are not equipped to answer. 

Methods:  We surveyed published literature over the past decade to identify the aims and findings of selected pilot studies in ASD. We also reviewed seminal papers by experts on the appropriate use of pilot studies in drug development.

Results:  Pilot trials paly an essential role in drug development by providing information on drug mechanism, drug dosing, time to clinical effect, adverse effects and acceptability in the ASD community. For example, in the absence of hype (secretin), new compounds with no prior use in in ASD may be greeted with skepticism in the autism community. Assumptions about the drug mechanism may result in narrow entry criteria that may hinder or even halt enrollment. On the other hand, pilot trials focused on mechanism should avoid overly broad entry criteria, which could result in a heterogeneous sample and a poor test of the mechanism.  Dosing strategies adopted from another clinical population may not be tolerated in ASD resulting in attrition.  That subjects with ASD do not tolerate medications as well as other clinical populations argues in favor of flexible rather than fixed-dose strategies.  The presentation will show that information on acceptability, tolerability and treatment compliance is more useful than information on efficacy in pilot trials. Indeed, focus on efficacy in pilot studies inevitably faces the possibility of false positive and false negative findings. Over-interpretation of false positive findings may lead to unnecessary expenditure of resources on compounds that do not warrant further study.  Premature acceptance of negative findings from a pilot study may lead to rejection of a potentially useful drug. The presentation will also provide examples of useful pilot trial designs.


The role of pilot studies is to assist with the critical judgment on whether a given compound warrants further study. Pilot trials are not about efficacy and not about estimating effect size for a larger clinical trial.