How Can Biomarkers Enhance Clinical Trials in Autism Spectrum Disorder?

Background:  Pharmacological clinical trials in autism spectrum disorder(ASD) have yielded mixed results in efficacy.  One challenge that may contribute to this situation is the heterogeneity in study participants, lack of biological targets associated with deficits in the core domains of ASD dysfunction, and the lack of objective biological measures of treatment response.

Objectives:  The goal of this presentation is to discuss the promise of biomarkers in stratifying subjects to enrich for the specific deficit being tested in the trial, and to highlight outcome measures that may inform whether a treatment is working. 

Methods:  NIMH has created a program, called the “Fast Fail Trials” which is designed to develop and test ASD biomarkers in clinical trials of investigational compounds.  Before initiating the first trial, we formed an advisory committee to establish compound selection criteria, identify compounds to test and then help inform the biomarker selection.

Results:  There is a wide range of biomarkers that are being tested in ASD but much fewer being incorporated into trial designs.  Examples of potential biomarkers include brain activity measures such as fMRI and EEG/MEG, peripheral measures that correlate with sympathetic nervous system activity, eye tracking, various cognitive assessments, actigraphy, sleep measures and peripheral blood measures.  This presentation will provide an overview of different biomarkers being studied in ASD that could be used in stratifying subjects and/or assessing treatment response. The significant effort needed to incorporate biomarkers in ASD trials will be emphasized.  To illustrate how this can be done, the presentation will provide 1-2 case examples on the application  of biomarkers  into an ASD trial, based on NIMH’s recent efforts in the Fast Fail Program

Conclusions:  

Different methodologies are becoming available to test how biomarkers can be used to stratify or enrich for subjects with specific phenotypes (deficits in a core domain of function) in ASD clinical trials.  However, many studies have used small numbers of subjects and broad inclusion criteria and therefore the measures have not been tested or validated for their ability to define subgroups with ASD that may benefit most from the intervention. Challenges remain as to the feasibility, specificity (developmental age, level of function, core symptom), reproducibility, sensitivity of measures to change over time, etc.  Especially in early treatment trials,  the ability to correlate a brain activity measure such as EEG or fMRI with cognitive outcomes may be more informative than reliance on behavioral measures or patient reported outcome measures.  The timing is right to begin incorporating quantitative biological measures into trials to test novel hypotheses for therapeutic interventions.