Lessons Learned from Phase I Proof-of-Mechanism and Biomarker Studies in ASD: Measurement and Trial Considerations

Background:  Next generation trials of drugs targeting ASD core domains will depend on precision of measurement of social dimensions. From studies investigating biomarkers and potential efficacy signals of antagonists of arginine vasopressin 1a receptor (AVPR1A), data enable comparative analyses of a multi-level panel of potential endpoints.  The presentation will present analyses comparing sensory, cognitive, and clinical endpoints from two studies: 1) a single administration study of effects of V1a receptor antagonist RG7713 in ASD; and, 2) a comparative biomarker study testing measures in patients and healthy controls. Possible relevance to RDoC domains will be discussed.

Objectives: To present analyses comparing sesnory, cognitive, and clinical endpoints from two studies. 

Methods:  In Study 1, 19 adults with ASD (mean age=23.4 ± 5.16 (SD) years (range: 18-40) participated in randomized, double-blind, placebo-controlled, cross-over study. Single doses of RG7713 or placebo were administered intravenously on two different days one week apart. Measures of eye-tracking, affective speech recognition (ASR), emotion processing (“Reading the Mind in the Eyes Test [RMET))”, olfactory recognition, social communication, parent-rated measures of social behavior (ABC-scale, 28 item adapted for the study with items 3,4 and 18 removed), and global improvement as measured by blind raters were obtained and compared. Additional comparisons with baseline assessments including the ADOS and Vineland are included. In Study 2, 19 healthy controls (mean age=26.68±4.33 (range 20-35)) and 19 adults with ASD (mean age=24.89±6.45 (range 19-39)) were assessed in a two-visit study of a similar battery used in Study 1. Separate and combined analyses were performed to examine associations; comparisons between post-treatment scores were calculated, as well as overall comparisons between both treatment conditions versus baseline assessments.

Results:  Several unique patterns of correlations were observed between measures. Correlations of medium-large (|r|>0.5) effect are emphasized.

For social communication measures, correlations were:

• positive between baseline ADOS social interaction scores and post-treatment ABC adapted (rs~-0.58 – 0.68)
• negative between baseline ADOS subtest scores and post-treatment Scripted Interaction (total) (rs~-0.56 – 0..93)
•  negative  between ABC adapted and Scripted Interaction total score (negative correlation, r~-0.8)
• negative between ABC adapted and VABS Social Communication (rs~-0.6 – 0.64)

For social perception measures, correlations were:

• positive between baseline RMET and ASR scores (r~ 0.48)
• positive between RMET % correct responses and Smell test scores (positive correlation, r~0.5-0.6). RMET % correct responses were also positively correlated with Verbal and Full scale IQ (rs~ 0.52 – 0.72).
• not seen between eye tracking and other social perception measures

Conclusions:  Results from these preliminary comparisons of multiple domains of social behavior, communication, and perception support their consideration as largely independent, though related, dimensions. The high proportion of shared variance of several measures, such as the ABC SW and Scripted interaction, with other well-established clinical tools such as the ADOS and VABS support their further use as ASD trial endpoints.