Immune Response to Pregnenolone Treatment in Adults with Autism Spectrum Disorder – Preliminary Analysis from an Open-Label Study

Friday, May 12, 2017: 5:00 PM-6:30 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
L. K. Fung1, J. Siebert2 and A. Y. Hardan1, (1)Psychiatry and Behavioral Sciences, Stanford University, Stanford, CA, (2)CytoAnalytics, Denver, CO

Pregnenolone (PREG) is the precursor of endogenous pharmacologically active neurosteroids. We recently reported the results of an open-label trial of PREG in the treatment of adults with autism spectrum disorder (ASD). We found that PREG reduced the levels of irritability and associated aggressive behaviors as measured by the Aberrant Behavior Checklist – Irritability subscale (ABC-I). PREG was also found to be well-tolerated by study participants. We hypothesize that responders of PREG will have a different immune biosignature, as compared to non-responders.


To explore the association between response to PREG and plasma concentrations of immune biomarkers before and after 12-week trial of oral PREG.


PREG was initiated at 50mg twice daily in weeks 1 and 2, then increased by 50mg twice daily every 2 weeks to a final dose of 250mg twice daily which was maintained from weeks 9 to 12. Primary outcome measure was the ABC-I. Response was defined as change of ABC-I of 7 or greater. The plasma levels of 62 cytokines and other immune biomarkers were measured by Luminex Multiplex Analyses (LMA; at the Human Immune Monitoring Center at Stanford University) before and after PREG treatment in our open-label trial in adults with ASD. LMA is a simultaneous analysis of multiple analytes from the same sample using differentially dyed beads. Two-tailed student t-tests were performed to compare selected pro-inflammatory cytokines between responders and non-responders. Multivariate analyses, including decision tree analyses and elastic net regression, of the immune biomarker data before and after PREG treatment were performed.


Twelve individuals with ASD (mean age 22.5 years) participated in this open-label study. PREG yielded improvement in the primary measure, ABC-I [17.4±7.4 at baseline; 11.2±7.0 at 12 weeks (p=0.028)]. Six participants were found to be responders to PREG treatment, while the rest of the six participants did not reach the responder criteria. Preliminary decision tree analysis suggested that participants can be cleanly classified into responders and non-responders based on baseline levels of IL-31 and fibroblast growth factor β (FGFβ). Targeting the cytokines known to have increased levels in individuals with ASD (IL-1β, IL-6, IL-8, IFN-γ, eotaxin and MCP-1), the levels of the selected cytokines at baseline and post-treatment were statistically indistinguishable. In exploring the plasma levels of other immune biomarkers in the open-label PREG study, the responders were found to have significantly lower levels of plasminogen activator inhibitor 1 (PAI-1) post-treatment, compared to non-responders (p=0.042). Elastic-net regression did not yield meaningful results on either pre- or post-treatment LMA data.


Preliminary LMA analysis found that specific immune biomarkers were associated with reduction of ABC-I. These findings suggest that effectiveness of PREG might be related to baseline levels of IL-31 and FGFβ. Additional research is needed to replicate these findings in a controlled trial.