A Folinic Acid Intervention for Autism Spectrum Disorders

Evidence-based interventions for autism spectrum disorders (ASD) are limited. Currently,
there is no FDA approved medical therapy that addresses either core ASD symptoms or
pathophysiological processes associated with ASD. For example, the two FDA approved
antipsychotic medications for treating ASD are indicated for symptoms associated with ASD,
not core ASD symptoms. Based on strong preliminary evidence, this study aims to improve core
ASD symptoms with folinic acid. Two biological mechanisms for response to the intervention
are proposed and investigated in this study.

Folate is an essential B vitamin required for normal neurodevelopment. Folinic acid is a
reduced form of folate that has been shown to improve both the metabolic and core symptoms
associated with ASD in an open-label study and large case-series. This study will extend this
preliminary evidence to demonstrate clinical efficacy of folinic acid in a double-blind
placebo-controlled manner while also evaluating the biological mechanisms associated with the
clinical response.

Preliminary evidence for the efficacy of folinic acid is two-fold: First, several independent
studies have demonstrated that a folinic acid intervention improves ASD core symptoms in ASD
patients diagnosed with cerebral folate deficiency - a metabolic disorder in which the
primary pathway for the transport of folate across the blood-brain barrier is dysfunctional.
The primary transport pathway for folate across the blood-brain barrier uses the folate
receptor α (FRα) and energy dependent endocytosis. A secondary pathway, the reduced folate
carrier (RFC), can transport reduced forms of folate, such as folinic acid, across the
blood-brain barrier. In case-series of children with ASD and cerebral folate deficiency,
folinic acid (0.5 to 2 mg/kg/day) improved communication, social interaction, attention, and
stereotypical behavior in some (3-6) and completely ameliorated ASD symptoms in others (6,
7).

The major cause of cerebral folate deficiency is an autoantibody that binds to the FRα and
interferes with its ability to bind and transport folate. Recently, Frye et al. found that
(a) autoantibodies to the FRα were present in approximately 75% of children with ASD, and (b)
an intervention with folinic acid (2mg/kg/day; max 50mg) in children with ASD and FRα
autoantibodies resulted in significant improvements in parental ratings of receptive and
expressive language, verbal communication, stereotypic behavior, and attention as compared to
parental rating for children that did not undergo any intervention (wait group) over a
similar time period. This proposal will extend these preliminary findings by documenting
response to a folinic acid intervention in a double-blind placebo-controlled manner and test
whether FRα autoantibody titers predict response to a folinic acid intervention. If FRα
autoantibody titers are found to predict response to the intervention, such titers could
provide a biomarker to identify a subset of children with ASD that may benefit from a folinic
acid intervention and may even predict the development of ASD symptoms in high risk siblings
during the presymptomatic period.

Second, the ratio of reduced-to-oxidized glutathione (GSH/GSSG) is an established measure of
systemic redox status and oxidative stress that has been shown in three independent
case-control studies to be significantly decreased in many ASD children. James et al. have
shown in a 3 month open-label clinical trial that an intervention of folinic acid and
methyl-B12 significantly improved GSH/GSSG and Vineland scores for expressive language,
receptive language, and socialization in ASD children. Important unanswered questions are (a)
whether this preliminary data can be confirmed in a larger double-blind placebo-controlled
study, (b) whether redox status before the folinic acid intervention can predict response to
intervention and (c) whether improved systemic redox status is associated with improvement in
core ASD symptoms.

This study will assess whether a folinic acid intervention can improve both core symptoms of
ASD (i.e., communication, socialization, stereotyped movements) and associated comorbid
symptoms (i.e., attention) in a double-blind placebo controlled clinical trial. This study
will address whether improvements in core ASD symptoms associated with the intervention are
related to biomarkers of either (or both) biological mechanisms proposed to be influenced by
the intervention (i.e., GSH/GSSG, FRα autoantibody titer). Thus, using these biomarkers, it
may be possible that children with ASD who optimally respond to the folinic acid intervention
can be readily identified early after diagnosis or even during the pre-symptomatic period.

This double blind, placebo-controlled (DBPC) study will evaluate the efficacy of Folinic Acid
for the treatment of language impaired autism spectrum disorder patients aged 3-14 years. The
study will consist of a single site trial with approximatley 48 entering the DBPC phase.
Subjects will be male or female with current or prospective diagnosis of Autism Spectrum
Disorder (ASD). Language Impairments will be defined by the CELF, or PLS when warranted.

Phase 1 will consist of approximately 57children aged 3-14 with a diagnosis of Autism
Spectrum Disorder (as defined by the Autism Diagnostic Observation Schedule or the Autism
Diagnostic Interview-Revised), until approximately 48 children with confirmed ASD are
enrolled into a randomized control trial (RCT; Phase 2). Enrollment will continue as needed
until approximately 43 complete Phase 2. Approximately 40 will enroll into phase 3.

Initial analyses will be undertaken to inspect data for errors, inconsistencies, and
incomplete information. This will include examining the data with simple frequency tables and
dot plots for univariate data and scatter plots and multi-way dot plots with bivariate and
multivariate data. To summarize bivariate relationships among predictors and between
predictors and outcomes, Spearman's rank correlation coefficient will be used. For reporting
inferential statistics, regression coefficients along with the 95 percent confidence
intervals will be used extensively to quantify degree of clinical importance.

The primary population will be the Intent-to-Treat (ITT) population and is defined as all
subjects who have completed at least one post treatment primary outcome measure. A secondary
analysis will be performed on the "as treated" population defined as those subjects who had
at least 75% compliance of prescribed medications and have not significantly deviated from
the protocol.

To determine whether the folinic acid intervention over a 12-week period improves core
symptoms of autism spectrum disorder in Phase 2, the procedure Proc Mixed in SAS will be used
to fit a mixed model analysis of covariance. For the multisite study, the full model with the
response variable language assessment will include predictors such as: intervention, time,
autoantibody, intervention by time interaction, autoantibody by intervention interaction,
redox status, redox status by time interaction, and baseline covariates. However, for the
single site study, the number of variables in the model will be limited due to the smaller
sample size and separate models will be performed on subgroups to compare the effects of the
treatment in various subgroups. The results from the mixed model will determine if the
biomarkers of physiological abnormalities, both autoantibodies and redox status, predict
intervention response. Similar analysis will be performed for the secondary outcomes.

Inclusion Criteria

- 1. Autism Spectrum Disorder (as defined by a gold standard measure for ASD diagnosis:
the Autism Diagnostic Observation Schedule and/or Autism Diagnostic
Interview-Revised). In an event where sufficient diagnostic information is lacking,
and the PI believes that the clients meet all other inclusion criteria and a
prospective diagnosis of an ASD is clinically warranted, and a formal diagnosis is
scheduled to occur within a reasonable time frame from the date of study entry but
before dispersing study drug/placebo, then the client may be considered as potentially
eligible. Furthermore, a research-reliable rater must complete the diagnosis.

- 2. 3 years to 14 years of age

- 3. Language Impairment

- 4. Ability to maintain complementary, traditional, and/or behavioral interventions and
to attempt to keep them constant during the study, when possible.

- 5. Unchanged complementary, traditional, and/or behavioral intervention for
approximately 8 weeks prior to study entry, when possible.

Exclusion Criteria

- 1. Currently taking Antipsychotic medication

- 2. Vitamin or Element Supplementation that exceeds the IOM Tolerable Upper Intake
Levels

- 3. Any moderate to severe positive response on that Aberrant Behavior Checklist
Irritability subscale on questions: Injures self on purpose, is aggressive to other
children or adults (verbally or physically), deliberately hurts himself/herself,
and/or does physical violence to self.

- 4. Prematurity (<34 weeks gestation) as determined by medical history

- 5. Current uncontrolled gastroesophageal reflux or ongoing significant kidney or liver
disease. The PI will determine whether any ongoing kidney or liver disease is
significant.

6. Drugs known to affect folate metabolism (e.g., methotrexate) and their derivatives.

7. Profound sensory deficits (e.g. hearing and vision deficits) that could interfere
with the interpretation of study results.

8. Any major genetic defect, or mutation, that is known to be associated with disease
or possibly related to disease that affects folate, methylation, and/or glutathione
metabolism. Questions regarding eligibility concerning this criterion will be
addressed with the lead site PI before enrollment into the trial.

9. Documented current or active seizures, as defined by a clinical seizure or abnormal
EEG within the past 6 months.

10. Children with major single-gene abnormalities, such as Fragile X, Rett's Syndrome,
etc., recognized chromosome syndromes, such as 15q11 microdeletion syndrome, or have
been diagnosed with other well recognized syndromes, such as fetal alcohol syndrome.
Children with copy number variants that represent known polymorphisms or benign
changes will not be excluded. Questions regarding eligibility concerning this
criterion should be addressed with the lead site PI before enrollment into the trial.

11. Children diagnosed with congenital brain malformations, acquired brain insults,
congenital or acquired microcephaly, or infection of the central nervous system.

12. Children with major well-defined metabolic disease, such as mitochondrial disease,
urea cycle disorders, succinic semialdehyde dehydrogenase deficiency, creatine
deficiency syndromes, etc.

13. Current therapies that could potentially interfere with interpretation of study
results.

14. Other conditions which, in the opinion of the study team, will place subjects at
unacceptable risk or result in inability to interpret the study data.

15. Unwillingness or inability to return for follow-up testing at specified interval.