The proposed study represents the first attempt to systematically investigate the use of DCS as an adjunct to behavioral intervention to address chronic food aversion through an randomized control drug trial in children treated at the Marcus Autism Center's Pediatrics Feeding Disorders Program. This pilot and feasibility study will involve a total of 16 participants randomly assigned to experimental conditions: behavioral intervention or behavioral intervention + DCS (8 in each group). All participants will receive the same behavioral protocol involving three treatment sessions per day (45 minutes in length), for a total of 15 sessions across five consecutive days. In addition, participants in the behavioral intervention + DCS group will receive a low dose (0.7 mg/kg) of the drug using an acute dosing methodology, which has been demonstrated to produce a nearly negligible side effect profile with comparable treatment outcomes to chronic dosing. Timing of dosing will occur 1 hour prior to behavioral intervention, in line with prior clinical studies. Study staff, with consultation from a psychiatrist, will observe administration of DCS to participants by caregivers via their preferred method of formula consumption (bottle, cup, or tube) in liquid form. Participants will be evaluated during each treatment session and at follow-up using trained observers to collect data on mealtime behaviors, including acceptance, swallowing, disruption, expulsion, and grams consumed. This type of data collection is standard practice in the feeding disorders program. It is hypothesized that participants who receive DCS as an adjunct to behavioral intervention will show greater improvement in mealtime behaviors as reflected by these measures.

• Extinction of fear is thought to use similar learning mechanisms as learning or conditioning of fear, and both are blocked by antagonists at the glutamatergic N-methyl-D-aspartate (NMDA) receptor. Furthermore, agonists at this site appear to augment some forms of learning in animal and human trials. The process of extinction of conditioned fear was initially shown to be facilitated by D-Cycloserine (DCS), an NMDA agonist, given in individual doses prior to extinction training in an animal model. There is growing evidence that a similar effect is found in human subjects undergoing controlled exposure therapy for specific phobia. This translational research proposal represents a randomized-control study with the goal of determining whether a drug that acutely enhances learning in both human and animal research will facilitate the extinction of fear that occurs with behavioral therapy for children with food aversion. Specifically, it is proposed that a single dose of DCS, given shortly before each of 5 days of behavioral therapy sessions utilizing escape extinction, will significantly enhance the rate of response and possibly the efficacy of treatment for pediatric food aversion. To this end, the investigators propose to enroll 16 participants with food aversion. Participants will be randomly assigned to receive behavioral intervention alone or in combination with 0.7mg/kg DCS. Behavioral ratings and outcomes will be assessed by independent assessors blind to subject condition. Participants will be assessed pre-treatment and at a one month follow-up session to assess long term effects.
• This type of combined treatment -- specific pharmacotherapeutic augmentation of behavioral therapy -- would be novel in pediatric populations with food aversion and would potentially be generalizable to many different forms of behavioral intervention for a wide range of pediatric disorders. The potential for detrimental health outcomes associated with feeding disorders, combined with their complex biopsychosocial etiology, intensifies the need to identify and refine effective treatments. If this translational research is successful, the ability of a relatively benign agent administered acutely before a behavioral therapy session to facilitate the extinction process could have important clinical, humanitarian, and economic advantages.

Specific Aims: The entire proposed study is designed to achieve the following specific aims:

1. To explore whether d-cycloserine (DCS), an NMDA partial agonist, facilitates extinction of food aversion in children using extinction based behavioral intervention.
2. To explore whether any facilitation in extinction produced by DCS in behavioral intervention evidenced within session and immediately post-treatment results in long-term gains in treatment response as compared to behavioral intervention alone.

2.2 Research Hypotheses: The following hypotheses will be tested:

1. DCS will facilitate extinction in children. It is predicted that the children who receive behavioral intervention combined with 0.7mg/kg DCS will evidence more rapid improvement in mealtime behavior (acceptance, swallowing, disruptions, expulsions, grams consumed) than children receiving behavioral intervention alone.
2. Facilitation in therapeutic response aided by DCS will result in long-term gains. It is predicted that the group of patients who receive behavioral intervention combined with 0.7 mg/kg DCS will evidence more improvement at the follow-up assessment as compared to baseline behavior (acceptance, swallowing, disruptions, expulsions, grams consumed) than the group who receives behavioral intervention alone.

Gender and Minority Participants' Inclusion Plan. Patients of all races and ethnic groups will be entered, both males and females. The investigators will recruit from the year long waiting list for behavioral services at the Feeding Disorders Program. The investigators will advertise via fliers at the Marcus Autism Center, as well as local pediatric and gastroenterological practices in the Atlanta area. (See Flyer Info to review content of the flyer). The racial composition of Atlanta is 71% white, 26% black, and 3% other (based on the 1990 US Census). Approximately 1% of these persons are of Hispanic origins. The gender composition of Atlanta is 51% female. This study will provide free treatment, thus ensuring equal opportunities for all to learn about and participate in this study. The treatment setting is located in a racially diverse county (DeKalb County), is easily accessible by public transportation and is wheelchair accessible. It is therefore expected that the study sample will closely approximate the demographic composition of Atlanta.

Treatment:

Behavioral Intervention: Treatment will commence following the pre-treatment assessment. Participants will be randomly assigned to one of two treatment groups: behavioral intervention alone or behavioral intervention plus medication. All participants will be treated for a period of 5 consecutive days. A total of 3, 45-minute meals will be held at regularly scheduled times (e.g., 9:00 a.m., 10:30 a.m., and 12:00 p.m.) each day for a total of 15 meals throughout treatment. Trained feeding therapists from the Marcus Autism Center's Pediatric Feeding Disorders Program will conduct sessions in treatment rooms equipped with one-way mirrors and an adjacent observation room to allow caregivers to watch all treatment sessions. Behavioral intervention targeting severe feeding disorders involves the combination of escape extinction and antecedent manipulation of food presentation to lessen the aversive quality of the meal. This allows treatment to address the consequences maintaining food refusal in the least restrictive environment while ameliorating possible side effects associated with extinction procedures. Please refer to the nonremoval of the spoon + representation + redistribution (NRS + REP + RED) protocol for a detailed example the proposed treatment. Data will be collected during each meal and all behavioral treatment sessions will be video recorded for the purpose of ensuring accurate protocol implementation and assessing reliability. Caregivers will be trained on the protocol during the last session of treatment to promote transition of the protocol into the home setting.

Medication: As described above, the investigators will compare behavioral intervention alone to behavioral intervention plus 0.7mg/kg DCS. Patients will be instructed to take the medication under the supervision of study personnel one hour prior to the first treatment session. A placebo will also be given to participants in the behavior only condition using a similar method of administration.

Design and Plan: The proposed design is to randomly assign 16 participants with food aversion to behavioral intervention or behavioral intervention plus 0.7mg/kg DCS to be taken acutely one hour prior to the onset of the first treatment session each day. All participants will receive three, 45 minute treatment sessions per day for a period of five days. Study medications will be administered in the clinic supervised by study personnel to ensure compliance. The investigators will be using the Children's Healthcare of Atlanta Investigational Drug Service (IDS) to provide both drug and placebo. Participants will be assessed pre-treatment and at one month following the termination of treatment by a blinded independent assessor.

Inclusion Criteria:

• Present with partial food refusal as evidenced by greater than 50% of caloric needs met by bottle, formula, or tube feedings, thus eliminating children whose lack of consumption is related to a skill deficit
• Have a medical history significant for an organic factor (e.g., gastrointestinal issues) which precipitated or played a role in the development of feeding concerns, thus capturing children whose food aversion mimics animal and human models of anxiety and aversion
• Between the ages of 18 months and 6 years
• Live within 2 hours of the Feeding Disorders Program at Marcus Autism Center to increase retention and maximize attendance
• English speaking

Exclusion Criteria:

• Patients with previous behavioral treatment for feeding disorder
• Patients with active medical conditions requiring ongoing hospitalization
• Patients unwilling to take study medication

Sixteen subjects (10 male subjects, 6 female subjects) were randomly assigned to DCS or placebo (Table 1). There were no significant group differences on demographic or clinical measures at baseline. Participants were enrolled beginning in April 2013 through July 2013, with final follow-up visits completed September 2013.

The sample size was small and analyses were underpowered to detect group differences.