Duodenal eosinophilia has been associated with dyspepsia in adults and the investigators have previously described the finding of duodenal mucosal eosinophilia in 71-79% of children undergoing diagnostic endoscopy. Previous studies in children have shown positive response to montelukast with approximately 50% finding complete relief and 20-30 percent showing no response.

There are a number of factors that have the potential to contribute to the observed variability in response to montelukast. These include variability in:

1. systemic drug exposure (drug absorption, biotransformation and/or elimination)
2. regulation of leukotriene biosynthesis
3. cysteinyl leukotriene receptors and downstream mediators
4. patient disease phenotype (e.g. Functional Gastrointestinal Disorder (FGID) disease classification, psychologic profile)

In this study, the investigators propose to utilize biopsy specimens stratified by drug response to identify candidate gene expression modules that will be validated in a prospective study design. The overall goal of this program is to develop a signature of montelukast response that can be applied not only to eosinophilic gastroenteritis, but more generally to other diseases, such as asthma, where the drug is widely used with variable success.

Inclusion Criteria:

• Ages 8 - 17 years, inclusive
• Abdominal pain of at least 8 weeks duration and fulfilling symptom- based criteria for functional dyspepsia
• Scheduled for endoscopy following failure to respond to acid-reduction therapy
• Evidence of written parental permission (consent) and subject assent

Exclusion Criteria:

• Previous treatment with montelukast
• Treatment with corticosteroids or oral cromolyn sodium in the four weeks prior to enrollment
• Prior history or clinical signs/symptoms of chronic disease requiring regular medical care (e.g., diabetes mellitus, juvenile idiopathic arthritis, cystic fibrosis or cancer)
• Exposure within the past two weeks to drugs or natural products that induce CYP2C8/9 or CYP3A4, including amprenavir, carbamazepine, lopinavir/ritonavir, nafcillin, nevirapine, oxcarbazepine, phenobarbital, phenytoin, rifampin, St. John's Wort, or that inhibit CYP2C8/9 or CYP3A4, such as ciprofloxacin, clarithromycin, erythromycin, fluconazole, fluvoxamine, grapefruit juice, paroxetine, sertraline, sulfamethoxazole, trimethoprim
• A Body Mass Index of 30 or greater
• Non-English speaking
• Those patients who will turn 18 during the duration of the study

Up to 50 subjects (allowance of 20 additional subjects for screen fails and those who will not complete Phase 3 of the study) will be recruited from a cohort of patients referred to the Abdominal Pain Clinic at Children's Mercy Hospitals and Clinics and enrolled prior to routine initial endoscopy.