A Trial of Montelukast for Maintenance Therapy of Eosinophilic Esophagitis in Children

Eosinophilic esophagitis (EE) is a disease diagnosed using both clinical and pathologic
results. Presentation varies among age groups with younger children suffering from
vomiting, failure to thrive, or food refusal and adolescents suffering from dysphagia,
odynophagia, or food impaction. It is more prevalent in males. EE is suggested when an
upper endoscopy yields eosinophils greater than 15 per high power field (hpf) in both the
middle and distal esophagus. Therapy for EE depends upon age and ability to be compliant.
Three main categories of treatment are options for families, including the use of medicines,
removal of the most allergenic foods, and the use of an elemental diet (formula) (1-2).
Treatment can be different within the two main phases of therapy for EE, initial and
maintenance therapy. Initial therapy occurs when the patient is first diagnosed and has
eosinophils over 15 per hpf and can last for years if the patient does not respond to
treatment. Maintenance therapy only occurs after the patient has responded to treatment
with a reduction of his/her eosinophil counts below 5 per hpf. Typically, medicinal therapy
with steroids (either oral prednisone or swallowed fluticasone) improves eosinophil counts
but has a large side effect profile suggesting that steroids are not ideal for maintenance
therapy. Dietary therapies are effective in both treatment phases but many patients find it
difficult to remain compliant with food restriction for a prolonged length of time.
Unfortunately, EE recurs upon withdrawal of oral therapy (3-7).

EE is believed to be an imbalance of the TH1/TH2 immune system and is more prevalent in
patients with asthma and allergies (8-9). Some EE centers have started treating patients
with targeted elimination diets and elemental diets and are having good success.
Unfortunately, most families and in particular teenagers, are noncompliant with the dietary
therapies. Formula for the elemental diet is not covered by most insurance and is very
costly as well. Therefore, finding a drug that would keep a patient with EE in maintenance
therapy and reduce both the long term consequences of steroids and potential for esophageal
strictures would be ideal.

Many adults with EE present with dysphagia or food impactions and subsequently are found to
have an esophageal stricture requiring multiple esophageal dilatations for treatment.
Esophageal subepithelial fibrosis has been found in the adult literature in patients with
EE. Chehade et al evaluated distal esophageal biopsy specimens in children with EE,
eosinophilic gastroenteritis, gastroesophageal reflux disease, and controls for the presence
of increased collagen deposition indicative of fibrosis. They found that fibrosis was
present in 57% of patients with EE, of which 42% had symptoms of dysphagia and 80% had food
impactions. The degree of fibrosis was not associated with eosinophil count but was
associated with eosinophilic degranulation (10). Upon degranulation, eosinophils release
major basic protein (MBP) and transforming growth factor (TGF-β). MBP changes smooth muscle
contractility by blocking M2 muscarinic receptors therefore changing the motility of the
esophagus. TGF-β is believed to cause an increase in the production of the extracellular
matrix and smooth muscle hypertrophy and hyperplasia leading to esophageal thickening and
fibrosis. TGF-β is a key growth factor that increases fibroblasts and subsequent fibrosis in
the lung as well. Corticosteroids have been shown to reverse esophageal remodeling, but no
large clinical trials have evaluated the treatment of esophageal fibrosis related to EE. To
quote Aceves and Ackerman: "[f]ibrosis likely contributes to multiple clinical aspects of
EE, including dysphagia symptoms, disease chronicity, and stricture formation." Fibrosis is
measured in esophageal tissue by evaluating the amount of collagen that is deposited.
Trichrome stains specifically for collagen (11).

Patients with EE have an abundance of inappropriate cells within the esophagus, including
eosinophils and mast cells. A normal esophagus is devoid of eosinophils. Mast cells likely
have an active role in the development of eosinophilic esophagitis; unfortunately the
precise mechanism is unknown. Mast cells are found in the majority of vascularized tissues
and are abundant in the gastrointestinal tract. They are located in the submucosa and
lamina propria. Within the esophagus specifically, they are found in the mucosa and
submucosa with rare expansion into muscularis layer. With severe eosinophilic inflammation,
mast cells can be seen in the muscularis layer. Biopsies from patients with EE have
demonstrated marked increase in mast cells in the mucosal layer of the esophagus. There has
been correlation of mast cell number with degree of eosinophilia. Mast cells within the GI
tract contain tryptase as well as other preformed mediators and can synthesize mediators
that promote inflammation. Tryptase may activate eosinophils and therefore induce
eosinophilic degranulation and cytokine secretion. Mast cells release chemokines and
leukotrienes specific to eosinophil recruitment. Ultimately mast cell mediators worsen
edema, create more inflammation, induce smooth muscle contractility and potentially affect
tissue fibrosis. Therefore, active mast cells are proinflammatory, modify esophageal
function, and encourage the development of strictures (12). Mast cells may be a key
component in the propagation of the inflammatory state as well as of significant importance
in the development of fibrosis. Evaluating tryptase levels could lead to a better
understanding of the inflammatory process and ultimate problem of fibrosis.

Eosinophils and mast cells contain leukotrienes in abundance. Cysteinyl leukotrienes are
inflammatory mediators derived from arachidonic acid. They are important in eosinophil
attraction and migration, smooth muscle bronchoconstriction, vascular permeability, and
mucus hypersecretion (13). Montelukast is a leukotriene receptor antagonist and blocks
specifically the leukotriene D4 (LTD4) receptor. Attwood et al treated 8 patients with EE
with Montelukast for which 6 reported complete subjective improvement and 5 remained
asymptomatic on a maintenance regimen. Treatment with Montelukast did not change the
density of eosinophils present in the esophagus however (14). Montelukast has been used in
mouse models to decrease fibrosis occurring in both the liver and the lung. El-Swefy and
Hassanen performed a randomized study in which they subjected mice to one of 4 groups, no
bile duct ligation plus saline, no bile duct ligation plus montelukast, bile duct ligation
plus montelukast and bile duct ligation plus saline. Bile duct ligation induced
inflammation leading to cirrhosis and fibrosis. They found that bile duct ligation
increased the amount of TGF-β, which correlated with the amount of fibrosis found. Of more
import is that montelukast significantly reduced the amount of fibrosis as well as decreased
the amount of TGF-β (15). Izumo, Kondo, and Nagai evaluated the use of montelukast in the
prevention of pulmonary fibrosis after exposure to bleomycin. They found that mice given
montelukast had significantly lower levels of TGF-β and fibrosis of the lung (16).

The investigator's research focus throughout fellowship included several retrospective
studies in eosinophilic esophagitis. One of the investiagotor's studies evaluated the
management of both initial and maintenance therapy for EE within our department. The
investigator wanted to know what medicines the majority of pediatric gastroenterologists at
the investigator's institution were using to treat EE and how often patients were being
reevaluated by endoscopy. The investigator also wanted to evaluate the use of montelukast
in the treatment of EE. The investigator reviewed 1500 charts of patients who had undergone
upper endoscopy and had a result of "esophagitis." Patients were then selected by the
presence of greater than or equal to 20 eosinophils per high power field in at least one
biopsy location. Of the 1500 charts the investigator evaluated, only 88 patients fulfilled
criteria for EE. The investigator discovered that there is not a consistent standard of
care within their department for the initial and maintenance treatment of EE. Many
different medicines are trialed initially and patients undergo esophageal biopsies at
different time intervals, contrary to what is recommended. The primary group of 88 patients
was then separated into 49 patients who reached maintenance therapy (defined by less than 5
eosinophils per high power field) and 39 patients who never experienced improvement in their
eosinophil counts. Of the 49 who reached maintenance therapy, 30 patients were treated with
montelukast and had follow up endoscopy. The investigator evaluated peak eosinophil counts
pre and post treatment, and contrary to their hypothesis, those patients treated with 5mg of
montelukast had lower post treatment eosinophil counts than those treated with higher doses
(10 and 20mg). This raised the possibility that the packaging differences of the two main
doses had an impact in treatment. The investigator considered the possibility that the
lower dosage of montelukast is protective since 5mg tablets are chewable and would yield a
topical effect on the esophagus whereas the 10mg tablets are swallowed, not chewed. In
addition, the investigator evaluated how often EE recurred and discovered there was a high
rate of histologic recurrence. Also interesting to the investigator was the lower
recurrence rate in patients given 5mg of montelukast compared to 10mg and 20mg doses. These
results were not expected and deserve further evaluation. If montelukast can prevent
fibrosis through a topical effect, then its therapeutic potential is huge in the treatment
of EE.

The investigator has developed a study that will investigate the use of montelukast in total
vs placebo as well as allow them to look for dosing effects within montelukast in the
treatment of eosinophilic esophagitis. The primary objective is to evaluate if montelukast
will keep eosinophil counts low enough for the patient to remain in remission and if so, is
there a particular dose of montelukast that will achieve this in a greater proportion. All
patients will be maintained on a proton pump inhibitor (PPI) at baseline so that patients
receiving placebo will still receive standard of care treatment. Many GI practitioners at
this institution will place patients with EE on montelukast as part of standard of care
treatment at varying doses without any good literature to support this use. The
investigator's goal is to determine if montelukast is a valuable therapy in the treatment of
EE. The investigator's study will also evaluate histological stains that may show an
improvement with the use of montelukast, even if the eosinophil counts do not change, as the
chemicals released by eosinophils and mast cells may be of primary import in the
continuation of EE.

Inclusion Criteria:

- Males and females aged 2-17

- Presence of more than 15 eosinophils per hpf on original endoscopy and less than 5
eosinophils/hpf on the most recent endoscopy

- Concurrent PPI for 1 month at 1-2mg/kg/dose prior to endoscopy or have a negative pH
study

- English speaking

- Ability to undergo a follow up endoscopy between 12 and 13 weeks after the start of
the study

- Procurement of written informed consent signed by the subject's legal guardian and
study investigator (s) and subject assent.

Exclusion Criteria:

- Subjects with eosinophils in stomach and duodenum on original endoscopy.

- Subjects requiring oral prednisone within 1 month of current endoscopy.

- Subjects with diagnosis of other co-morbid diseases such as heart disease, renal
disease, autoimmune disease, an immunodeficiency, diabetes, phenylketonuria, or
thyroid disease.

- Subjects using Montelukast within one month of current endoscopy

- Subjects with concurrent use of phenobarbital or rifampin