349 / 412
© 2013 by the American College of Gastroenterology
The
American Journal
of
GASTROENTEROLOGY
Fluticasone vs. Esomeprazole for Esophageal Eosinophilia
significant difference in eosinophilia between these two locations.
It is noteworthy that studies using oral viscous budesonide, which
can be administered as a slurry mixed with a sugar substitute or
as a swallowed nebulizer, demonstrated higher response rates
(72–80%) (17,18). This suggests that the delivery system, and not
the steroid type, may be the reason for low response rates. To fur-
ther study and explore this point, a recent study demonstrated that
oral viscous budesonide was more effective in reducing eosinophil
counts when compared with nebulized budesonide (19).
Interestingly, there was a greater number of acid and imped-
ance reflux episodes in the FP arm compared to the ESO
arm. However, the percentage of patients with abnormal acid
(>50) and impedance episodes (>73) was similar between
both treatment arms. Therefore, it is unclear whether an increased
number of reflux episodes had any role in the response to
topical steroids.
The rationale for using PPIs for the treatment of esophageal eosi-
nophilia is that it may help treat underlying GERD, whichmay con-
tribute to EoE (20). It may be difficult to completely separate these
two entities and it appears that many adult EoE patients respond
at least clinically to PPI therapy. In one study, 8 of 17 (47%) adult
patients presenting with food impaction and dense esophageal
eosinophilia (>20 eos/hpf) responded clinically to PPI therapy (5).
The mechanism for this is not clear, although PPIs may help heal a
disrupted epithelial barrier and therefore reduce further immune
activation (2). In our study, the four patients with esophageal eosi-
nophilia and GERD had a complete response to PPI therapy.
The histological response achieved in three patients with PPI
therapy despite having a negative 24-h pH study suggests these
patients may have PPI-responsive EoE, which stimulates an inter-
esting discussion regarding the pathophysiology of this entity. This
described response to esomeprazole in GERD-negative esophageal
eosinophilia patients can be attributed to anti-inflammatory prop-
erties of PPI independent of acid inhibition. Interestingly, PPIs
have been shown to exhibit anti-inflammatory properties by act-
ing directly on principal cytokines (IL-4 and IL-13) involved in the
recruitment of eosinophils in the esophagus (21). Additionally, PPI
therapy was shown to block the release of eotaxin-3, which has an
integral role in the pathogenesis of EoE (22). Other studies have
also reported a histological response to PPI therapy in EoE patients
even following a negative 24-h pH study. Peterson
et al.
(13) dem-
onstrated a 33% (4/12) histological response to PPI therapy. Simi-
lar results have been reported in children as well. In 43 pediatric
EoE patients, a subset of patients responded histologically to PPI
therapy following a normal pH study (23). These studies coupled
with our data suggest that a subset of patients whose clinical pres-
entation suggests EoE and have dense eosinophilia on esophageal
biopsies indeed appear to be PPI responsive.
Significant clinical improvement was seen in the majority of
patients with esophageal eosinophilia taking PPI even though only
a few had histological resolution. Other studies have also reported
a poor correlation between clinical remission and histological
response. In a randomized placebo controlled study, vomiting was
the only symptom that significantly improved in EoE children
treated with topical steroids and did not correlate with histological
response (11). Similar findings have been described by Molina
et al.
(24) in which clinical remission was achieved in the majority
of patients with food impaction or dysphagia despite persistence
of eosinophilia. Interestingly, a dissociation was recently reported
between clinical response and histological severity in EoE chil-
dren with the majority of patients reporting continued symptoms
despite histological remission (25). These data combined with ours
suggest one cannot use clinical response as a surrogate for histo-
logical change. However, whether achieving histological response
in EoE is an important outcome is yet to be determined. In the-
ory, normalization of the eosinophila may prevent further tissue
remodeling, fibrosis, and possibly stricture formation.
Esophageal dilation is one of the most effective treatments for
alleviating the symptoms of EoE even though it does not address
the underlying disease pathophysiology (26). In our cohort, all of
the 15 patients (8 in FP arm and 7 in ESO arm) with dysphagia and
esophageal eosinophilia who underwent dilation at index endo-
scopy had significant improvement in symptoms. Similar clinical
improvements have been noted in other studies in EoE (26,27).
In contrast, among the 27 patients who did not undergo dilation,
only those randomized to ESO had significant improvement in
symptoms. As a similar number of patients underwent dilation
in each treatment arm and dilation was not performed during the
8-week treatment course or at follow-up endoscopy, it is unlikely
that dilation had an effect on the change inMDQ scores before and
after therapy.
There are some limitations of our study. A placebo arm would
have helped better define the natural history of esophageal eosi-
nophilia. As we were comparing one medication administered as a
pill to another given as an inhaler, we found adding a placebo arm
to be challenging. Additionally the delivery system of aerosolized
fluticasone may have hampered the true assessment of steroid
response rates, and using another agent such as oral viscous budes-
onide may give us a different understanding of the true steroid
response. Another limitation of our study was the relatively small
sample size, which limited the subgroup analyses. Additionally, we
used the 2-week dysphagia score, which may not have accurately
reflected the symptoms in some EoE patients, particularly in those
with intermittent dysphagia. However, the 2-week MDQ was the
only validated questionnaire available at the time of this study. To
date, no dysphagia questionnaire developed exclusively for EoE
has been published.
In conclusion, histological response between ESO and FP were
similar in the treatment of esophageal eosinophilia, with neither
drug having overwhelming treatment success. On the other hand,
significant improvement in clinical symptoms was demonstrated
with PPI therapy. Further larger studies are needed to better define
the optimal treatment for patients with esophageal eosinophilia
and to better describe the subgroup and natural history of such
patients who respond to PPI therapy.
CONFLICT OF INTEREST
Guarantor of the article:
Fouad J. Moawad, MD.
Specific author contributions:
Study concept and design, subject
enrollment, acquisition of data, analysis and interpretation of data,
124