347 / 412
© 2013 by the American College of Gastroenterology
The
American Journal
of
GASTROENTEROLOGY
Fluticasone vs. Esomeprazole for Esophageal Eosinophilia
With regard to resolution of endoscopic findings, no signifi-
cant difference was seen between FP and ESO (
Table 2
).
Dilation was performed on 15 patients on index endoscopy pri-
marily for a dominant stricture seen or a coexisting Schatzki ring
(8 patients taking FP and 7 patients taking ESO). No dilations were
performed during the treatment period or on follow-up endos-
copy post therapy. There was significant improvement in clinical
symptoms based on a mean decrease in the MDQ score seen in
both treatment groups who underwent dilation (FP= −10.6±10.5,
P
=0.027 and ESO= −14.3±14.0,
P
=0.027). Among patients who
did not undergo dilation (
n
=27) on index endoscopy, there was
a significant decrease in MDQ score in the ESO group (−20±24,
P
=0.005) but not in the FP group (−1.9±21.5,
P
=0.721).
To further explore the relationshipbetween reflux andeosinophilia,
we examined the association between impedance reflux episodes
and eosinophil counts. There was no significant association between
impedance reflux episodes and eosinophil counts in the proximal
(
r
s
=0.263,
P
=0.116) or distal esophagus (
r
s
=0.162,
P
=0.359).
In both treatment arms, adherence to therapy was very good.
There was no significant difference in adherence to treatment in
both groups (FP=86±24% vs. ESO=92±10%,
P
=0.977).
Two patients randomized to the FP arm discontinued treatment
during the study period. One patient had worsening of migraine
headaches, which he attributed to FP. Another patient had bother-
some GERD-related symptoms and discontinued the steroid, and
began treatment with a PPIs. Both patients were analyzed as inten-
tion to treat. One patient in the fluticasone arm developed esopha-
geal candidiasis. He was asymptomatic during the study period
and this was discovered on follow-up endoscopy and confirmed
on esophageal biopsies. He was treated with a course of oral fluco-
nazole. No adverse events occurred in the PPI arm.
DISCUSSION
This randomized controlled single-blinded study demonstrated a
similar histological response between esomeprazole and flutica-
sone treatment groups in patients with esophageal eosinophilia.
With regard to clinical improvement, based on a validated symp-
tom questionnaire, ESO was significantly better than FP regard-
less of a concomitant GERD diagnosis.
We defined our patient population who had clinical symptoms
of esophageal dysfunction and elevated eosinophil counts on biop-
sies as having esophageal eosinophilia rather than EoE, based on
the most recent updated consensus statement (2). Although swal-
lowed FP is commonly accepted as the first-line treatment for EoE,
in this study we examined its efficacy in patients who had pheno-
typic appearance of EoE and elevated eosinophil counts yet were
PPI naive. Our data demonstrated a 19% histological response in
patients treated with FP.
Table 1.
Baseline patient characteristics
Variables
Fluticasone
(
n
=21)
Esomeprazole
(
n
=21)
P
value
Mean age±s.d., years
37.0±11.1
38.0±8.8
0.771
Male,
n
(%)
19 (90.5)
19 (90.5)
1.000
Race,
n
(%)
0.766
Caucasian
17 (81.0)
17 (81.0)
—
Hispanic
3 (14.3)
2 (9.5)
—
African American
1 (4.7)
2 (9.5)
—
Coexisting allergies,
n
(%)
Any atopic disease
11 (52.4)
15 (71.4)
0.341
Seasonal allergies
7 (33.3)
7 (33.3)
1.000
Asthma
2 (9.5)
2 (9.5)
1.000
Eczema
2 (9.5)
0 (0.0)
0.488
Food allergies
3 (14.3)
9 (42.9)
0.085
Concomitant use of allergy medications,
n
(%)
Antihistamines
4 (19.0)
4 (19.0)
1.000
Nasal steroid spray
3 (14.3)
1 (4.7)
0.606
Leukotriene antagonist
1 (4.7)
1 (4.7)
1.000
GERD by pH score
4 (19.0)
4 (19.0)
1.000
Erosive esophagitis
a
3 (14.3)
4 (19.0)
1.000
pH reflux episodes
No. of episodes
45.8±40.6
25.5±19.3
0.045
% (
x
/
n
) abnormal patients
28.5% (6/21) 9.5% (2/21)
0.238
Impedance reflux episodes
No. of episodes
63.6±23.1
44.8±21.4
0.012
% (
x
/
n
) abnormal patients
38% (8/21)
14.2% (3/21) 0.159
Mayo dysphagia score
17.1±17.8
19.5±20.7
0.691
Pretreatment eosinophil count
Proximal biopsy
39.1±33.2
32.9±19.4
0.473
Distal biopsy
38.4±22.3
34.2±25.2
0.593
GERD, gastroesophageal reflux disease.
a
All cases of erosive esophagitis were Los Angeles Grade A.
0
20
40
60
80
100
Before
After
0
20
40
60
80
100
Before
After
Esomeprazole (
n
=21)
P
=0.102
Fluticasone (
n
=21)
P
=0.174
Figure 1.
There was no significant change in eosinophil count post treat-
ment in either arm regardless of coexisting gastroesophageal reflux disease
(GERD). Dashed lines indicate GERD-positive patients.
122