348 / 412
The
American Journal
of
GASTROENTEROLOGY
therapy. Interestingly, no significant clinical improvement was
noted when compared with placebo.
The relatively poor response seen in many randomized stud-
ies may at least partially be due to a lack of uniformity in medi-
cation dosage, duration of therapy, the definition of response,
and delivery method. Additionally, variability in the definition
of EoE exists in the literature (15) and in many studies EoE
is often synonymous with esophageal eosinophilia. Alexander
et al.
(14) used a higher dose of swallowed fluticasone (880mcg
twice daily), which may reflect the higher response rate seen in
their study compared with other randomized controlled studies
using topical fluticasone for EoE. Five patients developed esopha-
geal candidiasis, compared with one patient in our study, suggest-
ing this may be dose dependent. Another reason our response rate
may have been different when compared with other randomized
studies may be due to the definition of response. Variability exists
in the literature with some studies using a change in histological
grade (12) and even a decrease in percent eosinophils in compari-
son with baseline (14,16). The majority of the studies have used
absolute changes in eosinophil count as an end point to include
1 eos/hpf (11), 5 eos/hpf (13,17), and <7 eos/hpf (18). We chose to
use <7 eos/hpf as a target end point. However, we also examined
our data using <5 eos/hpf and obtained identical results (19%
fluticasone and 33% for esomeprazole).
Another explanation for the low response to FP may be related
to the mechanism of drug delivery. FP is an aerosolized medication
administered through a metered dose inhaler, which is intended
for the airways, and it is unclear how much drug is actually deliv-
ered to the esophagus. One would assume that response may be
superior in the proximal esophagus than the distal esophagus with
this type of delivery. However, when we compared the histological
response between the proximal and distal esophagus, there was no
Response to topical steroids for EoE has varied in the literature
with small case series and retrospective studies reporting response
rates in up to 80% (1,10). In contrast, data from prospective con-
trolled studies have demonstrated lower response rates. Konikoff
et al.
(11) reported a 50% efficacy (10/20) in a pediatric population
with histological response defined as
≤
1 eos/hpf. Again, complete
histological response was seen in 50% (18/36) of patients after
4 weeks of therapy with FP (12). In another study comparing the
efficacy of topical steroids to PPI therapy in adults with EoE, only
15% achieved complete histological response with FP, defined as
5 eos/hpf (13). A recent study by Alexander
et al.
(14) reported a
62% histological response to fluticasone, defined as a more than
90% decrease in mean levels of eosinophils following 6 weeks of
0
20
40
60
80
0
20
40
60
80
Before
After
Before
After
Esomeprazole (
n
=21)
P
=0.162
Fluticasone (
n
=21)
P
=0.001
Figure 2.
Change in Mayo dysphagia questionnaire (MDQ) score following
treatment. There was significant clinical improvement in ESO but not in the
FP treatment arm. Dashed lines indicate gastroesophageal reflux disease
(GERD)-positive patients.
Table 2.
Data presented for percent (number) of patients with histological and endoscopic findings in pretreatment and posttreatment
Fluticasone
Esomeprazole
Pre
Post
% Improve
a
Pre
Post
% Improve
a
Histological findings
Basal cell hyperplasia
100% (21)
81% (17)
19% (4/21)
100% (21)
52% (11)
43% (9/21)
Intercellular edema
100% (21)
76% (16)
24% (5/21)
86% (18)
57%(12)
35% (7/18)
Eosinophilic microabscess
86% (18)
52% (11)
44% (8/18)
71% (15)
38% (8)
67% (10/15)
Eosinophilic degranulation
95% (20)
52% (11)
50% (10/20)
76% (16)
43% (9)
50% (8/16)
Eosinophilic distribution
b
95% (20)
76% (16)
25% (5/20)
90% (19)
48%(10)
47% (9/19)
Endoscopic findings
Stenosis on index endoscopy
24% (5)
14% (3)
80% (4/5)
24% (5)
10% (2)
80% (4/5)
Concentric rings
76% (16)
76% (16)
13% (2/16)
76% (16)
52% (11)
44% (7/16)
Longitudinal furrows
81% (17)
76% (16)
18% (3/17)
81% (17)
52% (11)
41% (7/17)
White plaques
19% (4)
29% (6)
50% (2/4)
24% (5)
0% (0)
100% (5/5)
% improve, % improvement; pre, pretreatment; post, posttreatment.
Percent improvement is among a subgroup of patients who had a pretreatment finding. There was no statistically significant difference in improvement between
treatment arms.
a
Percent improvement is among patient who had a pretreatment finding.
b
Involving entire thickness of epithelium.
123