Clinical Gastroenterology and Hepatology 2016;14:809–815

ALIMENTARY TRACT
Proton Pump Inhibitors Increase Incidence of Nonsteroidal
Anti-Inflammatory Drug–Induced Small Bowel Injury:
A Randomized, Placebo-Controlled Trial
Ema Washio,* Motohiro Esaki,* Yuji Maehata,* Masashi Miyazaki,‡ Hiroyuki Kobayashi,‡
Hideki Ishikawa,§ Takanari Kitazono,* and Takayuki Matsumotok

*Department of Medicine and Clinical Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan;
‡
Division of Gastroenterology, International University of Health and Welfare Fukuoka Sanno Hospital, Fukuoka, Japan;
§
Department of Molecular-Targeting Cancer Prevention, Graduate School of Medical Science, Kyoto Prefectural University of
Medicine, Osaka, Japan; and kDivision of Gastroenterology, Department of Internal Medicine, School of Medicine,
Iwate Medical University, Morioka, Japan

BACKGROUND & AIMS: Some studies have reported a high incidence of small bowel injuries in 60%–80% of subjects
who take nonselective nonsteroidal anti-inflammatory drugs and PPIs simultaneously. We
performed a randomized, double-blind, controlled study to determine whether proton pump
inhibitors (PPIs) exacerbate nonsteroidal anti-inflammatory drug–induced small bowel injury.

METHODS: Fifty-seven healthy subjects were randomly assigned groups given the cyclooxygenase (COX)
2 inhibitor celecoxib (200 mg, twice daily) plus placebo for 2 weeks (COX-2 D placebo group,
n [ 30), or celecoxib plus the PPI rabeprazole (20 mg, once daily) for 2 weeks (COX-2 D PPI
group, n [ 27). The study was performed from October 2012 through September 2013 at a
tertiary medical center in Japan. All subjects were evaluated by capsule endoscopy at the start
of the study and then after 2 weeks administration of celecoxib with rabeprazole or placebo.
The incidence rates and the numbers of small bowel injuries (ulcers and erosions) that were
observed under capsule endoscopy were compared between groups. The primary endpoint was
the incidence of mucosal injuries at the second capsule endoscopy examination.

RESULTS: A significantly higher proportion of subjects in the COX-2 D PPI group developed small bowel
injury (12 of 27 subjects; 44.4%) than in the COX-2 D placebo group (5 of 30 subjects; 16.7%;
P [ .04). Subjects in the COX-2 D PPI group had a significant increase in risk of small bowel
injury compared with the COX-2 D placebo group (relative risk, 2.67; 95% confidence interval,
1.08–6.58). The number of erosions in each member of the COX-2 D PPI group was greater
than in each member of the COX-2 D placebo group (P [ .02). The number of ulcers did not
differ between groups. Twenty-six percent of subjects in the COX-2 D PPI group developed
mucosal injury in the jejunum, compared with none of the subjects in the COX-2 D placebo
group (P [ .003); no such trend was found in the ileum.

CONCLUSIONS: In a randomized, controlled trial, PPIs increased the risk of short-term nonsteroidal
anti-inflammatory drug–induced small bowel injury. UMIN clinical trial registry number:
UMIN000008883.

Keywords: Intestine; Damage; Cyclooxygenase-2 Inhibitor; COX-2.

(CE) and balloon-assisted enteroscopy have revealed that

See editorial on page 816.
onsteroidal anti-inflammatory drugs (NSAIDs) are
N one of the most frequently prescribed types of
medication worldwide. Despite their favorable analgesic
effect, the use of NSAIDs is associated with high incidence
of gastrointestinal (GI) toxicity. Gastroduodenal ulcers
have been found to occur in 20%–30% of chronic NSAIDs
users.1,2 Moreover, the applications of capsule endoscopy
NSAIDs induce small bowel injuries more frequently than
had previously been believed.3–6
Abbreviations used in this paper: CE, capsule endoscopy; COX, cyclo-
oxygenase; GI, gastrointestinal; NSAID, nonsteroidal anti-inflammatory
drug; PPI, proton pump inhibitor.
Most current article
© 2016 by the AGA Institute
1542-3565/$36.00
http://dx.doi.org/10.1016/j.cgh.2015.10.022
810 Washio et al
Antiacid secretory drugs, especially proton pump in-
hibitors (PPIs), have become the mainstay prophylactic
treatment against NSAID-induced GI toxicities. The re-
sults of animal studies have suggested PPIs exert pro-
phylactic effects against small bowel injuries.7–9
However, recent clinical trials using CE have shown
considerably high incidence of small bowel injuries
(60%–80%) in subjects who take nonselective NSAIDs
and PPIs simultaneously.4,10,11
We recently carried out a randomized clinical trial to
evaluate small bowel toxicity in association with cele-
coxib, a clinically available cyclooxygenase (COX)-2
selective inhibitor, in healthy subjects and identified
small bowel injuries in 43% of the study population.12
Although the incidence of the mucosal injuries was
conspicuously higher in comparison with the previous
studies of Goldstein et al,13,14 we were not able to arrive
at provisional explanation for our discordant results.
However, Wallace et al15 subsequently demonstrated
PPI use exacerbated NSAID-induced small bowel injury
in a rodent model. A review of the study protocols
revealed a difference: PPI was simultaneously adminis-
tered in our previous study, but not in the studies of
Goldstein et al. These observations strongly suggested
that PPIs might be closely associated with the develop-
ment of NSAIDs-induced small bowel mucosal injuries in
humans.
To examine the influence of concomitant use of PPIs
on NSAID-induced small bowel injury in humans,
we performed a prospective, double-blind, randomized
controlled study.
Methods
Study Design
This study was a prospective, double-blind, random-
ized controlled trial. Before randomization, all subjects
underwent laboratory tests (complete blood cell count,
serum chemistry, and Helicobacter pylori antibodies), an
electrocardiogram, and a baseline CE. Subjects with
abnormal laboratory test results or those who had
abnormal electrocardiogram findings were excluded from
the study. Subjects in whom small bowel bleeding, ulcers,
or erosions were detected in baseline CE were also
excluded. Under a computer-generated randomization
system using a minimization method, the study subjects
were allocated into 1 of 2 groups: the COX-2 SI group
received COX-2 selective inhibitor and placebo, whereas
the COX-2 SI þ PPI received COX-2 selective inhibitor
plus rabeprazole. The following patient characteristics
were chosen as stratification factors: age (<30 years or
30 years), sex, and the hemoglobin level (<14.5 g/dL or
14.5 g/dL). Subjects in the COX-2 SI group received
celecoxib (200 mg, twice daily) plus placebo (lactose 20
mg, once daily); the subjects in the COX-2 SI þ PPI group
received celecoxib (200 mg, twice daily) plus rabeprazole
Clinical Gastroenterology and Hepatology Vol. 14, No. 6
(20 mg, once daily) for 2 weeks. The dosages of celecoxib
and rabeprazole, which were also applied to other Japa-
nese clinical trials,16,17 were determined based on the
dosages approved by the Japanese Ministry of Health and
Welfare. Rabeprazole and placebo were prepared in
dummy capsules. The study subjects were instructed to
take a capsule once per day for 2 weeks. The use of other
NSAIDs, aspirin, and antiulcer drugs was strictly pro-
hibited during the study period. After 2 weeks of medi-
cation, the subjects completed a questionnaire about GI
symptoms, underwent repeated laboratory tests, and
underwent second CE.
The study protocol was approved by the institutional
review board of the International University of Health
and Welfare, Fukuoka Sanno Hospital (FS30), and the
study was conducted in accordance with the Helsinki
Declaration. This trial was registered as the PPI-NSAID
Kyushu University study (PINK study) in the University
Hospital Medical Information Network Clinical Trials
Registry under registration number UMIN000008883
(https://upload.umin.ac.jp/cgi-open-bin/ctr/ctr.cgi?function
¼brows&action¼brows&recptno¼R000010425&type¼
summary&language¼E). Written informed consent was
provided by each subject before entry into the study. All
authors had access to the study data and reviewed and
approved the final manuscript.
Subjects
Healthy volunteers with normal physical examination
and laboratory test results were eligible. The exclusion
criteria were as follows: (1) a history of peptic ulcers, (2)
a history of recent (within 1 month) NSAID or aspirin
use, (3) a history of aspirin-induced asthma, (4) an
allergy to sulfonamide, (5) a history of recent treatment
with antiulcer drugs, (6) stenosis of the GI tract, (7) a
history of ileus, (8) pregnant or nursing women, and (9)
the presence of other disorders regarded to be inap-
propriate for participation in the present study.
Capsule Endoscopy
The baseline and second CEs were performed using
an Endo Capsule (Olympus Medical Systems, Tokyo,
Japan). After a 12-hour overnight fast, each subject was
prepared with sensor arrays and a data recorder, and
instructed to swallow the capsule with a small amount of
water. When the real-time viewer VE-1 (Olympus Co,
Tokyo, Japan) identified the capsule reaching the cecum,
we removed the sensor array and data recorder from the
subject, and all of the digital video image streams were
downloaded to the Olympus WS-1 (Olympus Co).
Two observers (E.W. and Y.M.) who remained blinded
to the subjects’ group independently assessed the CE
images. We defined mucosal injuries as positive CE find-
ings; these injuries were classified as either ulcers or
erosions as has been described in our previous study.12
June 2016
Figure 1. Positive CE findings. (A) Ulcer (arrow). (B) Erosion.
A circumscribed mucosal defect with obvious whitish
mucous that was estimated to be 3 mm or larger in diam-
eter was defined as an ulcer (Figure 1A). Although it is
sometimes difficult to distinguish a small ulcer from an
area of erosion, a small mucosal break surrounded by
redness was regarded as an erosion (Figure 1B).18 The
small intestine was divided equally into the jejunum and
the ileum by the small bowel transit time. If the CE findings
were discordant between the 2 observers, they discussed
the case until a consensus opinion could be obtained.
Endpoints
The primary endpoint was the incidence of mucosal
injuries at second CE. The secondary endpoints were the
PPIs and NSAID-Induced Small Bowel Injury 811
incidence of CE findings in the jejunum and in the ileum,
the number of ulcers or erosions in subjects with posi-
tive CE findings, and the prevalence of GI symptoms and
anemia. GI symptoms were assessed at the end of the
medication period with use of the Gastrointestinal
Symptom Rating Scale.19 Anemia was defined as a
decrease in the hemoglobin level of 2.0 g/dL or more
from the baseline value.
Statistical Analysis
The incidence of small bowel injury caused by
administration of celecoxib for 2 weeks has been shown
to range from 6% to 16%.13,14 Although the incidence of
small bowel injury caused by celecoxib plus rabeprazole
is unknown, our previous trial revealed the incidence of
small bowel injury caused by 2-week administration of
celecoxib and omeprazole was 43%.12 Thus, the sample
size of the present study was calculated on the
assumption that the incidence of small bowel injury
would be 10% in the COX-2 SI group and 43% in the
COX-2 SI þ PPI group. Thirty subjects per group were
required to detect this difference with a .05 significance
level and a statistical power of 80%.
The number of mucosal injuries and Gastrointestinal
Symptom Rating Scale score were expressed as the me-
dian (range); other values were expressed as the mean
(standard deviation). The numerical data were compared
between the groups using the Mann-Whitney U test, and
the categorical data were compared by Fisher exact
probability test or the chi-square test. P values of < .05
were considered to be statistically significant in each of
the tests.
Results
Subjects
The study was conducted from October 2012 to
September 2013. During the study period, 61 subjects
were enrolled. A flow chart of the study subjects is shown
in Figure 2. One subject was excluded because he was
suspected of having a small bowel tumor at the baseline
CE. The remaining 60 subjects were then randomly
assigned to the 2 groups. One subject was excluded from
the COX-2 SI þ PPI group because he took another NSAID
without permission during the study period. The second
CE enabled total enteroscopy in 59 subjects. However, we
were not able to examine CE videos of 2 subjects belonging
to the COX-2 SI þ PPI group because of technical issues
with the Olympus WS-1. Consequently, the COX-2 SI group
and COX-2 SI þ PPI groups were composed of 30 subjects
and 27 subjects, respectively.
Table 1 shows the demographic data of the 2 groups.
The prevalence of H pylori infection tended to be higher in
the COX-2 SI þ PPI group than in the COX-2 SI group
(5 subjects [18.5%] vs 1 subject [3%], respectively;
812 Washio et al Clinical Gastroenterology and Hepatology Vol. 14, No. 6

Figure 2. Flow chart of the study subjects. ECG,

electrocardiogram.
P ¼ .09). Two subjects in COX-2 SI group continued to take
concurrent medications, 1 for essential hypertension
(losartan potassium, 50 mg/day), the other for chronic
pancreatitis (camostat mesilate, 300 mg/day). One
subject in the COX-2 SI þ PPI group took medication
for orthostatic hypotension (amezinium metilsulfate,
20 mg/day).
The Capsule Endoscopy Findings
The 2 observers reported concordant results for each
of the subjects with regard to the incidence of positive
findings on the second CE. Figure 3 shows the incidence
of mucosal injuries in the 2 groups. Eleven subjects in the
COX-2 SI þ PPI group had erosions. Among them, ulcers
were also found in 5 subjects. Another subject in the
group had ulcers only. In the COX-2 SI group, 4 subjects
had erosions, 2 of whom also had ulcers. Another subject
in the group had ulcers only. Consequently, the incidence
of small bowel injuries was significantly higher in the
COX-2 SI þ PPI group than in the COX-2 SI group (44.4%
vs 16.7%; P ¼ .04). These results indicated the subjects
in the COX-2 SI þ PPI group were at an increased risk of
small bowel injury in comparison with the COX-2 SI
Figure 3. The incidence of small bowel injury in the COX-2 SI
and COX-2 SI þ PPI groups.
group (relative risk, 2.67; 95% confidence interval,
1.08–6.58).
We compared the number of mucosal injuries between
the 2 groups. The crude data on the mucosal injuries
among the subjects with positive CE findings are shown in
Supplementary Table 1. The incidence of mucosal injuries
was significantly higher in the COX-2 SI þ PPI group than
in the COX-2 SI group (0 [0–31] vs 0 [0–7]; P ¼ .02). In the
comparison of the numbers of each type of mucosal injury,
we found the number of erosions was significantly higher
in the COX-2 SI þ PPI group than that in the COX-2 SI group
(0 [0–29] vs 0 [0–3]; P ¼ .02). The number of ulcers did not
differ to a statistically significant extent (0 [0–3] vs
0 [0–7]; P ¼ .19).
We then compared the number of mucosal injuries in
each segment of the small bowel between the 2 groups
(Figure 4). Jejunal mucosal injuries, all of which were
erosions, were found only in the COX-2 SI þ PPI group
(P ¼ .003). There was no difference between the 2
groups in the incidence of mucosal injuries in the ileum
(37% vs 17%; P ¼ .11); nor was there any difference in
the incidence of ulcers (22% and 33%) or erosions (10%
and 13%).

Table 1. Demographic Data of COX-2 SI and COX-2 SI þ PPI

Groups
Symptoms, Laboratory Data,
and Complications
COX-2 SI COX-2 SI þ P
group PPI group value During the 2-week medication period, 2 subjects in
Number of subjects 30 27
each group experienced diarrhea. Among these 4 sub-
Age, y 32 (8.5) 34 (8.3) .67 jects, 1 subject in the COX-2 SI group also complained of
Gender (female/male) 13/17 10/17 .62 epigastric pain, and another subject in the COX-2 SI þ
Body weight, kg 59 (9.9) 59 (9.8) .72 PPI group reported abdominal pain. Consequently,
Helicobacter pylori infection 1 5 .09 Gastrointestinal Symptom Rating Scale score of the 2
Concurrent medications 2 1 .61
groups after the medication period did not differ
(Table 2). There were no cases of anemia at the end of
NOTE. Values are expressed as the mean (standard deviation). the medication period.
June 2016 PPIs and NSAID-Induced Small Bowel Injury 813

Among various NSAIDs, we chose celecoxib, a COX-2

selective inhibitor, to minimize upper GI toxicity caused
by NSAIDs and to refrain from administering antiacid
secretory medication to the control subjects. As a
consequence, we found small bowel mucosal injuries in
16.7% of the subjects in the COX-2 SI group, which was
slightly higher than rate reported by Goldstein et al
(7%).14 These results seem to imply COX-2-dependent
pathway is involved in the maintenance of mucosal
integrity and in the development of NSAID-induced small
bowel injury. However, because celecoxib possesses
COX-1 inhibitory properties at higher doses,23 it remains
uncertain whether the results represent the significance
of COX-2 inhibition in the pathogenesis of NSAID-induced
small bowel injury.
In the present study, the incidence of small bowel
injury in the COX-2 SI þ PPI group was as high as 44.4%,
which was equivalent to the value in patients taking
Figure 4. The incidence of jejunal and ileal mucosal injuries.
celecoxib and omeprazole in our previous study.12
Intestinal motility,24 mucous content,25 and indigestive
Discussion solid food content26 have previously been shown to be
affected by antiacid secretory drugs. The alteration of
PPIs have been shown to be effective for the pre- bacterial flora in small intestine by the marked sup-
vention of NSAID-induced upper GI mucosal injury.20 pression of gastric acid secretion has been also
Although the possibly beneficial effects of PPIs against indicated.27–29 Wallace et al15 recently demonstrated a
NSAID-induced small bowel injury have been shown in significant reduction of Actinobacteria and Bifidobacteria
animal experiments,9,20,21 such effects have been rigor- species in the jejunum in parallel with the exacerbation
ously disputed since the report by Wallace et al,15 in of NSAID-induced small bowel injury under concomitant
which exacerbation of small bowel injury via the mech- use of PPIs. Because PPIs alone cause neither small
anisms of dysbiosis was shown in a rodent experiment. A bowel mucosal damage nor inflammation,15 the alter-
recent Japanese cross-sectional study using CE identified ation of luminal environment by PPIs seems to be an
use of antiacid secretory drugs (PPIs and histamine H2 exacerbating factor in NSAID-induced small bowel
receptor antagonists) as independent risk factor for injuries. The obvious difference in the mucosal injuries of
severe NSAID-induced small bowel damage.22 However, the jejunum in the 2 groups further supports the sig-
no clinical trials have focused on the effects of concom- nificance of PPIs in NSAID-induced small bowel injury.
itant use of PPIs on the development of NSAID-induced The luminal microflora of the proximal small bowel seem
small bowel injury. We thus conducted this prospec- to be more profoundly altered by gastric acid suppres-
tive, double-blind, randomized controlled trial, and sion than the distal small bowel.
demonstrated an exacerbating effect on NSAID-induced When considering the influence of gastric acid sup-
small bowel injury. pression on NSAID-induced small bowel injuries, the
types of antiacid secretory drugs need to be taken into
account. Satoh et al30 recently demonstrated protective
Table 2. Abdominal Symptoms and Laboratory Data effect of lansoprazole against small bowel injuries in
COX-2 SI group COX-2 SI þ PPI P indomethacin-treated rats; such an effect was not found
(n ¼ 30) group (n ¼ 27) value with any other antiacid secretory drugs. However, more
recent experiments have demonstrated exacerbating
GSRS pre 1.2 (1.1–1.4) 1.3 (1.1–1.7) .43 effects of PPIs, including lansoprazole.15,31 Such contro-
GSRS post 1.2 (1.0–1.6) 1.1 (1.0–1.5) .92
versy in the findings regarding the effects of PPIs may be
Hemoglobin level 14.4 (1.3) 15.0 (1.5) .19
pre, g/dL a consequence of differences in the experimental
Hemoglobin level 14.3 (1.4) 14.7 (1.6) .60 methods, especially with regard to the dosage of antiacid
post, g/dL secretory drugs and in the control of microflora. Further
Anemia, n (%) 0 (0) 0 (0) 1.0 investigations should examine the correlation between
intraluminal pH and small bowel injuries.
NOTE. The GSRS values are expressed as the median (range). The hemoglobin Various guidelines32 recommend concomitant
levels are expressed as the mean (standard deviation). Anemia was defined as administration of COX-2 selective inhibitors plus PPIs for
positive with a decrease of at least 2.0 g/dL in the hemoglobin level. the prevention of upper GI complications in patients with
GSRS, Gastrointestinal Symptom Rating Scale; pre, before taking celecoxib
plus rabeprazole or a placebo; post, after taking celecoxib plus rabeprazole or high GI and low cardiovascular risks. In contrast, results
a placebo. of our present and previous studies12 seem to suggest
814 Washio et al
that the risk of small bowel injury increases with
concomitant use of PPIs and COX-2 selective inhibitors.
However, our results should be interpreted cautiously,
because none of the subjects in the present study
showed anemia, whereas there was significant difference
in the incidence of mucosal injuries in the 2 groups.
Further investigations with clinically significant end-
points seem to be necessary to reach a conclusion
regarding the effects of the combined administration of
NSAIDs and PPIs on mucosal injuries.
The present study had some limitations. First, it was
conducted in a relatively young and healthy population
under short-term administration of NSAIDs. Therefore,
our results may not be applied to patients who are at
increased risk for NSAID-induced GI toxicity. Second, we
were unable to take H pylori status, which is known to be
associated with GI blood loss in patients taking NSAIDs
or COX-2 SI,33 into consideration because there were
only 6 H pylori–positive subjects. Third, the small sample
size might have contributed to a higher incidence of
small bowel injury in our COX-2 SI group. However,
because the incidence of small bowel injury in the COX-2
SI þ PPI group was equivalent to that in our previous
study,12 and because the protocol of the present study
was similar to that of Goldstein et al, the incidence of the
injury in the COX-2 SI group may be representative of the
Asian population. Furthermore, even though the inci-
dence in the COX-2 SI group was high, it was significantly
lower than that in the COX-2 SI þ PPI group.
In conclusion, our prospective study demonstrated
that the incidence of small bowel injury was higher in
subjects treated with celecoxib plus rabeprazole than
those treated with celecoxib alone. It thus can be sug-
gested PPIs increase the risk of small bowel injury by
NSAIDs including COX-2 selective inhibitor, whereas its
clinical significance remains to be clarified.
Supplementary Material
Note: To access the supplementary material accom-
panying this article, visit the online version of Clinical
Gastroenterology and Hepatology at www.cghjournal.org,
and at http://dx.doi.org/10.1016/j.cgh.2015.10.022.
References
1. Graham DY, Agrawal NM, Roth SH. Prevention of NSAID-
induced gastric ulcer with misoprostol: multicenter, double-
blind, placebo-controlled trial. Lancet 1988;2:1277–1280.
2. Allison MC, Howatson AG, Torrance CJ, et al. Gastrointestinal
damage associated with the use of nonsteroidal anti-
inflammatory drugs. N Engl J Med 1992;327:749–754.
3. Graham DY, Opekun AR, Willingham FF, et al. Visible small-
intestinal mucosal injury in chronic NSAID users. Clin Gastro-
enterol Hepatol 2005;3:55–59.
4. Maiden L, Thjodleifsson B, Theodors A, et al. A quantitative
analysis of NSAID-induced small bowel pathology by capsule
enteroscopy. Gastroenterology 2005;128:1172–1178.
Clinical Gastroenterology and Hepatology Vol. 14, No. 6
5. Matsumoto T, Kudo T, Esaki M, et al. Prevalence of non-
steroidal anti-inflammatory drug-induced enteropathy deter-
mined by double-balloon endoscopy: a Japanese multicenter
study. Scand J Gastroenterol 2008;43:490–496.
6. Bjarnason I, Hayllar J, MacPherson AJ, et al. Side effects of
nonsteroidal anti-inflammatory drugs on the small and large
intestine in humans. Gastroenterology 1993;104:1832–1847.
7. Kuroda M, Yoshida N, Ichikawa H, et al. Lansoprazole, a proton
pump inhibitor, reduces the severity of indomethacin-induced
rat enteritis. Int J Mol Med 2006;17:89–93.
8. Pozzoli C, Menozzi A, Grandi D, et al. Protective effects of
proton pump inhibitors against indomethacin-induced lesions in
the rat small intestine. Naunyn Schmiedebergs Arch Pharmacol
2007;374:283–291.
9. Higuchi K, Yoda Y, Amagase K, et al. Prevention of NSAID-
induced small intestinal mucosal injury: prophylactic potential
of lansoprazole. J Clin Biochem Nutr 2009;45:125–130.
10. Fujimori S, Gudis K, Takahashi Y, et al. Distribution of small
intestinal mucosal injuries as a result of NSAID administration.
Eur J Clin Invest 2010;40:504–510.
11. Niwa Y, Nakamura M, Ohmiya N, et al. Efficacy of rebamipide for
diclofenac-induced small-intestinal mucosal injuries in healthy
subjects: a prospective, randomized, double-blinded, placebo-
controlled, cross-over study. J Gastroenterol 2008;43:270–276.
12. Maehata Y, Esaki M, Morishita T, et al. Small bowel injury
induced by selective cyclooxygenase-2 inhibitors: a prospec-
tive, double-blind, randomized clinical trial comparing celecoxib
and meloxicam. J Gastroenterol 2012;47:387–393.
13. Goldstein JL, Eisen GM, Lewis B, et al. Video capsule endos-
copy to prospectively assess small bowel injury with celecoxib,
naproxen plus omeprazole, and placebo. Clin Gastroenterol
Hepatol 2005;3:133–141.
14. Goldstein JL, Eisen GM, Lewis B, et al. Small bowel mucosal injury
is reduced in healthy subjects treated with celecoxib compared
with ibuprofen plus omeprazole, as assessed by video capsule
endoscopy. Aliment Pharmacol Ther 2007;25:1211–1222.
15. Wallace JL, Syer S, Denou E, et al. Proton pump inhibitors
exacerbate NSAID-induced small intestinal injury by inducing
dysbiosis. Gastroenterology 2011;141:1314–1322.
16. Sakamoto C, Soen S. Efficacy and safety of the selective
cyclooxygenase-2 inhibitor celecoxib in the treatment of rheu-
matoid arthritis and osteoarthritis in Japan. Digestion 2011;
83:108–123.
17. Mizokami Y. Efficacy and safety of rabeprazole in non-steroidal
anti-inflammatory drug-induced ulcer in Japan. World J Gas-
troenterol 2009;15:5097–5102.
18. Niwa Y, Nakamura M, Miyahara R, et al. Geranylgeranylacetone
protects against diclofenac-induced gastric and small intestinal
mucosal injuries in healthy subjects: a prospective randomized
placebo-controlled double-blind cross-over study. Digestion
2009;80:260–266.
19. Svedlund J, Sjödin I, Dotevall G. GSRS–a clinical rating scale for
gastrointestinal symptoms in patients with irritable bowel
syndrome and peptic ulcer disease. Dig Dis Sci 1988;33:
129–134.
20. Umegaki E, Yoda Y, Tokioka S, et al. Protective effect of rox-
atidine against indomethacin-induced small intestinal mucosal
injury in rats. J Gastroenterol Hepatol 2010;25(Suppl 1):S35–S40.
21. Watanabe T, Higuchi K, Kobata A, et al. Non-steroidal anti-
inflammatory drug-induced small intestinal damage is Toll-like
receptor 4 dependent. Gut 2008;57:181–187.
June 2016
22. Watanabe T, Tanigawa T, Nadatani Y, et al. Risk factors for
severe nonsteroidal anti-inflammatory drug-induced small in-
testinal damage. Dig Liver Dis 2013;45:390–395.
23. Warner TD, Giuliano F, Vojnovic I, et al. Nonsteroid drug se-
lectivities for cyclo-oxygenase-1 rather than cyclo-oxygenase-2
are associated with human gastrointestinal toxicity: a full in vitro
analysis. Proc Natl Acad Sci USA 1999;96:7563–7568.
24. Takeuchi K, Miyazawa T, Tanaka A, et al. Pathogenic impor-
tance of intestinal hypermotility in NSAID-induced small intes-
tinal damage in rats. Digestion 2002;66:30–41.
25. Kunikata T, Tanaka A, Miyazawa T, et al. 16,16-Dimethyl prosta-
glandin E2 inhibits indomethacin-induced small intestinal lesions
through EP3 and EP4 receptors. Dig Dis Sci 2002;47:894–904.
26. Satoh H. Role of dietary fiber in formation and prevention of
small intestinal ulcers induced by nonsteroidal anti-inflammatory
drug. Curr Pharm Des 2010;16:1209–1213.
27. Verdu E, Viani F, Armstrong D, et al. Effect of omeprazole on
intragastric bacterial counts, nitrates, nitrites, and N-nitroso
compounds. Gut 1994;35:455–460.
28. Lombardo L, Foti M, Ruggia O, et al. Increased incidence of
small intestinal bacterial overgrowth during proton pump inhib-
itor therapy. Clin Gastroenterol Hepatol 2010;8:504–508.
29. Jacobs C, Coss Adame E, Attaluri A, et al. Dysmotility and
proton pump inhibitor use are independent risk factors for small
intestinal bacterial and/or fungal overgrowth. Aliment Pharmacol
Ther 2013;37:1103–1111.
PPIs and NSAID-Induced Small Bowel Injury 815
30. Satoh H, Amagase K, Takeuchi K. Exacerbation of nonsteroidal
anti-inflammatory drug-induced small intestinal lesions by anti-
secretory drugs in rats: the role of intestinal motility.
J Pharmacol Exp Ther 2012;343:270–277.
31. Lim YJ, Phan TM, Dial EJ, et al. In vitro and in vivo protection
against indomethacin-induced small intestinal injury by proton
pump inhibitors, acid pump antagonists, or indomethacin-
phosphatidylcholine. Digestion 2012;86:171–177.
32. Lanza FL, Chan FK, Quigley EM. Guidelines for prevention of
NSAID-related ulcer complications. Am J Gastroenterol 2009;
104:728–738.
33. Lanas A, Goldstein JL, Chan FK, et al. Risk factors associated
with a decrease 2 g/dL in hemoglobin and/or 10% hemat-
ocrit in osteoarthritis patients taking celecoxib or a nonselective
NSAID plus a PPI in a larger randomized controlled trial
(CONDOR). Aliment Pharmacol Ther 2012;36:485–492.
Reprint requests
Address requests for reprints to: Takayuki Matsumoto, MD, Division of
Gastroenterology, Department of Internal Medicine, School of Medicine,
Iwate Medical University, Uchimaru 19-1, Morioka 020-8505, Japan. e-mail:
tmatsumo@iwate-med.ac.jp; fax: þ81-19-907-4327.
Conflicts of interest
The authors disclose no conflicts.
Funding
This study was sponsored by Eisai Inc.
815.e1 Washio et al Clinical Gastroenterology and Hepatology Vol. 14, No. 6

Supplementary Table 1. Crude Numbers of Small Bowel Injuries Among the Subjects With Positive CE Findings

Erosion Ulcer

Jejunum Ileum Jejunum Ileum Total

COX-2 SI group
1 0 2 0 1 3
2 0 1 0 0 1
3 0 2 0 1 3
4 0 0 0 7 7
5 0 3 0 0 3
COX-2 SI þ PPI group
1 0 5 0 2 7
2 1 0 0 0 1
3 0 2 0 3 5
4 1 0 0 0 1
5 0 2 0 2 4
6 5 24 0 2 31
7 0 1 0 0 1
8 1 2 0 0 3
9 5 3 0 2 10
10 1 1 0 0 2
11 4 1 0 0 5
12 0 0 0 3 3