D

Ri

THYa

b

a

ARAA

KACP

1

mrdmdt

st5

sJ

1h

Digestive and Liver Disease 45 (2013) 390– 395

Contents lists available at SciVerse ScienceDirect

Digestive and Liver Disease

jou rn al h om epage: www.elsev ier .com/ locate /d ld

igestive endoscopy

isk factors for severe nonsteroidal anti-inflammatory drug-induced smallntestinal damage�

oshio Watanabea,∗, Tetsuya Tanigawaa, Yuji Nadatania, Yasuaki Nagamia, Satoshi Sugimoria,irotoshi Okazakia, Hirokazu Yamagamia, Kenji Watanabea, Kazunari Tominagaa,asuhiro Fujiwaraa, Tatsuya Koikeb, Tetsuo Arakawaa

Department of Gastroenterology, Osaka City University Graduate School of Medicine, JapanDepartment of Rheumatosurgery, Osaka City University Graduate School of Medicine, Japan

r t i c l e i n f o

rticle history:eceived 5 November 2012ccepted 10 December 2012vailable online 18 January 2013

eywords:nemiaapsule endoscopyroton pump inhibitor

a b s t r a c t

Background: Few studies have assessed the risk factors associated with nonsteroidal anti-inflammatorydrugs (NSAIDs)-induced small intestinal damage.Aims: To evaluate the risk factors for NSAID-induced enteropathy in patients with rheumatoid arthritis.Methods: A cross-sectional study using capsule endoscopy was conducted. A total of 113 patients whotook NSAIDs for over 3 months underwent capsule endoscopies. Endoscopic findings were scored as (0)normal, (1) red spots, (2) 1–4 erosions, (3) >4 erosions, or (4) large erosions/ulcers. Initial scores weregrouped into 3 categories: No damage (0–1), mild damage (2), and severe damage (3–4), and the potentialrisk factors for damage development were assessed.Results: Five patients were excluded because of incomplete visualization of the entire small intes-tine. Fifty-two (47.2%) and 27 (25%) patients had no damage and mild damage, respectively, while theremaining 30 patients (27.8%) had severe damage and significantly decreased hemoglobin levels. In a

multivariate logistic regression analysis, ages of 65 years or more (odds ratio [OR], 4.16; 95% confidenceinterval [CI], 1.51–11.47), proton pump inhibitor usage (OR, 5.22; 95% CI, 1.36–20.11), and histamineH2 receptor antagonist usage (OR, 3.95; 95% CI, 1.28–12.25) were independent risk factors for severedamage.Conclusions: Elderly patients and acid suppressant users are more likely to develop severe NSAID-induced

Gast

enteropathy.

© 2012 Editrice

. Introduction

Nonsteroidal anti-inflammatory drugs (NSAIDs) are one of theost commonly prescribed categories of drugs. They are used to

educe fever and treat pain or inflammation caused by chronic con-itions such as rheumatoid arthritis (RA) and osteoarthritis. Theost common side effect of NSAIDs is upper gastrointestinal (GI)

amage which results in peptic ulceration, bleeding and perfora-ion.

Recent advances in GI endoscopy, especially wireless video cap-

ule endoscopy (VCE) [1], have shown that NSAIDs frequently injurehe small intestine. A small bowel mucosal break was induced in5% of healthy volunteers given naproxen for 2 weeks [2], and

� Grant, equipment, or drugs support: None.∗ Corresponding author at: Department of Gastroenterology, Osaka City Univer-

ity Graduate School of Medicine, 1-4-3 Asahimachi, Abeno-ku, Osaka 545-8585,apan. Tel.: +81 6 6645 3811; fax: +81 6 6645 3813.

E-mail address: watanabet@med.osaka-cu.ac.jp (T. Watanabe).

590-8658/$36.00 © 2012 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevierttp://dx.doi.org/10.1016/j.dld.2012.12.005

roenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.

another study demonstrated that a 2-week ingestion of slow-release diclofenac resulted in macroscopic small bowel pathologiesin 68% of healthy volunteers [3]. This ulcerogenic effect of NSAIDson the small intestine was not transient and did not disappear withlong-term therapy. Graham et al. reported that small bowel injurywas seen in 71% of arthritis patients who took NSAIDs for more than3 months [4], and we reported that mucosal breaks were detected inapproximately 80% of patients with RA taking NSAIDs for more than1 year [5]. These clinical results, together with the fact that NSAID-induced small intestinal damage can cause complications such asovert bleeding, perforation, stricture, hypoalbuminemia, and occultbleeding which may lead to the development of iron-deficiencyanemia [6–8], has focused much attention on the treatment andprevention of NSAID-induced enteropathy as well as the patho-genesis of this disorder.

Risk assessment is very important for managing NSAID-induced

GI damage. A variety of studies have identified risk factors for thedevelopment of peptic ulceration and upper GI complications withNSAID use, and several factors, such as advanced age, history of pep-tic ulcers, high doses of NSAIDs, comorbidities, and the concomitant

Ltd. All rights reserved.

nd Liv

ube(mtistNcd

Ni

2

2

saNtae

parmdbiltPlitU

2

(bwlaelt

2

iwsnewAd

T. Watanabe et al. / Digestive a

se of corticosteroids and anticoagulants, have been reported toe associated with an increased risk of NSAID-related upper GIvents [9]. Consequently, cotherapy with proton pump inhibitorsPPIs), (high-dose) histamine-2-receptor antagonists (H2RAs), or

isoprostol (a synthetic prostaglandin E1 analog) and/or substitu-ion of a traditional NSAID with a selective cyclooxygenase (COX)-2nhibitor has been recommended as a gastroprotective strategy forubjects at increased risk for NSAID-induced upper GI complica-ions [10]. However, few studies have addressed the risk factors forSAID-induced lower GI damage, and no data on the risks asso-iated with the development of NSAID-induced small intestinalamage have been reported.

In this study, we evaluated the prevalence and risk factors ofSAID-induced small intestinal damage in RA patients who chron-

cally took NSAIDs with capsule endoscopy.

. Methods

.1. Subjects

This was a single center, open-label, endoscopist-blind, cross-ectional study that was performed to evaluate the prevalencend risk factors of small-bowel injury in RA patients on long-termSAID treatment by capsule endoscopy. A total of 113 consecu-

ive RA patients (22 males and 91 females; mean age, 60.7 years;ge range, 24–83) who took NSAIDs for more than 3 months werenrolled in this study.

The main exclusion criteria included swallowing disorders,regnancy, history of abdominal surgery, aspirin use, andny serious central nervous system, psychiatric, cardiovascular,espiratory, musculoskeletal, or intestinal disease. Patients onisoprostol were excluded because the preventive effects of this

rug on NSAID-induced small intestinal damage were confirmedy a capsule endoscopic study [11]. Patients who were tak-

ng low-dose aspirin (≤325 mg/day) were also excluded becauseow-dose aspirin alone can induce multiple endoscopic ulcers inhe small intestine [12,13], while other drugs including H2RAs,PIs, corticosteroids, disease-modifying antirheumatic drugs, bio-ogics, sulphasalazine, and bisphosphonates were allowed. Writtennformed consent was obtained from all patients. The protocol forhe study was approved by the Ethics Committee of the Osaka Cityniversity Graduate School of Medicine.

.2. Procedure

Capsule endoscopy was performed with a PillCam SB1 or SB2Given Imaging, Ltd, Yoqneam, Israel). Patients fasted for 12 hefore swallowing the capsule. Fluids were allowed 2 h later, whichas followed by a light meal another 2 h later. Data were col-

ected for up to 8 h after capsule ingestion. After 8 h, the sensorrray and recording device were removed. In addition to capsulendoscopy, routine laboratory tests, including measurements of theevels of hemoglobin and albumin, were performed at the time ofhe endoscopy.

.3. Data interpretation

The VCE digital image stream was reviewed and interpretedndependently by 2 experienced endoscopists (N.Y. and S.S.) who

ere blind to patient treatment. Capsule endoscopic findings werecored according to the following method of Graham et al. [4]: (0)ormal; (1) red spots; (2) 1–4 erosions; (3) >4 erosions; (4) large

rosions/ulcers. A red spot was defined as a mucosal disruption thatas denuded of villi with red areas but no clear mucosal breaks.

small erosion was defined as a circumscribed area of mucosalisruption that was denuded of villi with or without exudates or

er Disease 45 (2013) 390– 395 391

red color and that involved, at most, a diameter that was equiva-lent to those of valvulae conniventes. Large erosions were definedas circumscribed breaks in the mucosa that were larger than theequivalent diameter of a valvulae conniventes. Ulcers were definedas large erosions with a central area with exudates, typically witha surrounding border of elevated mucosa. The categories for smallbowel injury were modified from the previous report [4]. The initialscores were grouped to yield 3 categories: No damage (0–1), milddamage (2), and severe damage (3–4). Typical examples of smallbowel pathologies are shown in Fig. 1. If the endoscopic findingswere different between the two interpreters, they then discussedthe case until consensus was achieved.

2.4. Statistics

Data are summarized as mean ± standard deviation for contin-uous variables and proportions (%) for discrete variables. A listof baseline clinical variables was specified for the assessment ofpotential risk factors for NSAID-induced small intestinal damage.Chi-squared tests (or Fisher’s exact test when necessary because ofsmall sample size) were used to assess the differences in risk fac-tors between subjects with and without small intestinal damage.The effects of the clinical variables on the development of NSAID-induced small intestinal damage were estimated by calculatingodds ratios (ORs) and 95% confidence intervals (CIs) using logisticregression analyses. Risk factors that were determined to be sig-nificant by univariate analysis were then subjected to multivariateanalyses. Analysis of variance was conducted to test for the signif-icance of the differences in the levels of hemoglobin, and serumtotal protein and albumin between groups, followed by Tukey’sHonestly Significant Difference test. Statistical analyses were per-formed using IBM SPSS Statistics 19 (IBM Corporation, Armonk, NY).All statistical tests were two-sided, and P values less than 0.05 wereconsidered to be statistically significant.

3. Results

3.1. Patient characteristics

A total of 113 patients who were enrolled between April 2006and December 2011 underwent VCE. No side effects were reportedby any patient during the procedure. Five patients were excludedbecause of incomplete visualization of the entire small bowel(incomplete transit through the small bowel [n = 4], poor visu-alization [n = 1]). Thus, 108 patients had their comparative dataanalyzed. The characteristics of the 108 patients examined in thestudy protocol are shown in Table 1. The patients included 21 malesand 87 females with a mean age of 60.2 years. A history of pepticulcer was present in 33 patients. Fifty-two patients, 84 patients,15 patients, and 22 patients were taking low-dose corticosteroids,methotrexate, salazosulfapyridine, and bisphosphonate, respec-tively. Regarding acid suppressants, 20 patients were taking PPIs(rabeprazole [n = 12], lansoprazole [n = 5], and omeprazole [n = 3])and 34 patients were taking H2RAs (famotidine [n = 14], lafutidine[n = 12], ranitidine [n = 3], cimetidine [n = 3], nizatidine [n = 1], androxatidine [n = 1]).

Among the 108 chronic users of NSAIDs, 87 patients were tak-ing traditional NSAIDs, and the remaining 21 patients were takingthe selective COX-2 inhibitor, celecoxib. Seventy patients tookstandard doses of NSAIDs, while 38 patients did lower doses ofNSAIDs. No patients took higher doses of NSAIDs. The NSAIDs taken

by the patients in this study were as follows: loxoprofen (n = 56,60–180 mg/day), etodolac (n = 14, 200–400 mg/day), diclofenac(n = 5, 37.5–75 mg/day), sulindac (n = 5, 150–300 mg/day), meloxi-cam (n = 5, 10 mg/day), indomethacin farnesil (n = 1, 400 mg/day),

392 T. Watanabe et al. / Digestive and Liver Disease 45 (2013) 390– 395

F al ante

cd

3

add51

TB

P

ig. 1. Capsule endoscopic images of small bowel injuries induced by nonsteroidrosion, (D) ulcer.

elecoxib (n = 21, 200–400 mg/day), and the combined therapy oficlofenac plus etodolac (n = 1).

.2. Prevalence of NSAIDs-induced small intestinal damage

Out of the 108 patients, 57 (52.8%) had small intestinal dam-ge (endoscopic score, 2–4), with 27 patients (25.0%) having mild

amage (endoscopic score, 2) and 30 patients (27.8%) having severeamage (endoscopic score of 3 [n = 14] or 4 [n = 16]). The remaining1 patients (47.2%) had no damage (endoscopic score of 0 [n = 7] or

[n = 44]).

able 1aseline characteristics of the subjects.

Total No d(n = 108) (n =

Female 87 (80.6%) 44 (8Age (years)

Mean (range) 60.2 (24–83) 57.6≥65 38 (35.2%) 14 (2

History of peptic ulcers 33 (30.6%) 15 (2Smoker 14 (13.0%) 9 (1Alcohol intake 14 (13.0%) 6 (1Low-dose corticosteroids use 52 (48.1%) 25 (4Methotrexate use 84 (77.8%) 41 (8Salazosulfapyridine use 15 (13.9%) 10 (1Bisphosphonate use 22 (20.4%) 12 (2Acid suppresant

PPI use 20 (18.6%) 7 (1H2RA use 34 (31.5%) 13 (2

Standard dose of NSAIDs 70 (64.8%) 32 (6Traditional NSAID use 87 (80.6%) 40 (7

PI, proton pump inhibitor; H2RA, histamine-2-receptor antagonist; NSAID, nonsteroidal

i-inflammatory drugs (NSAIDs). (A) Red spot (arrow), (B) small erosion, (C) large

3.3. Risk factors for NSAIDs-induced small intestinal damage

No variables, including age, sex, smoking habit, alcohol intake,and the use of anti-RA drugs, acid suppressants, NSAIDs dose, orselective COX-2 inhibitors, were associated with the developmentof small intestinal damage (Table 2). However, when small intesti-nal injuries were divided into the 2 categories of mild damage and

severe damage according to the endoscopic findings, an age of 65years or more and the use of PPIs or H2RAs were associated withthe development of severe small intestinal damage in the univariate

amage Mild damage Severe damage51) (n = 27) (n = 30)

6.3%) 22 (81.5%) 21 (70%)

(24–80) 58.4 (31–83) 66.4 (47–80)7.5%) 6 (22.2%) 18 (60%)9.4%) 6 (22.2%) 12 (40%)7.6%) 3 (11.1%) 2 (6.7%)1.8%) 4 (14.8%) 4 (13.3%)9%) 14 (51.9%) 13 (43.3%)0.4%) 19 (70.4%) 24 (80%)9.6%) 1 (3.7%) 4 (13.3%)3.5%) 2 (7.4%) 8 (26.7%)

3.7%) 5 (18.5%) 8 (26.7%)5.5%) 7 (25.9%) 14 (46.7%)2.7%) 20 (74.1%) 18 (60%)8.4%) 23 (85.1%) 24 (80%)

anti-inflammatory drug.

T. Watanabe et al. / Digestive and Liver Disease 45 (2013) 390– 395 393

Table 2Univariate analysis of risk factors for NSAID-induced small intestinal damage.

Risk factor Reference Any damage (Scores 2–4) Mild damage (Score 2) Severe damage (Scores 3–4)

OR (95% CI) p-Value OR (95% CI) p-Value OR (95% CI) p-Value

Female Male 0.59 (0.22–1.62) 0.30 0.92 (0.24–3.45) 0.90 0.37 (0.12–1.13) 0.08Age, ≥65 Age, <65 2.04 (0.90–4.61) 0.09 0.76 (0.25–2.26) 0.62 3.96 (1.53–10.30) 0.01History of peptic ulcers No ulcer 0.99 (0.44–2.27) 0.99 0.69 (0.23–2.04) 0.50 1.60 (0.62–4.12) 0.33Smoker Non-smoker 0.41 (0.13–1.32) 0.13 0.58 (0.14–2.4) 0.45 0.33 (0.07–1.66) 0.18Alcohol intake No intake 1.58 (0.49–5.08) 0.44 1.30 (0.33–5.09) 0.70 1.14 (0.30–4.47) 0.84Low-dose corticosteroids use Non-user 0.94 (0.44–2.01) 0.88 1.12 (0.44–2.85) 0.81 0.80 (0.32–1.97) 0.80Methotrexate use Non-user 0.82 (0.33–2.06) 0.68 0.58 (0.20–1.70) 0.58 0.98 (0.32–3.02) 0.98Salazosulfapyridine use Non-user 0.50 (0.17–1.53) 0.22 0.16 (0.02–1.31) 0.09 0.63 (0.18–2.22) 0.47Bisphosphonate use Non-user 0.63 (0.25–1.61) 0.33 0.26 (0.05–1.26) 0.09 1.18 (0.42–3.33) 0.75PPI use Non-user of acid

suppressants2.32 (0.80–6.73) 0.12 1.48 (0.40–5.43) 0.56 4.43 (1.23–15.90) 0.02

H2RA use Non-user of acidsuppressants

2.29 (0.95–5.55) 0.07 1.11 (0.37–3.37) 0.85 4.17 (1.41–12.33) 0.01

Standard doses of NSAIDs Lower doses ofNSAIDs

0.96 (0.44–2.13) 0.92 1.70 (0.61–4.76) 0.32 0.89 (0.35–2.25) 0.81

Selective COX-2 inhibitor use TraditionalNSAID use

0.89 (0.34–2.32) 0

OR, odds ratio; CI, confidence interval; PPI, proton pump inhibitor; H2RA, histamine-2-rece

Table 3Multivariate analysis of risk factors for NSAID-induced severe small intestinaldamage.

Risk factor Reference OR (95% CI) p-Value

Age, ≥65 Age, <65 4.16 (1.51–11.47) 0.01PPI use Non-user of acid

suppressants5.22 (1.36–20.11) 0.02

H2RA use Non-user of acidsuppressants

3.95 (1.28–12.25) 0.02

NSAID, nonsteroidal anti-inflammatory drug; OR, odds ratio; CI, confidence interval;OR, odds ratio; CI, confidence interval; PPI, proton pump inhibitor; H2RA, histamine-2-receptor antagonist.

Table 4Comparison of the levels of hemoglobin and serum total protein and albumin accord-ing to severity of NSAID-induced small intestinal damage.

No damage(n = 51)

Mild damage(n = 27)

Severe damage(n = 30)

Hemoglobin (g/dL) 12.97 ± 1.22 12.51 ± 0.98 12.16 ± 1.84*

Total protein (g/dL) 7.02 ± 0.47 7.10 ± 0.49 6.95 ± 0.35Albumin (g/dL) 4.11 ± 0.30 4.05 ± 0.28 3.96 ± 0.38

Data are expressed as mean ± standard deviation.

N

ads(1dTmn(sb

4

dddi

* p < 0.05 vs no damage group.SAID, nonsteroidal anti-inflammatory drug.

nalysis (Table 2). Female sex had a tendency to decrease the inci-ence of severe damage, but the differences did not reach statisticalignificance. In the multivariate analysis, an age of 65 years or moreOR, 4.16; 95% CI, 1.51–11.47) and the use of PPIs (OR, 5.22; 95% CI,.36–20.11) or H2RAs (OR, 3.95; 95% CI, 1.28–12.25) were indepen-ent risk factors for the development of severe damage (Table 3able 3). No variables were associated with the development ofild damage. In addition, patients with severe damage had sig-

ificantly lower hemoglobin levels than those with no damageTable 4). Serum albumin levels were also lower in RA patients withevere damage than in those with no damage, but the differenceetween the 2 group did not reach statistical significance (P = 0.10).

. Discussion

Recent studies using VCE have shown that NSAIDs frequently

amage the small intestine, but the risk factors associated withamage have not been evaluated because most studies were con-ucted in young, healthy volunteer subjects with low risks. Even

n the case of clinical trials that were conducted in chronic NSAIDs

.82 0.63 (0.18–2.22) 0.47 0.91 (0.30–2.78) 0.91

ptor antagonist; NSAID, nonsteroidal anti-inflammatory drug; COX, cyclooxygenase.

users, the number of subjects was small in each trial. In this study,we attempted to identify the risk factors for NSAID-induced smallintestinal damage by performing VCE in over 100 RA patients whochronically used NSAIDs and found that old age and the use of PPIsand H2RAs were associated with an increased risk of severe NSAID-induced small intestinal damage with ORs of 4.16, 5.22, and 3.95,respectively. To the best of our knowledge, this is the first reporton the risk factors for NSAID-induced small intestinal damage.

Importantly, we demonstrated that more than a quarter ofRA patients who chronically used NSAIDs developed severe smallintestinal damage and that the patients with severe damage hadsignificantly lower hemoglobin concentrations and an insignificanttrend for lower serum albumin levels compared to those withoutdamage, suggesting that the prevalence of NSAID-induced severeintestinal damage that was categorized by the VCE findings in thisstudy was high, and that severe enteropathy may be a clinically sig-nificant condition that can result in anemia and hypoalbuminemia.

We found that the use of acid suppressants, including PPIs andH2RA, which are key drugs for preventing and treating NSAID-induced upper GI ulceration and complications, were associatedwith the development of severe enteropathy. Consistent withour results, the possibility that acid suppressants may aggravateNSAID-induced enteropathy has been suggested in several clini-cal studies using VCE. Goldstein et al. reported that the incidenceof small intestinal damage in healthy volunteers who were givenboth naproxen and omeprazole for 2 weeks was 55% [2]. Similarly,a very high incidence of damage in the small intestine in healthyvolunteers who were given NSAIDs concomitant with PPIs wasreported by Maiden et al. [3] and Graham et al. [4] Recent experi-mental studies have demonstrated that enterobacteria, especiallygram-negative bacteria, play a crucial role in the development ofNSAID-induced small intestinal damage [14–16]. Because gastricacid can kill bacteria in the stomach and the small intestine, acidsuppression by PPIs and H2RAs affects the bacterial flora of the GItract and thereby could lead to the aggravation of NSAID-inducedenteropathy. More recently, Wallace et al. demonstrated that PPIs,such as omeprazole and lansoprazole, exacerbated NSAID-inducedsmall intestinal injury by inducing dysbiosis in rats [17], stronglysupporting our hypothesis.

In fact, many studies have demonstrated changes in bacterialflora in the GI tract with acid suppressants. Lombardo et al.

reported that small intestinal bacterial overgrowth (as assessed byglucose hydrogen breath tests) was detected in 50% of 200 patientswith gastroesophageal reflux disease who were using PPIs for atleast 2 months [18]. In addition, in a randomized study, gastric

3 nd Liv

o1pfire

darftisticdemtt

madaf[aeotsNt

ifrtittmsiilMeogTcgtwtg

sisei

[

[

[

[

[

[

[

[

[

[

[

[

94 T. Watanabe et al. / Digestive a

r duodenal bacterial overgrowth or both was present in 10 of9 patients (53%) taking omeprazole for 4 weeks and in 3 of 18atients (17%) taking cimetidine for a similar period [19]. Thesendings strongly suggest that PPIs and H2RAs may increase theisk of severe NSAID-induced enteropathy through their inhibitoryffects on gastric acid secretion.

It is noteworthy that the selective COX-2 inhibitor, celecoxib,id not decrease the risk of NSAID-induced small intestinal dam-ge, including mild and severe damage. Three previous studieseported that selective COX-2 inhibitors were associated withewer mucosal lesions of the small bowel than were nonselec-ive traditional NSAIDs [2,20,21], but in these studies, the smallntestinal toxicity of COX-2 inhibitors had been assessed after thehort-term (14–16 days) administration of these drugs. In con-rast, Maiden et al. reported no difference in the incidence of smallntestinal injury between chronic users of traditional NSAIDs andhronic users of selective COX-2 inhibitors [22]. The reason for thisiscrepancy remains unclear, but these results suggest that the ben-ficial effects of selective COX-2 inhibitors on the small intestineay be abolished by their long-term use and that switching tradi-

ional NSAIDs to selective COX-2 inhibitors in order to prevent orreat NSAID-induced enteropathy may be ineffective.

In combination with NSAIDs, RA patients take many types ofedication, including low-dose corticosteroids, disease-modifying

ntirheumatic drugs (such as methotrexate and salazosulfapyri-ine), and bisphosphonates. Among these drugs, corticosteroidsnd bisphosphonates have been shown to be independent riskactors for NSAID-induced upper GI ulceration and complications9,23,24], but we found that these 2 classes of medication did notffect the risks associated with NSAID-induced enteropathy. It isvident that differences in the risk factors for the developmentf the damage between the gastroduodenum and the small intes-ine exist in chronic users of NSAIDs, although our present resultsuggested that old age is a common independent risk factor forSAID-induced damage in the entire digestive system, including

he upper, middle, and lower GI tracts [9,25].There are three limitations to this study. First, since this study

s a cross-sectional study, there is no baseline VCE evaluation per-ormed before induction of NSAID therapy. Therefore we cannotule out the possibility that NSAID users with severe enteropa-hy are more likely to use acid suppressants. In other words, thencreased risk of severe enteropathy related to the concomitantreatment of NSAIDs and acid suppressants may simply indicatehat patients with dyspeptic symptoms are also those who are

ore likely to develop severe enteropathy. Therefore, a prospectivetudy to evaluate the adverse effects of acid suppressants on NSAID-nduced enteropathy is urgently needed. Second, loxoprofen, whichs the most commonly used NSAID in this study, is available inimited numbers of countries including Brasil, India, Japan, and

exico. However, animal studies demonstrated that ulcerogenicffects of this drug on the small intestine were similar to thosef indomethacin [26], indicating that in human, intestinal ulcero-enicity of loxoprofen is also similar to other traditional NSAIDs.hird, since we did not perform esophagogastroduodenoscopy andolonoscopy, we cannot rule out the possibility that other patholo-ies developed in upper GI and/or lower GI tract during NSAIDherapy may be responsible for the hemoglobin drop in patientsith severe small intestinal damage. Further studies to assess the

rue clinical relevance of severe intestinal damage that was cate-orized by the VCE findings in this study are warranted.

In conclusion, these results suggested that the prevalence ofevere small intestinal damage is high in RA patients who chron-

cally take NSAIDs and that elderly patients and users of aciduppressants may be more likely to develop severe NSAID-inducednteropathy. Because a quarter of chronic NSAID users develop clin-cally significant severe enteropathy with decreased hemoglobin

[

er Disease 45 (2013) 390– 395

levels, close attention should be paid to patients with this condi-tion. Although medical cotherapy with PPIs or misoprostol and/orthe use of selective COX-2 inhibitors have been recommended todecrease GI clinical events in patients at high GI risk, such as thoseof old age [10], a PPI-based strategy or a switch to COX-2 inhibitorsmay have no benefit to the small intestine, and the former can evenexacerbate enteropathy. Because misoprostol has been reported toprevent small bowel pathologies that were induced by a 2-weekadministration of diclofenac in healthy volunteers [11], the effi-cacy of this drug on enteropathy in chronic NSAID users shouldbe urgently examined in order to establish a strategy for NSAID-induced gastroenteropathy.

References

[1] Iddan G, Meron G, Glukhovsky A, et al. Wireless capsule endoscopy. Nature2000;405:417.

[2] Goldstein JL, Eisen GM, Lewis B, et al. Video capsule endoscopy to prospec-tively assess small bowel injury with celecoxib, naproxen plus omeprazole,and placebo. Clinical Gastroenterology and Hepatology 2005;3:133–41.

[3] Maiden L, Thjodleifsson B, Theodors A, et al. A quantitative analysis of NSAID-induced small bowel pathology by capsule enteroscopy. Gastroenterology2005;128:1172–8.

[4] Graham DY, Opekun AR, Willingham FF, et al. Visible small-intestinal mucosalinjury in chronic NSAID users. Clinical Gastroenterology and Hepatology2005;3:55–9.

[5] Sugimori S, Watanabe T, Tabuchi M, et al. Evaluation of small bowel injuryin patients with rheumatoid arthritis by capsule endoscopy: effects of anti-rheumatoid arthritis drugs. Digestion 2008;78:208–13.

[6] Morris AJ, Wasson LA, MacKenzie JF. Small bowel enteroscopy in undiagnosedgastrointestinal blood loss. Gut 1992;33:887–9.

[7] Smale S, Tibble J, Sigthorsson G, et al. Epidemiology and differential diagnosisof NSAID-induced injury to the mucosa of the small intestine. Best Practice andResearch Clinical Gastroenterology 2001;15:723–38.

[8] Adebayo D, Bjarnason I. Is non-steroidal anti-inflammaory drug (NSAID)enteropathy clinically more important than NSAID gastropathy. PostgraduateMedical Journal 2006;82:186–91.

[9] Wolfe MM, Lichtenstein DR, Singh G. Gastrointestinal toxicity of nonsteroidalantiinflammatory drugs. New England Journal of Medicine 1999;340:1888–99.

10] Lanza FL, Chan FK, Quigley EM. Guidelines for prevention of NSAID-related ulcercomplications. American Journal of Gastroenterology 2009;104:728–38.

11] Fujimori S, Seo T, Gudis K, et al. Prevention of nonsteroidal anti-inflammatorydrug-induced small-intestinal injury by prostaglandin: a pilot randomizedcontrolled trial evaluated by capsule endoscopy. Gastrointestinal Endoscopy2009;69:1339–46.

12] Watanabe T, Sugimori S, Kameda N, et al. Small bowel injury by low-doseenteric-coated aspirin and treatment with misoprostol: a pilot study. ClinicalGastroenterology and Hepatology 2008;6:1279–82.

13] Smecuol E, Pinto Sanchez MI, Suarez A, et al. Low-dose aspirin affects the smallbowel mucosa: results of a pilot study with a multidimensional assessment.Clinical Gastroenterology and Hepatology 2009;7:524–9.

14] Konaka A, Kato S, Tanaka A, et al. Roles of enterobacteria, nitric oxide and neu-trophil in pathogenesis of indomethacin-induced small intestinal lesions inrats. Pharmacological Research 1999;40:517–24.

15] Watanabe T, Higuchi K, Kobata A, et al. Non-steroidal anti-inflammatorydrug-induced small intestinal damage is Toll-like receptor 4 dependent. Gut2008;57:181–7.

16] Watanabe T, Nishio H, Tanigawa T, et al. Probiotic Lactobacillus casei strainShirota prevents indomethacin-induced small intestinal injury: involvementof lactic acid. American Journal of Physiology, Gastrointestinology and LiverPhysiology 2009;297:G506–13.

17] Wallace JL, Syer S, Denou E, et al. Proton pump inhibitors exacerbateNSAID-induced small intestinal injury by inducing dysbiosis. Gastroenterology2011;141:1314–22.

18] Lombardo L, Foti M, Ruggia O, et al. Increased incidence of small intestinalbacterial overgrowth during proton pump inhibitor therapy. Clinical Gastroen-terology and Hepatology 2010;8:504–8.

19] Thorens J, Froehlich F, Schwizer W, et al. Bacterial overgrowth during treatmentwith omeprazole compared with cimetidine: a prospective randomised doubleblind study. Gut 1996;39:54–9.

20] Goldstein JL, Eisen GM, Lewis B, et al. Small bowel mucosal injury is reduced inhealthy subjects treated with celecoxib compared with ibuprofen plus omepra-zole, as assessed by video capsule endoscopy. Alimentary Pharmacology andTherapeutics 2007;25:1211–22.

21] Hawkey CJ, Ell C, Simon B, et al. Less small-bowel injury with lumiracoxibcompared with naproxen plus omeprazole. Clinical Gastroenterology and Hep-

atology 2008;6:536–44.

22] Maiden L, Thjodleifsson B, Seigal A, et al. Long-term effects of nonsteroidalanti-inflammatory drugs and cyclooxygenase-2 selective agents on the smallbowel: a cross-sectional capsule enteroscopy study. Clinical Gastroenterologyand Hepatology 2007;5:1040–5.

nd Liv

[

[

[

T. Watanabe et al. / Digestive a

23] Laine L, Bombardier C, Hawkey CJ, et al. Stratifying the risk of NSAID-relatedupper gastrointestinal clinical events: results of a double-blind outcomes study

in patients with rheumatoid arthritis. Gastroenterology 2002;123:1006–12.

24] Miyake K, Kusunoki M, Shinji Y, et al. Bisphosphonate increases risk ofgastroduodenal ulcer in rheumatoid arthritis patients on long-term nonste-roidal antiinflammatory drug therapy. Journal of Gastroenterology 2009;44:113–20.

[

er Disease 45 (2013) 390– 395 395

25] Laine L, Curtis SP, Langman M, et al. Lower gastrointestinal events in a double-blind trial of the cyclo-oxygenase-2 selective inhibitor etoricoxib and the

traditional nonsteroidal anti-inflammatory drug diclofenac. Gastroenterology2008;135:1517–25.

26] Amagase K, Ochi A, Sugihara T, et al. Protective effect of lafutidine, a histamineH2 receptor antagonist, against loxoprofen-induced small intestinal lesions inrats. Journal of Gastroenterology and Hepatology 2010;25(Suppl. 1):S111–8.