sign in sign up
<<
Home , Colitis, Plesiomonas shigelloides, Ulcerative colitis

Article 1

see related editorial on page x Enteric Are Common in Patients with Flares of Infammatory Bowel Disease

Jordan E. Axelrad, MD, MPH1,2, Andrew Joelson, MD2, Peter H. R. Green, MD2, Garrett Lawlor, MD2, Simon Lichtiger, MD2, Ken Cadwell, PhD3,4 and Benjamin Lebwohl, MD, MS2

OBJECTIVES: Few studies have examined the role of non-Clostridium diffcile enteric infections in fares of infammatory bowel disease (IBD). Our objective was to investigate enteric detected by multiplex PCR stool testing in patients with IBD.

Methods: We performed a cross-sectional analysis of 9403 patients who underwent 13,231 stool tests with a gastrointestinal pathogen PCR panel during a diarrheal illness from March 2015 to May 2017. Our primary outcome was the presence of an infection. Secondary outcomes included endoscopic and histologic predictors of infection, and IBD outcomes following testing.

Results: A total of 277 patients with Crohn’s disease (CD), 300 patients with ulcerative (UC), and 8826 patients without IBD underwent 454, 503, and 12,275 tests, respectively. Compared to patients without IBD, patients with IBD were less likely to test positive (CD 18.1%, UC 16.1%, no IBD 26.6%,

p < 0.001). Compared to patients without IBD, CD had a higher prevalence of norovirus (p = 0.05) Bowel D i sease Inflammatory and Campylobacter (p = 0.043), whereas UC had a lower prevalence of norovirus (p = 0.001) and a higher prevalence of Campylobacter (p = 0.013), Plesiomonas (p = 0.049), and species (p < 0.001). Of 77 patients who underwent endoscopy, there were no major endoscopic or histologic predictors of a positive test. Patients who tested negative were more likely to have IBD therapy escalated (p = 0.004). Enteric infection did not impact IBD outcomes following testing (log-rank 0.224).

Conclusions: Non-Clostridium diffcile enteric infections were identifed in 17% of symptomatic patients with IBD. Endoscopic and histologic fndings may not differentiate fare from infection. Norovirus and E.coli may play an important role in fare of IBD. SUPPLEMENTARY MATERIAL accompanies this paper at https://doi.org/10.1038/s41395-018-0211-8

Am J Gastroenterol https://doi.org/10.1038/s41395-018-0211-8

Introduction [8, 11, 12]. IBD patients with Infammatory bowel disease (IBD), comprising Crohn’s disease Clostridium difcile (C. difcile) infection had more pronounced dys- (CD) and (UC), develops from a combination biosis and signifcantly worse clinical outcomes compared to patients of genetic susceptibility and environmental factors that elicit an without IBD, including longer hospital stays, higher deleterious infammatory response [1]. Te intestinal microbiota rates, higher recurrence rates, and increased mortality [7, 13–16]. regulates mucosal immunity through a number of pathways and In addition, several studies have implicated enteric infection in is thought to be a major environmental factor in the relapse of known IBD, including viral pathogens [7, 17–20]. pathogenesis and maintenance of IBD [2–5]. Enteric infection is a In a recent cross-sectional study on stool polymerase chain common cause of dysbiosis and is frequently identifed in patients reaction (PCR) testing during an exacerbation of symptoms in with IBD [6]. patients with known IBD, we identifed enteric infection in 26.8% Several observational studies have demonstrated a link between of tests with C. difcile, the most common (12.9%) followed by enteric infection, functionally altered commensal , and the Escherichia coli (E.coli) species (8.1%) and viruses (5.1%) [7]. subsequent development of IBD [7–10]. An increased risk of IBD Tis study, however, was limited by lack of a control group consist- was observed in patients with a previous episode of Salmonella or ing of patients without IBD, meaningful measures of clinical IBD

1Infammatory Bowel Disease Center, Division of , Department of Medicine, NYU Langone Health, New York, NY, USA. 2Division of Digestive and , Department of Medicine, Columbia University Medical Center, New York, NY, USA. 3Kimmel Center for Biology and Medicine at the Skirball Institute, New York University School of Medicine, New York, NY, USA. 4Department of Microbiology, New York University School of Medicine, New York, NY, USA. Correspondence: J.E.A. (email: [email protected]) Received 8 February 2018; accepted 29 June 2018

© 2018 The American College of Gastroenterology The American Journal of Gastroenterology Inflammatory Bowel Disease 2 The Campylobacter stool and including 4parasites 5viruses, 13bacteria, including Te gastrointestinal pathogen panel PCRtests for 22 analytes in Enteric pathogentesting patients were correctly classifed. PCRtestcodes, dates, and results. sampled Ofthose patients, all to confrmassessed that identifed records had correct diagnoses with IBD and arandom sample of 25patients without IBD were of aPCRtest were classifed as IBD patients. 577patients All patient were excluded. Patients with IBD within 7days diagnosed tests within 6months and any repeat positive tests on same the diagnosis, and IBD subtype. Repeat gastrointestinal pathogen PCR International Classifcation (ICD) of date coding, Diseases of IBD zation, and ), race, ethnicity, sex, presence of IBD using (e.g., emergency department, outpatient visit, inpatient hospitali- of PCRtest, PCRresults, date of zip birth, place code, of PCRtest spanning March 2015through May 2017. UT) during an of episode during 26-month the period pathogentinal PCR panel (BioFire City, Diagnostics, Salt Lake tients underwent stool who testing with aFilmArray gastrointes- from more distant regions. We outpatients identifed all and inpa- tristate region (NY, NJ, and CT),as well as care people seeking patientsCity that from serves surrounding the urban area, the versity Medical Center, a quaternary care institution in New York recordscal of patients at New York Presbyterian-Columbia Uni- We across-sectional performed study using electronic the - medi Study populationandvariables Methods [ tests aslogical arapid and accurate means of approaching diarrhea multiplex assays have to started replace conventional microbio- unknown. In addition, sensitive highly and molecular specifc diferentiating non-C. difcile enteric infection from fare remain tion. As such, importance clinical the and diagnostic approach to ist as a complicating factor, or represent asymptomatic coloniza- formay an etiology sole the exacerbation be insymptoms,- coex substantial to barriers diagnosis and treatment. Enteric infection and laboratory fndings inrelapse IBD and enteric infection pose patients without IBD. In addition, presentations similar clinical the distributionthe of infections inpatients with IBD compared to infection inpatients with fare of IBD, little is known regarding fare from enteric infection. outcomes, and endoscopic or histologic fndings to help elucidate Axelrad etal. patients with and without enteric infection. histologic associations, and outcomes, clinical insymptomatic IBD objective was toTe compare secondary clinical, endoscopic, and testing insymptomatic patients with CD, UC,and without IBD. non- isms not previously and readily diagnosable by clinician. the 21, We recorded following the values from record: medical the Date Despite growing research regarding prevalence the of enteric Te objective of study this was to evaluate distribution the of American Journal American ]. Tese assays22]. Tese allow for identifcation the of organ specifc - C. difcile enteric infections by detected multiplex PCRstool (jejuni, coli, and upsaliensis of Gastroenterology ), Clostridium difcile Clostridium

did not diferentiate fndings that would suggest enteric infection discretion at individual the gastroenterologist. In addition, we to IBD patients underwent stool who testing, was this done at the tural distortion, and presence the of . ity of infammation, including , chronic, and mixed, architec- reportspathology for documentation the of , colitis, chronic- or andfocal difuse, abiopsy whether was obtained. We assessed withtis severity, , focality of endoscopic infammation, either weendoscopy assessed tract, reports for presence the of ileitis, coli- H.pylori. without tritis chronic active without H.pylori, and chronic inactive gas - tion of , chronic active gastritis with was obtained. We reports pathology for assessed nota the - erosions,as , or hemorrhage), petechial and a whether gastropathytis, withlocation, gastric duodenopathy (as defned weendoscopy assessed tract, reports for presence the of esophagi- and length of stay. doscopy, histologic fndings, requirement for hospitalization, denoscopy (EGD),ileoscopy, , or fexible sigmoi- 2 weeks of testing, endoscopic fndings on esophagogastroduo- tomography (CT)or magnetic resonance (MR)imaging within C-reactive protein, active infammation noted on computed testing sedimentation including rate erythrocyte (ESR) and days before and at PCRtesting, infammatory markers at PCR current of use aproton pump inhibitor, antibiotic within 90 use biologics, and other exposures such as previous and (5-ASA) agents,salicylate , immunomodulators, duration of IBD, medication exposures, including 5-Amino- testinal pathogen panel PCRtest, IBD we phenotype, collected evaluation within 30days before or afer submitting agastroin- In asubset of patients with IBD underwent endoscopic who Endoscopic evaluation Cryptosporidium coli enterotoxigenic oaggregative ytica, Giardiaytica, lamblia colitica (Toxin shigelloides A/B),Plesiomonas ity is 94.5–100%for targets all [22,23]. cess approximately takes 1 samples Blair transport Te media. multiplex inCary PCRpro - from multiple and parasites viruses, bacteria, directly from stool simultaneousthe detection and identifcation of nucleic acids excluded. Te gastrointestinal pathogen panel PCRis capable of PCR within 7days of agastrointestinal pathogen panel PCRwere were not examined study. inthis Patients with apositive C.difcile not reported with gastrointestinal the pathogen panel PCRand C. difcile Sunnyvale,, Cephid, CA)and as such, results these are institution, an alternative PCRtest isfor utilized C.difcile (Xpert ovirus GI/GII,rotavirus A,and sapovirus (I,II,IV, and In V). our As there was no established protocol for endoscopic the approach For endoscopic and fndings on examination of lower the GI For endoscopic and fndings on examination of upper the GI (STEC), E.coli O157,Shigella/enteroinvasive E. coli (EIEC), , Vibrio (parahaemolyticus, vulnifcus, andcholerae E. coli (EAEC),enteropathogenic E. coli (EPEC), spp., Cyclospora cayetanensis E. coli (ETEC), Shiga-like toxin-producing E. , adenovirus (AdV) F40/41,, nor h. Te sensitivity clinical and- specifc , Salmonella, Yersinia entero- , pylori Helicobacter Entamoeba histol www.nature.com/ajg ), enter - - - ,

infection (p IBD, patients with IBD were less likely to test positive for an enteric were positive for 4431pathogens. Compared to patients without total of 115pathogens. In patients without IBD, 3263(26.6%)tests © 2018The AmericanCollegeof Gastroenterology (Table CD, 82(18.1%)tests were positive for atotal of 122 pathogens pathogen panel PCRand were excluded. Among patients with C. difcile PCRsubmitted at same the as time agastrointestinal patients with IBD and 461patients without IBD had apositive 503, and 12,275 stool tests, respectively (Table with UC,and 8826patients without IBD underwent 454, who of diarrhea, wesode identifed 277patients with CD, 300patients tests with a gastrointestinal pathogen PCR panel during an epi- Of 9403inpatients and outpatients underwent 13,231stool who Results Columbia University Medical Center Institutional Board. Review statisticalto analyses.perform all Te study was approved by the a two-sided pvalue less than 0.05.SPSS sofware (IBM)was used and using log-rank the test. tests All were considered signifcant at pathogentinal PCR panel by constructing Kaplan–Meier curves We evaluated to time IBD-related outcomes afer agastrointes- ings inpatients with IBD and presence the of enteric infections. test for association the endoscopic between and histologic fnd- t-test for continuous variables. We Chi-square used analysis to results Chi-square the via test for categorical variables and the ables, presence the of IBD, and IBD subtype with PCRstool test and IBD subtype. We measured for associations vari - between intestinal pathogen PCRpanel according to presence the of IBD outcome primary Our was presence the of any positive gastro- Outcomes andstatisticalanalyses intervention.cal cation (e.g., stricture, fstula, obstruction, and abscess), and surgi- visit, steroids prescription, other IBD therapy escalation, compli- related data including hospitalization, emergency department evaluation, end of study the , or period, following the IBD- frst stool PCRtesting with patients censored at loss to follow-up ing. Follow-up for time patients all from was of time accrued the and change inIBD management afer results the of PCRtest- We data collected regarding treatment for positive PCRresults follow-up (through April 30,2018)to assess course their of IBD. symptom resolution until last the date of outpatient or inpatient patientsAll with IBD were retrospectively followed afer initial IBD outcomesfollowingentericpathogentesting and reports pathology for analysis. versus fare, but rather, recorded data all reported inendoscopy vs 30.6years for non-IBD, p with IBD were older without than those IBD (median 33.0for CD rus (17.4% vs24.6%,p rus without IBD, patients with CDhadprevalence ahigher of norovi- p

= In terms of distribution the of infections, compared to patients

0.043), and alower prevalence of parasites (6.1%vs0.8%, 2 ). In patients with UC, 81 (16.1%)tests were positive for a

<

0.001). Among with apositive those test, patients

= 0.05) and 0.05)

=

0.045; 49.6for UC,p Campylobacter (7.6% vs 13.1%, (7.6%vs13.1%, 1). Forty-eight

< 0.001).

(7.8% vs17.4%,p asites compared to patients without IBD (0.9%vs6.1%,p (82.6% vs62.1%,p and viruses (16.5%vs31.9%,p and viruses p 12.4%, positive testing was more common in patients with CD (20.7% vs vs 7.6%,p genic gative E. coli (64.3%vs47.6%,p species p (16.5% vs32.8%,p had alower prevalence compared of viruses to patients with CD Campylobacter norovirus, and parasites, and specifcally of odds ahigher bacteria, parasites; patients with UChad alower specifcally of odds viruses, patients of with CDhadodds ahigher norovirus and lower of odds ( ( the severity of severity the colitis (p ( Norovirus (5,15.2%;Supplementary Table positive for 33enteric pathogens. E.coli (14,42.4%) and species days of agastrointestinal pathogen PCRpanel, 26(33.8%)tested pies, 61colonoscopies, and 13fexible sigmoidoscopies within 30 ing gastropathy (p were no other endoscopic predictors of apositive test includ- in patients with an infection (44.4%vs7.7%,p Table in age (p infection underwent endoscopic who evaluation were similar mon pathogens detected. Patients with and without an enteric tis ( tis underwent biopsy,who no histologic fnding, including esophagi- distortion (p with and 460IBD patients without an enteric infection (log-rank, outcome including above the variables 117IBD patients between IBD (log-rank, (log-rank, 0.575),or 0.765),surgery acomposite other IBD therapy escalation (log-rank, 0.221), of ment visit (log-rank, 0.777),steroid prescription (log-rank, 0.968), IBD-related hospitalization (log-rank, 0.268),emergency- depart resolution of fare,initial the there were no diferences in totime (81.3% vs50.0%;p IBD held or no changes to IBD their management positive gastrointestinal pathogen panel were more likely to have oradded up-titrated (50.0%vs18.6%),whereas patients with a pathogen PCRpanel were more likely to have IBD medications antimicrobial therapy. IBD patients with a negative gastrointestinal test. tested Of87who positive for 46(52.9%)received abacteria, and bacteria. viruses between with apositive pathogen test result. fndings Tese did not difer p p p

= <

Of 77 patients underwent who 22 upper ,- 2 ileosco During amedianDuring follow-up of 10.5months (range, 0–36.3)afer Of 577patients with IBD, 117tested positive on frst the stool = = = 0.01, Table 0.01,

0.001). Afer adjusting for age, race, ethnicity, sex, and season, p

0.998), colitis (p 0.693), or focality of the colitis (p 0.43), colitis (p

= E. coli (EPEC; 33.9%vs22.1%,p 3 E. coli (EAEC;20.9%vs13.5%,p p ). Endoscopic fndings of esophagitis were more common

0.795), gastritis (p

=

=

0.041). Patients with UChad alower prevalence of par =

0.601), sex (p 0.601), sex 0.013),

= 2 , and E.coli (Supplementary species Table 1

0.65), or granulomata (p ). Among patients with apositive test result, repeat enteric infectionininflammatory Bowel Disease

= = =

=

= <

0.001).

0.004), specifcally norovirus0.004), specifcally (7.8%vs24.6%, Plesiomonas 0.019), and prevalence ahigher of bacteria =

0.58), the distribution0.58), the of colitis (p

0.001), specifcally specifcally 0.001), 0.251), duodenopathy (p

The 0.082), chronicity (p

=

American Journal American

= =

0.273), the presence0.273), the of colonic ulcers

0.084), and IBD subtype (p 0.106), (p

(2.6%vs0.7%,p <

=

0.001), including enteroaggre-

0.007), specifcally norovirus0.007), specifcally

=

= = 0.693; Table 0.693;

0.004). Patients with UC 0.028), and enteropatho-

= of Campylobacter 2

Gastroenterology 0.46) were associated

) were most the com- = .1, architectural 0.21),

=

0.043), but there

=

=

.9) ileitis 0.595), = 4 .4) ileitis 0.142), ). In patients .4) and 0.049),

=

=

= (13.9% (13.9% ). 0.585), 0.585), 0.014) 0.014) 0.806; 0.806; -

3 Inflammatory Bowel Disease Inflammatory Bowel Disease 4 Non-Hispanic Asian/Pacifc Race Female Male Sex Winter Fall Spring Summer Season

Emergency Inpatient Outpatient Place oftest

50–69 n ew York City Residential zipcode 30–49 Non-Hispanic Non-Hispanic 18–29 Other/unknown American Median ageat Mean ageat Hispanic Hispanic ethnicity Age group Other Surrounding White Black test(range) test (years) stool tests Number of during anepisodeofdiarrhea Table 1 c b a Crohn’s diseasecompared toulcerativecolitis Crohn’s diseasecomparedtonoinfammatory boweldisease Ulcerative colitiscompared tonoinfammatoryboweldisease The Axelrad etal. Islander Endoscopy unit room >70 <18 Indian/Alaskan Tri-State area American Journal American Characteristics of9403inpatientsandoutpatientswhounderwent13,231stooltestswithagastrointestinalpathogenPCRpanel

566 6266 6008 3175 2771 3545 2783 147 1336 7242 3549 2132 3660 7580 2180 10087 1319 5884 2187 2983 250 4444 20 4477 1327 47.3 42.9 12,275 (0–102) (n el disease matory bow No infam Total = 8826) of

Gastroenterology - - 423 (74.7%) 4710 (75.2%) 4301 (71.6%) 2281 (71.8%) 2093 (75.5%) 2674 (75.4%) 1963 (70.5%) 114 (77.6%) 1019 (76.3%) 5287 (73%) 2591 (73%) 1754 (82.3%) 2958 (80.8%) 5400 (71.2%) 1613 (74%) 7472 (74.1%) 925 (70.1%) 4092 (69.5%) 1539 (70.4%) 1761 (59%) 203 (81.2%) 3408 (76.7%) 10 (50%) 3436 (76.7%) 1050 (79.1%) 46.3 9011 (73.4%) 51.7 (0–102) PCR negative

143 (25.3%) 1556 (24.8%) 1707 (28.4%) 894 (28.2%) 678 (24.5%) 817 (24.6%) 820 (29.5%) 33 (22.4%) 317 (23.7%) 1955 (27%) 958 (27%) 378 (17.7%) 702 (19.2%) 2180 (28.8%) 567 (26%) 2615 (25.9%) 394 (29.9%) 1792 (30.5%) 648 (29.6%) 1222 (41%) 47 (18.8%) 10 (50%) 1041 (23.3%) 277 (20.9%) 33.7 3263 (26.6%) 30.6 (0–100) 1036 (23.3%) PCR positive 11 222 232 104 122 124 104 5 30 271 148 42 98 262 197 338 110 173 116 97 0 230 40 37.7 454 12 (3–92) 31.6 Total 180 277) (n disease Crohn’s =

7 (63.6%) 185 (82.9%) 188 (81%) 90 (86.5%) 95 (77.9%) 100 (80.6%) 87 (83.7%) 4 (80%) 21 (70%) 229 (84.5%) 118 (79.7%) 37 (88.1%) 78 (79.6%) 214 (81.7%) 82 (76.6%) 284 (84%) 96 (87.3%) 135 (78%) 88 (75.9%) 79 (81.4%) 11 (91.7%) 0 194 (84.3%) 36 (90%) 37.8 372 (81.9%) 31.3 (3–92) PCR 147 (81.7%) negative

4 (36.4%) 38 (17.1%) 44 (19%) 14 (13.5%) 27 (22.1%) 24 (19.4%) 17 (16.3%) 1 (20%) 9 (30%) 42 (15.5%) 30 (20.3%) 5 (11.9%) 20 (20.4%) 48 (18.3%) 25 (23.4%) 54 (16%) 14 (12.7%) 38 (22%) 28 (24.1%) 18 (18.6%) 1 (8.3%) 0 36 (15.7%) 4 (10%) 37 82 (18.1%) 33 (3–81) 33 (18.3%) PCR positive

0.888 0.362 0.459 0.002 0.002 0.185 0.001 0.045 0.266 value p a 4 289 214 110 115 132 146 5 50 274 174 86 165 276 108 418 87 223 85 57 503 0 232 44 47.1 (2–93) 49.7 219 8 Ulcerative Total (n colitis = 300)

3 (75%) 249 (86.2%) 173 (80.8%) 98 (89.1%) 96 (83.5%) 109 (82.6%) 119 (81.5%) 5 (100%) 41 (82%) 231 (84.3%) 145 (83.3%) 76 (88.4%) 135 (81.8%) 229 (83%) 90 (83.3%) 358 (85.6%) 72 (82.8%) 190 (85.2%) 64 (75.3%) 49 (86%) 422 (83.9%) 0 189 (81.5%) 40 (90.9%) 47.3 50.1 (2–93) PCR 186 (84.9%) 7 (87.5%) negative

1 (25%) 40 (13.8%) 41 (19.2%) 12 (10.9%) 19 (16.5%) 23 (17.4%) 27 (18.5%) 0 9 (18%) 43 (15.7%) 29 (15.7%) 10 (11.6%) 30 (18.2%) 47 (17%) 18 (16.7%) 60 (14.4%) 15 (17.2%) 33 (14.8%) 21 (24.7%) 8 (14%) 81 (16.1%) 0 43 (18.5%) 4 (9.1%) 45.9 49.6 (8–90) 33 (15.1%) 1 (12.5%) PCR positive

www.nature.com/ajg 0.842 0.392 0.439 0.001 0.227 0.001 0.002 0.001 value p 0.229 b 0.119 0.177 0.359 0.251 0.920 0.498 0.094 0.001 value p 0.632

c Figure 1 antimicrobial therapy (41,47.1%;log-rank, 0.442; Supplementary follow-upthe compared period to patients didnot who receive (46, 52.9%) were asequally likely to remain in remission during positive received who for those antimicrobial abacteria, therapy © 2018The AmericanCollegeof Gastroenterology tests in patients with CD, UC, and without IBD, respectively. enteric infection in18.1%,16.1%,and was detected 26.8%of a gastrointestinal pathogen panel PCRtest, non-C. difcile In cross-sectional this analysis of patients underwent who Discussion 0.224; Fig. Table 2 c b a Total pathogensidentifed Viruses

Astrovirus norovirusGI/GII

Bacteria

Enteroaggregative Enteropathogenic Enterotoxigenic  E. coli0157 Shigella/enteroinvasive Parasites Cryptosporidium

Multiple pathogens Crohn’s diseasecomparedtoulcerativecolitis Crohn’s diseasecomparedtonoinfammatorybowel Ulcerative colitiscomparedtonoinfammatoryboweldisease Adenovirus F40/41 Rotavirus A Sapovirus (I,II,IV, V) Campylobacter species Plesiomonas shigelloides Salmonella species Vibrio species Escherichia colispecies Cyclospora cayetanesis Giardia lamblia STX/ST2 Shiga-like toxin-producing ). Distribution ofpathogensinpatientswithCrohn’s disease,ulcerativecolitis,andwithoutIBD 1 ). In exploratory analyses, of 87 patients tested who E. coli(L E. coli(EPEC) E. coli(EAEC) E. coli(EIEC) T/ST) E. coli No IBD 4431 1412 (31.9%) 106 (2.4%) 113 (2.6%) 771 (17.4%) 204 (4.6%) 218 (4.9%) 2750 (62.1%) 338 (7.6%) 30 (0.7%) 166 (3.7%) 93 (2.1%) 12 (0.3%) 4 (0.1%) 2107 (47.6%) 598 (13.5%) 981 (22.1%) 186 (4.2%) 147 (3.3%) 24 (0.5%) 171 (3.9%) 269 (6.1%) 107 (2.4%) 18 (0.4%) 2 (0.1%) 142 (3.2%) 871/3263 (26.7%) Crohn’s disease 122 40 (32.8%) 2 (1.6%) 1 (0.8%) 30 (24.6%) 3 (2.5%) 4 (3.3%) 81 (66.4%) 16 (13.1%) 0 5 (4.1%) 1 (0.8%) 1 (0.8%) 1 (0.8%) 57 (46.7%) 17 (13.9%) 28 (23%) 4 (3.3%) 3 (2.5%) 1 (0.8%) 4 (3.3%) 1 (0.8%) 0 0 0 1 (0.8%) 29/82 (35.4%)

CD, UC, and without IBD. In CD, distribution of infections symptomatic between patients with UC [25]. ranging from 0.88%inpatients with CDto 2.5%inpatients with infection using limited and heterogeneous diagnostic methods, data suggesting amuch lower rate of non-C. difcile bacterial utilizing amultiplex PCRpanel, difer signifcantly from recent than in symptomatic patients without IBD [7, mon in symptomatic patients with IBD, understandably lower reported pilot data demonstrating that enteric infection was com- of IBD from enteric infection. Tis study confrms previously Endoscopic and histologic fndings didnot diferentiate fare In present the study, there were signifcant diferences inthe p value 0.844 0.999 0.373 0.05 0.279 0.427 0.639 0.043 0.999 0.805 0.520 0.298 0.127 0.920 0.920 0.862 0.632 0.799 0.494 0.999 0.010 0.119 0.999 0.999 0.594 0.100 enteric infectionininflammatory Bowel Disease a Ulcerative colitis 115 19 (16.5%) 0 2 (1.7%) 9 (7.8%) 4 (3.5%) 4 (3.5%) 95 (82.6%) 16 (13.9%) 3 (2.6%) 0 2 (1.7%) 0 0 74 (64.3%) 24 (20.9%) 39 (33.9%) 4 (3.5%) 4 (3.5%) 0 3 (2.6%) 1 (0.9%) 0 0 0 1 (0.9%) 23/81 (28.4%) The American Journal American norovirus of p value 0.007 0.115 0.999 0.019 0.584 0.498 0.001 0.013 0.049 0.022 0.999 0.999 0.999 0.001 0.028 0.004 0.999 0.797 0.999 0.803 0.014 0.115 0.999 0.999 0.269 0.791 Gastroenterology and and ]. Our fndings,24]. Our b Campylobacter p value 0.004 0.499 0.614 0.001 0.716 0.999 0.276 0.888 0.117 0.061 0.614 0.999 0.999 0.157 0.524 0.310 0.999 0.716 0.999 0.999 0.999 — — — 0.999 0.527 c

5 Inflammatory Bowel Disease Inflammatory Bowel Disease 6 The Axelrad etal. PCR stooltestingandendoscopy Table 3

no activeinfam- CT orMRimagingattesting

Infammatory markersattesting Antibiotics

Immunomodulator Biologics Immunomodulators Corticosteroids 5-ASA Medications attesting niitc exposure Antibiotic  Previous   Other/unknown Asian Black Hispanic Caucasian Race/ethnicity Female Male Gender n colonic Any sltd ileal/upper Isolated IBD phenotype ulcerativecolitis IBD subtype within 90days colectomy Previous partialortotal diffcile infection (years) Average ageattest at test(years) Average durationofIBD American Journal American Active infammation mation Median C-RP Median ESR Elevated ESRorC-RP Proton pumpinhibitor with biologic involvement GI only Crohn’s disease Characteristics of77patientswithIBDwhounderwentGI Clostridium of Gastroenterology 24/28 (85.7%) 4/28 (14.3%) 25.9 35 41 (80.4%) 5 (9.8%) 6 (11.8%) 2 (3.9%) 8 (15.7%) 5 (9.8%) 9 (17.6%) 16 (31.4%) 7 (13.7%) 8 (15.7%) 3 (5.9%) 38.7 6.8 2 (3.9%) 0 5 (9.%) 15 (29.4%) 29 (56.9%) 22 (43.1%) 29 (56.9%) 48 (94.1%) 3 (5.9%) 26 (51%) 25 (49%) testn pathogen PCR gastrointestinal Negative ± 8.7 ± 21.8 = 51 11/15 (73.3%) 4/15 (26.7%) 29.2 43 18 (81.8%) 3 (11.5%) 4 (15.4%) 1 (3.8%) 3 (11.5%) 3 (11.5%) 9 (34.6%) 8 (30.8%) 7 (26.9%) 3 (11.5%) 4 (15.4%) 36.2 6.7 2 (7.7%) 1 (3.8%) 1 (3.8%) 10 (38.5%) 12 (46.2%) 6 (23.1%) 20 (76.9%) 22 (84.6%) 4 (15.4%) 14 (53.8%) 12 (46.2%) testn pathogen PCR trointestinal Positive gas- ± 8.2 ± 17.1 = 26 0.320 0.223 0.480 0.887 0.814 0.655 0.987 0.623 0.814 0.096 0.957 0.156 0.623 0.170 0.601 0.960 0.403 0.084 0.170 0.806 p value patients with IBD fare complicated by Campylobacter therapy. Previous studies have demonstrated worse outcomes in IBD and none have evaluated impact the of targeted antimicrobial signifcance clinical the ined of enteric infection inpatients with pies or added escalated [7 patients with an enteric infection are less likely to have IBD- thera tion testing signifcantly impacts IBD management such that patients without IBD. prevalence of E.coli EPEC and species EAECinUCcompared to patients with UCis limited. We are frst the to report an increased tributing factor [35].In addition, data regardingE. coli infection in regionsparticular of gastrointestinal the may acon be tract - also TH1 signature compared with UC[31– associationselective with CD, typically which displays astronger TH1 immune responses provides apotential explanation for the UC remains unclear. Te ability of norovirus infection to induce although signifcance the of prevalence decreased the in of viruses for evidence role the further of enteric the virome infares of CD, ATG16L1 produces CDinmice[27, ting of apolymorphism CDsusceptibility inthe autophagy gene licated inanimal models where norovirus- infection set inthe but inexacerbations also of IBD. fndings Tese have rep- been implicatedovirus has been not only pathogenesis in the of IBD, pathogenesis inthe of viruses of CD[26– response inpatients with UC. may play an important role in modulating mucosal the immune pathogens,patients bacterial such with CDwhile as E.coli species, an important role inmodulating mucosal the immune response in fndingsTese suggest such that as norovirus, viruses, may play nella overall, including Campylobacter were more common, were whereas inUC,bacteria more common cations, and surgical interventions. In exploratory analyses, our its, steroids prescriptions, other IBD therapy escalations, compli- infection, including hospitalizations, emergency department vis- symptom resolution patients between with and without enteric no diference in relevant long-term IBD outcomes afer initial potential pathogens. scopic, and histologic fndings, and PCR stool testing may identify infection and non-infectious fare may elicit similar clinical, endo- by of use the immunosuppression results [38].Our suggest that of endoscopic pseudomembranes low to be and not accounted for C. difcile inpatients with IBD, where studies have shown rate the data may consistent be with existing endoscopic literature on infection. Although our sample and was small non-random, these histologic fndings did not diferentiate fare of IBD from enteric underwent endoscopy near PCRtesting. endoscopic, Clinical, and tion inIBD, we examined data from a subset of patients who non-infectious fare [25]. in remission within 1year compared with C.difcile to those or patients with non-C As we and other groups have previously reported, enteric infec- Tere have studies several examining been potential the role In addition to endoscopic and histologic fndings, we found In an efort to evaluate signifcance clinical the of enteric infec- [ 36, 37]. Conversely, more recent data have suggested that . difcile infection are more likely to remain , 25]. However, few studies have exam- , Plesiomonas ]. Our fndings31]. Our may yield 34]. Norovirus tropism for 30]. In particular, nor , and E.coli species. www.nature.com/ajg and Salmo and - -

© 2018The AmericanCollegeof Gastroenterology IBD whounderwentGIPCRstooltestingandendoscopy Table 4 Focality Right-sided colitis Pancolitis Location Severe Moderate Mild Severity Colitis Lower GItract Ileitis    Duodenitis Gastritis Esophagitis histologicfnding Upper GItract Duodenopathy Entire stomach Antrum Body Cardia/fundus Gastropathy Upper GItract Esophagitis (range) from GIPCRtoendoscopy Median numberofdays Colonoscopy Ileoscopy upperendoscopy Endoscopic procedures Endoscopic fndings Endoscopic fndings Flexible gastritis withHP Chronic active gastritis Chronic active gastritis Chronic inactive Left-sided colitis Endoscopic andhistologicfndingsin77patientswith

8/45 (17.8%) 16/45 (35.6%) 21/45 (46.7%) 18/45 (40%) 18/45 (40%) 9/45 (20%) 21/45 (46.7%) 45/51 (88.2%) 11/51 (21.6%) 51 0 2/9 (22.2%) 6/11 (54.5%) 7/9 (77.8%) 9/11 (81.8%) 3/11 (50.3%) 11 3/13 (23.1%) 4/7 (57.1%) 2/7 (28.6%) 1/7 (14.3%) 0 7/13 (53.8%) 1/13 (7.7%) 13 2.5 (0–23) 44 (86.3%) 1 (2%) 13 (25.5%) 6 (11.8%) testn pathogen PCR trointestinal Negative gas- = 51 7/24 (29.2%) 9/24 (37.5%) 14/24 (58.3%) 5/24 (20.8%) 13/24 (54.2%) 6/24 (25%) 8/24 (33.3%) 24/25 (96%) 9/25 (36%) 25 2/5 (40%) 0 2/9 (22.2%) 3/5 (60%) 5/9 (55.6%) 2/9 (22.2%) 9 3/9 (33.3%) 3/7 (42.9%) 4/7 (57.1%) 0 0 7/9 (77.8%) 4/9 (44.4%) 9 5.9 (0–29) 17 (65.4%) 1 (3.8%) 9 (34.6%) 7 (26.9%) testn pathogen PCR trointestinal Positive gas- = 26 0.693 0.508 0.585 0.580 0.273 0.431 0.142 0.106 0.795 0.595 0.364 0.251 0.043 0.745 0.153 value p symptomatic patients with IBD, and distribution the of pathogens procedures. and other factors may have infuenced decision-making inthese endoscopy and biopsy were inanon-random performed manner, tution, we were unable to analyze C.difcile infection. Moreover, ing importance inIBD, and given testing constraints inour insti- presencethe of apathogen (CMV), of increas- mArray gastrointestinal pathogen PCRpanel not does assess for probabilitythe of asymptomatic carriage is likely low.- Te Fil multiplex assays [39].However, asis this asymptomatic cohort, regarding interpretation clinical and cost-efectiveness of such asymptomatic colonization, and there is considerable uncertainty PCR testing fails to discriminate active infection and between ity of patients resided Northeast inthe . In addition, population and was ethnically geographically diverse, major the and as such, follow-up was time limited. Although patient the manuscript. Te PCRtest available became to clinicians in2015 distribution inthis infections of detected, as described particular however, we donot would this believe infuence particularly the populationsin specifc may infuence rates of overall infection; went endoscopic evaluation. In addition, thresholds for testing analysis.full Tere may bias selection inpatients be under who tions, and management afer stool testing was not available for exposures, recent travel, behavior, sexual other comorbid condi- mation concerning precise presenting symptoms, medication symptoms of IBD, and enteric infections. Individual patient infor and-efect relationship diarrhea, an between exacerbation in a retrospective study analyses Our donot design. prove acause- therapeutic beneft. data Tese require study. further ing a non-C. difcile infectious source may not ofer any long-term infection received who antimicrobial therapy suggests that target- fnding of equivalent outcomes inpatients enteric with bacterial Granulomas Architecture distortion Acute and chronic Chronic Acute Chronicity Colitis Ileitis histologicfndings Lower GItract Diffuse Focal In summary, enteric infection was common, identifed in17%of Tere are limitations several to current the study inherent to enteric infectionininflammatory Bowel Disease The American Journal American 4/46 (8.7%) 21/46 (45.7%) 11/46 (23.9%) 29/46 (63%) 6/46 (13%) 46/47 (97.9%) 11/11 (100%) 28/45 (62.2%) 17/45 (37.8%) testn pathogen PCR trointestinal Negative gas- = 51 of 1/22 (4.5%) 10/22 (45.5%) 3/22 (13.6%) 14/22 (63.6%) 5/22 (22.7%) 22/25 (88%) 9/9 (100%) 13/24 (54.2%) 11/24 (45.8%) Gastroenterology testn pathogen PCR trointestinal Positive gas- = 26 0.460 0.646 0.082 0.211 0.998 value p - - - 7 Inflammatory Bowel Disease Inflammatory Bowel Disease 8 The comes, however, more robust data is lacking regarding clinical the enteric infection didnot appear to impact long-term IBD out- apparent relapse of IBD. Te identifcation and treatment of an ing should considered be as adiagnostic step inpatients with an didnot diferentiateogy fare from infection. As such, PCRtest- PCR testing impacted IBD management, endoscopy and histol- difered signifcantly from patients without IBD. Although broad complication ofIBD(elog-rank0.765),surgeryf0.575),oracompositeoutcomeallendpoints g log-rank0.224) (a log-rank0.268),emergencydepartmentvisitb0.777),steroidprescription c log-rank0.968),otherIBDtherapyescalation(d0.221), Fig. 1 Axelrad etal. No infection No infection No infection American Journal American g d ab Infection Infection Infection Proportion in IBD remission Proportion free from IBD therapy escalation Proportion free from hospitalization Time toIBD-relatedoutcomesafterresolutionoftheinitialfarebetweenpatientswith andwithoutentericinfectionincludingtimetohospitalization 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 117 460 117 460 117 460 0 0 0 Time toIBDtherapyescalation(days) 227 200 221 200 200 158 58 57 43 Time tohospitalization(days) of Gastroenterology 145 400 143 400 400 39 34 81 21 Time (days) 600 600 600 20 82 17 82 44 9 Enteric infectio No entericinfectio Enteric infectio No entericinfectio Enteric infectio No entericinfectio 800 800 800 34 36 19 7 9 4 P =0.224 P =0.268 P =0.221 n n n n n n 1000 1000 1000 1 3 1 0 0 1 No infection No infection ef Infection Infection Proportion free from IBD complication Proportion free from emergency department visit 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 117 460 117 460 0 0 Time toemergencydepartmentvisit(days) 200 260 200 237 66 62 Time tocomplicationofIBD(days) 400 183 400 160 43 35 strategies and outcomes. coursethe of IBD, and impact the GI PCR testing on treatment is required to evaluate impact the of various enteric infections on and phenotypic the presentations of IBD. In addition, study further interaction enteric specifc between infections, mucosal immunity, in patients with IBD. Further study is required to investigate the relevance and outcomes of non- specifc C. difcile enteric infection 600 110 600 24 15 84 Enteric infection No entericinfectio Enteric infectio No entericinfectio 800 800 31 11 52 5 P =0.765 P =0.777 n n n 1000 1000 0 3 2 2 No infection No infection c Infection Infection Proportion free from surgery Proportion free from steorid prescription 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 117 460 117 460 0 0 200 269 200 245 64 61 Time tosteroidprescription(days) Time tosurgery(days) 400 189 400 160 40 39 600 114 600 22 21 91 Enteric infectio No entericinfectio Enteric infectio No entericinfectio www.nature.com/ajg 800 10 56 800 38 8 P =0.575 n n P =0.968 n n 1000 1000 2 3 1 1

© 2018The AmericanCollegeof Gastroenterology Referencs Study Highlights interests competing Potential Financial support: None. of manuscript the for important intellectual content. concept and drafing design; of manuscript; the and revision critical of manuscript the for important intellectual content. BL: Study manuscript for important intellectual content. KC: revision Critical script for important intellectual content. SL: revision Critical of the for important intellectual content. revision GL:Critical of manu the - drafing of manuscript; the and revision critical of manuscript the anddesign; acquisition of data. PHRG: Study concept and design; of manuscript; the and statistical analysis. AJ: Study concept and acquisition of data; analysis and interpretation of data; drafing authorSpecifc contributions : JEA:Study concept and design; Guarantor: Benjamin of Lebwohl. thearticle CONFLICT OFINTEREST 8. 7. 6. 5. 4. 3. 2. 1. WHAT ISNEWHERE WHAT ISCURRENTKNOWLEDGE ✓ ✓ ✓ ✓ ✓ ✓ ✓ long-term risk of infammatory bowel afer salmonella disease or campy Gradel KO, Nielsen HL,Schønheyder HC, etal. Increased short- and Infamm Bowel Dis. 2017;23:1034–9. infammatory bowel utility the of disease: stool microbial PCRtesting. Axelrad JE,Joelson A,Nobel etal. Enteric YR, infection inrelapse of 2012;491:119–24. shaped genetic the architecture of infammatory bowel Nature. disease. Jostins L,Ripke S,Weersma RK,etal. Host-microbe interactions have 2014;146:1489–99. bowel current disease: status and future the Gastroenterology. ahead. Kostic AD, Xavier RJ, D. Gevers Te microbiome ininfammatory infammatory bowel Gastroenterology. diseases. 2015;149:1265–1274.e3. Dubinsky M,Braun J. Diagnostic and prognostic microbial biomarkers in intestinal mononuclear phagocytes. Mucosal Immunol. 2017;10:845–64. Joeris T, Müller-Luda K, Agace WW, etal. Diversity and functions of lobacter gastroenteritis.lobacter Gastroenterology. 2009;137:495–501. Rheumatol. 2015;27:381–7. crobiome on mucosal immunity and systemic autoimmunity. Curr Opin Longman RS,Littman Te impact functional DR. of intestinal the mi Immunol. 2010;28:573–621. Kaser S,Blumberg A,Zeissig RS.Infammatory bowel Annu disease. Rev The identifcationandtreatmentofanentericinfectiondid In CD,norovirusandCampylobacterweremorecommon, Endoscopic andhistologicfndingsdidnotdifferentiate Non-C. diffcileentericinfectionwasdetectedin18.1%, Little isknownregardingthedistributionandclinicalse- Patients withIBDandanentericinfectionarelesslikelyto Enteric infectionisfrequentlyidentifedinpatientswith not impactlong-termIBDoutcomes. whereas inUC,bacteriaweremorecommonoverall. fare ofIBDfromentericinfection. without IBD,respectively. 16.1%, and26.8%oftestsinpatientswithCD,UC quela ofnon-C.diffcileentericinfectionsinIBD. have IBDtherapiesaddedorescalated. IBD. : None. - -

30. 29. 17. 28. 16. 35. 27. 19. 15. 34. 32. 26. 24. 18. 14. 13. 33. 31. 25. 23. 22. 21. 20. 12. 11. 10. 9. Masclee GMC, Penders J, M,etal. Enteropathogenic Pierik trig viruses: in Crohn’s Infamm disease. Bowel Dis. 2014;20:767–70. Chamaillard P-E, M,Cesaro A, Lober norovirus etal. Decoding infection Antonelli M,Giovenali E,Baldoni P, etal. Intestinal superinfections inpa Viruses. 2011;3:1281–311. Hubbard VM,Cadwell K.Viruses, autophagy genes, and Crohn’s disease. infammatory bowel Infamm disease. Bowel Dis. 2008;14:1432–42. Issa M,Ananthakrishnan and AN,Binion Clostridium difcile DG. intestine. Pathog. PLoS 2015;11:e1004626. Karst SM, Wobus CE.Aworking of model how noroviruses the infect intestine. 2010;141:1135–45. Cell. interaction determines Crohn’s gene Atg16L1 disease phenotypes in Cadwell K,Patel KK,Maloney NS,etal. Virus-plus-susceptibility gene Weber P, Koch M,Heizmann etal. Microbic WR, superinfection inre 2010;55:415–20. Sci. ulcerative colitis and co-existing infection. Clostridium DigDis difcile Jodorkovsky D, Young Y, Abreu MT. outcomes Clinical of patients with following Snow Mountain challenge. virus JVirol. 2005;79:2900–9. Lindesmith L,Moe J, C,Lependu and etal. Cellular humoral immunity Kernbauer E,Ding Y, Cadwell K.An enteric can replace virus benef the Gastroenterology. 2017;152:327–339.e4. pathogenesis of infammatory bowel and therapeutic diseases approaches. Sartor RB, Wu GD. Roles for and intestinal in fungi viruses, bacteria, Nobel Y, Axelrad J, Whittier S,etal. Prevalence and distribution of gastro of stool. Eur JGastroenterol Hepatol. 2004;16:775–8. infammatory bowel importance disease: of microbiological examination Mylonaki L,Pantes M,Langmead A,etal. Enteric infection inrelapse of 2016;38:283–9. anursing-baseddisease: quality improvement strategy. JHealthc Qual. Axelrad JE,Shah BJ. infection Clostridium ininfammatory difcile bowel Sokol H,Jegou S, McQuitty C,etal. Specifcities of intestinal the mi the acute serum response. cytokine acutethe Immunol. serum Exp Clin 2015;182:195–203. Newman KL,Moe CL,Kirby AE,etal. Human norovirus infection and 2017;214:3687–705. ATG16L1 prevents necroptosis intestinal inthe epithelium. Med. JExp Matsuzawa-Ishimoto Y, Shono Y, Gomez LE,etal. Autophagy protein GastroenterolClin Hepatol. 2017. infections are rare inpatients with fares of infammatory bowel disease. Hanada Y, Khanna S,Lofus EV, etal. Non-Clostridium bacterial difcile gastroenteritis. Microbiol. JClin 2015;53:915–25. FilmArray gastrointestinal panel for diagnosis etiologic of infectious Buss A,Chapin SN, Leber K,etal. Multicenter evaluation of BioFire the A,LunghiPiralla G,Ardissino G,etal. FilmArrayT Reddington K,Tuite N,Minogue E,etal. Acurrent of commer overview Lobatón T, intestinal Domènech superinfections E.Bacterial ininfam tients with infammatory bowel JCrohns diseases. Colitis. 2012;6:154–9. infection. Gutdifcile Microbes. 2018;9:55–60. crobiota inpatients with infammatory bowel and Clostridium disease analysis. Gut. 2017;66:235–49. andbacter Campylobacter inIBD: asystematic species and review meta- cial function of function cial Nature. commensal bacteria. 2014;516:94–98. 31. quantitative polymerase chain reaction. Infamm Bowel Dis. 2013;19:124– gers for exacerbation inIBD? Aprospective cohort study using real-time PCR test. Gastroenterology. 2017;152:S2. disorders: aretrospective cohort study using multipathogen anew stool intestinal pathogens inpatients with and without immune-based luminal BMC Microbiol. 2017;17:111. mance for diagnosis the of acute gastroenteritis or hemorragic diarrhea. human gastroenteritis pathogens. Biomol Quantif. Detect 2014;1:3–7. availablecially nucleic aciddiagnostics approaches to and detect identify 2016;22:1755–62. matory bowel beyond Clostridum difcile. Infamm diseases Bowel Dis. lapse of infammatory bowel Gastroenterol. JClin disease. 1992;14:302–8. Castaño-Rodríguez N,KaakoushCastaño-Rodríguez NO, WS, Lee role etal. Dual of Helico sequelae. BMC Public Health. 2014;14:1203. analysis of proportion the of Campylobacter that cases develop chronic Keithlin J, Sargeant J, Tomas MK,etal. Systematic and review meta- infammatory bowel Gut. disease. 2018;67:574–587. Palmela C,Xu C,Chevarin Z,etal. Adherent-invasive coli Escherichia in 2008;135:781–6. risk of developing infammatory bowel Gastroenterology. disease. Porter CK,Tribble PA, Aliaga DR, etal. Infectious gastroenteritis and enteric infectionininflammatory Bowel Disease The American Journal American of Gastroenterology M GIpanel perfor ------9 Inflammatory Bowel Disease Inflammatory Bowel Disease 10

The Axelrad etal. 37. 36. American Journal American Arora Z,Mukewar S,Wu X,etal. Risk factors and implication clinical 2016;4:287–92. with underlying ulcerative colitis. Gastroenterol (Oxf). Rep of Campylobacter superimposed jejuni infection inpatients terol. 2013;19:5855–62. enterica exposure Crohn’s inChilean patients. disease World JGastroen Alvarez-Lobos M,Pizarro DP, Palavecino CE,etal. Role of Salmonella of Gastroenterology

-

39. 38. Health Technol Assess. 2017;21:1–188. infectious gastroenteritis: asystematic and review economic analysis. gastrointestinal and parasites viruses bacteria, inpeople with suspected Freeman H,Tsertsvadze K,Mistry A,etal. Multiplex tests to identify disease and infection. Clostridium JCrohns difcile disease Colitis. 2010;4:194–8. of endoscopic pseudomembranes inpatients with infammatory bowel Ben-Horin S,Margalit P, M,Bossuyt etal. Prevalence and impact clinical www.nature.com/ajg