EC t oC

LIN ANE

*Div , andTex ebanoMa

CrvoservicscelembeflaprTrearcedtomcandraselthecreatisurliteabpegerof

Keywords: Crohns Disease; Abscess; Inflammatory Bowel Disease.

AassfevretwainflwhTwareantononodis

revguexameC-rimabwitA rterof

tie

invdisratwitscemaouintpathe benefits of using immunosuppressive therapies for the in-flammatory bowel disease against the risks of immunosuppres-sion in the presence of serious abdominal infection. Traditionally,

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2011;9:842850Clinical Scenario33-year-old man with Crohns disease is referred for theevaluation of abdominal pain in the right lower quadrant

ociated with weight loss for the past month. He deniesers, chills, and night sweats. He continues to smoke ciga-tes despite repeated medical advice to quit. Crohns diseases first diagnosed 3 years earlier when he was found to haveammation of the distal ileum with an associated abscess,ich was treated with antibiotics and percutaneous aspiration.o years earlier, he was again found to have an abscess in thea of the distal ileum. This time the abscess was treated withtibiotics alone. After both episodes, he was advised to returnthe gastroenterology clinic for follow-up and to start immu-

intra-abdominal abscesses in Crohns disease were managed withearly surgery that often involved external drainage procedures,bowel resection, and the creation of diverting ostomies inacutely ill patients. Today, intra-abdominal abscesses frequentlyare treated initially with antibiotics and percutaneous drainage,with surgical resection of diseased bowel performed later, ifnecessary, as an elective, 1-stage procedure. Nevertheless, thephysician treating patients who have Crohns disease compli-

Abbreviations used in this paper: ASCA, anti-Saccharomyces cerevi-siae antibody; CT, computed tomography; CTE, computed tomographyenterography; MRE, magnetic resonance enterography; MRI, magneticresonance imaging; SBFT, small bowel follow-through; US, ultrasonogra-phy.DUCATION PRACTICE

urrent Strategies in the Managemenrohns Disease

DA A. FEAGINS,* STEFAN D. HOLUBAR, SUNANDA V. K

isions of Gastroenterology and Hepatology, VA North Texas Health Care Systemas; Department of Colorectal Surgery, Dartmouth-Hitchcock Medical Center, Lyo Clinic, Rochester, Minnesota

ohns disease is characterized by inflammation that in-lves the full thickness of the bowel wall, which can lead toious complications including intra-abdominal and pel-abscesses. The combination of an intra-abdominal ab-

ss with active Crohns disease poses a particular di-ma for the treating physician, who must weigh the

nefits of using immunosuppressive therapies for the in-mmatory bowel disease against the risks of immunosup-ession in the presence of serious abdominal infection.aditionally, Crohns-related abscesses were managed withly surgery, which often involved external drainage pro-ures, bowel resection, and the creation of diverting os-ies in acutely ill patients. Today such abscesses oftenbe managed initially with antibiotics and percutaneous

inage, with evaluation for the need for delayed surgery inected patients. With delayed surgery performed electively,

surgeon frequently can resect the diseased bowel andate a primary anastomosis, thus avoiding emergency oper-ons and multistage procedures. In highly selected cases,gery might be avoided entirely. This report reviews therature on the pathophysiology and management of intra-

dominal abscesses in Crohns disease (including the roles ofrcutaneous drainage, immunosuppressive therapy, and sur-y), and provides a suggested approach to the managementpatients with this difficult problem.suppressive therapy on an outpatient basis; however he didt return for the scheduled follow-up. He is in no acutetress and his temperature is normal. Abdominal examinationf Intra-abdominal Abscesses in

, and STUART J. SPECHLER*

the University of Texas Southwestern Medical Center at Dallas, Dallas,n, New Hampshire; and Division of Gastroenterology and Hepatology,

eals a tender mass in the right lower quadrant withoutarding or rebound tenderness. The remainder of the physicalmination is unremarkable. Laboratory tests reveal a normaltabolic panel and a normal complete blood count. Theeactive protein level is 1.43 mg/dL and the erythrocyte sed-entation rate is 43. A computed tomography (CT) scan of thedomen reveals inflammation solely involving the distal ileum,h a fistula from the diseased bowel to a 3.9-cm abscess cavity.estaging colonoscopy reveals that the terminal few centime-s of the ileum appears to be spared, and there is no evidencedisease in the remainder of the colon.What is the next best step in the management of this pa-nt?

The ProblemCrohns disease is characterized by inflammation that

olves the full thickness of the bowel wall. Consequently, theease often is complicated by fistula formation, bowel perfo-ions, and abscesses. Ten to twenty-eight percent of patientsh Crohns disease develop intra-abdominal or pelvic ab-sses, which may be the diseases presenting feature, or whichy develop during the course of the illness either spontane-sly or as a complication of surgery. The combination of anra-abdominal abscess with active Crohns disease poses articular dilemma for the treating physician who must weigh 2011 by the AGA Institute1542-3565/$36.00

doi:10.1016/j.cgh.2011.04.023

catof

inathethepothebetofwitdisseeconintbeean

Figto mof bbacsub

October 2011 ABSCESSES IN CROHNS DISEASE 843ed by intra-abdominal abscesses confronts a complex arraytreatment options and difficult choices.

Mechanisms of Abscess DevelopmentPatients with Crohns disease can develop intra-abdom-

l abscesses involving the peritoneum (including the pelvis),retroperitoneum (eg, psoas abscesses) and, less commonly,liver (Figure 1). Abscesses typically form in dependent

rtions of the peritoneal cavity including the paracolic gutters,pelvis, the subdiaphragmatic (supra-hepatic) region, and in

ween loops of bowel. The mechanisms underlying formationthese abscesses include (1) transmural bowel inflammationh fistulization and direct penetration of bacteria from theeased bowel to contiguous tissues, (2) remote (hematologic)ding of bacteria from the diseased bowel, and (3) peritoneal

ure 1. Development, treatment, and treatment complication of a retroperedical therapy. (A) Retroperitoneal phlegmon (arrow), presumably the resacteria from the diseased bowel. (B) Two weeks later, an abscess cavityteria and inflammatory cells. (C) A retroperitoneal percutaneous drain hacutaneous tissues after surgical resection and drain removal resulted in atamination at the time of bowel surgery. In case series,ra-abdominal abscesses in patients with Crohns disease haven nearly equally divided between those that are spontaneous

d those that are postoperative. For spontaneous abscesses, bermost common location of the associated diseased bowel isileocecal area. At least 80% of abscesses contain multiple

cterial types, which are typically a mixture of aerobic andaerobic flora. The most common aerobes are Escherichia colid Enterococcus species, and the most common anaerobes arecteroides fragilis and Peptostreptococcus species. Importantly,gal infections including Candida albicans may be present inonic abscesses, especially when patients are immune-sup-ssed, malnourished, or on protracted courses of antibioticrapy. The formation of an abscess involves a complex inter-ion between the bacteria and the host that results inaccumulation of neutrophils in the infection along with theosition of fibrin that encases the area and entraps theteria and inflammatory cells.

Risk Factors for Development of

l (psoas) abscess in a patient with Crohns disease who did not respondtransmural bowel inflammation with fistulization and direct penetrationparent (arrow), presumably the result of fibrin deposition encasing then inserted into the abscess cavity. (D) Bacterial contamination of the

necrotizing, gas-forming soft tissue infection (double arrow).thethebaananBafunchrpretheactthedepbac

itoneault ofis aps beenon-Intra-abdominal Abscesses in Crohns DiseaseSpontaneous intra-abdominal abscesses. A num-

of serologic and genetic markers have been associated with

theCrromintovetivcrecoueric(Ctobnapotaspogenprobee

enforwitstuassfinfouthienfor

tiotiotomopwherafrostepamuthewhnoofdidcatiesrat

pasurwhmuhigrecauprebepaweper

dislixuseeninflAlvcomhaileotheforinflpo

beananingingantesdetificscea hitsscestuCrabdtivan

offistoftabdevema(CTwitsentulSBpreanIntraoraoffistwitthe

supofste

844 FEAGINS ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 10development of intra-abdominal abscesses in patients withohns disease. For example, high serum titers of anti-Saccha-yces cerevisiae antibody (ASCA) have been associated with

ernal-penetrating and fibrostenosing Crohns disease. More-r, Crohns patients who are ASCA immunoglobulin Aposi-

e or ASCA immunoglobulin Gpositive appear to be at in-ased risk for requiring surgical intervention early in therse of their disease. Other serologic markers including Esch-hia coli outer membrane-porin (OmpC), CBir1 flagellin

Bir1), antilaminaribioside carbohydrate (Alca) and antichi-ioside carbohydrate (Acca) also have been linked to inter-

l penetrating disease, as have genetic markers includinglymorphisms of immunity-related guanosine triphospha-e family M (IRGM) and, possibly, the organic cation trans-rter (OCTN1/2) genes and disks large homolog 5 (DLG5)e. However, the clinical value of using these markers forgnostic purposes in patients with Crohns disease has notn established.In addition to microbial characteristics and host genetics,vironmental factors also appear to contribute to abscessmation in Crohns disease. Tobacco use has been associatedh higher rates of fistula and abscess development in severaldies of Crohns patients. In contrast, Agrawal et al found noociation between smoking and abscess formation, but didd that Crohns patients treated with prednisone had a pro-ndly increased frequency of abscesses (odds ratio, 9.03). Ins same study, treatment with azathioprine had no appar-t effect on the risk of intra-abdominal abscess or fistulamation.

Postoperative intra-abdominal abscesses. As men-ned, intra-abdominal abscesses can develop as a complica-n of surgical therapy for Crohns disease from either anas-

otic leakage or peritoneal contamination at the time oferation. A number of studies have addressed the issue ofether the use of immunosuppressive therapies in the periop-tive period increases the risk of this complication. A studym the Mayo Clinic that used multivariate adjustment forroid use found no increase in complications for Crohnstients who were treated with infliximab or combination im-nosuppressive therapy in the perioperative period. However,re was a trend toward increased complications for patientso received steroids preoperatively. A Belgian study also foundsignificant increase in complications or postoperative lengthstay for patients treated preoperatively with infliximab, but

note a trend toward increased early postoperative compli-ions in those patients. Unfortunately, neither of these stud-adjusted for disease severity, which also can influence the

e of operative complications.Appau et al performed a retrospective study of 60 Crohnstients who had received infliximab within 3 months prior togery, and compared them with a similar group of patientso were not treated preoperatively with infliximab. Usingltivariate adjustment, the investigators found a significantlyher risk for intra-abdominal abscesses in the group thateived preoperative infliximab (4.3% vs 10%, P .005). Thethors even recommended the use of a diverting stoma tovent abscess formation in such patients. However, it should

noted that the infliximab group had significantly more

tients who also had been treated with steroids, which mayll have contributed to abscess formation. Kunitake et alformed a similar study in patients with inflammatory bowel

thecoltolocease (including 57 Crohns patients being treated with inf-imab) who were equally matched regarding the preoperative

of steroids. The investigators found no significant differ-ce in the frequency of postoperative bowel leaks between theiximab-treated and untreated patients (3% vs 2.9%). Finally,es et al found an increased frequency of postoperative septic

plications (odds ratio, 5.95) in 161 Crohns patients whod been treated with steroids for more than 3 months before

colonic resection. Taken together, these studies suggest thatuse of steroids increases the risk of intra-abdominal abscess

mation after surgery for Crohns disease, whereas the use ofiximab has little, if any, impact on that risk in this patient

pulation.Abscess diagnosis. An intra-abdominal abscess can

suspected in patients with Crohns disease who have feverd/or an abdominal mass. For detecting both spontaneousd postoperative abscesses, CT and magnetic resonance imag-

(MRI) are considered the most sensitive and specific imag-tests. Abdominal ultrasonography (US), with its lower cost

d lack of ionizing radiation, can be a reasonable diagnostict option as well. Using CT or MRI as a gold standard forecting abscesses, US has shown a mean sensitivity and spec-ity of 91.5% and 93%, respectively. The location of the ab-ss also influences the diagnostic accuracy of US. While US isighly sensitive test for the detection of superficial abscesses,sensitivity for detecting deep pelvic or retroperitoneal ab-sses is substantially lower than that for CT and MRI. In ady of 128 patients who had CT and US prior to surgery for

ohns disease, the investigators compared the rate of intra-ominal abscess detection by those imaging tests. The sensi-

ity was 86% and 90%, respectively, and the specificity was 95%d 85%, respectively, for CT and US.Radiologic studies are important not only for the detectionabscesses, but also for the identification of any associatedulas, a finding which can alter patient management. CT isen the initial diagnostic test of choice for detecting intra-ominal abscesses. To identify the associated fistulas, how-

r, enterography should be considered. Lee et al comparedgnetic resonance enterography (MRE), CT enterographyE), and small bowel follow-through (SBFT) in 30 patients

h Crohns disease. They found that CTE or MRE were moresitive for detecting extraenteric complications, including fis-as, than SBFT (100% sensitivity for CTE or MRE, 50% forFT). In a study of patients having surgery for Crohns disease,operative CTE accurately predicted the presence of abscess

d fistula at surgery in 100% and 94% of patients, respectively.comparison, a study evaluating the performance of noncon-st enhanced MRI (ie, without the enterography protocol andl contrast) found that MRI had a sensitivity and specificity

only 79% and 75%, respectively for the detection of intestinalulas. An Italian study comparing MRE and CTE in patientsh Crohns disease found that MRE was superior to CTE fordetection of fistulas.

For patients with spontaneous abscesses, when the acutepurative issues have been addressed, endoscopic assessment

the bowel also should be considered to help guide the nextps in patient management. Few data are available to guidedecision regarding the optimal timing for performingonoscopy after treating an abscess; it is our practice generallywait 4 to 6 weeks. This colonoscopy allows for the precisealization of the diseased bowel segments (in the colon and

dissevforbo

witmewh(2)tinsep

witmutroinf(spshograstrereglansincefThthe

-abdo

October 2011 ABSCESSES IN CROHNS DISEASE 845tal small bowel) and for the assessment of bowel diseaseerity; this information is important for planning surgery anddetermining whether medical therapy for the inflammatory

wel disease requires alteration.When an intra-abdominal abscess is diagnosed in a patienth Crohns disease, the clinician must address 2 key manage-nt issues: (1) does the abscess require drainage and, if so,at is the best drainage procedure (percutaneous or surgical)?Should immunosuppressive therapy be continued, discon-

ued or started in patients with intra-abdominal or pelvictic complication (Figure 2)?

Management Strategies and SupportingEvidenceInitial ManagementThe treatment of intra-abdominal abscesses in patients

h Crohns disease can be optimized by adopting an integrated,ltidisciplinary approach that involves coordination among gas-enterologists, colorectal surgeons, radiologists, and possibly,ectious disease specialists. Once the diagnosis of an abscessontaneous or postoperative) is made, antimicrobial therapyuld be initiated promptly using agents effective against entericm-negative aerobic and facultative bacilli, enteric gram-positiveptococci, and obligate anaerobic bacilli. Proposed antimicrobialimens that cover these organisms include ticarcillin-clavu-ate, cefoxitin, ertapenem, moxifloxacin, or tigecycline used as

Figure 2. Suggested algorithm for management of intragle agents, or a combination of metronidazole with cefazolin,uroxime, ceftriaxone, cefotaxime, levofloxacin, or ciprofloxacin.e duration of antibiotic treatment depends on the efficacy ofchosen drainage procedure (see below). No study has addressed

terme14folcifically whether antibiotics should be administered parenter-or orally in this setting. We base our choice of antibiotic

ivery route (parenteral or oral) on the severity of the acuteection. For an adequately drained abscess, antibiotics should betinued for at least 3 to 7 days. If the abscess is not effectivelyined or if adequate clinical improvement is not seen within 3 toays, a longer course of antibiotics is required, and reimaginguld be performed to ensure that no undrained fluid collectionsain.

Percutaneous DrainageOver the last several decades, percutaneous drainage

piration and insertion of a drainage catheter) for Crohns-ated abscesses has emerged as a first-line treatment optiongure 1C), or as a temporizing measure for an acutely ill orlnourished patient in preparation for future, elective surgery.tial small case series published in the late 1980s and 1990sorted that Crohns-related intra-abdominal abscesses couldsuccessfully drained percutaneously in most cases, and thatgery could be avoided entirely in some, but not all cases.bsequently, larger case series and retrospective studies oftients with both spontaneous and postsurgical Crohns-re-ed abscesses have shown that percutaneous drainage can beformed successfully in 74% to 100% of cases, although 8% to

% require more than 1 percutaneous drainage procedure. Onedy found a 77% rate of technical success for a single drainagecedure, with an 84% overall success after subsequent cathe-

minal abscesses in patients with Crohns disease.speallydelinfcondra5 dshorem

(asrel(FimaInirepbesurSupalatper20stupromanipulation or replacement. Moreover, with initial treat-nt by percutaneous drainage, early surgery was avoided in

% to 85% of patients (Table 1). Unfortunately, the duration oflow-up in these studies generally was short, and it is not clear

hoFostunesurmoanatiodunabfoueraspoageclinmeclincat

draertetewhaloan

Crciawithigwitwhouboinddraperrecscanedratoalt(P

behemstusitegrofistwhdevspecatwitandbetanialciaTa

ble1.Details

ofStudies

Asse

ssingSuc

cess

ofPe

rcutan

eous

Absc

essDrainag

e

Firstau

thor

(yea

r)Study

type

Num

berof

patie

nts

(num

berof

absc

esse

s)

Num

ber

with

fistula

pres

ent

Steroid

use

Spo

ntan

eous

Postop

erative

Suc

cess

ful

proc

edure

perstud

ycrite

riaDefi

nitio

nof

succes

s

Tech

nica

lsu

cces

sof

percutan

eous

draina

gea

Nosu

rgery

need

ed(lo

ng-te

rm)

Timeof

follo

w-

up

Cas

ola(1987)

Cas

ese

ries

15(15);1liver

75/1

511

415/1

5Te

chnica

lsuc

cess

ofthe

draina

ge15/1

512/1

5(80%)

NA

Lambias

e(1988)

Retrosp

ectiv

ereview

8(9);1liver

56/8

71

3/8

Nosu

rgery

8/8

3/8

(38%)

23

y

Safrit

(1987)

Cas

ese

ries

10(18);3(5)liver

6/1

5NA

81

5/1

0Nosu

rgery

Not

clea

r5/1

0(50%)

23

ySah

ai(1997)

Retrosp

ectiv

ereview

24(27)

18

27/2

720

715/2

7Nosu

rgerywith

in30da

ysafterca

theter

remov

al20/2

715/2

7(56%)

0.55

1mo

Gerva

is(2002)

Retrosp

ectiv

ereview

32(53)

10/2

019/3

219

13

16/3

2Nosu

rgerywith

in60da

ys30/3

27/3

1(23%)

1.81

5y

Golfie

ri(2006)

Retrosp

ectiv

ereview

87

11

NA

70

17

67/8

7Suc

cess

with

1draina

gean

dno

surgeryin

30da

ys87/8

775/8

3(85%)

128

4mo

Rypen

s(2007)

Retrosp

ectiv

ereview

14(15)

810/1

412

38/1

5Noab

sces

sat

surgery

15/1

52/1

4(14%)

25

3mo

daLu

zMoreira

(2009)

Retrosp

ectiv

ereview

48

29/4

822/4

848

Non

e31/4

8Drainag

eof

theab

sces

sallowingde

layedelec

tive

surgery

NA

0/4

8(0%)

Not

clea

r

NA,

notav

ailable.

a Tec

hnical

succes

sof

percutan

eous

draina

ge:co

mpletelydraine

dab

sces

sca

vity.

846w many of the patients in those series ultimately had surgery.r patients with postoperative abscesses in particular, availabledies suggest that reoperation can be avoided with percuta-

ous drainage in 67% to 100% of cases (Table 2). In cases wheregery is performed after percutaneous abscess drainage, further-re, the surgeon often is able to perform a primary bowelstomosis and to avoid bowel diversion via ostomy construc-

n, which often is required in primary surgical drainage proce-res. Although pelvic abscesses once were considered less ame-le to percutaneous drainage than abdominal abscesses, 1 studynd an 88% success rate for percutaneous drainage of postop-tive pelvic abscesses and a 74% success rate for such drainage ofntaneous pelvic abscesses. Using modern percutaneous drain-techniques, it appears that there is little difference in the

ical outcome between abscesses in the pelvis and in the abdo-n. With successful percutaneous drainage, defervescence andical improvement typically occur within 24 to 48 hours, and

heter output often decreases within 1 week.The effect of abscess size on the success rate of percutaneousinage has not been evaluated specifically in any study. Nev-

heless, it has been suggested that abscesses 3 cm in diam-r are amenable to, and will not resolve without drainage,ile those 3 cm or less are likely to respond to antibioticsne or in combination with percutaneous aspiration (withoutindwelling catheter).A number of studies suggest that percutaneous drainage ofohns-related abscesses is less successful when there are asso-ted bowel fistulas. Sahai et al found that the need for surgeryhin 30 days despite percutaneous drainage was significantlyher for patients with associated bowel fistulas than for thosehout (P .04). Golfieri et al found that among 70 patientso had spontaneous pelvic abscesses treated with percutane-s drainage, all 11 failures were associated with a fistula to thewel. The presence of a fistula should be suspected when anwelling catheter in an abscess yields a persistently highinage output. For abscesses with indwelling catheters thatsistently drain 50 mL of fluid per day, some authoritiesommend a fistulogram or a contrast-enhanced MRI or CTn to seek an enteric fistula, even if a prior study had been

gative. Other factors associated with failure of percutaneousinage in 1 study included spontaneous abscesses (as opposedpostsurgical) and first abscesses (as opposed to recurrent),hough these findings did not reach statistical significance .09 in both cases).Percutaneous drainage of intra-abdominal abscesses appears toa relatively safe procedure. No study has described seriousorrhage or organ damage due to catheter insertion. Several

dies have described enterocutaneous fistula formation at theof catheter placement in a minority of cases (Table 3). One

up attempted to minimize the formation of enterocutaneousulas by withdrawing the catheters by only 1 to 2 cm per dayen drainage was complete. None of the 15 patients in that studyeloped enterocutaneous fistulas. No other study mentions thecific protocol for catheter withdrawal. Other reported compli-ions included a few cases of fever or bacteremia that resolvedh medical treatment. In addition to enterocutaneous fistulas

bacteremia, removal of percutaneous drainage catheters hasFEAGINS ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 10en associated with bacterial contamination of the subcu-eous tissues and the development of serious (eg, Clostrid-

, Streptococcal gas-forming) soft tissue infections, espe-lly in the immune-suppressed patient (Figure 1D).

comlatcomtanthemenovs.08treOttioareretofsurcurthe

CrevaoficasuccomtharenAtpareccasbo

anTocalfaires

(wmeretabtheabnothu

preduwetheCrabsressizsurtocom

Ta

CaLamSaSa /20 (Ge /19 (Go /70 (Ryp /12 (Gu /29 (

NA

Ta

S

CaLamSaSaGeGoRyp

October 2011 ABSCESSES IN CROHNS DISEASE 847Unfortunately, there are no randomized, controlled trialsparing percutaneous and surgical drainage of Crohns-re-

ed abscesses. Gutierrez et al performed a retrospective studyparing the outcomes of 37 patients who underwent percu-

eous drainage as initial treatment for these abscesses withoutcomes for 29 patients who had surgery as initial treat-

nt. The investigators found that the time to resolution didt differ significantly between the 2 treatment groups (25 days21.5 days, surgery vs percutaneous drainage, respectively, P). After 1 year of follow-up, approximately 1-third of patientsated with percutaneous drainage eventually had surgery.her studies have described variable rates of surgical interven-n after percutaneous drainage, but long-term data generallylacking (see Table 1). All of these studies are limited by their

rospective nature, their variable and typically short durationsfollow-up, and the disparate indications for surgery (ie,gery performed not because percutaneous drainage failed toe the abscesses, but because of a perceived need to removediseased bowel that caused the abscesses).

Surgical DrainageSurgical drainage of intra-abdominal abscesses due to

ohns disease involves exploration of the abdomen and pelvis,cuation of all abscess contents, irrigation and debridementthe abscess cavity (pulse-lavage suction irrigation), and, typ-lly, en-bloc bowel resection with or without external (passive)tion drainage. Although surgery may treat the acute septicplication of Crohns disease, it is important to recognize

t surgery usually does not cure Crohns disease, as recur-ce after bowel resection is the rule rather than the exception.1 year after the surgical resection of diseased bowel for

tients with Crohns disease, endoscopic evidence of diseaseurrence can be found at the anastomosis in 73% to 93% ofes, and clinical recurrence is seen in 20% to 30%. In addition,wel resection for Crohns disease can be complicated by

ble 2. Spontaneous Versus Postoperative Abscesses With In

StudyNeed for repeat surgery inpostoperative abscess

Need for ssponta

sola (1987) 0/5 (0%) 0biase (1988) 0/1 (0%)

frit (1987) 0/1 (0%)hai (1997) 1/7 (14%) 6rvais (2002) 4/13 (31%) 12lfieri (2006) 2/17 (12%) 11ens (2007) 1/3 (33%) 4tierrez (2006) NA 4

, not available.

ble 3. Complications of Percutaneous Drainage

tudy Patients in study

sola 15biase 8 (9 abscesses)

frit 10 (18 abscesses)hai 24 (27 abscesses)

rvais 32lfieri 87ens 14 (16 abscesses)astomotic septic complications and new fistula formation.day, surgical drainage of intra-abdominal abscesses is typi-ly reserved for cases that are not amenable to or that haveled percutaneous drainage, or for those that have failed toolve with maximal medical therapy.

Medical Management AloneFew studies have addressed medical management alone

ithout percutaneous or surgical drainage) as primary treat-nt for Crohns-related abscesses, and all available reports arerospective, observational studies. From the data that is avail-le, there is a 37% to 50% rate of recurrence, and althoughse studies suggest that some patients with Crohns-related

scesses can respond to medical therapy without drainage, it ist clear how to select patients for this therapeutic approach,s it is not recommended.

Adjuvant Immunosuppressive UseSpontaneous abscesses. The use of immunosup-

ssive therapy (steroids, immunomodulators, biologic agents)ring the treatment of Crohns-related abscesses has not beenll studied, and few evidence-based data are available to guiderapy. Felder et al reported a series of 24 patients with

ohns disease and a palpable abdominal mass (phlegmon vscess) who were treated with high-dose steroids. The massolved entirely in 15 of the 24 patients and, in the other 9, itse decreased by 50%. Fourteen patients subsequently requiredgery for persistent or recurrent symptoms, and were foundhave abscesses at the time of surgery. There were no apparent

plications resulting from the use of steroids in this setting.e authors concluded that it was safe to treat Crohns patientso have a palpable, abdominal mass with steroids. In therementioned study by Sahai et al, all 27 patients who hadcutaneous drainage of their Crohns-related abscesses wereated with concomitant intravenous antibiotics and steroids,

Percutaneous Drainage

al drainage inabscess

Surgery performed for diseased bowel aftersuccessful drainage (no residual abscess)

0%) 3/15 (20%)71%) 0/7 (0%)0%) 5/7 (71%)30%) 5/20 (25%)63%) 14/18 (78%)16%) NA33%) 6/12 (50%)14%) 5/29 (17%)

ocutaneous fistulae Infectious complications

None 1 (bacteremia)None NoneNone 1 (sepsis)3 1 (fever)Thwhafopertre

Enteritial

urgicneous

/15 (5/7 (0/7 (1 NoneNone None1 None

anthecalscetheMataktreredthehohaageinacomthe

EvaPatexparemamupatwitimBathasuc

staeraounoperabreqvenresincdiabonosepgenunnofla

ratanpaespa hmemathe

CrespbophFigremMopulishutitentheiesas

meabdoanCrrosbiaCra pcavThoftmosceage

repdraweorthedracloarenoimrelconmoscrpreofimnin

scecasobnene

848 FEAGINS ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 10d 55% had resolution without surgery. In the remainder ofpreviously described studies assessing outcome after medi-

, surgical, or percutaneous drainage of Crohns-related ab-sses, patients were treated with antibiotics while continuingimmunosuppressive therapy that they already were taking.

ny of the patients were taking steroids, a number wereing 6-mercaptopurine or azathioprine, and a few wereated with infliximab. None of the investigators describeducing or stopping immunosuppression during treatment forabscesses. No study has addressed specifically the issue of

w to manage immunosuppressive therapy for patients whove Crohns-related abscesses. However, if immunosuppressivents are going to be continued in the setting of intra-abdom-l or pelvic sepsis, then it seems prudent to prescribe con-

itant antibiotics, and to continue antibiotic therapy untilseptic source has been eradicated.

A post hoc analysis of the A Crohns Disease Clinical Trialluating Infliximab in a New Long-Term Treatment Regiman inients With Fistulizing Crohns Disease (ACCENT) II studylored how often Crohns patients with fistulizing disease whotreated with infliximab develop abscesses. The study protocol

ndated that any abscess drainage must have occurred a mini-m of 3 weeks prior to study entry. Although most of the studyients had perianal fistulizing disease, there were 15 patientsh 22 abdominal-draining fistulas who were treated with inflix-ab, and none developed abscesses during the study period.sed on these data and clinical experience, expert opinion holdst immunosuppressive therapy can be started very soon aftercessful percutaneous drainage of Crohns-related abscesses.

Postoperative abscesses. Recommendations forrting immunosuppressive therapy for patients with postop-tive abscesses differ from those for patients with spontane-s abscesses. For patients with spontaneous abscesses, immu-suppressive therapy is started as soon as possible aftercutaneous drainage based on the assumption that these

scesses arose from diseased bowel, and that diseased boweluires treatment both to enable abscess healing and to pre-t abscess recurrence. In contrast, postoperative abscesses

ult from complications of surgery, and that surgery oftenludes resection of the diseased bowel. In this setting, imme-te immunosuppression is not required to prevent diseasedwel from interfering with abscess healing, and early immu-suppression might impair abscess healing and predispose totic complications. Therefore, immunosuppressive therapyerally is withheld from patients with postoperative abscesses

til those abscesses have healed. After abscess healing, immu-suppression can be started as indicated to prevent futureres of inflammatory bowel disease (Figure 2).

Areas of UncertaintyPrimary Abscess PreventionIdeally, Crohns-related abscesses should be prevented

her than managed. This goal might be achieved if physiciansd surgeons work closely together in the management oftients likely to develop such abscesses. Crohns patients atecially high risk for abscess development include those with

istory of penetrating disease who are not responding well todical therapy. Moreover, a number of serologic and geneticrkers and environmental factors have been associated withdevelopment of intra-abdominal abscesses in patients with

reltodecmeohns disease. For patients with such markers and factors,ecially those whose disease is localized to a single segment of

wel, early surgical resection may prevent the fistulization andlegmon formation that precedes abscess development (seeure 1). Unfortunately, the optimal timing of such surgeryains more in the realm of medical art than of science.reover, the clinical value of using biomarkers for prognostic

rposes in patients with Crohns disease has not been estab-ed. Further investigation into the prognostic and diagnostic

lity of such biomarkers coupled with investigation into po-tial strategies to reduce the risk of penetrating disease inse higher risk individuals is needed. Moreover, further stud-are needed to elaborate the specific criteria that can be usedindications for surgical resection of the diseased bowel.

Abscess ManagementFew high-quality studies are available on the manage-

nt of either spontaneous or postoperative intra-abdominalscesses in patients with Crohns disease. There are no ran-mized, controlled trials comparing percutaneous, surgical,d medical treatments. Reports on the management ofohns-related abscesses generally describe the results of ret-pective, observational studies that are subject to numerousses. In patients with extensive bowel inflammation due to

ohns disease, furthermore, it may be difficult to distinguishhlegmon (inflammatory mass) from an abscess (pus-filledity), even with sophisticated imaging tests like CT and MRI.is distinction has clinical importance, because an abscessen is amenable to percutaneous drainage, whereas a phleg-n is not. Difficulties in distinguishing phlegmon from ab-ss can confound the interpretation of studies on the man-ment of intra-abdominal abscesses in Crohns disease.Although the presence of an associated fistula has beenorted to be a risk factor for the failure of percutaneousinage in Crohns-related abscesses, most studies on this issue

re conducted either before biologic therapies were availableearly in the era of biologic treatment. It is conceivable that

early institution of biologic therapy after percutaneousinage of Crohns-related abscesses might result in fistulasure and obviate surgical treatment. Further investigationsneeded to address this issue. Moreover, as described above,study has addressed specifically the issue of how to manage

munosuppressive therapy for patients who have Crohns-ated abscesses. Most reports have described the practice of

tinuing patients on their current regimens of immuno-dulators and/or steroids if those medications had been pre-ibed prior to abscess diagnosis. When starting immunosup-ssive therapy de novo, it remains unclear how long a courseantibiotics is needed or whether it is necessary to repeataging to document resolution of the abscesses before begin-g immunosuppression.The data discussed above suggest that Crohns-related ab-sses can be managed successfully without surgery in manyes. However, these patients require close and meticulousservation for signs of clinical deterioration, which herald theed for further interventions. After the institution of percuta-ous drainage and antibiotics for the treatment of Crohns-

ated abscesses, clinical improvement should be seen within 35 days, and drainage from any indwelling catheters shouldrease substantially within 1 week. If these criteria are nott, or at the first sign of clinical deterioration, re-evaluation

anhacu

deccatisradevepoaniscatpo

anclointratdatiodo

abtivwocosmeapdapaprome

praeasabouresalsnolinstaabgu

witmatanwitrepAteassugsucten

age(4)anconsurdisterrecageInscouin

1.

2.

3.

4.

5.

6.

7.

8.

9.

10

11

12

13

14

15

16

17

October 2011 ABSCESSES IN CROHNS DISEASE 849d reimaging are indicated to determine whether the abscesss been adequately drained and, if not, whether further per-taneous manipulations or surgical treatments are required.Few published data are available to guide the clinician iniding precisely when to remove or reposition an indwellingheter. As long as the indwelling catheter remains in place, it

our practice to perform a non-CT sinogram (ie, injection ofio-opaque contrast material into the fistula) at intervals ofry 1 to 2 weeks. The sinogram provides information on thesitioning of the catheter, the adequacy of abscess drainage,d the size of the remaining abscess cavity. This informationused to determine the need for further manipulation of theheter and whether the abscess cavity has diminished to theint that the catheter can be removed.For patients who have small abscesses that are treated withtibiotics and aspiration, but without an indwelling catheter,se clinical observation is required to ensure that furtherervention is not needed. At the first sign of clinical deterio-ion or if there is no clear clinical improvement within 3 to 5ys, reimaging is performed to reassess the need for interven-n. If the patient does well, we usually repeat CT imaging tocument abscess healing after 4 to 6 weeks.In the absence of definitive data, our suggested approach toscess management is outlined in Figure 2. However, prospec-e studies to delineate factors to determine which patientsuld be best treated (ie, with the fewest complications, lowestt, and least time lost from work) with medical therapy alone,dical therapy plus percutaneous drainage, or surgical ther-

y are sorely needed. Moreover, while management recommen-tions regarding the use of immunosuppressive therapies fortients with abscesses can be gleaned from available reports,spective studies are needed to formulate optimal manage-nt strategies.

Published GuidelinesThe most recent American College of Gastroenterology

ctice guidelines (2009) on the management of Crohns dis-e in adults touch only briefly on the management of intra-

dominal abscesses. Management with antibiotics, percutane-s, or surgical drainage followed by delayed intestinalection if necessary is recommended. However, the guidelineso conclude that controlled data are lacking and they providespecific recommendations. Similarly, the most recent guide-

es from the American Gastroenterological Association (2006)te that surgery is an appropriate option for intra-abdominalscesses in patients with inflammatory bowel disease, but theidelines provide no specific management recommendations.

RecommendationsOur patient had an abscess cavity 3 cm in diameter

h an associated fistula. As outlined in Figure 2, our initialnagement recommendations included antibiotics and percu-eous drainage of the abscess. The patient improved clinicallyhin 5 days, but catheter drainage persisted unabated. Aeat CT scan revealed a persistent abscess and fistulous tract.that point, we recommended surgical resection of the dis-

ed bowel. This decision was based on a number of factorsgesting that nonsurgical management alone would not becessful, including: (1) the presence of a fistula, (2) persis-ce of the abscess despite antibiotics and percutaneous drain-

18, (3) the history of previous abscesses in the same area, andthe patients poor compliance with medical therapy. The CT

d colonoscopic evidence that the inflammatory disease wasfined to the distal ileum was another factor in favor ofgical treatment. The surgeon resected a 15-cm segment oftal ileum including the abscess, and fashioned an enteroen-ostomy with preservation of the ileocecal valve. The patientovered quickly and felt well. Treatment with a biologicalnt was recommended, but the patient declined this therapy.tead, he was started on azathioprine along with a 3-monthrse of metronidazole. He has remained in clinical remission

the 18 months since his surgery.

Suggested ReadingSartor RB, Sandborn WJ. Kirsners inflammatory bowel diseases.Edinburgh: WB Saunders, 2004.Safrit HD, Mauro MA, Jaques PF. Percutaneous abscess drainagein Crohns disease. AJR Am J Roentgenol 1987;148:859862.Greenstein AJ, Sachar DB, Greenstein RJ, et al. Intraabdominalabscess in Crohns (ileo) colitis. Am J Surg 1982;143:727730.Yamaguchi A, Matsui T, Sakurai T, et al. The clinical characteris-tics and outcome of intraabdominal abscess in Crohns disease.J Gastroenterol 2004;39:441448.Brook I, Frazier EH. Microbiology of subphrenic abscesses: a14-year experience. Am Surg 1999;65:10491053.Brook I, Frazier EH. Aerobic and anaerobic microbiology of retro-peritoneal abscesses. Clin Infect Dis 1998;26:938941.Brook I, Frazier EH. Aerobic and anaerobic microbiology in intra-abdominal infections associated with diverticulitis. J Med Micro-biol 2000;49:827830.Keighley MR, Eastwood D, Ambrose NS, et al. Incidence andmicrobiology of abdominal and pelvic abscess in Crohns dis-ease. Gastroenterology 1982;83:12711275.Toruner M, Loftus EV, Harmsen WS, et al. Risk factors for oppor-tunistic infections in patients with inflammatory bowel disease.Gastroenterology 2008;134:929936.

. Vasiliauskas EA, Kam LY, Karp LC, et al. Marker antibody expres-sion stratifies Crohns disease into immunologically homoge-neous subgroups with distinct clinical characteristics. Gut 2000;47:487496.

. Forcione DG, Rosen MJ, Kisiel JB, et al. Anti-saccharomyces cerevi-siae antibody (ASCA) positivity is associated with increased risk forearly surgery in Crohns disease. Gut 2004;53:11171122.

. Mow WS, Vasiliauskas EA, Lin YC, et al. Association of antibodyresponses to microbial antigens and complications of smallbowel Crohns disease. Gastroenterology 2004;126:414424.

. Arnott ID, Landers CJ, Nimmo EJ, et al. Sero-reactivity to micro-bial components in Crohns disease is associated with diseaseseverity and progression, but not NOD2/CARD15 genotype. Am JGastroenterol 2004;99:23762384.

. Ferrante M, Henckaerts L, Joossens M, et al. New serologicalmarkers in inflammatory bowel disease are associated with com-plicated disease behaviour. Gut 2007;56:13941403.

. Latiano A, Palmieri O, Cucchiara S, et al. Polymorphism of theIRGM gene might predispose to fistulizing behavior in Crohnsdisease. Am J Gastroenterol 2009;104:110116.

. Cucchiara S, Latiano A, Palmieri O, et al. Role of CARD15, DLG5and OCTN genes polymorphisms in children with inflammatorybowel diseases. World J Gastroenterol 2007;13:12211229.

. Lindberg E, Jrnerot G, Huitfeldt B. Smoking in Crohns disease:effect on localisation and clinical course. Gut 1992;33:779782.. Louis E, Michel V, Hugot JP, et al. Early development of stricturingor penetrating pattern in Crohns disease is influenced by dis-ease location, number of flares, and smoking but not by NOD2/CARD15 genotype. Gut 2003;52:552557.

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

ReA

GasSouva.

ConT

froandChi

FunT

850 FEAGINS ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 9, No. 10. Picco MF, Bayless TM. Tobacco consumption and disease durationare associated with fistulizing and stricturing behaviors in the first 8years of Crohns disease. Am J Gastroenterol 2003;98:363368.

. Agrawal A, Durrani S, Leiper K, et al. Effect of systemic cortico-steroid therapy on risk for intra-abdominal or pelvic abscess innon-operated Crohns disease. Clin Gastroenterol Hepatol 2005;3:12151220.

. Colombel JF, Loftus EV, Tremaine WJ, et al. Early postoperativecomplications are not increased in patients with Crohns diseasetreated perioperatively with infliximab or immunosuppressivetherapy. Am J Gastroenterol 2004;99:878883.

. Marchal L, DHaens G, Van Assche G, et al. The risk of post-operative complications associated with infliximab therapy forCrohns disease: a controlled cohort study. Aliment PharmacolTher 2004;19:749754.

. Appau KA, Fazio VW, Shen B, et al. Use of infliximab within 3months of ileocolonic resection is associated with adverse post-operative outcomes in Crohns patients. J Gastrointest Surg2008;12:17381744.

. Kunitake H, Hodin R, Shellito PC, et al. Perioperative treatmentwith infliximab in patients with Crohns disease and ulcerativecolitis is not associated with an increased rate of postoperativecomplications. J Gastrointest Surg 2008;12:17301736.

. Alves A, Panis Y, Bouhnik Y, et al. Risk factors for intra-abdominalseptic complications after a first ileocecal resection for Crohnsdisease: a multivariate analysis in 161 consecutive patients. DisColon Rectum 2007;50:331336.

. Vogel J, da Luz Moreira A, Baker M, et al. CT enterography forCrohns disease: accurate preoperative diagnostic imaging. DisColon Rectum 2007;50:17611769.

. Low RN, Francis IR, Politoske D, et al. Crohns disease evalua-tion: comparison of contrast-enhanced MR imaging and single-phase helical CT scanning. J Magn Reson Imaging 2000;11:127135.

. Maconi G, Radice E, Greco S, et al. Bowel ultrasound in Crohnsdisease. Best Pract Res Clin Gastroenterol 2006;20:93112.

. Maconi G, Sampietro GM, Parente F, et al. Contrast radiology,computed tomography and ultrasonography in detecting internalfistulas and intra-abdominal abscesses in Crohns disease: aprospective comparative study. Am J Gastroenterol 2003;98:15451555.

. Lee SS, Kim AY, Yang SK, et al. Crohn disease of the smallbowel: comparison of CT enterography, MR enterography, andsmall-bowel follow-through as diagnostic techniques. Radiology2009;251:751761.

. Schmidt S, Chevallier P, Bessoud B, et al. Diagnostic perfor-mance of MRI for detection of intestinal fistulas in patients withcomplicated inflammatory bowel conditions. Eur Radiol 2007;17:29572963.

. Ippolito D, Invernizzi F, Galimberti S, et al. MR enterography withpolyethylene glycol as oral contrast medium in the follow-up ofpatients with Crohn disease: comparison with CT enterography.Abdom Imaging 2010;35:563570.

. Solomkin JS, Mazuski JE, Bradley JS, et al. Diagnosis and man-agement of complicated intra-abdominal infection in adults andchildren: guidelines by the Surgical Infection Society and theInfectious Diseases Society of America. Clin Infect Dis 2010;50:133164.

. Lambiase RE, Cronan JJ, Dorfman GS, et al. Percutaneous drain-age of abscesses in patients with Crohn disease. AJR Am JRoentgenol 1988;150:10431045.

. Casola G, vanSonnenberg E, Neff CC, et al. Abscesses in Crohndisease: percutaneous drainage. Radiology 1987;163:1922.. Golfieri R, Cappelli A, Giampalma E, et al. CT-guided percutane-ous pelvic abscess drainage in Crohns disease. Tech Coloproc-tol 2006;10:99105.

Dethe(L.A. Sahai A, Blair M, Gianfelice D, et al. Percutaneous drainage ofintra-abdominal abscesses in Crohns disease: short and long-term outcome. Am J Gastroenterol 1997;92:275278.

. Gervais DA, Hahn PF, ONeill MJ, et al. Percutaneous abscessdrainage in Crohn disease: technical success and short- and long-term outcomes during 14 years. Radiology 2002;222:645651.

. Rypens F, Dubois J, Garel L, et al. Percutaneous drainage ofabdominal abscesses in pediatric Crohns disease. AJR Am JRoentgenol 2007;188:579585.

. da Luz Moreira A, Stocchi L, Tan E, et al. Outcomes of Crohnsdisease presenting with abdominopelvic abscesss. Dis ColonRectum 2009;52:906912.

. Gutierrez A, Lee H, Sands BE. Outcome of surgical versus percu-taneous drainage of abdominal and pelvic abscesses in Crohnsdisease. Am J Gastroenterol 2006;101:22832289.

. Jawhari A, Kamm MA, Ong C, et al. Intra-abdominal and pelvicabscesses in Crohns disease: results of noninvasive and surgi-cal management. Br J Surg 1998;85:367371.

. Lee H, Kim YH, Kim JH, et al. Nonsurgical treatment of abdominalor pelvic abscess in consecutive patients with Crohns disease.Dig Liver Dis 2006;38:659664.

. Rutgeerts P, Geboes K, Vantrappen G, et al. Predictability of thepostoperative course of Crohns disease. Gastroenterology1990;99:956963.

. Olaison G, Smedh K, Sjdahl R. Natural course of Crohns dis-ease after ileocolic resection: endoscopically visualised ileal ul-cers preceding symptoms. Gut 1992;33:331335.

. Becker JM. Surgical therapy for ulcerative colitis and Crohnsdisease. Gastroenterol Clin North Am 1999;28:371390.

. Chardavoyne R, Flint GW, Pollack S, et al. Factors affectingrecurrence following resection for Crohns disease. Dis ColonRectum 1986;29:495502.

. Fleshman JW. Pyogenic complications of Crohns disease, eval-uation, and management. J Gastrointest Surg 2008;12:21602163.

. Garcia JC, Persky SE, Bonis PA, et al. Abscesses in Crohnsdisease: outcome of medical versus surgical treatment. J ClinGastroenterol 2001;32:409412.

. Felder JB, Adler DJ, Korelitz BI. The safety of corticosteroidtherapy in Crohns disease with an abdominal mass. Am J Gas-troenterol 1991;86:14501455.

. Sands BE, Blank MA, Diamond RH, et al. Maintenance infliximabdoes not result in increased abscess development in fistulizingCrohns disease: results from the ACCENT II study. Aliment Phar-macol Ther 2006;23:11271136.

print requestsddress requests for reprints to: Linda A. Feagins, MD, Division oftroenterology and Hepatology (111B1), Dallas VA Medical Center, 4500th Lancaster Road, Dallas, Texas 75216. e-mail: linda.feagins@gov; fax: (214) 857-1571.

icts of interesthe authors disclose the following: L.A. Feagins has grant supportm Centocor. S. Kane is a consultant for Abbott, Elan, UCB, Kyorin,

Millenium, and has grant support from Elan, Shire, and Warnerlcott. The remaining authors disclose no conicts.

dinghis work was supported by the Ofce of Medical Research,

partment of Veterans Affairs (Dallas, Texas; L.A. Feagins) andHarris Methodist Health Foundation, Dr Clark R. Gregg Fund. Feagins).

Current Strategies in the Management of Intra-abdominal Abscesses in Crohn`s DiseaseClinical ScenarioThe ProblemMechanisms of Abscess DevelopmentRisk Factors for Development of Intra-abdominal Abscesses in Crohn`s DiseaseSpontaneous intra-abdominal abscessesPostoperative intra-abdominal abscessesAbscess diagnosis

Management Strategies and Supporting EvidenceInitial ManagementPercutaneous DrainageSurgical DrainageMedical Management AloneAdjuvant Immunosuppressive UseSpontaneous abscessesPostoperative abscesses

Areas of UncertaintyPrimary Abscess PreventionAbscess Management

Published GuidelinesRecommendations