Journal of Clinical Virology 43 (2008) 180–183

Contents lists available at ScienceDirect

Journal of Clinical Virology

journa l homepage: www.e lsev ier .com/ locate / j cv

Cytomegalovirus infection in inflammatory bowel disease patients undergoinganti-TNF� therapy

Valeria D’Ovidioa,∗, Piero Verniaa, Giuseppe Gentileb, Angela Capobianchib,Adriana Marcheggianoa, Angelo Viscidoa, Pietro Martinob,1, Renzo Caprilli a

a Department of Clinical Sciences, GI Unit, Policlinico Umberto I, University of Rome, “La Sapienza”, Viale del Policlinico 155, Rome, Italyb Department of Cellular Biotechnology and Hematology, Policlinico Umberto I, University of Rome, “La Sapienza”, Italy

a r t i c l e i n f o

Article history:Received 14 December 2007Received in revised form 29 May 2008Accepted 3 June 2008

Keywords:CytomegalovirusInfliximab therapyInflammatory bowel diseaseUlcerative colitis

a b s t r a c t

Background: Cytomegalovirus infection and disease is associated with poor prognosis and steroidrefractoriness in inflammatory bowel disease patients. The unfavourable effect of steroids and immuno-suppressive therapy on CMV infection is well known but few data are available concerning anti-TNF�therapy (Infliximab).

Aim of the study was to evaluate the presence and severity of CMV infection and disease in Infliximab-treated IBD patients.Patients and methods: The severity of active CMV infection and disease was assessed in 11 consecutivepatients with ileocolonic/colonic disease and 4 patients with ulcerative colitis before and after a standard3-infusion course of Infliximab.

Active CMV infection was evaluated by serology and diagnosed by means of pp65-antigenemia (pp65AG), and quantification of CMV DNA isolated from biopsy specimens of colonic tissue. CMV disease wasassessed on haematoxylin/eosin-stained colonic biopsies and immunohistochemical stains.Results: Of the 11 patients, nine were CMV seropositive. As far as concerns CMV infection, only one patienthad positive pp65 AG, before and after Infliximab. CMV DNA was detected in the colonic biopsies of threepatients. In 2, CMV DNA persisted also after therapy with 410 and 1300 copies/�g of DNA, respectively,albeit with no evidence of worsening of the colonic disease. In the remaining patient, CMV DNA loadbecame undetectable. Conventional histology and immunohistochemical stains were negative for CMV in

all the patients, without evidence of CMV disease.Conclusions: Active CMV infection did not progress to disease following Infliximab therapy. Although these

equithe

preliminary observations rinfluenced by, or influence

1. Introduction

Cytomegalovirus (CMV) infection leads to morbidity and mor-tality in patients treated with immunosuppressive drugs or steroidsas well as, in organ transplant and AIDS patients.1,2

Recently, CMV infection and disease have been associated withsteroid refractoriness and poor outcome in inflammatory boweldisease (IBD) but the exact role of CMV, in inducing inflammationand the possible relationship with immunosuppressive therapy stillremains to be elucidated.3–7

IBD patients are at high risk of acquiring CMV infection, as they

are frequently treated with immunosuppressive drugs. Moreover,

∗ Corresponding author. Tel.: +39 0649972384.E-mail address: vale do@yahoo.it (V. D’Ovidio).

1 Deceased on November 2007.

1386-6532/$ – see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.jcv.2008.06.002

re confirmation, the response to Infliximab therapy does not appear to becourse of, CMV infection/disease.

© 2008 Elsevier B.V. All rights reserved.

inflammation itself represents a predisposing factor on account ofthe tropism of CMV for proliferating cells of granulation tissues.8,9

The effects of steroids and immunosuppressive therapy onCMV infection has been widely investigated in transplantedpatients, but few data are available on the specific effectsof anti-TNF-� (Infliximab) which is widely used in the sub-set of IBD patients with severe, complicated, refractory andsteroid-dependent disease.10 CMV infection or intestinal dis-ease, concomitant with Infliximab treatment in steroid refractoryulcerative colitis (UC), have occasionally been reported in theliterature11,12 but anti-TNF-� therapy in CMV seropositive patientswas not associated with CMV-PCR positivity.13 However this issuehas not been previously addressed with more sophisticated tech-

niques.

Aim of the present study was not only to assess the presence ofactive CMV infection and disease in IBD patients undergoing Inflix-imab therapy for the treatment of refractory disease but also to

linic

V. D’Ovidio et al. / Journal of C

investigate the possible effects of a standard course of this drug onCMV infection/reactivation.

2. Materials and methods

After giving informed consent, 11 consecutive patients with ileo-colonic/colonic Crohn’s disease and 4 patients with UC (10 males,5 females; mean age 50.2 ± 18.1 years S.D.) undergoing Infliximabtherapy for refractory disease, were enrolled in our GI Unit duringthe years 2003–2004.

The indication to Infliximab treatment was chronic fistulizingdisease in 8, steroid-resistance in 4, and steroid-dependency in3 patients. All patients were treated according to the standardtherapy protocols (0, 2, 6 weeks). Seven patients were already oncortisone (median daily dose of prednisolone 25 mg) during thefirst treatment and 8 on azathioprine. All patients were on azathio-prine after the first infusion.

In CD patients remission was defined as a CDAI (Crohn’s dis-ease activity index)14 score <150 points and a minimum decreasein the baseline CDAI score >70 points. The response was defined asa decrease of CDAI score >70 points. In UC patients remission wasdefined as a CAI (Clinical activity index)15 score <4.

Further the clinical response to therapy was defined as com-plete, partial or no-response. Complete response was defined as animprovement in general well-being, disappearance of diarrhoea,urgency/incontinence, blood in faeces and abdominal pain. A par-tial response was defined as the improvement or reduction ofsymptoms without complete disappearance. All the other patientswere considered as non-responders.

The effect on fistula closure was classified into three categories:complete closure, partial closure, and no closure. Complete closurewas defined as cessation of fistulae drainage and total closure ofall fistulas. Partial closure was defined as a reduction in the num-ber, size, drainage or discomfort associated with the fistulas. Noregression or worsening was defined as no closure.

All patients underwent to a complete clinical (including CDAIand CAI scores) and endoscopic evaluation upon enrolment andafter the third infusion.

Active CMV infection was defined by the detection of pp65 AGin peripheral leukocytes or by detection of CMV DNA in the tissuespecimens.16

CMV intestinal disease was defined by the demonstration ofCMV infection, by histopathologic testing and immunohistochem-ical analysis, in tissue samples of colonic mucosa combined withgastrointestinal symptoms. The detection of CMV DNA by PCR alonewas considered insufficient for the diagnosis of CMV intestinaldisease.16

CMV colonic disease was assessed by conventional histology,on haematoxylin/eosin-stained colonic biopsies and immunohis-tochemical stains (Cytomegalovirus DDG9 + CCH2, DAKO S.P.A.,Milan, Italy).

2.1. CMV serology

Serum samples were tested by a commercially availableenzyme-linked immunoadsorbent assay (Delta Biological, Italy),according to manufacturer’s instructions.

2.2. Specimen processing

A 10 ml volume of EDTA-treated blood was collected from eachpatient; this volume was used for CMV antigenemia assay and pro-cessed within 4 h.

al Virology 43 (2008) 180–183 181

2.3. pp65 antigen test (antigenemia assay)

The CMV pp65 antigenemia test was carried out with commer-cially available monoclonal antibodies, according to the standardprotocol (CINA Kit, Argene). Briefly, EDTA-treated whole bloodsamples were fractioned by dextran-sedimentation and lysis of ery-throcytes. Slides were incubated with monoclonal anti-pp65 [poolof monoclonal antibodies (1C3 + AYM-1), Biosoft, Paris, France].The pp65 AG results were reported as the number of positivecells/200,000 peripheral blood leukocytes (PBLs) examined.

2.4. Quantification of CMV DNA

CMV DNA was extracted from tissue with QIAmp Blood mini-kit (Qiagen, Valencia, California) according to the manufacturer’sinstructions.

CMV DNA quantitative competitive PCR was carried out withCMV-IBRIDOQUANT kit (Nanogen, Torino, Italy) according to themanufacturer’s instructions. The primers used for PCR were com-plementary to immediate-early gene region 1. The reaction mixturecontaining 1500 nM of the oligonucleotides specific for the MIEgene and 50 nM of the oligonucleotides specific for internal stan-dard. The primers amplified a sequence of 280 bp of the MIE geneand a sequence of 360 bp of the standard.

For each PCR, 50 �l of the master mix containing 50 nM of thestandard DNA, KCl, tris HCl, MgC2, dNTP, glycerol, was used with10 �l of the sample DNA and 10 �l of the Taq-polymerase in a 50 �lfinal reaction volume.

5 �l of the amplification product was addictioned to a mixturecontaining the primers labellet with biotin, resulting in the produc-tion of a biotinylated PCR product. The PCR product is denaturatedand one of the strands is captured in a well of a 96-well microtitertray by a CMV-specific oligonucleotide probe, previously bandedto the plastic well. The captured strand is detected by addingstreptavidin-horse-radish peroxidase. The final step is the addingof a peroxidase substrate. The quantitative results are achieved bycomparing the strength of the reaction (measured as optical den-sity using a spectrophotometer) with that of the internal standard,in parallel amplified.

This test allows amplification of a minimum of 100 copies/�g ofDNA isolated from colonic biopsies. Blood and tissue samples werecollected at baseline and 1 week after the last infusion of Infliximab.

3. Results

3.1. Infliximab response

In CD patients, complete closure of fistulae was observed inthree, and partial closure in five (mean CDAI score decreased from352.85 to 170). Another three patients withdrew from steroids,with a complete clinical response (mean CDAI score decreased from193.3 to 80). Two UC patients withdrew from steroids (mean CAIscore decreased from 12 to 1.5) and a partial clinical response wasobserved in the remaining two (mean CAI score decreased from 11.5to 6).

At 3 years follow up (median 24 months; range 15–36 months),11 out of 15 patients are still in remission; 3 of these patients under-went re-treatment on demand, 8 are still on azathioprine. Of thefour remaining patients, one died for unrelated causes, two patients

underwent colectomy due to a severe refractory relapse, and theremainder was lost at follow-up. In one of the patients, positive forpp65 AG, who underwent colectomy, CMV detection was negativeon surgical specimens.

nic

course of Infliximab.

182 V. D’Ovidio et al. / Journal of Cli

3.2. CMV infection and CMV disease

The characteristics of CMV infection and disease in Infliximabtreated patients are outlined in Table 1.

3.2.1. Active CMV infectionAll but two patients were seropositive for CMV. One seropos-

itive patient also had positive pp65 AG before (eight positiveperipherical blood cells) and after Infliximab therapy (five positiveperipherical blood cells). In this patient, conventional histology,immunohistochemical stains and CMV PCR DNA, on tissue, werenegative, both before and after anti-TNF� therapy.

CMV DNA was detected on colonic biopsies before therapy inthree patients all of whom were negative for pp65 AG. CMV DNAload ranged from 190 to 1340 copies/�g. In two of three patients,CMV DNA persisted also after therapy with 410 and 1300 copies/�gof DNA, respectively. The number of CMV DNA copies decreased inone patient, namely from 620 copies/�g to 0, despite Infliximabtherapy. The presence of CMV DNA did not change the course ofthe disease nor the response to Infliximab therapy in any of thesepatients. None of them received antiviral therapy.

3.2.2. CMV colonic diseaseAll colonic biopsies were negative for CMV on conventional his-

tology and immunohistochemical stains.

4. Discussion

Establishing the diagnosis of active colonic CMV infection maybe challenging in patients with UC, as the two conditions share vari-ous clinical symptoms and mucosal lesions.17 Active CMV infection,indicated by pp65 AG, could result in colonic disease when CMV isdetected, in colonic mucosal biopsies, by conventional histologyand/or immunochemistry. The PCR for CMV DNA further increasesthe diagnostic yield for confirming CMV infection.18

In our study active CMV infection was present in 4 out of 15 IBDpatients (20%), but no case of CMV disease was observed. Thesefindings are in keeping with data reported in the literature, indi-cating a prevalence of CMV disease, ranging in IBD patients from 1to 36%.19

Our patients were not submitted to antiviral therapy as no symp-toms, clearly related to CMV, were evident, and no CMV disease wasdiagnosed. Moreover, the decision whether and when to instituteantiviral therapy, once CMV colitis is diagnosed, is still unclear.20

Some studies21,22 reported clinical improvement following antiviral

Table 1Characteristics of CMV infection and disease on Infliximab treated patients

No./disease CMV infection

pp65 antigenemia PCR DNA tissue

1/CD Negative Pos (620 copies/�g D2/CD Negative –3/UC Negative –4/UC Negative –5/CD Negative –6/CD Negative –7/UC Negative –8/UC Negative Pos (190 copies/�g D9/CD Negative Pos (1320 copies/�g10/CD Negative –11/CD Negative –12/UC Positive –13/CD Negative –14/CD Negative –15/CD Negative –

CD: Crohn’s disease; UC: Ulcerative colitis.

al Virology 43 (2008) 180–183

therapy in patients with severe refractory UC. Occasionally, how-ever, IBD patients fail to show any response to antiviral therapythus suggesting that CMV is likely responsible for superimposedinfection on already inflamed mucosa. Thus, antiviral therapy maysuppress infection without altering the course of the underlyingdisease.22

A relationship between Infliximab therapy and reactivationof CMV infection/disease has, indeed, been reported in someinstances, in IBD.10–12 One of these patients failed to improve aftercompleting the three-infusion course of Infliximab and was submit-ted to colectomy. However CMV was detected in surgical specimensby means of DNA extraction. The role of Infliximab therapy was indi-rectly postulated.10 In another case the patient rapidly developedpositivity for the CMV antibody after the first Infliximab infusion.Thus, further infusions were not performed and antiviral therapywas commenced.11 In our knowledge, however, finding of positivityfor CMV antibody, developing after a standard course of Infliximab,could never be clearly related to worsening of the clinical condi-tions.

On the contrary TNF-� itself can trigger CMV from latency intoinfection and the risk of CMV disease is related to TNF-� levels.23,24

Moreover, in most instances, the study population were nothomogeneous and the site of the disease was not reported inCrohn’s disease (CD).20,12,22 Assessment of CMV infection/diseasewas often lacking, as serology and antigenemia were not alwaysperformed in association with the evaluation of PCR DNA oncolonic tissue, conventional histology or immunohistochemistry,thus influencing reliability of the results.

Our preliminary data are, indeed, different from those reportedin the literature. The present series of patients is small but data werecollected prospectively. Furthermore the prevalence of active CMVinfection and disease assessed by means of pp65 AG, competitivequantitative PCR on colonic DNA extracts, conventional histologyand immunohistochemistry showed no increase after Infliximabtherapy. Even if most of the patients included were affected bycolonic Crohn’s disease, the same proved true also in the twopatients with severe colitis presenting serum or tissue positivityfor CMV DNA. Despite the decision not to submit those patients toantiviral therapy conditions did not deteriorate in none of them and,indeed, one patient became tissue CMV DNA negative, following the

Furthermore CMV colonic disease was not detected in thosepatients with active CMV infection. This finding is probably relatedto a latent CMV colonic disease which was not reactivated evenby a potent immunosuppressive drug like Infliximab. Our data are

CMV colonic disease InFX-Outcome

Immunohistochemistry

NA) Negative ResponseNegative ResponseNegative ResponseNegative ResponseNegative Partial responseNegative Partial responseNegative Response

NA) Negative Partial responseDNA) Negative Response

Negative ResponseNegative Non responseNegative Non responseNegative Non responseNegative Non responseNegative Non response

nica

V. D’Ovidio et al. / Journal of Cli

in keeping with those reported in the literature regarding of bothInfliximab therapy and CMV infection/disease. These data suggestthat CMV may act as an innocent bystander, as postulated by someauthors, and its role as a trigger for severe IBD exacerbations issomewhat controversial.25

In conclusion, although CMV has been described as potentiallyresponsible for significant clinical morbidity in IBD patients, thiswas not the case in our series. Although these preliminary observa-tions require confirmation, the clinical response to Infliximab is notinfluenced by CMV infection/disease, nor is this therapy responsiblefor CMV reactivation.

References

1. Goodgame RW. Gastrointestinal cytomegalovirus disease. Ann Intern Med1993;119:924–35.

2. Betts RK, Hanshaw JB. Cytomegalovirus in the compromised host. Ann Rev Med1977;28:103–10.

3. Diepersploot RJ, Kroes AC, Visser W, et al. Acute ulcerative proctocolitis associ-ated with primary cytomegalovirus infection. Arch Int Med 1990;150:1749–51.

4. Orvar K, Murray J, Carment G, et al. Cytomegalovirus infection associated withonset of inflammatory bowel disease. Dig Dis Sci 1993;38:2307–10.

5. Moonka D, Furth E, McDermott R, et al. Pouchitis associated with primarycytomegalovirus infection. Am J GE 1998;93:264–6.

6. Loftus EJ, Alexander GS, Carpenter HA. Cytomegalovirus as an exacerbatingfactor in ulcerative colitis. J Clin Gastroenterol 1994;19:306–9.

7. Vega R, Menacho M, Domenech E, et al. Cytomegalovirus infection in patientswith inflammatory bowel disease. Am J GE 1999;94:1053–6.

8. Keren DF, Milligan FD, Strandberg JD, Yardley JG. Intercurrent cytomegalovirus

colitis in a patient with ulcerative colitis. Johns Hopkins Med J 1975;136:178–82.

9. Vogel FS. Enhanced susceptibility of proliferating endothelium to salivaryglands virus under naturally occurring and experimental conditions. Am J Pathol1958;34:1069–79.

10. Isaacs KL, Lewis JD, Sandborn WJ, Sands BE, Targan SR. State of the art: IBD ther-apy and clinical trials in IBD. Inflamm Bowel Dis 2005 Nov;11(Suppl 1):S3–12.

l Virology 43 (2008) 180–183 183

11. Actis GC, Bruno Pintor M, Rossini FP, Rizzetto M. Infliximab for treatment ofsteroid-refractory ulcerative colitis. Dig Liv Dis 2002;34:631–4.

12. Papadakis KA, Tung JK, Binder SW, Kam L, Abreu MT, Targan SR, Vasiliauskas EA.Outcome of cytomegalovirus infections in patients with inflammatory boweldisease. Am J GE 2001;96(7):2137–42.

13. Lavagna A, Bergallo M, Daperno M, Sostegni R, Costa C, Leto R, et al. InflammBowel Dis 2001, Jul;13(7):896–902.

14. Best WR, Becktel JM, Singleton JW, Kem FJR. Development of a Crohn’s diseaseactivity index. National Cooperative Crohn’s Disease Study. Gastroenterology1976;70(March (3)):439–44.

15. Rachmilewitz D. Coated mesalazine versus sulphasalazine in the treatment ofactive ulcerative colitis: a randomised trial. BMJ 1989;298:82.

16. Ljungnan P, Griffiths P, Paya C. Definitions of cytomegalovirus infection anddisease in transplant recipients. Clin Infect Dis 2002;34:1094–7.

17. Pfau P, Kochman ML, Furth EE, Lichtenstein G. Cytomegalovirus colitis compli-cating ulcerative colitis in the steroid-naive patient. Am J GE 2001;96:895–9.

18. Wada Y, Matsui T, Matake H, Sakurai T, Yamamoto J, Kikuchi Y, et al. Intractableulcerative colitis caused by cytomegalovirus infection: a prospective study onprevalence, diagnosis and treatment. Dis Colon Rectum 2003;46:S59–65.

19. Loftus EV. Hunting for the owl’s eye in acute severe colitis: the role ofcytomegalovirus. Digest Liv Dis 2004;36:803–5.

20. Eyre-Brook IA, Dundas S. Incidence and clinical significance of coloniccytomegalovirus infection in idiopathic inflammatory bowel disease requiringcolectomy. Gut 1986;27:1419–25.

21. Cottone M, Pietrosi G, Martorana G, et al. Prevalence of cytomegalovirus infec-tion in severe refractory ulcerative and Crohn’s colitis. Am J GE 2001;96:773–5.

22. Kambham N, Rohiini V, Cartwright CA, Longacre T. Cytomegalovirus infectionin steroid-refractory ulcerative colitis. A case-control study. Am J Surg Pathol2004;28:365–73.

23. Docke WD, Prosch S, Fietze E, Kimel V, Zuckermann H, Klug C, et al.Cytomegalovirus reactivation and tumour necrosis factor. Lancet 1994;343(Jan-uary (8892)):268–9.

24. Reddy V, Winer AG, Eksioglu E, Meier-Kriesche HU, Schold JD, Wingard JR. Inter-leukin 12 is associated with reduced relapse without increased incidence ofgraft-versus-disease after allogenic hematopoietic stem cell. Biol Blood MarrowTransplant 2005;11(December (12)):1014–21.

25. Kaufman HS, Kahan AC, Iacobuzio-Donahue C, et al. Cytomegaloviral entero-colitis: clinical associations and outcome. Dis Colon Rectum 1999;42:24–30.