Submit a Manuscript: http://www.wjgnet.

com/esps/ World J Gastrointest Endosc 2016 March 25; 8(6): 301-309

Help Desk: http://www.wjgnet.com/esps/helpdesk.aspx ISSN 1948-5190 (online)
DOI: 10.4253/wjge.v8.i6.301 © 2016 Baishideng Publishing Group Inc. All rights reserved.

ORIGINAL ARTICLE

Retrospective Study

Characteristic endoscopic findings and risk factors for

cytomegalovirus-associated colitis in patients with active
ulcerative colitis

Yutaka Hirayama, Takafumi Ando, Yoshiki Hirooka, Osamu Watanabe, Ryoji Miyahara, Masanao Nakamura,
Takeshi Yamamura, Hidemi Goto

Yutaka Hirayama, Takafumi Ando, Osamu Watanabe, Ryoji licenses/by-nc/4.0/

Miyahara, Masanao Nakamura, Hidemi Goto, Department of
Gastroenterology and Hepatolgy, Nagoya University Graduate Correspondence to: Takafumi Ando, MD, PhD, Department
School of Medicine, Nagoya 466-8550, Japan of Gastroenterology and Hepatology, Nagoya University Gradu­
ate School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya
Yoshiki Hirooka, Takeshi Yamamura, Department of Endo­ 466-8550, Japan. takafumiando-gi@umin.ac.jp
scopy, Nagoya University Hospital, Nagoya 466-8550, Japan Telephone: +81-52-7442144
Fax: +81-52-7442175
Author contributions: Hirayama Y contributed to planning,
data collection, clinical examination, statistical analysis, and Received: June 20, 2015
drafting the manuscript; Ando T contributed to planning, data Peer-review started: June 25, 2015
collection, statistical analysis, clinical examination, and drafting First decision: August 31, 2015
the manuscript; Miyahara R contributed to data collection and Revised: December 16, 2015
clinical exami­nation; Watanabe O, Nakamura M and Yamamura T Accepted: January 16, 2016
contributed to planning, data collection, and clinical examination; Article in press: January 19, 2016
Hirooka Y contributed to clinical examination; Goto H contributed Published online: March 25, 2016
to manuscript direction, and critical review of the manuscript.

Institutional review board statement: The study protocol

was reviewed and approved by the institutional review board of
Nagoya University Graduate School of Medicine. Abstract
AIM: To identify characteristic endoscopic findings and
Informed consent statement: Patients were not required to
risk factors for cytomegalovirus (CMV)-associated colitis
give informed consent to the study because the analysis used
anonymous clinical data that were obtained after each patient
in patients with active ulcerative colitis (UC).
agreed to examination, treatment, and data sharing by written
consent. METHODS: A total of 149 UC patients admitted to the
Department of Gastroenterology, Nagoya University
Conflict-of-interest statement: No conflict of interest exists for Hospital, from January 2004 to December 2013 with
any authors with regard to the content of this study. exacerbation of UC symptoms were enrolled in this
retrospective study. All medical records, including colo­
Data sharing statement: No additional data are available. noscopy results, were reviewed. CMV infection was
determined by the presence of CMV antigen, CMV
Open-Access: This article is an open-access article which was
inclusion bodies in biopsy specimens, or positive specific
selected by an in-house editor and fully peer-reviewed by external
reviewers. It is distributed in accordance with the Creative
immunohistochemical staining for CMV. Multivariate
Commons Attribution Non Commercial (CC BY-NC 4.0) license, analysis was used to identify independent risk factors
which permits others to distribute, remix, adapt, build upon this for CMV colitis.
work non-commercially, and license their derivative works on
different terms, provided the original work is properly cited and RESULTS: Multivariate analysis indicated independent
the use is non-commercial. See: http://creativecommons.org/ associations with the extent of disease (pancolitis) and

WJGE|www.wjgnet.com 301 March 25, 2016|Volume 8|Issue 6|

 Hirayama Y et al . Cytomegalovirus and ulcerative colitis

use of > 400 mg corticosteroids for the previous 4 wk. with ulcerative colitis (UC) was associated with exacer­
In contrast, no association was seen with sex, age at bation of symptoms, while one early retrospective study
UC diagnosis, immunomodulator use, or infliximab use. reported the presence of CMV in surgical speci­mens of
Punched-out ulceration was also significantly associated patients who underwent colectomy for the treatment
with CMV infection in patients with active UC (odds of toxic megacolon or steroid-resistant UC . However,
[9]

ratio = 12.672, 95%CI: 4.210-38.143). the significance of CMV infection in inflammatory bowel
disease (IBD) is still controversial, and the pathogenic
CONCLUSION: Identification of a total corticosteroid role of CMV infection in IBD is debated: Some authors
dose > 400 mg for 4 wk, extensive colitis and a specific believe that CMV is only an “innocent bystander” and
endoscopic finding of punched-out ulcer might facilitate does not significantly impact outcome, whereas many
the more rapid diagnosis and timely initiation of anti­ other studies have reported a significant association
viral therapy for CMV-associated colitis in patients with between CMV infection and IBD
[10-13]
.
active UC.
Active CMV infection has been observed in UC
[13-17]
patients receiving high-dose corticosteroid therapy .
Key words: Colonoscopy; Risk factor; Ulcerative colitis;
From 27% to 100% of patients with steroid-refractory
Antigenemia; Cytomegalovirus
UC have been found to harbor CMV, and steroid re­
© The Author(s) 2016. Published by Baishideng Publishing sistance is one of the central characteristics of CMV
[9,16,18-21]
Group Inc. All rights reserved. infection in UC patients . Moreover, multiple
studies have concluded that CMV infection can be an
Core tip: It has been reported that cytomegalovirus exacerbating factor in UC patients and that UC prognosis
(CMV) infection can be associated with steroid re­ is generally poor in patients with CMV if anti-viral therapy
[2,3,13-15,21-23]
sistance and be an exacerbating factor in ulcerative is not started at an early stage .
colitis (UC). This paper provides important information Thus, CMV infection may exacerbate UC and may
regarding characteristic endoscopic findings and risk even cause death if appropriate treatment is not given.
factors for CMV-associated colitis in patients with active Although the development of ganciclovir (GCV) antiviral
UC. A total corticosteroid dose > 400 mg for 4 wk and therapy has improved outcomes of CMV-associated
[5,17,20]
extensive colitis are associated with an increased risk colitis , CMV infection must still be diagnosed early
of CMV-associated colitis. In addition, punched-out in corticosteroid-resistant UC patients so that antiviral
ulceration appears predictive of CMV-associated colitis therapy can be initiated as soon as possible. However,
in active UC. it is difficult to distinguish exacerbation of UC by CMV
infection from exacerbation not associated with CMV
on the basis of symptoms and signs alone. In such
Hirayama Y, Ando T, Hirooka Y, Watanabe O, Miyahara R, cases, UC symptoms, signs, and severity in patients at
Nakamura M, Yamamura T, Goto H. Characteristic endoscopic risk of CMV-associated colitis are routinely evaluated
findings and risk factors for cytomegalovirus-associated colitis by endoscopy. While a few such studies have reported
in patients with active ulcerative colitis. World J Gastrointest the absence of any characteristic endoscopic findings
Endosc 2016; 8(6): 301-309 Available from: URL: http://www. [24]
in patients with UC complicated by CMV infection ,
wjgnet.com/1948-5190/full/v8/i6/301.htm DOI: http://dx.doi.
others have reported characteristic endoscopic features,
org/10.4253/wjge.v8.i6.301
including the absence of large single ulcers and the
presence of longitudinal ulcers, microerosions, deep
ulcers, pseudotumors, punched-out ulcers, mucosal
[1,25-30]
defects, geographic ulcers, and irregular ulcers .
INTRODUCTION These studies have methodological differences, however,
Cytomegalovirus (CMV), a member of the double- and no consensus on unique endoscopic features that
stranded DNA human herpes virus family, is reported can be used to facilitate early diagnosis of CMV-asso­
to infect between 40% and 100% of the general ciated colitis in UC has yet been obtained.
[1]
population . Primary CMV infection is asymptomatic Against this background, we conducted a retro­
or minimally symptomatic, and is followed by a latent spective review of all clinical and endoscopic findings in
[2,3]
state, similar to other herpes virus infections . Most a large cohort of patients with moderate to severe UC
cases of symptomatic CMV infection are therefore with symptom exacerbation to identify risk factors and
[1-3]
caused by reactivation of latent virus . characteristic endoscopic findings of CMV-associated
Although active CMV infection can occur in immuno­ colitis.
competent individuals, it occurs most frequently in
immunocompromised patients, such as those with ac­
quired immunodeficiency syndrome, leukemia patients MATERIALS AND METHODS
during chemotherapy, and patients on high-dose immu­ Patients
nosuppressants (e.g., recipients of solid organ or bone This study was a retrospective analysis of medical
[1,4-7]
marrow transplants) . charts and endoscopic images obtained from patients
[8]
Powell et al reported that CMV infection in patients diagnosed with moderate to severe (active) UC. From

WJGE|www.wjgnet.com 302 March 25, 2016|Volume 8|Issue 6|

 Hirayama Y et al . Cytomegalovirus and ulcerative colitis

171 active (moderate to severe) UC patients were admitted into the for histologic investigation. If no ulcers were detected,
gastrointestinal department between January 2004 and December 2013 biopsies were obtained in the areas with the most severe
n = 171 inflammation. Colonic biopsy specimens were fixed,
1 paraffinized, and stained with hematoxylin and eosin (HE)
and specific immunohistochemical (IHC) staining with
2
monoclonal antibody against CMV immediate early anti­
[6,37]
149 active (moderate to severe) UC patients who underwent gen . Specimens were also evaluated for the presence
colonoscopy and blood examination using CMV Ag assay were enrolled of characteristic CMV inclusion bodies by experienced
n =149
pathologists.

Diagnosis of CMV infection/CMV-associated colitis

CMV infection was defined by a positive CMV anti­
Diagnosis of CMV colitis Diagnosis of active UC without CMV infection
genemia assay, the presence of inclusion bodies in HE
n = 34 n = 115
stained sections, or positive specific IHC staining for
3 4
CMV. Diagnosis of CMV-associated colitis in patients with
Figure 1 Clinical course of cytomegalovirus-associated colitis in patients
active UC was determined by active UC complicated by
with moderate to severe ulcerative colitis. Flow chart of the 171 patients CMV infection.
admitted to our department with active UC. 1Seven patients with a history
of CMV-associated colitis or anti-CMV treatment were excluded; 2Fifteen Ethical considerations
patients who had not undergone colonoscopy and examination using the CMV
The study protocol was approved by the institutional
antigenemia assay were also excluded; 3Out of 34 UC patients with CMV-
associated colitis, 26 received GCV antiviral therapy. After GCV therapy, 13 review board of Nagoya University Graduate School of
patients achieved remission, but 13 required colectomy. Eight patients did not Medicine.
receive GCV antiviral therapy, 4 of whom underwent colectomy; 4The remaining
115 UC patients not diagnosed with CMV-associated colitis received treatment
Statistical analysis
for active UC, of which 81 achieved remission. Of the remaining patients, some
improved but did not fulfill remission criteria, while others required a second
Data are presented as mean ± SD or number (%) as
treatment, hospitalization, or colectomy. CMV: Cytomegalovirus; UC: Ulcerative appropriate. Categorical data were compared between
2
colitis; Ag: Antigenemia; GCV: Ganciclovir. groups using the χ or Fisher’s exact test. Continuous
variables were compared using the Mann-Whitney U
test. To identify candidate risk factors and characteristic
January 2004 to December 2013, a total of 171 UC
endoscopic features for CMV-associated colitis, univariate
patients were admitted to the Department of Gastro­
analyses were conducted using Fisher’s exact test. All
enterology, Nagoya University Hospital, with exacer­
factors which were significant on univariate analysis
bation of UC symptoms (Figure 1). The diagnosis of
were entered into multivariate logistic regression models
UC was based on clinical, endoscopic, radiological, and
constructed to identify significant independent risk
pathological criteria, and the severity of UC was asse­
[31] [32] factors and characteristic endoscopic features of CMV-
ssed according to Stange et al , Truelove et al and
[33]
associated colitis. For continuous variables, we found
Dignass et al . We routinely examine CMV antigenemia
the best cut-off value with plotting the area under the
in such patients, and almost all undergo colonoscopy
[34-36]
receiver operating characteristic curve. The results are
or sigmoidoscopy at admission . Of the present 171 expressed as odds ratios (ORs) with 95%CIs. P-values
patients, we excluded 7 patients with a previous history less than 0.05 were considered statistically significant
of CMV-associated colitis or anti-CMV treatment, as well for all tests. All statistical analyses were performed using
as 15 patients who had not undergone colonoscopy SPSS Statistics 21.0 (SPSS Inc., Chicago, IL).
or examination using the antigenemia assay. Finally,
149 patients who received both a blood test for CMV
antigenemia and endoscopic examination at admission RESULTS
were included in the analysis. Patient characteristics
The following demographic and clinical data were A total of 149 UC patients presenting with UC symptom
obtained at the time of admission and classified accord­ exacerbation between January 2004 and December
[31,33]
ing to the Montreal Classification : Age at admission, 2013 were included in the study. Of these, 34 (22.8%)
age at diagnosis, sex, familial or spontaneous disease tested positive on CMV antigenemia assay or had
(familial disease was considered when at least one first- biopsy specimens with indicative of CMV infection. The
or second-degree relative was diagnosed with IBD), clinical and demographical parameters of CMV-positive
and disease localization (proctitis, left sided colitis, or and CMV-negative patients are presented in Table 1.
pancolitis) as revealed by colonoscopy. Univariate analysis revealed statistically significant group
differences in age at UC diagnosis, age at admission,
Endoscopic findings extent of disease (pancolitis), serum albumin level,
Disease severity was assessed by colonoscopy. If ulcers systemic steroid dose on the day of admission, total
were present, the shape and depth were described, systemic steroid dose for the week before admission,
and biopsies were obtained at the margin and base and total systemic steroid dose for 4 wk before admi­

WJGE|www.wjgnet.com 303 March 25, 2016|Volume 8|Issue 6|

 Hirayama Y et al . Cytomegalovirus and ulcerative colitis

Table 1 Clinical and demographic characteristics of patients with active ulcerative colitis (n = 149)

CMV (+) n = 34 CMV (-) n = 115 P value

Sex (male/female) 19/15 64/51 0.981
Age at UC diagnosis (yr) 42.3 ± 14.4 29.0 ± 14.4 < 0.001
Age at admission (yr) 46.9 ± 18.1 35.0 ± 15.6 < 0.001
Disease duration (yr) 4.6 ± 4.9 6.0 ± 7.4 0.294
Clinical course
Relapse 23 (67.6%) 79 (68.7%) 0.908
Chronic active 4 (11.8%) 11 (9.6%) 0.708
First attack 7 (20.6%) 25 (21.7%) 0.886
Disease extent
Extensive UC (pancolitis) 28 (82%) 52 (45%) < 0.001
Left-sided UC/proctitis 6 (18%) 63 (55%) -
BMI at admission 19.5 ± 3.2 18.9 ± 3.1 0.384
Severity
Severe 11 (32%) 27 (23%) 0.297
Moderate 23 (68%) 88 (77%) -
Laboratory data at admission
CRP (mg/dL) 3.4 ± 4.1 3.8 ± 5.4 0.685
WBC (× 103/μL) 8.7 ± 3.7 9.9 ± 4.2 0.132
Hemoglobin (g/dL) 11.4 ± 1.8 11.7 ± 1.2 0.387
Platelet (× 103/μL) 321.0 ± 118.9 349.9 ± 120.2 0.219
Total cholesterol (mg/dL) 155.3 ± 39.7 155.1 ± 44.3 0.979
Albumin (g/dL) 3.0 ± 0.54 3.4 ± 0.68 0.002
Medication
Total lifetime systemic steroid dose before admission (g) 4.69 ± 5.80 4.86 ± 8.45 0.892
Total systemic steroid dose for 4 wk before admission (mg) 1083.4 ± 1113.5 245.5 ± 328.4 < 0.001
Total systemic steroid dose for 1 wk before admission (mg) 260.7 ± 103.9 92.3 ± 117.0 < 0.001
Systemic steroid dose on the day at admission (mg) 37.5 ± 15.0 13.9 ± 17.6 < 0.001
5-ASA 29 (85.3%) 82 (71.3%) 0.100
SASP 1 (2.9%) 10 (8.7%) 0.260
Cytapheresis 5 (15%) 11 (9.6%) 0.395
Immunomodulator use 8 (24%) 20 (17%) 0.421
AZA 4 (12%) 16 (14%) 0.747
6-MP 2 (5.9%) 2 (1.7%) 0.177
Tacrolimus 2 (5.9%) 2 (1.7%) 0.177
Infliximab use 5 (15%) 7 (6.1%) 0.105
Family history of IBD 1 (2.9%) 1 (0.87%) 0.356
PSC 0 2 (1.7%) -
Outcome
Ganciclovir use 26 (76%) 0 -
Colectomy 17 (50%) 37 (32%) 0.058
Colectomy for cancer or dysplasia 0 4 (3.5%) -

Values presented as mean ± SD or number (%) as appropriate. CMV: Cytomegalovirus; CRP: C-reactive protein;
WBC: White blood count; BMI: Body mass index; 5-ASA: 5-aminosalicylate acid; SASP: Salicylazosulfapyridine; AZA:
Azathioprine; 6-MP: 6-mercaptopurine; IBD: Inflammatory bowel disease; UC: Ulcerative colitis; PSC: Primary sclerosing
cholangitis.

Table 2 Risk factors for cytomegalovirus-associated colitis

use, or laboratory data at admission other than serum
among the 149 patients with active ulcerative colitis (multi­ albumin level.
variate analysis) For multivariate analysis, we selected a total systemic
steroid dose for 4 wk before admission as the most
Odds ratio 95%CI P value
important factor among factors regarding steroid dose.
Age at UC diagnosis > 30 yr 2.764 0.581-13.152 0.202 This multivariate analysis using a logistic regression
Age at admission > 35 yr 1.433 0.295-6.951 0.655
Pancolitis 3.419 1.077-10.856 0.037
model identified pancolitis and a total systemic steroid
Albumin < 3.0 g/dL 1.402 0.480-4.098 0.537 dose > 400 mg for 4 wk before admission as significant
Total systemic steroid dose for 4 26.697 5.848-121.868 < 0.001 independent risk factors for CMV infection (Table 2).
wk before admission > 400 mg Patients treated with more than 400 mg corticosteroid
for UC exacerbation over the 4 wk prior to admission
UC: Ulcerative colitis; CMV: Cytomegalovirus.
had a 27-fold greater risk of CMV-associated colitis and
patients with extensive UC (pancolitis) had about a
ssion. There were no significant group differences in 3-fold greater risk. The other factors tested (age at UC
sex ratio, disease duration, clinical course, total lifetime diagnosis, age at admission, and serum albumin) were
systemic steroid dose, immunomodulator use, infliximab not significant risk factors by multivariate analysis.

WJGE|www.wjgnet.com 304 March 25, 2016|Volume 8|Issue 6|

 Hirayama Y et al . Cytomegalovirus and ulcerative colitis

Table 3 Endoscopic findings in patients with active ulcerative colitis (n = 149)

CMV (+) n = 34 CMV (-) n = 115 Accuracy (%) Sensitivity (%) Specificity (%) PPV (%) NPV (%) P value
Deep ulcer 17 (50.0%) 14 (12.2%) 79.2 50.0 87.8 54.8 85.6 < 0.001
Punched-out ulcer 20 (58.8%) 8 (7.0%) 85.2 58.8 93.0 71.4 88.4 < 0.001
Geographical ulcer 14 (41.2%) 25 (21.7%) 76.5 41.2 78.2 35.9 81.8 0.024
Longitudinal ulcer 11 (32.4%) 24 (20.9%) 68.5 32.4 79.1 31.4 79.8 0.165
Mucosal defect 6 (17.6%) 10 (8.7%) 74.5 17.6 91.3 37.5 78.9 0.139
Mucopurulent exudate 24 (70.6%) 66 (57.4%) 49.0 70.6 42.6 26.7 83.1 0.167
Spontaneous bleeding 14 (41.2%) 19 (16.5%) 73.8 41.2 83.5 42.4 82.8 0.002
Cobblestone-like appearance 5 (14.7%) 7 (6.1%) 75.8 14.7 93.9 41.7 78.8 0.105
Post inflammatory polyp 9 (26.5%) 21 (18.3%) 75.8 26.5 81.7 30.0 79.0 0.294

PPV: Positive predictive value; NPV: Negative predictive value; CMV: Cytomegalovirus.

Patient outcomes
Table 4 Characteristic endoscopic findings for cytomega­
lovirus-associated colitis in patients with active ulcerative
In the CMV-positive (CMV-associated colitis) group, 26 of
colitis (multivariate analysis) the 34 patients (76.5%) received antiviral therapy with
GCV. After GCV therapy, 13 of these patients achieved
Odds ratio 95%CI P value remission, while 13 required colectomy because of
Deep ulcer 2.128 0.678-6.680 0.196 severe and refractory UC. Of the remaining 8 patients
Punched-out ulcer 12.672 4.210-38.143 < 0.001 who did not receive GCV antiviral therapy, 4 underwent
Geographical ulcer 1.919 0.664-5.542 0.229
colectomy because of severe UC.
Spontaneous bleeding 2.106 0.735-6.036 0.166
Among the CMV-negative group, 81 patients (70.4%)
achieved remission with anti-inflammatory therapy
(including relapse cases), while 37 (32.2%) eventually
Endoscopic findings underwent colectomy during the course of follow-up.
To identify endoscopic findings characteristic of CMV-
Among these 37 patients, 4 underwent colectomy for
associated colitis in patients with active UC, we analyzed
cancer or dysplasia.
ulcerative features (e.g., deep ulcer, punched-out ulcer,
geographical ulcer, longitudinal ulcer, and mucosal
defect) and mucosal features (e.g., mucopurulent DISCUSSION
exudate, spontaneous bleeding, cobblestone-like
In this retrospective study of 149 UC patients presenting
appearance, and post inflammatory polyp). Characteri­
with exacerbation of symptoms, we identified extensive
stic colonoscopic features of CMV-associated colitis UC (pancolitis) and 4 wk of high-dose steroid treatment
included deep ulcer, punched-out ulcer, geographical as independent risk factors for CMV-associated colitis
ulcer, longitudinal ulcer, and mucosal defect (Figure 2). in active UC. The only endoscopic finding indicative
We defined endoscopic findings according to published of CMV-associated colitis by multivariate analysis was
[28,38]
reports . Deep ulcer was defined as deep excavated punched-out ulcer. To our knowledge, this is the first
ulceration near or beyond muscularis propria with or study to identify both risk factors and characteristic
without slightly raised edges. Punched-out ulcer was endoscopic findings for CMV-associated colitis in patients
defined as ulceration with an almost round shape and with moderate to severe UC. These factors may help
clear demarcation. Geographical ulcer was defined as facilitate both the timely diagnosis and treatment of UC
ulceration with an irregular pattern and a branched complicated by CMV infection.
shape. Longitudinal ulcer was defined as ulceration We evaluated total systemic steroid dose over the
with a longitudinal spread along the lumen of the colon. patient’s lifetime, as well as dose over the 4 wk before
Mucosal defect was defined as a wide area of defect with admission, over the previous week before admission,
a longitudinal and/or transverse spread, indicating that and on the day of admission. Between CMV-positive and
more than one-fourth of the mucosa in the endoscopic CMV-negative patients, total systemic steroid dose over
field was defective. The accuracy, sensitivity, specificity, the 4 wk prior to admission (total dose > 400 mg) was
positive predictive value, and negative predictive value an independent risk factor for CMV-associated colitis
for each of these features were determined. Univariate in active UC patients. Furthermore, neither immuno­
analysis revealed that deep ulcer, punched-out ulcer, modulator nor infliximab use was associated with CMV-
geographical ulcer, and spontaneous bleeding were more associated colitis. However, this study included only a few
frequent in CMV-positive patients than in CMV-negative cases treated by immunomodulators or infliximab, and
patients (Table 3). additional studies are required to confirm these results.
Multivariate analysis showed that only punched-out Nonetheless, the finding that immunomodulator and
ulcer was a significant independent predictor of CMV infliximab use did not alter the risk of CMV-associated
colitis (OR = 12.672, 95%CI: 4.210-38.143) (Table 4). colitis is important, because it suggests an alternative

WJGE|www.wjgnet.com 305 March 25, 2016|Volume 8|Issue 6|

 Hirayama Y et al . Cytomegalovirus and ulcerative colitis

A B

C D

E F

G H

I J

WJGE|www.wjgnet.com 306 March 25, 2016|Volume 8|Issue 6|

 Hirayama Y et al . Cytomegalovirus and ulcerative colitis

K L

Figure 2 Endoscopic images of cytomegalovirus-associated colitis in patients with active ulcerative colitis. A-C: Deep ulcer; D-G: Punched-out ulcer; H-J:
Geographical ulcer; K: Longitudinal ulcer; L: Mucosal defect.

treatment regimen for patients with moderate to severe antigenemia. Histology including IHC is considered the
UC rather than using high-dose corticosteroids for objective standard for the diagnosis of CMV infection. In
corticosteroid-refractory cases or corticosteroid-resistant our study, however, among the 34 patients with CMV-
cases. Given that tumor necrosis factor (TNF)-α from associated colitis whose biopsy specimens were stained
monocytes and dendritic cells plays an important role in with HE and a CMV antibody, only 8 patients were
the reactivation of CMV and that infliximab is a potent positive by histology. Only 7 were positive by both CMV
blocker of TNF-α, we consider that this combination antigenemia and hitology. We therefore suggest that our
[7,39]
therapy may be particularly effective . However, the combination of CMV antigenemia and histology including
efficacy of infliximab for UC patients with concomitant IHC for CMV is an appropriate strategy for diagnosis
CMV infection remains controversial, as there have been of CMV infection/CMV-associated colitis in active UC
few case reports and no controlled clinical trials. patients.
Pancolitis was significantly associated with CMV Colonoscopy is usually performed in patients with
infection in active UC, consistent with the theory that exacer­bation of UC symptoms because direct observation
[9]
CMV is prone to proliferate in granulation tissue . Some of the colonic mucosa provides detailed information on
studies reported that CMV was readily found in granu­ disease status and is useful for judging disease severity
lation tissue and tissue from deep ulcers, suggesting that and treatment efficacy. The rapid and accurate diagnosis
CMV can penetrate inflamed mucosa via mononuclear of CMV-associated colitis in UC patients is critical,
[2,9,40,41]
cells and then proliferate in the mucosa . It is thus because its treatment strategy differs markedly from that
possible that a more extensive UC lesion may lead to for UC exacerbation not associated with CMV infection.
wider CMV infection. A few reports have documented the endoscopic findings
In general, there is no clear consensus on the dia­ of CMV-associated colitis, but several failed to find
gnostic criteria for CMV infection in active UC. There are features able to rapidly distinguish CMV-associated colitis
several methods of detecting CMV infection, including from unrelated active UC. Endoscopic findings of UC
histology with IHC, serology, CMV culture, polymerase concomitant with CMV infection can range from normal
chain reaction (PCR) detection of the CMV genome, appearing mucosa to mucosal erosion or ulceration,
[6,34-37,42]
and CMV antigenemia . Each method offers which can be difficult to distinguish from active UC
advantages and disadvantages in the precise diagnosis unrelated to CMV infection. In our study, punched-out
of CMV infection. For example, histological examination ulceration was significantly more frequent in UC patients
is a relatively easy method, but its sensitivity is lower with CMV infection, consistent with reports that CMV
(10%-87%) than PCR. In contrast, PCR for CMV genes is tends to localize to the colon mucosa and granulation
[2,9,40,41]
highly sensitive, but the method is time-consuming and tissue in deep ulcers . Regardless of etiology,
its selectivity is low given the ubiquity of CMV infection. we suggest that a finding of punched-out ulceration
CMV culture is too slow. In contrast, CMV antigenemia is may facilitate the rapid and accurate diagnosis of CMV-
relatively sensitive (60%-100%) and easy to measure associated colitis in UC patients.
within a short period, and has also been used to monitor The limitations of this study include its retrospective
CMV infection in heart transplant recipients and for nature and evaluation of patients at a single institution.
the early diagnosis of CMV infection in renal transplant This study also involved a relatively small number of
[43]
recipients . Moreover, results of CMV antigenemia are patients, which limits its statistical power.
[44,45]
good indication for antiviral therapy . In conclusion, this study suggests that a total corti­
Accordingly, we adopted CMV antigenemia and costeroid dose > 400 mg for 4 wk and extensive colitis
histology, including IHC for CMV, to detect CMV infection are associated with an increased risk of CMV-associated
in our analysis. Results showed that 33 of the 34 CMV- colitis in patients with moderate to severe UC. In
associated colitis patients (97.1%) were positive for CMV addition, punched-out ulceration appears predictive of

WJGE|www.wjgnet.com 307 March 25, 2016|Volume 8|Issue 6|

 Hirayama Y et al . Cytomegalovirus and ulcerative colitis

CMV-associated colitis associated with UC. These clinical 16/0002-9343(61)90105-X]

predictors and specific endoscopic findings may facilitate 9 Cooper HS, Raffensperger EC, Jonas L, Fitts WT. Cytome­
galovirus inclusions in patients with ulcerative colitis and toxic
rapid diagnosis and antiviral treatment.
dilation requiring colonic resection. Gastroenterology 1977; 72:
1253-1256 [PMID: 192627]
10 Orvar K, Murray J, Carmen G, Conklin J. Cytomegalovirus
COMMENTS
COMMENTS infection associated with onset of inflammatory bowel disease.
Dig Dis Sci 1993; 38: 2307-2310 [PMID: 8261839 DOI: 10.1007/
Background
BF01299914]
Although it has been reported that cytomegalovirus (CMV) infection can be
11 Matsuoka K, Iwao Y, Mori T, Sakuraba A, Yajima T, Hisamatsu T,
associated with steroid resistance and be an exacerbating factor in ulcerative
Okamoto S, Morohoshi Y, Izumiya M, Ichikawa H, Sato T, Inoue
colitis (UC), the relationship between CMV and UC is not well studied.
N, Ogata H, Hibi T. Cytomegalovirus is frequently reactivated and
disappears without antiviral agents in ulcerative colitis patients.
Research frontiers Am J Gastroenterol 2007; 102: 331-337 [PMID: 17156136 DOI:
The aim of this study was to identify characteristic endoscopic findings and risk 10.1111/j.1572-0241.2006.00989.x]
factors for CMV-associated colitis in patients with active UC. 12 Lawlor G, Moss AC. Cytomegalovirus in inflammatory bowel
disease: pathogen or innocent bystander? Inflamm Bowel Dis 2010;
Innovations and breakthroughs 16: 1620-1627 [PMID: 20232408 DOI: 10.1002/ibd.21275]
This is one of a few retrospective studies focused on important information 13 Cottone M, Pietrosi G, Martorana G, Casà A, Pecoraro G, Oliva
regarding characteristic endoscopic findings and risk factors for CMV-associated L, Orlando A, Rosselli M, Rizzo A, Pagliaro L. Prevalence of
colitis in patients with active UC. cytomegalovirus infection in severe refractory ulcerative and
Crohn’s colitis. Am J Gastroenterol 2001; 96: 773-775 [PMID:
11280549 DOI: 10.1111/j.1572-0241.2001.03620.x]
Applications 14 Kaufman HS, Kahn AC, Iacobuzio-Donahue C, Talamini MA,
This study suggests that a total corticosteroid dose > 400 mg for 4 wk and Lillemoe KD, Hamilton SR. Cytomegaloviral enterocolitis: clinical
extensive colitis are associated with an increased risk of CMV-associated colitis associations and outcome. Dis Colon Rectum 1999; 42: 24-30
in patients with moderate to severe UC. In addition, punched-out ulceration [PMID: 10211516 DOI: 10.1007/BF02235178]
appears predictive of CMV-associated colitis associated with UC. These clinical 15 Berk T, Gordon SJ, Choi HY, Cooper HS. Cytomegalovirus
predictors and specific endoscopic findings may facilitate rapid diagnosis and infection of the colon: a possible role in exacerbations of inflamma­
antiviral treatment. tory bowel disease. Am J Gastroenterol 1985; 80: 355-360 [PMID:
2859801]
Peer-review 16 Wada Y, Matsui T, Matake H, Sakurai T, Yamamoto J, Kikuchi
An interesting article dealing with clinically relevant subject of risk factors in Y, Yorioka M, Tsuda S, Yao T, Yao S, Haraoka S, Iwashita A.
ulcerative colitis. There is a solid number of patients and good experimental Intractable ulcerative colitis caused by cytomegalovirus infection:
and clinical design. Data are good and discussion is a good representation of a prospective study on prevalence, diagnosis, and treatment. Dis
the problem. Colon Rectum 2003; 46: S59-S65 [PMID: 14530660]
17 Kambham N, Vij R, Cartwright CA, Longacre T. Cytomegalovirus
infection in steroid-refractory ulcerative colitis: a case-control
study. Am J Surg Pathol 2004; 28: 365-373 [PMID: 15104299
REFERENCES DOI: 10.1097/00000478-200403000-00009]
1 Goodgame RW. Gastrointestinal cytomegalovirus disease. Ann 18 Kuwabara A, Okamoto H, Suda T, Ajioka Y, Hatakeyama K.
Intern Med 1993; 119: 924-935 [PMID: 8215005 DOI: 10.7326/00 Clinicopathologic characteristics of clinically relevant cytome­
03-4819-119-9-199311010-00010] galovirus infection in inflammatory bowel disease. J Gastroenterol
2 Hommes DW, Sterringa G, van Deventer SJ, Tytgat GN, Weel 2007; 42: 823-829 [PMID: 17940835 DOI: 10.1007/­s00535-007-2
J. The pathogenicity of cytomegalovirus in inflammatory bowel 103-3]
disease: a systematic review and evidence-based recommendations 19 Domènech E, Vega R, Ojanguren I, Hernández A, Garcia-
for future research. Inflamm Bowel Dis 2004; 10: 245-250 [PMID: Planella E, Bernal I, Rosinach M, Boix J, Cabré E, Gassull MA.
15290919 DOI: 10.1097/00054725-200405000-00011] Cytomegalovirus infection in ulcerative colitis: a prospective,
3 Surawicz CM, Myerson D. Self-limited cytomegalovirus colitis comparative study on prevalence and diagnostic strategy. Inflamm
in immunocompetent individuals. Gastroenterology 1988; 94: Bowel Dis 2008; 14: 1373-1379 [PMID: 18452205 DOI: 10.1002/
194-199 [PMID: 2826283] ibd.20498]
4 Dieterich DT, Rahmin M. Cytomegalovirus colitis in AIDS: 20 Roblin X, Pillet S, Oussalah A, Berthelot P, Del Tedesco E, Phelip
presentation in 44 patients and a review of the literature. J Acquir JM, Chambonnière ML, Garraud O, Peyrin-Biroulet L, Pozzetto
Immune Defic Syndr 1991; 4 Suppl 1: S29-S35 [PMID: 1848619] B. Cytomegalovirus load in inflamed intestinal tissue is predictive
5 Papadakis KA, Tung JK, Binder SW, Kam LY, Abreu MT, Targan of resistance to immunosuppressive therapy in ulcerative colitis.
SR, Vasiliauskas EA. Outcome of cytomegalovirus infections in Am J Gastroenterol 2011; 106: 2001-2008 [PMID: 21788989 DOI:
patients with inflammatory bowel disease. Am J Gastroenterol 10.1038/ajg.2011.202]
2001; 96: 2137-2142 [PMID: 11467645 DOI: 10.1111/­j.1572-0241. 21 Kojima T, Watanabe T, Hata K, Shinozaki M, Yokoyama T,
2001.03949.x] Nagawa H. Cytomegalovirus infection in ulcerative colitis. Scand
6 de la Hoz RE, Stephens G, Sherlock C. Diagnosis and treatment J Gastroenterol 2006; 41: 706-711 [PMID: 16716970 DOI: 10.108
approaches of CMV infections in adult patients. J Clin Virol 2002; 0/00365520500408584]
25 Suppl 2: S1-12 [PMID: 12361752 DOI: 10.1016/­S1386-6532(0 22 Nakase H, Matsumura K, Yoshino T, Chiba T. Systematic review:
2)00091-4] cytomegalovirus infection in inflammatory bowel disease. J
7 Pereyra F, Rubin RH. Prevention and treatment of cytome­ Gastroenterol 2008; 43: 735-740 [PMID: 18958541 DOI: 10.1007/
galovirus infection in solid organ transplant recipients. Curr Opin s00535-008-2246-x]
Infect Dis 2004; 17: 357-361 [PMID: 15241082 DOI: 10.1097/01. 23 Nakase H, Yoshino T, Ueno S, Uza N, Mikami S, Matsuura
qco.0000136933.67920.dd] M, Chiba T. Importance of early detection of cytomegalovirus
8 Powell RD, Warner NE, Levine RS, Kirsner JB. Cytomegalic infection in refractory inflammatory bowel disease. Inflamm Bowel
inclusion disease and ulcerative colitis; report of a case in a young Dis 2007; 13: 364 [PMID: 17206718 DOI: 10.1002/ibd.20033]
adult. Am J Med 1961; 30: 334-340 [PMID: 13737621 DOI: 10.10 24 Franzin G, Muolo A, Griminelli T. Cytomegalovirus inclusions in

WJGE|www.wjgnet.com 308 March 25, 2016|Volume 8|Issue 6|

 Hirayama Y et al . Cytomegalovirus and ulcerative colitis

the gastroduodenal mucosa of patients after renal transplan­tation. Bone Marrow Transplant 2004; 33: 431-434 [PMID: 14676775
Gut 1981; 22: 698-701 [PMID: 6271652 DOI: 10.1136/­gut.22.9.698] DOI: 10.1038/sj.bmt.1704369]
25 Battaglino MP, Rockey DC. Cytomegalovirus colitis presenting 36 Nagata N, Kobayakawa M, Shimbo T, Hoshimoto K, Yada T,
with the endoscopic appearance of pseudomembranous colitis. Gotoda T, Akiyama J, Oka S, Uemura N. Diagnostic value of
Gastrointest Endosc 1999; 50: 697-700 [PMID: 10536332 DOI: antigenemia assay for cytomegalovirus gastrointestinal disease in
10.1016/S0016-5107(99)80025-X] immunocompromised patients. World J Gastroenterol 2011; 17:
26 Nishimoto Y, Matsumoto T, Suekane H, Shimizu M, Mikami Y, 1185-1191 [PMID: 21448424 DOI: 10.3748/wjg.v17.i9.1185]
Iida M. Cytomegalovirus infection in a patient with ulcerative 37 Beaugerie L, Cywiner-Golenzer C, Monfort L, Girard PM,
colitis: colonoscopic findings. Gastrointest Endosc 2001; 53: Carbonnel F, Ngô Y, Cosnes J, Rozenbaum W, Nicolas JC, Châtelet
816-818 [PMID: 11375602 DOI: 10.1067/mge.2001.114955] FP, Gendre JP. Definition and diagnosis of cytomegalovirus colitis
27 Falagas ME, Griffiths J, Prekezes J, Worthington M. Cytome­ in patients infected by human immunodeficiency virus. J Acquir
galovirus colitis mimicking colon carcinoma in an HIV-negative Immune Defic Syndr Hum Retrovirol 1997; 14: 423-429 [PMID:
patient with chronic renal failure. Am J Gastroenterol 1996; 91: 9170416 DOI: 10.1097/00042560-199704150-00005]
168-169 [PMID: 8561127] 38 Annese V, Daperno M, Rutter MD, Amiot A, Bossuyt P, East J,
28 Suzuki H, Kato J, Kuriyama M, Hiraoka S, Kuwaki K, Yamamoto Ferrante M, Götz M, Katsanos KH, Kießlich R, Ordás I, Repici A,
K. Specific endoscopic features of ulcerative colitis complicated Rosa B, Sebastian S, Kucharzik T, Eliakim R; European Crohn's
by cytomegalovirus infection. World J Gastroenterol 2010; 16: and Colitis Organisation. European evidence based consensus for
1245-1251 [PMID: 20222169 DOI: 10.3748/wjg.v16.i10.1245] endoscopy in inflammatory bowel disease. J Crohns Colitis 2013; 7:
29 Omiya M, Matsushita M, Tanaka T, Kawamata S, Okazaki K. The 982-1018 [PMID: 24184171 DOI: 10.1016/j.crohns.2013.09.016]
absence of large ulcer predicts latent cytomegalovirus infection 39 Nakase H, Chiba T. TNF-alpha is an important pathogenic factor
in ulcerative colitis with positive mucosal viral assay. Intern contributing to reactivation of cytomegalovirus in inflamed mucosa
Med 2010; 49: 2277-2282 [PMID: 21048360 DOI: 10.2169/ of colon in patients with ulcerative colitis: lesson from clinical
internalmedicine.49.3657] experience. Inflamm Bowel Dis 2010; 16: 550-551 [PMID: 19637380
30 Iida T, Ikeya K, Watanabe F, Abe J, Maruyama Y, Ohata A, DOI: 10.1002/ibd.210]
Teruyuki S, Sugimoto K, Hanai H. Looking for endoscopic 40 Yoshino T, Nakase H, Ueno S, Uza N, Inoue S, Mikami S,
features of cytomegalovirus colitis: a study of 187 patients with Matsuura M, Ohmori K, Sakurai T, Nagayama S, Hasegawa S,
active ulcerative colitis, positive and negative for cytomegalovirus. Sakai Y, Chiba T. Usefulness of quantitative real-time PCR assay
Inflamm Bowel Dis 2013; 19: 1156-1163 [PMID: 23619714 DOI: for early detection of cytomegalovirus infection in patients with
10.1097/MIB.0b013e31828075ce] ulcerative colitis refractory to immunosuppressive therapies.
31 Stange EF, Travis SP, Vermeire S, Reinisch W, Geboes K, Inflamm Bowel Dis 2007; 13: 1516-1521 [PMID: 17828781 DOI:
Barakauskiene A, Feakins R, Fléjou JF, Herfarth H, Hommes DW, 10.1002/ibd.20253]
Kupcinskas L, Lakatos PL, Mantzaris GJ, Schreiber S, Villanacci V, 41 Pfau P, Kochman ML, Furth EE, Lichtenstein GR. Cytome­
Warren BF. European evidence-based Consensus on the diagnosis galovirus colitis complicating ulcerative colitis in the steroid-naive
and management of ulcerative colitis: Definitions and diagnosis. patient. Am J Gastroenterol 2001; 96: 895-899 [PMID: 11280572
J Crohns Colitis 2008; 2: 1-23 [PMID: 21172194 DOI: 10.1016/­j. DOI: 10.1111/j.1572-0241.2001.03672.x]
crohns.2007.11.001] 42 Kishore J, Ghoshal U, Ghoshal UC, Krishnani N, Kumar S, Singh
32 Truelove SC, Witts LJ. Cortisone in ulcerative colitis; final report M, Ayyagari A. Infection with cytomegalovirus in patients with
on a therapeutic trial. Br Med J 1955; 2: 1041-1048 [PMID: inflammatory bowel disease: prevalence, clinical significance and
13260656 DOI: 10.1136/bmj.2.4947.1041] outcome. J Med Microbiol 2004; 53: 1155-1160 [PMID: 15496396
33 Dignass A, Eliakim R, Magro F, Maaser C, Chowers Y, Geboes DOI: 10.1099/jmm.0.45629-0]
K, Mantzaris G, Reinisch W, Colombel JF, Vermeire S, Travis 43 Bernabeu-Wittel M, Pachón-Ibáñez J, Cisneros JM, Cañas E,
S, Lindsay JO, Van Assche G. Second European evidence-based Sánchez M, Gómez MA, Gentil MA, Pachón J. Quantitative
consensus on the diagnosis and management of ulcerative colitis pp65 antigenemia in the diagnosis of cytomegalovirus disease:
part 1: definitions and diagnosis. J Crohns Colitis 2012; 6: 965-990 prospective assessment in a cohort of solid organ transplant
[PMID: 23040452 DOI: 10.1016/j.crohns.2012.09.003] recipients. J Infect 2005; 51: 188-194 [PMID: 16230214 DOI:
34 Mazzulli T, Drew LW, Yen-Lieberman B, Jekic-McMullen D, 10.1016/j.jinf.2004.10.014]
Kohn DJ, Isada C, Moussa G, Chua R, Walmsley S. Multicenter 44 Manteiga R, Martino R, Sureda A, Labeaga R, Brunet S, Sierra J,
comparison of the digene hybrid capture CMV DNA assay (version Rabella N. Cytomegalovirus pp65 antigenemia-guided pre-emptive
2.0), the pp65 antigenemia assay, and cell culture for detection treatment with ganciclovir after allogeneic stem transplantation:
of cytomegalovirus viremia. J Clin Microbiol 1999; 37: 958-963 a single-center experience. Bone Marrow Transplant 1998; 22:
[PMID: 10074509] 899-904 [PMID: 9827819 DOI: 10.1038/sj.bmt.1701439]
35 Mori T, Mori S, Kanda Y, Yakushiji K, Mineishi S, Takaue Y, 45 Boeckh M, Gooley TA, Myerson D, Cunningham T, Schoch G,
Gondo H, Harada M, Sakamaki H, Yajima T, Iwao Y, Hibi T, Bowden RA. Cytomegalovirus pp65 antigenemia-guided early
Okamoto S. Clinical significance of cytomegalovirus (CMV) treatment with ganciclovir versus ganciclovir at engraftment after
antigenemia in the prediction and diagnosis of CMV gastrointestinal allogeneic marrow transplantation: a randomized double-blind
disease after allogeneic hematopoietic stem cell transplantation. study. Blood 1996; 88: 4063-4071 [PMID: 8916975]

P- Reviewer: Landsman MJ, Ma XP, Vetvicka V

S- Editor: Song XX L- Editor: A E- Editor: Liu SQ

WJGE|www.wjgnet.com 309 March 25, 2016|Volume 8|Issue 6|

 Published by Baishideng Publishing Group Inc
8226 Regency Drive, Pleasanton, CA 94588, USA
Telephone: +1-925-223-8242
Fax: +1-925-223-8243
E-mail: bpgoffice@wjgnet.com
Help Desk: http://www.wjgnet.com/esps/helpdesk.aspx
http://www.wjgnet.com

© 2016 Baishideng Publishing Group Inc. All rights reserved.