- Email: [email protected]
Impact of mucosal inflammation on risk of colorectal neoplasia in patients with ulcerative colitis: a systematic review and meta-analysis
Accepted Manuscript Impact of Mucosal Inflammation on Risk of Colorectal Neoplasia in Patients with Ulcerative Colitis: A Systematic Review and Meta-Analysis Brisas M. Flores, Anthony OʼConnor, Alan C. Moss PII:
S0016-5107(17)32121-1
DOI:
10.1016/j.gie.2017.07.028
Reference:
YMGE 10675
To appear in:
Gastrointestinal Endoscopy
Received Date: 4 May 2017 Revised Date:
0016-5107 0016-5107
Accepted Date: 14 July 2017
Please cite this article as: Flores BM, OʼConnor A, Moss AC, Impact of Mucosal Inflammation on Risk of Colorectal Neoplasia in Patients with Ulcerative Colitis: A Systematic Review and Meta-Analysis, Gastrointestinal Endoscopy (2017), doi: 10.1016/j.gie.2017.07.028. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Impact of Mucosal Inflammation on Risk of Colorectal Neoplasia in Patients with
Brisas M. Flores1, Anthony OʼConnor2, Alan C Moss1
RI PT
Ulcerative Colitis: A Systematic Review and Meta-Analysis
1. Division of Gastroenterology, Beth Israel Deaconess Medical Center, Boston, MA
SC
2. Leeds Gastroenterology Institute, St James's University Hospital, Leeds
Alan C Moss MD Division of Gastroenterology
M AN U
Contact:
Beth Israel Deaconess Medical Center 330 Brookline Ave
MA 02215 617-667-3197
TE D
Boston
AC C
EP
[email protected]
ACCEPTED MANUSCRIPT 1
Impact of Mucosal Inflammation on Risk of Colorectal Neoplasia in Patients with
Brisas M. Flores1, Anthony OʼConnor2, Alan C Moss1
RI PT
Ulcerative Colitis: A Systematic Review and Meta-Analysis
1. Division of Gastroenterology, Beth Israel Deaconess Medical Center, Boston, MA
Contact: Alan C Moss MD Division of Gastroenterology
M AN U
SC
2. Leeds Gastroenterology Institute, St James's University Hospital, Leeds
Beth Israel Deaconess Medical Center
Boston MA 02215
EP
617-667-3197
TE D
330 Brookline Ave
AC C
[email protected]
ACCEPTED MANUSCRIPT 2 ABSTRACT Background and Aims: Long-standing ulcerative colitis is an established risk factor for colorectal neoplasia. A number of observational studies have suggested that evidence
RI PT
of mucosal inflammation (endoscopic or histologic) is associated with a greater risk for colorectal neoplasia than is mucosal healing. Our goal was to systematically analyze
inflammation to better inform surveillance strategies.
SC
the risk of colorectal neoplasia in ulcerative colitis patients with ongoing mucosal
Methods: We performed a systematic review and meta-analysis of the effect of
M AN U
endoscopic and/or histologic inflammation on the risk of colorectal neoplasia in cohort and case-control studies. Sensitivity analyses for study setting and case definition were performed.
Results: Six studies met the inclusion criteria, incorporating outcomes in 1,443 patients.
TE D
No study used a single validated measure for mucosal inflammation. The pooled odds ratio for colorectal neoplasia was 3.5 (95% CI, 2.6 - 4.8; p < 0.001) in those with any mucosal inflammation, and 2.6 (95% CI, 1.5-4.5; p = 0.01) in those with histologic
EP
inflammation, when compared to those with mucosal healing. Overall quality of the studies was good.
AC C
Conclusions: The presence of objective evidence of mucosal inflammation during follow-up in patients with ulcerative colitis is associated with a greater risk of subsequent colorectal neoplasia than in those with mucosal healing. This risk factor should be considered in guidelines on surveillance intervals for these patients.
ACCEPTED MANUSCRIPT 3 Keywords: inflammatory bowel disease; ulcerative colitis; histologic inflammation; colorectal neoplasia INTRODUCTION
RI PT
It is well-established that inflammatory bowel disease (IBD), including both ulcerative colitis (UC) and Crohn’s disease, puts one at increased risk of colorectal neoplasia (CRN). For ulcerative colitis, the cumulative risk of colorectal cancer (CRC) is estimated
SC
at 18% after 30 years1. Recognition of this adverse event has led to the incorporation of colonoscopic surveillance for dysplasia into clinical practice in IBD1-4. During
M AN U
surveillance, the prevalence of low-grade dysplasia (LGD) is approximately 10%, and 3% of patients with LGD progress to high-grade dysplasia or cancer5. The guidelines on this practice have not included specific recommendations based on the presence or absence of mucosal inflammation during follow-up.
TE D
Risk factors for the development of neoplasia in IBD are multiple and include duration of disease, extent of disease, family history of CRC, and history of primary sclerosing cholangitis1,3,4. Endoscopic findings associated with cancer risk include
EP
multiple pseudopolyps, strictures, and low-grade dysplasia on biopsy3,6. A number of retrospective studies have also concluded that the presence of any endoscopic or
AC C
histologic inflammation confers a greater risk of CRC by itself7-12. However, most of these studies were relatively small, and provided varying estimates of risk. This issue warrants further attention, as recent studies by our group and others have reported that greater than 40% of patients with UC have ongoing endoscopic or histologic inflammation at the time of colonoscopy, despite being in clinical remission13. If these
ACCEPTED MANUSCRIPT 4 factors have implications for prognosis and cancer risk stratification, then a quantification of their magnitude may be important. The aim of this paper is to use systematic review and meta-analysis to quantify
RI PT
the risk of colorectal neoplasia in the setting of endoscopic and/or histologic inflammation in patients with UC. We hypothesize that the independent risk colonic inflammation has on the development of CRC is significant, and could be used to adjust
SC
surveillance patterns accordingly in at risk individuals. It is currently recommended that surveillance colonoscopy begin after 8 to 10 years of disease and then be done every 1
M AN U
to 2 years, whether there is objective evidence of ongoing chronic inflammation or not1.
METHODS
Literature Search
TE D
This systematic review and meta-analysis was conducted with guidance provided by the Cochrane Handbook for Systematic Reviews14. It is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines15. A
language,
for
EP
literature search was performed to identify all published and unpublished studies, in any consideration
of
inclusion
studies
with
information
regarding
AC C
clear histologic or endoscopic inflammatory scores and the development of colorectal neoplasia in patients with UC. A systematic search of the following databases was performed: MEDLINE [PubMed], Web of Science, Allied Health Literature, and EMBASE. The following search strategy was constructed by using a combination of MeSH
subject
headings
and
text-words: "inflammation,"
"histologic
inflammation," "inflammatory bowel disease," "ulcerative colitis," "risk," "colorectal
ACCEPTED MANUSCRIPT 5 dysplasia," "colorectal neoplasia." There were no non-English articles or abstracts that required translation.
RI PT
Inclusion and Exclusion Criteria
We included observational studies in this meta-analysis that met the following inclusion criteria:
Participants: studies performed in patients with ulcerative colitis.
(2)
Risk Factor: described histologic or endoscopic histologic inflammatory activity.
(3)
Comparators: patients with UC with no histologic or endoscopic inflammatory
M AN U
SC
(1)
activity. (4)
Outcomes: odds ratio (OR) and 95% confidence intervals (CI) of development of
either colorectal cancer or non-cancer dysplasia.
Studies: reported with clear definitions on rates of colorectal dysplasia and
TE D
(5)
cancer, and provided sufficient data to allow estimation of effect size. Inclusion was not otherwise restricted by study size, language, or publication type. When there were
EP
multiple publications from the same cohort, data from the most recent comprehensive
AC C
report were included.
Study Selection
Two authors independently scanned the abstract of every trial identified by the search to determine eligibility. Full articles were further assessed if they met inclusion criteria based upon the abstract, or if the abstract was unclear and the full article was needed to determine if it met inclusion criteria. Articles not meeting the inclusion criteria were
ACCEPTED MANUSCRIPT 6 excluded. All final included and excluded studies were reviewed by the senior
Data extraction and quality assessment
RI PT
investigator [ACM].
The following data were retrieved from the included studies: number of patients in the study, type of setting, inflammation criteria and grading system (histologic and/or
SC
endoscopic), odds ratio or hazards ratio for neoplasia risk, and duration of follow-up. Assessment of quality was performed according to the Newcastle-Ottawa Scale (NOS)
M AN U
guidelines, with scores greater than 5 being deemed acceptable16.
Statistical analysis
Our analysis focused on the risk of the development of CRN as defined by cancer or
TE D
dysplasia among patients with UC, based on whether or not they had evidence of mucosal inflammation on colonoscopy (endoscopic and/or histologic). The random effects model as proposed by DerSimonian and Laird was used to calculate pooled OR
EP
and 95% confidence intervals17. The effect estimate used was the OR, as given the rare
AC C
nature of CRN among the UC cohort, it was considered to be a suitable estimate of relative risk. Potential publication bias was examined by funnel plots. Heterogeneity was examined by the Q statistic and the I2 statistic18. The Q and I2 statistics were used to test statistical heterogeneity among
studies18. For the Q statistic, a P value of less than 0.1 is considered representative of statistically significant heterogeneity. An I2 index of around 25% is considered to demonstrate low levels of heterogeneity, 50% medium, and 75% high18. Sensitivity
ACCEPTED MANUSCRIPT 7 analyses were conducted, omitting each study in turn to screen for outliers with respect to results, study population, study design, or type of 5-ASA exposure. Analyses were performed using the STATA statistical package (Version 12.0; STATA Corporation,
RI PT
College Station, Tex, USA), and RevMan (Review Manager, Version 5.3. Copenhagen:
RESULTS Literature Search
SC
The Nordic Cochrane Centre, The Cochrane Collaboration, 2014).
M AN U
The search yielded 134 articles of potential relevance. After further text review (Figure 1), six met the inclusion criteria, incorporating outcomes in 1,443 patients7-12. All 6 articles were published as full articles7-12. The characteristics of the included articles are described in Table 1. Two of the 6 studies were retrospective observational cohort
TE D
studies7,8, and 4 of the 6 were retrospective case control studies9-12. There was no consensus score used for histologic inflammation grading, and each study used its own criteria to be categorized as having “inflammation.” The reasons for exclusion of the
EP
other 128 studies were outline in Supplementary Table 1. The quality assessment
AC C
scores are detailed in Supplementary Table 2.
Risk of development of colorectal neoplasia in the setting of inflammation Four studies reported raw data on the rates of CRN in patients with or without mucosal inflammation8-11. There were 1025 patients included in these studies, with a range of mean duration of disease from 7 to 20 years. “Mucosal inflammation” included any histologic inflammation or endoscopic inflammation. The pooled odds ratio for colorectal
ACCEPTED MANUSCRIPT 8 neoplasia (dysplasia or cancer) in patients with any mucosal inflammation at time of colonoscopy (Figure 2) was 3.5 (95% CI, 2.6 - 4.8; p<0.001). There was no significant statistical heterogeneity (I2=58%, p=0.07).
RI PT
Five studies reported odds ratios for CRN in patients on the basis of histologic inflammation only (Figure 3)7-11. The pooled odds ratio was 2.6 (95% CI, 1.5 – 4.5) across the studies. Notably, there was significant statistical heterogeneity (I2=69.2%,
SC
p=0.011). The two studies with the greatest odds ratio for this outcome differed from other studies in that they reported histologic inflammation as a continuous variable11 or
M AN U
only the highest degree of inflammation9. In contrast, the other 3 studies considered patients with any histologic abnormalities as having “mucosal inflammation.”7,8,10. Supplementary Figure 1 shows the funnel plot of the log of odds ratio versus standard error for all studies.
TE D
There were insufficient studies included to perform meaningful sensitivity analyses, or analyze the risk of CRN based on endoscopic scores alone. No studies adjusted odds ratios for other CRN risk factors such as primary sclerosing cholangitis,
EP
family history, or smoking status. Only the paper by Rubin et al adjusted for medication use in their analyses10. In the studies of histological grade, there was heterogeneity in
AC C
the methods of reporting severity of inflammation; mean and upper limits, per-unit increase, or ordinal scales were used7-11.
DISCUSSION
There has been much focus in recent years on “mucosal healing” as a goal of therapy for patients with IBD. The rationale for this is the association between endoscopic
ACCEPTED MANUSCRIPT 9 evidence of inflammation and adverse outcomes such as disease relapse, hospitalization, surgery, and cancer risk in patients with ulcerative colitis19-21. In particular, the severity of mucosal inflammation, or some of its features, has been
RI PT
independently associated with the risk of colorectal neoplasia in some studies7,12. Unfortunately, these studies were mostly retrospective studies from hospital based cohorts, and subject to the associated biases of this methodology and selection.
SC
Colman and Rubin (2016) recently published a systematic review emphasizing the relationship between histologic inflammation and colorectal neoplasia in patients with
M AN U
ulcerative colitis, however no meta-analysis was done22.
This meta-analysis emphasizes that in those with ulcerative colitis, histologic or endoscopic inflammation puts one at increased risk for developing colorectal neoplasia. Our results suggest that the pooled risk of CRN is more than 2-fold greater in patients
TE D
with established UC who have mucosal inflammation during follow-up colonoscopy than in those who have no or minimal inflammation during colonoscopy. The consistency of this effect persists across studies, and the lack of statistical heterogeneity provides
EP
confidence in this conclusion. When inflammation is based on histological abnormalities only, the risk of CRN remains elevated when compared with patients with normal or mild
AC C
histological abnormalities.
Our analysis does have its limitations. All studies were retrospective, and there
was no universal definition of histologic or endoscopic inflammation. There was statistical heterogeneity in the pooled odds ratio, which reflects these differences in methodology. Nieminen et al included both patients with ulcerative colitis and Crohn’s disease, further adding to potential heterogeneity9. The consistent impact, regardless of
ACCEPTED MANUSCRIPT 10 method of describing inflammation, suggests a biologically robust effect. All of the studies were done in a tertiary hospital setting, with associated bias in population risk of cancer. This may limit generalizing these findings to the general IBD population in the
alone to calculate risk and draw meaningful conclusions.
RI PT
community. In addition, there were insufficient data regarding endoscopic inflammation
Despite the limitations, this analysis confirms the implication that inflammation
SC
confers a true risk in the development of neoplasia in ulcerative colitis patients. This may be clinically relevant, as current guidelines recommend colorectal cancer screening
M AN U
every one to two years, regardless of factors such as inflammation severity1. The risk of colorectal neoplasia in patients with ulcerative colitis is up to 18% after having the disease for 30 years1, and as evidence continues to arise that inflammation increases the risk for progression to neoplasia, physicians may consider adjusting screening
TE D
intervals in individuals who lack mucosal healing at surveillance colonoscopy. In addition, this finding may further encourage physicians to aim for resolution of mucosal inflammation, rather than just clinical remission. Whether achieving mucosal healing via
EP
maintenance therapy will reduce an individual’s risk of cancer is unknown, although mesalamine appears to have a modest benefit in this regard23. Future studies on this
AC C
topic should clearly define “inflammation” using validated histological or endoscopic indices of inflammation. Risks of either dysplasia or cancer should be adjusted for factors such as duration of disease, presence of primary sclerosing cholangitis, family history, and chemoprevention. In particular, research should explore how particular treatments for UC impact the rates of colorectal neoplasia in those who have achieved remission of inflammation, when compared with those who have not.
ACCEPTED MANUSCRIPT 11 In conclusion, this meta-analysis confirms a pooled effect of ongoing colonic mucosal inflammation on risk of colorectal neoplasia in patients with ulcerative colitis. This association should inform treatment goals and surveillance strategies for this
RI PT
patient population.
References
SC
1. Farraye FA, Odze RD, Eaden J et al. AGA medical position statement on the diagnosis and management of colorectal neoplasia in inflammatory bowel disease.
M AN U
Gastroenterology. 2010;138(2):738-745.
2. Laine L, Kaltenbach T, Barkun A et al. SCENIC international consensus statement on surveillance
and
management
of
dysplasia
in
inflammatory
bowel
disease.
Gastroenterology. 2015;148(3):639-651.e28.
clinicopathologic
TE D
3. Rutter MD, Riddell RH. Colorectal dysplasia in inflammatory bowel disease: A perspective.
2014;12(3):359-367.
Clinical
gastroenterology
and
hepatology.
EP
4. Ananthakrishnan AN, Cagan A, Cai T et al. Colonoscopy is associated with a reduced risk for colon cancer and mortality in patients with inflammatory bowel
AC C
diseases. Clinical gastroenterology and hepatology. 2015;13(2):322-329. 5. Thomas T, Abrams KA, Robinson RJ et al. Meta-analysis: cancer risk of low-grade dysplasia in chronic ulcerative colitis. Alimentary Pharmacology and Therapeutics. 2007;25(6):657-68.
6. Dyson J, Rutter M. Colorectal cancer in inflammatory bowel disease: What is the real magnitude of the risk? World journal of gastroenterology. 2012;18(29):3839-3848.
ACCEPTED MANUSCRIPT 12 7. Gupta RB, Harpaz N, Itzkowitz S et al. Histologic inflammation is a risk factor for progression
to
colorectal
neoplasia
in
ulcerative
colitis:
a
cohort
study.
Gastroenterology. 2007;133(4):1099-1105.
carcinoma
in
extensive
ulcerative
colitis.
World
2014;20(17):4980-4986.
RI PT
8. Korelitz BI, Sultan K, Kothari M et al. Histological healing favors lower risk of colon journal
of
gastroenterology.
SC
9. Nieminen U, Jussila A, Nordling S et al. Inflammation and disease duration have a cumulative effect on the risk of dysplasia and carcinoma in IBD: A case–control
M AN U
observational study based on registry data. International journal of cancer. 2014;134:189–196.
10. Rubin DT, Huo D, Kinnucan JA et al. Inflammation is an independent risk factor for colonic neoplasia in patients with ulcerative colitis: a case-control study. Clinical
TE D
gastroenterology and hepatology. 2013;11(12):1601-8.
11. Rutter M, Saunders B, Wilkinson K et al. Severity of inflammation is a risk factor for colorectal neoplasia in ulcerative colitis. Gastroenterology. 2004;126(2):451-459.
ulcerative
EP
12. Rutter MD, Saunders BP, Wilkinson KH et al. Cancer surveillance in longstanding colitis:
endoscopic
appearances
help
predict
cancer
risk.
Gut.
AC C
2004;53(12):1813-1816.
13. Rosenberg L, Nanda KS, Zenlea T et al. Histologic markers of inflammation in patients with ulcerative colitis in clinical remission. Clinical gastroenterology and hepatology. 2013;11(8):991. 14. Higgins JPT. Cochrane handbook for systematic reviews of interventions. http://www.cochrane-handbook.org. Updated 2011.
ACCEPTED MANUSCRIPT 13 15. Moher D, Liberati A, Tetzlaff J et al. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. British medical journal. 2009;339(4):b2535.
RI PT
16. Wells G, Shea B, O'Connell J et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analysis. 2011. [Website]. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. Available from: URL:
SC
http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp.
17. DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled clinical trials.
M AN U
1986;7(3):177-88.
18. Higgins JPT, Thompson SG, Deeks JJ et al. Measuring inconsistency in metaanalyses. British medical journal. 2003;327(7414):557-560.
19. Ardizzone S, Cassinotti A, Duca P et al. Mucosal healing predicts late outcomes
TE D
after the first course of corticosteroids for newly diagnosed ulcerative colitis. Clinical gastroenterology and hepatology. 2011;9(6):483-489. 20. Bessissow T, Lemmens B, Ferrante M et al. Prognostic value of serologic and
EP
histologic markers on clinical relapse in ulcerative colitis patients with mucosal healing. American journal of gastroenterology. 2012;107(11):1684-92.
AC C
21. Bitton A, Peppercorn MA, Antonioli DA et al. Clinical, biological, and histologic parameters
as
predictors
of
relapse
in
ulcerative
colitis.
Gastroenterology.
2001;120(1):13-20.
22. Colman RJ, Rubin DT. Histological inflammation increases the risk of colorectal neoplasia in ulcerative colitis: a systematic review. Intestinal Research. 2016;14(3):202-10.
ACCEPTED MANUSCRIPT 14 23. OʼConnor A, Packey CD, Akbari M et al . Mesalamine, but not sulfasalazine, reduces the risk of colorectal neoplasia in patients with inflammatory bowel disease: an
2015;21(11):2562-9.
Figure 1. Results of the literature search.
SC
FIGURE LEGENDS
RI PT
agent-specific systematic review and meta-analysis. Inflammatory bowel disease.
M AN U
Figure 2. Pooled odds ratio (OR) for colorectal neoplasia if there is presence of either histologic or endoscopic inflammation8-11. Odds ratio along with confidence intervals (CI) are provided.
Figure 3. Pooled odds ratio for colorectal neoplasia if there is presence of histologic
AC C
EP
TE D
inflammation7-11. Odds ratio along with confidence intervals (CI) are provided.
ACCEPTED MANUSCRIPT 15
TABLES Table 1. Characteristics of included studies. Type
Design
N
n (cases)
Dysplasia Definition
Inflammation Score
Inflammation definition
Gupta
2007
Paper
Retrospective observational cohort
418
224
Histologic Activity Index: Grade 0-3
Score of 1 or greater
Korelitz8
2014
Paper
Retrospective observational cohort
68
20
Low grade dysplasia, high grade dysplasia, cancer High grade dysplasia, cancer
2014
Paper
Retrospective case-control
506
168
Dysplasia, cancer
Microscopic inflammation in the absence of gross inflammation Score of 1 or greater
>20 years
Nieminen9
Rubin
10
2013
Paper
Retrospective case-control
200
59
Neoplasia, cancer
Score of 2 or greater
11 years
Rutter11
2004
Paper
Retrospective case-control
204
68
Adenoma, low grade dysplasia, high grade dysplasia, cancer
Microscopic inflammation in the absence of gross inflammation Histologic Activity Index: Grade 0-3 Histologic Inflammatory Activity: Grade 0-5 Histologic Inflammation: Grade 0-4
Unclear, assess at 1 unit increases
14 years
Presence or absence of scarring, postinflammatory polyps, backwash ileitis, shortened colon, tubular appearance, featureless colon
14 years
Paper
Retrospective case-control
EP
2004
AC C
Rutter12
204
SC
M AN U
7
68
Adenoma, low grade dysplasia, high grade dysplasia, cancer
Duration of followup 6.7 years
RI PT
Year
TE D
Author
Colonoscopic Inflammation: Grade 0-4 Presence of scarring, postinflammatory polyps, backwash ileitis, shortened colon, tubular appearance, featureless colon
12 years
ACCEPTED MANUSCRIPT 16
SUPPLEMENTARY TABLES Supplementary Table 1. Excluded studies and reason for exclusion.
1981 No HIA
TE D
No HIA; No OR/RR of CA; Review Mixed CD and UC; No HIA No HIA No HIA Mixed CD and UC; No HIA No HIA Mixed CD and UC; No HIA No HIA No HIA No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA Mixed CD and UC; No HIA No HIA; No OR/RR of CA; Review No HIA; Review No HIA Mixed CD and UC; No HIA; Review No HIA No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA
EP
2014 2008 1989 1986 2009 2002 2007 2005 2013 2004 2014 2014 1984 2012 2004 1989 1992 2011 2010 2010 2014
SC
Reason for exclusion No HIA; No OR/RR of CA No HIA No HIA; No OR/RR of CA No HIA No HIA Mixed CD and UC; No HIA Mixed CD and UC; No HIA No HIA No HIA
M AN U
Year 2015 1989 2010 2009 2011 2010 2006 1995 1989
AC C
Study Aalykke et al Albert et al Allen et al Awais et al Baars et al Bergeron et al Bernstein Biasco et al Binder Blackstone et al Bressenot et al Breynaert et al Broström Broström Campbell et al Campbell et al Cendan et al Chen et al Chen YX et al Clevers Connelly et al DeRoche et al Dobbins Dyson et al Eaden Farmer Fischbach Frego et al Fujita et al Fujita T et al Gallinger et al
RI PT
OR* odds ratio. RR* relative risk. CA* cancer. HIA* histologic inflammatory activity. UC* ulcerative colitis. CD* Crohn’s disease.
ACCEPTED MANUSCRIPT
2000 2012 2000 2000 2014 2013
No HIA No HIA No HIA Mixed CD and UC; No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA
2015 No HIA; No OR/RR of CA 2004 No HIA; Review 2004 No HIA
TE D
M AN U
No HIA No HIA No HIA No HIA Mixed CD and UC; No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA Mixed CD and UC; No HIA Mixed CD and UC; No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA Mixed CD and UC; No HIA No HIA No HIA No HIA No HIA No HIA
EP
1993 1994 2007 2006 2002 2013 2010 2007 2005 2010 2014 2001 1995 2004 2006 1989 1991
SC
2007 Mixed CD and UC; No HIA 2007 Mixed CD and UC; No HIA; Review
AC C
Geboes Gong W Gorfine et al Greenstein Harpaz et al Harpaz et al Henriksen et al Herszeny et al Herszényi et al Itzkowitz et al Itzkowitz et al Jablonská et al Jacobsen et al Jess et al Jess et al Judge et al Julka et al Kallesøe et al Katsanos et al Katsanos et al Kekilli et al Kiran et al Konecný et al Konishi F et al Krok et al Lakatos et al Lashner et al Leidenius et al LennardJones LennardJones et al Lindberg et al Löfberg et al Loftus Lukas Lukas M et al Mathy et al Matsui et al
1983 No HIA
1977 2006 1989 2006 2010 2010 2003 1993
RI PT
17
No HIA No HIA No HIA No HIA; Review No HIA; No OR/RR of CA No HIA; No OR/RR of CA; Review No HIA No HIA
ACCEPTED MANUSCRIPT 18 No HIA No HIA Mixed CD and UC; No HIA; Review No HIA
2013 1995 2014 1979 1994 2010 2000 2008
No HIA No HIA No HIA; OR/RR of CA from dysplasia only No HIA No HIA No HIA No HIA No HIA; No OR/RR of CA; Review
1985 2013 2011 1984 2003 2011 2012 2014 2008 2006 2014 2003 2000 1999
No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA No HIA No HIA No HIA; Review No HIA No HIA; No OR/RR of CA No HIA No HIA; Review No HIA; No OR/RR of CA Mixed CD and UC ; No HIA; Review No HIA No HIA
EP
TE D
M AN U
SC
RI PT
1999 1985 2003 2014
2013 No HIA; No OR/RR of CA 2013 No HIA; No OR/RR of CA
AC C
Mayer et al Morson Munkholm Murphy et al Navaneethan et al Nixon et al Nowacki Nugent et al Paris et al Pekow et al Pohl et al Potack et al Ransohoff et al Ren et al Ren et al Riddell Riegler et al Rizzo et al Rogle Rogler G Rubin Rubin et al Scarpa Shanahan Shinozaki et al Shinozaki et al Shivakumar et al Shivakumar et al Shivakumar et al Sipos Sjöqvist et al Snapper et al Stolwijk JA Taylor et al Terdiman et al Terhaar et al Thomas et al
2013 2004 2004 1998 2013 1992 2007 2006 2007
No HIA; No OR/RR of CA No HIA No HIA No HIA No HIA No HIA No HIA No HIA No HIA; Meta-analysis
ACCEPTED MANUSCRIPT 19
2005 No HIA
M AN U
2013 No HIA 2009 Mixed CD and UC; No HIA 1991 No HIA
RI PT
Review Mixed CD and UC; No HIA No HIA No HIA Mixed CD and UC; No HIA No HIA No HIA No HIA No HIA
SC
2009 2002 2011 2003 2002 2004 2005 2006 2005
1994 No HIA 2012 No HIA; Review 1983 No HIA
No HIA No HIA; Review No HIA No HIA; No OR/RR of CA; Review No HIA; No OR/RR of CA No HIA; No OR/RR of CA No HIA; No OR/RR of CA; Review
TE D
1999 2001 1992 2008 2014 2012 2008
EP
Triantafillidis et al Ullman Ullman et al Ullman et al Ullman et al Usaj et al van Staa et al Velayos et al Velayos et al Venkataraman et al Venkatesh et al Viennot et al Vilien et al von Herbay et al Vu et al Waye Werner M et al Wong et al Woolrich et al Xie J et al Yashiro Zisman et al Zisman et al
AC C
Supplementary Table 2. Quality measures of included studies16. 1 2 3 4 5a 5b 6 7 Quality Measures Gupta Y N Y Y N N Y Y Korelitz Y Y Y Y Y Y Y Y Nieminen Y Y Y Y Y Y Y Y Rubin Y Y Y Y Y Y Y Y 10 Rutter Y Y Y Y Y Y Y Y Rutter11 Y Y Y Y Y Y Y Y
8
Total
N N N N N N
5 8 8 8 8 8
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Impact of Mucosal Inflammation on Risk of Colorectal Neoplasia in Patients with Ulcerative Colitis: A Systematic Review and Meta-Analysis
1. IBD: inflammatory bowel disease 2. UC: ulcerative colitis 3. CRN: colorectal neoplasia
SC
4. CRC: colorectal cancer
RI PT
List of Acronyms
guidelines 6. OR: odds ratio 7. CI: confidence interval
M AN U
5. PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses
8. HIA: histologic inflammatory activity
10. CA: cancer
TE D
9. RR: relative risk
AC C
EP
11. CD: Crohn’s disease
S0016-5107(17)32121-1
DOI:
10.1016/j.gie.2017.07.028
Reference:
YMGE 10675
To appear in:
Gastrointestinal Endoscopy
Received Date: 4 May 2017 Revised Date:
0016-5107 0016-5107
Accepted Date: 14 July 2017
Please cite this article as: Flores BM, OʼConnor A, Moss AC, Impact of Mucosal Inflammation on Risk of Colorectal Neoplasia in Patients with Ulcerative Colitis: A Systematic Review and Meta-Analysis, Gastrointestinal Endoscopy (2017), doi: 10.1016/j.gie.2017.07.028. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Impact of Mucosal Inflammation on Risk of Colorectal Neoplasia in Patients with
Brisas M. Flores1, Anthony OʼConnor2, Alan C Moss1
RI PT
Ulcerative Colitis: A Systematic Review and Meta-Analysis
1. Division of Gastroenterology, Beth Israel Deaconess Medical Center, Boston, MA
SC
2. Leeds Gastroenterology Institute, St James's University Hospital, Leeds
Alan C Moss MD Division of Gastroenterology
M AN U
Contact:
Beth Israel Deaconess Medical Center 330 Brookline Ave
MA 02215 617-667-3197
TE D
Boston
AC C
EP
[email protected]
ACCEPTED MANUSCRIPT 1
Impact of Mucosal Inflammation on Risk of Colorectal Neoplasia in Patients with
Brisas M. Flores1, Anthony OʼConnor2, Alan C Moss1
RI PT
Ulcerative Colitis: A Systematic Review and Meta-Analysis
1. Division of Gastroenterology, Beth Israel Deaconess Medical Center, Boston, MA
Contact: Alan C Moss MD Division of Gastroenterology
M AN U
SC
2. Leeds Gastroenterology Institute, St James's University Hospital, Leeds
Beth Israel Deaconess Medical Center
Boston MA 02215
EP
617-667-3197
TE D
330 Brookline Ave
AC C
[email protected]
ACCEPTED MANUSCRIPT 2 ABSTRACT Background and Aims: Long-standing ulcerative colitis is an established risk factor for colorectal neoplasia. A number of observational studies have suggested that evidence
RI PT
of mucosal inflammation (endoscopic or histologic) is associated with a greater risk for colorectal neoplasia than is mucosal healing. Our goal was to systematically analyze
inflammation to better inform surveillance strategies.
SC
the risk of colorectal neoplasia in ulcerative colitis patients with ongoing mucosal
Methods: We performed a systematic review and meta-analysis of the effect of
M AN U
endoscopic and/or histologic inflammation on the risk of colorectal neoplasia in cohort and case-control studies. Sensitivity analyses for study setting and case definition were performed.
Results: Six studies met the inclusion criteria, incorporating outcomes in 1,443 patients.
TE D
No study used a single validated measure for mucosal inflammation. The pooled odds ratio for colorectal neoplasia was 3.5 (95% CI, 2.6 - 4.8; p < 0.001) in those with any mucosal inflammation, and 2.6 (95% CI, 1.5-4.5; p = 0.01) in those with histologic
EP
inflammation, when compared to those with mucosal healing. Overall quality of the studies was good.
AC C
Conclusions: The presence of objective evidence of mucosal inflammation during follow-up in patients with ulcerative colitis is associated with a greater risk of subsequent colorectal neoplasia than in those with mucosal healing. This risk factor should be considered in guidelines on surveillance intervals for these patients.
ACCEPTED MANUSCRIPT 3 Keywords: inflammatory bowel disease; ulcerative colitis; histologic inflammation; colorectal neoplasia INTRODUCTION
RI PT
It is well-established that inflammatory bowel disease (IBD), including both ulcerative colitis (UC) and Crohn’s disease, puts one at increased risk of colorectal neoplasia (CRN). For ulcerative colitis, the cumulative risk of colorectal cancer (CRC) is estimated
SC
at 18% after 30 years1. Recognition of this adverse event has led to the incorporation of colonoscopic surveillance for dysplasia into clinical practice in IBD1-4. During
M AN U
surveillance, the prevalence of low-grade dysplasia (LGD) is approximately 10%, and 3% of patients with LGD progress to high-grade dysplasia or cancer5. The guidelines on this practice have not included specific recommendations based on the presence or absence of mucosal inflammation during follow-up.
TE D
Risk factors for the development of neoplasia in IBD are multiple and include duration of disease, extent of disease, family history of CRC, and history of primary sclerosing cholangitis1,3,4. Endoscopic findings associated with cancer risk include
EP
multiple pseudopolyps, strictures, and low-grade dysplasia on biopsy3,6. A number of retrospective studies have also concluded that the presence of any endoscopic or
AC C
histologic inflammation confers a greater risk of CRC by itself7-12. However, most of these studies were relatively small, and provided varying estimates of risk. This issue warrants further attention, as recent studies by our group and others have reported that greater than 40% of patients with UC have ongoing endoscopic or histologic inflammation at the time of colonoscopy, despite being in clinical remission13. If these
ACCEPTED MANUSCRIPT 4 factors have implications for prognosis and cancer risk stratification, then a quantification of their magnitude may be important. The aim of this paper is to use systematic review and meta-analysis to quantify
RI PT
the risk of colorectal neoplasia in the setting of endoscopic and/or histologic inflammation in patients with UC. We hypothesize that the independent risk colonic inflammation has on the development of CRC is significant, and could be used to adjust
SC
surveillance patterns accordingly in at risk individuals. It is currently recommended that surveillance colonoscopy begin after 8 to 10 years of disease and then be done every 1
M AN U
to 2 years, whether there is objective evidence of ongoing chronic inflammation or not1.
METHODS
Literature Search
TE D
This systematic review and meta-analysis was conducted with guidance provided by the Cochrane Handbook for Systematic Reviews14. It is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines15. A
language,
for
EP
literature search was performed to identify all published and unpublished studies, in any consideration
of
inclusion
studies
with
information
regarding
AC C
clear histologic or endoscopic inflammatory scores and the development of colorectal neoplasia in patients with UC. A systematic search of the following databases was performed: MEDLINE [PubMed], Web of Science, Allied Health Literature, and EMBASE. The following search strategy was constructed by using a combination of MeSH
subject
headings
and
text-words: "inflammation,"
"histologic
inflammation," "inflammatory bowel disease," "ulcerative colitis," "risk," "colorectal
ACCEPTED MANUSCRIPT 5 dysplasia," "colorectal neoplasia." There were no non-English articles or abstracts that required translation.
RI PT
Inclusion and Exclusion Criteria
We included observational studies in this meta-analysis that met the following inclusion criteria:
Participants: studies performed in patients with ulcerative colitis.
(2)
Risk Factor: described histologic or endoscopic histologic inflammatory activity.
(3)
Comparators: patients with UC with no histologic or endoscopic inflammatory
M AN U
SC
(1)
activity. (4)
Outcomes: odds ratio (OR) and 95% confidence intervals (CI) of development of
either colorectal cancer or non-cancer dysplasia.
Studies: reported with clear definitions on rates of colorectal dysplasia and
TE D
(5)
cancer, and provided sufficient data to allow estimation of effect size. Inclusion was not otherwise restricted by study size, language, or publication type. When there were
EP
multiple publications from the same cohort, data from the most recent comprehensive
AC C
report were included.
Study Selection
Two authors independently scanned the abstract of every trial identified by the search to determine eligibility. Full articles were further assessed if they met inclusion criteria based upon the abstract, or if the abstract was unclear and the full article was needed to determine if it met inclusion criteria. Articles not meeting the inclusion criteria were
ACCEPTED MANUSCRIPT 6 excluded. All final included and excluded studies were reviewed by the senior
Data extraction and quality assessment
RI PT
investigator [ACM].
The following data were retrieved from the included studies: number of patients in the study, type of setting, inflammation criteria and grading system (histologic and/or
SC
endoscopic), odds ratio or hazards ratio for neoplasia risk, and duration of follow-up. Assessment of quality was performed according to the Newcastle-Ottawa Scale (NOS)
M AN U
guidelines, with scores greater than 5 being deemed acceptable16.
Statistical analysis
Our analysis focused on the risk of the development of CRN as defined by cancer or
TE D
dysplasia among patients with UC, based on whether or not they had evidence of mucosal inflammation on colonoscopy (endoscopic and/or histologic). The random effects model as proposed by DerSimonian and Laird was used to calculate pooled OR
EP
and 95% confidence intervals17. The effect estimate used was the OR, as given the rare
AC C
nature of CRN among the UC cohort, it was considered to be a suitable estimate of relative risk. Potential publication bias was examined by funnel plots. Heterogeneity was examined by the Q statistic and the I2 statistic18. The Q and I2 statistics were used to test statistical heterogeneity among
studies18. For the Q statistic, a P value of less than 0.1 is considered representative of statistically significant heterogeneity. An I2 index of around 25% is considered to demonstrate low levels of heterogeneity, 50% medium, and 75% high18. Sensitivity
ACCEPTED MANUSCRIPT 7 analyses were conducted, omitting each study in turn to screen for outliers with respect to results, study population, study design, or type of 5-ASA exposure. Analyses were performed using the STATA statistical package (Version 12.0; STATA Corporation,
RI PT
College Station, Tex, USA), and RevMan (Review Manager, Version 5.3. Copenhagen:
RESULTS Literature Search
SC
The Nordic Cochrane Centre, The Cochrane Collaboration, 2014).
M AN U
The search yielded 134 articles of potential relevance. After further text review (Figure 1), six met the inclusion criteria, incorporating outcomes in 1,443 patients7-12. All 6 articles were published as full articles7-12. The characteristics of the included articles are described in Table 1. Two of the 6 studies were retrospective observational cohort
TE D
studies7,8, and 4 of the 6 were retrospective case control studies9-12. There was no consensus score used for histologic inflammation grading, and each study used its own criteria to be categorized as having “inflammation.” The reasons for exclusion of the
EP
other 128 studies were outline in Supplementary Table 1. The quality assessment
AC C
scores are detailed in Supplementary Table 2.
Risk of development of colorectal neoplasia in the setting of inflammation Four studies reported raw data on the rates of CRN in patients with or without mucosal inflammation8-11. There were 1025 patients included in these studies, with a range of mean duration of disease from 7 to 20 years. “Mucosal inflammation” included any histologic inflammation or endoscopic inflammation. The pooled odds ratio for colorectal
ACCEPTED MANUSCRIPT 8 neoplasia (dysplasia or cancer) in patients with any mucosal inflammation at time of colonoscopy (Figure 2) was 3.5 (95% CI, 2.6 - 4.8; p<0.001). There was no significant statistical heterogeneity (I2=58%, p=0.07).
RI PT
Five studies reported odds ratios for CRN in patients on the basis of histologic inflammation only (Figure 3)7-11. The pooled odds ratio was 2.6 (95% CI, 1.5 – 4.5) across the studies. Notably, there was significant statistical heterogeneity (I2=69.2%,
SC
p=0.011). The two studies with the greatest odds ratio for this outcome differed from other studies in that they reported histologic inflammation as a continuous variable11 or
M AN U
only the highest degree of inflammation9. In contrast, the other 3 studies considered patients with any histologic abnormalities as having “mucosal inflammation.”7,8,10. Supplementary Figure 1 shows the funnel plot of the log of odds ratio versus standard error for all studies.
TE D
There were insufficient studies included to perform meaningful sensitivity analyses, or analyze the risk of CRN based on endoscopic scores alone. No studies adjusted odds ratios for other CRN risk factors such as primary sclerosing cholangitis,
EP
family history, or smoking status. Only the paper by Rubin et al adjusted for medication use in their analyses10. In the studies of histological grade, there was heterogeneity in
AC C
the methods of reporting severity of inflammation; mean and upper limits, per-unit increase, or ordinal scales were used7-11.
DISCUSSION
There has been much focus in recent years on “mucosal healing” as a goal of therapy for patients with IBD. The rationale for this is the association between endoscopic
ACCEPTED MANUSCRIPT 9 evidence of inflammation and adverse outcomes such as disease relapse, hospitalization, surgery, and cancer risk in patients with ulcerative colitis19-21. In particular, the severity of mucosal inflammation, or some of its features, has been
RI PT
independently associated with the risk of colorectal neoplasia in some studies7,12. Unfortunately, these studies were mostly retrospective studies from hospital based cohorts, and subject to the associated biases of this methodology and selection.
SC
Colman and Rubin (2016) recently published a systematic review emphasizing the relationship between histologic inflammation and colorectal neoplasia in patients with
M AN U
ulcerative colitis, however no meta-analysis was done22.
This meta-analysis emphasizes that in those with ulcerative colitis, histologic or endoscopic inflammation puts one at increased risk for developing colorectal neoplasia. Our results suggest that the pooled risk of CRN is more than 2-fold greater in patients
TE D
with established UC who have mucosal inflammation during follow-up colonoscopy than in those who have no or minimal inflammation during colonoscopy. The consistency of this effect persists across studies, and the lack of statistical heterogeneity provides
EP
confidence in this conclusion. When inflammation is based on histological abnormalities only, the risk of CRN remains elevated when compared with patients with normal or mild
AC C
histological abnormalities.
Our analysis does have its limitations. All studies were retrospective, and there
was no universal definition of histologic or endoscopic inflammation. There was statistical heterogeneity in the pooled odds ratio, which reflects these differences in methodology. Nieminen et al included both patients with ulcerative colitis and Crohn’s disease, further adding to potential heterogeneity9. The consistent impact, regardless of
ACCEPTED MANUSCRIPT 10 method of describing inflammation, suggests a biologically robust effect. All of the studies were done in a tertiary hospital setting, with associated bias in population risk of cancer. This may limit generalizing these findings to the general IBD population in the
alone to calculate risk and draw meaningful conclusions.
RI PT
community. In addition, there were insufficient data regarding endoscopic inflammation
Despite the limitations, this analysis confirms the implication that inflammation
SC
confers a true risk in the development of neoplasia in ulcerative colitis patients. This may be clinically relevant, as current guidelines recommend colorectal cancer screening
M AN U
every one to two years, regardless of factors such as inflammation severity1. The risk of colorectal neoplasia in patients with ulcerative colitis is up to 18% after having the disease for 30 years1, and as evidence continues to arise that inflammation increases the risk for progression to neoplasia, physicians may consider adjusting screening
TE D
intervals in individuals who lack mucosal healing at surveillance colonoscopy. In addition, this finding may further encourage physicians to aim for resolution of mucosal inflammation, rather than just clinical remission. Whether achieving mucosal healing via
EP
maintenance therapy will reduce an individual’s risk of cancer is unknown, although mesalamine appears to have a modest benefit in this regard23. Future studies on this
AC C
topic should clearly define “inflammation” using validated histological or endoscopic indices of inflammation. Risks of either dysplasia or cancer should be adjusted for factors such as duration of disease, presence of primary sclerosing cholangitis, family history, and chemoprevention. In particular, research should explore how particular treatments for UC impact the rates of colorectal neoplasia in those who have achieved remission of inflammation, when compared with those who have not.
ACCEPTED MANUSCRIPT 11 In conclusion, this meta-analysis confirms a pooled effect of ongoing colonic mucosal inflammation on risk of colorectal neoplasia in patients with ulcerative colitis. This association should inform treatment goals and surveillance strategies for this
RI PT
patient population.
References
SC
1. Farraye FA, Odze RD, Eaden J et al. AGA medical position statement on the diagnosis and management of colorectal neoplasia in inflammatory bowel disease.
M AN U
Gastroenterology. 2010;138(2):738-745.
2. Laine L, Kaltenbach T, Barkun A et al. SCENIC international consensus statement on surveillance
and
management
of
dysplasia
in
inflammatory
bowel
disease.
Gastroenterology. 2015;148(3):639-651.e28.
clinicopathologic
TE D
3. Rutter MD, Riddell RH. Colorectal dysplasia in inflammatory bowel disease: A perspective.
2014;12(3):359-367.
Clinical
gastroenterology
and
hepatology.
EP
4. Ananthakrishnan AN, Cagan A, Cai T et al. Colonoscopy is associated with a reduced risk for colon cancer and mortality in patients with inflammatory bowel
AC C
diseases. Clinical gastroenterology and hepatology. 2015;13(2):322-329. 5. Thomas T, Abrams KA, Robinson RJ et al. Meta-analysis: cancer risk of low-grade dysplasia in chronic ulcerative colitis. Alimentary Pharmacology and Therapeutics. 2007;25(6):657-68.
6. Dyson J, Rutter M. Colorectal cancer in inflammatory bowel disease: What is the real magnitude of the risk? World journal of gastroenterology. 2012;18(29):3839-3848.
ACCEPTED MANUSCRIPT 12 7. Gupta RB, Harpaz N, Itzkowitz S et al. Histologic inflammation is a risk factor for progression
to
colorectal
neoplasia
in
ulcerative
colitis:
a
cohort
study.
Gastroenterology. 2007;133(4):1099-1105.
carcinoma
in
extensive
ulcerative
colitis.
World
2014;20(17):4980-4986.
RI PT
8. Korelitz BI, Sultan K, Kothari M et al. Histological healing favors lower risk of colon journal
of
gastroenterology.
SC
9. Nieminen U, Jussila A, Nordling S et al. Inflammation and disease duration have a cumulative effect on the risk of dysplasia and carcinoma in IBD: A case–control
M AN U
observational study based on registry data. International journal of cancer. 2014;134:189–196.
10. Rubin DT, Huo D, Kinnucan JA et al. Inflammation is an independent risk factor for colonic neoplasia in patients with ulcerative colitis: a case-control study. Clinical
TE D
gastroenterology and hepatology. 2013;11(12):1601-8.
11. Rutter M, Saunders B, Wilkinson K et al. Severity of inflammation is a risk factor for colorectal neoplasia in ulcerative colitis. Gastroenterology. 2004;126(2):451-459.
ulcerative
EP
12. Rutter MD, Saunders BP, Wilkinson KH et al. Cancer surveillance in longstanding colitis:
endoscopic
appearances
help
predict
cancer
risk.
Gut.
AC C
2004;53(12):1813-1816.
13. Rosenberg L, Nanda KS, Zenlea T et al. Histologic markers of inflammation in patients with ulcerative colitis in clinical remission. Clinical gastroenterology and hepatology. 2013;11(8):991. 14. Higgins JPT. Cochrane handbook for systematic reviews of interventions. http://www.cochrane-handbook.org. Updated 2011.
ACCEPTED MANUSCRIPT 13 15. Moher D, Liberati A, Tetzlaff J et al. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. British medical journal. 2009;339(4):b2535.
RI PT
16. Wells G, Shea B, O'Connell J et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analysis. 2011. [Website]. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. Available from: URL:
SC
http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp.
17. DerSimonian R, Laird N. Meta-analysis in clinical trials. Controlled clinical trials.
M AN U
1986;7(3):177-88.
18. Higgins JPT, Thompson SG, Deeks JJ et al. Measuring inconsistency in metaanalyses. British medical journal. 2003;327(7414):557-560.
19. Ardizzone S, Cassinotti A, Duca P et al. Mucosal healing predicts late outcomes
TE D
after the first course of corticosteroids for newly diagnosed ulcerative colitis. Clinical gastroenterology and hepatology. 2011;9(6):483-489. 20. Bessissow T, Lemmens B, Ferrante M et al. Prognostic value of serologic and
EP
histologic markers on clinical relapse in ulcerative colitis patients with mucosal healing. American journal of gastroenterology. 2012;107(11):1684-92.
AC C
21. Bitton A, Peppercorn MA, Antonioli DA et al. Clinical, biological, and histologic parameters
as
predictors
of
relapse
in
ulcerative
colitis.
Gastroenterology.
2001;120(1):13-20.
22. Colman RJ, Rubin DT. Histological inflammation increases the risk of colorectal neoplasia in ulcerative colitis: a systematic review. Intestinal Research. 2016;14(3):202-10.
ACCEPTED MANUSCRIPT 14 23. OʼConnor A, Packey CD, Akbari M et al . Mesalamine, but not sulfasalazine, reduces the risk of colorectal neoplasia in patients with inflammatory bowel disease: an
2015;21(11):2562-9.
Figure 1. Results of the literature search.
SC
FIGURE LEGENDS
RI PT
agent-specific systematic review and meta-analysis. Inflammatory bowel disease.
M AN U
Figure 2. Pooled odds ratio (OR) for colorectal neoplasia if there is presence of either histologic or endoscopic inflammation8-11. Odds ratio along with confidence intervals (CI) are provided.
Figure 3. Pooled odds ratio for colorectal neoplasia if there is presence of histologic
AC C
EP
TE D
inflammation7-11. Odds ratio along with confidence intervals (CI) are provided.
ACCEPTED MANUSCRIPT 15
TABLES Table 1. Characteristics of included studies. Type
Design
N
n (cases)
Dysplasia Definition
Inflammation Score
Inflammation definition
Gupta
2007
Paper
Retrospective observational cohort
418
224
Histologic Activity Index: Grade 0-3
Score of 1 or greater
Korelitz8
2014
Paper
Retrospective observational cohort
68
20
Low grade dysplasia, high grade dysplasia, cancer High grade dysplasia, cancer
2014
Paper
Retrospective case-control
506
168
Dysplasia, cancer
Microscopic inflammation in the absence of gross inflammation Score of 1 or greater
>20 years
Nieminen9
Rubin
10
2013
Paper
Retrospective case-control
200
59
Neoplasia, cancer
Score of 2 or greater
11 years
Rutter11
2004
Paper
Retrospective case-control
204
68
Adenoma, low grade dysplasia, high grade dysplasia, cancer
Microscopic inflammation in the absence of gross inflammation Histologic Activity Index: Grade 0-3 Histologic Inflammatory Activity: Grade 0-5 Histologic Inflammation: Grade 0-4
Unclear, assess at 1 unit increases
14 years
Presence or absence of scarring, postinflammatory polyps, backwash ileitis, shortened colon, tubular appearance, featureless colon
14 years
Paper
Retrospective case-control
EP
2004
AC C
Rutter12
204
SC
M AN U
7
68
Adenoma, low grade dysplasia, high grade dysplasia, cancer
Duration of followup 6.7 years
RI PT
Year
TE D
Author
Colonoscopic Inflammation: Grade 0-4 Presence of scarring, postinflammatory polyps, backwash ileitis, shortened colon, tubular appearance, featureless colon
12 years
ACCEPTED MANUSCRIPT 16
SUPPLEMENTARY TABLES Supplementary Table 1. Excluded studies and reason for exclusion.
1981 No HIA
TE D
No HIA; No OR/RR of CA; Review Mixed CD and UC; No HIA No HIA No HIA Mixed CD and UC; No HIA No HIA Mixed CD and UC; No HIA No HIA No HIA No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA Mixed CD and UC; No HIA No HIA; No OR/RR of CA; Review No HIA; Review No HIA Mixed CD and UC; No HIA; Review No HIA No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA
EP
2014 2008 1989 1986 2009 2002 2007 2005 2013 2004 2014 2014 1984 2012 2004 1989 1992 2011 2010 2010 2014
SC
Reason for exclusion No HIA; No OR/RR of CA No HIA No HIA; No OR/RR of CA No HIA No HIA Mixed CD and UC; No HIA Mixed CD and UC; No HIA No HIA No HIA
M AN U
Year 2015 1989 2010 2009 2011 2010 2006 1995 1989
AC C
Study Aalykke et al Albert et al Allen et al Awais et al Baars et al Bergeron et al Bernstein Biasco et al Binder Blackstone et al Bressenot et al Breynaert et al Broström Broström Campbell et al Campbell et al Cendan et al Chen et al Chen YX et al Clevers Connelly et al DeRoche et al Dobbins Dyson et al Eaden Farmer Fischbach Frego et al Fujita et al Fujita T et al Gallinger et al
RI PT
OR* odds ratio. RR* relative risk. CA* cancer. HIA* histologic inflammatory activity. UC* ulcerative colitis. CD* Crohn’s disease.
ACCEPTED MANUSCRIPT
2000 2012 2000 2000 2014 2013
No HIA No HIA No HIA Mixed CD and UC; No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA
2015 No HIA; No OR/RR of CA 2004 No HIA; Review 2004 No HIA
TE D
M AN U
No HIA No HIA No HIA No HIA Mixed CD and UC; No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA Mixed CD and UC; No HIA Mixed CD and UC; No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA Mixed CD and UC; No HIA No HIA No HIA No HIA No HIA No HIA
EP
1993 1994 2007 2006 2002 2013 2010 2007 2005 2010 2014 2001 1995 2004 2006 1989 1991
SC
2007 Mixed CD and UC; No HIA 2007 Mixed CD and UC; No HIA; Review
AC C
Geboes Gong W Gorfine et al Greenstein Harpaz et al Harpaz et al Henriksen et al Herszeny et al Herszényi et al Itzkowitz et al Itzkowitz et al Jablonská et al Jacobsen et al Jess et al Jess et al Judge et al Julka et al Kallesøe et al Katsanos et al Katsanos et al Kekilli et al Kiran et al Konecný et al Konishi F et al Krok et al Lakatos et al Lashner et al Leidenius et al LennardJones LennardJones et al Lindberg et al Löfberg et al Loftus Lukas Lukas M et al Mathy et al Matsui et al
1983 No HIA
1977 2006 1989 2006 2010 2010 2003 1993
RI PT
17
No HIA No HIA No HIA No HIA; Review No HIA; No OR/RR of CA No HIA; No OR/RR of CA; Review No HIA No HIA
ACCEPTED MANUSCRIPT 18 No HIA No HIA Mixed CD and UC; No HIA; Review No HIA
2013 1995 2014 1979 1994 2010 2000 2008
No HIA No HIA No HIA; OR/RR of CA from dysplasia only No HIA No HIA No HIA No HIA No HIA; No OR/RR of CA; Review
1985 2013 2011 1984 2003 2011 2012 2014 2008 2006 2014 2003 2000 1999
No HIA No HIA; No OR/RR of CA No HIA; No OR/RR of CA No HIA No HIA No HIA; Review No HIA No HIA; No OR/RR of CA No HIA No HIA; Review No HIA; No OR/RR of CA Mixed CD and UC ; No HIA; Review No HIA No HIA
EP
TE D
M AN U
SC
RI PT
1999 1985 2003 2014
2013 No HIA; No OR/RR of CA 2013 No HIA; No OR/RR of CA
AC C
Mayer et al Morson Munkholm Murphy et al Navaneethan et al Nixon et al Nowacki Nugent et al Paris et al Pekow et al Pohl et al Potack et al Ransohoff et al Ren et al Ren et al Riddell Riegler et al Rizzo et al Rogle Rogler G Rubin Rubin et al Scarpa Shanahan Shinozaki et al Shinozaki et al Shivakumar et al Shivakumar et al Shivakumar et al Sipos Sjöqvist et al Snapper et al Stolwijk JA Taylor et al Terdiman et al Terhaar et al Thomas et al
2013 2004 2004 1998 2013 1992 2007 2006 2007
No HIA; No OR/RR of CA No HIA No HIA No HIA No HIA No HIA No HIA No HIA No HIA; Meta-analysis
ACCEPTED MANUSCRIPT 19
2005 No HIA
M AN U
2013 No HIA 2009 Mixed CD and UC; No HIA 1991 No HIA
RI PT
Review Mixed CD and UC; No HIA No HIA No HIA Mixed CD and UC; No HIA No HIA No HIA No HIA No HIA
SC
2009 2002 2011 2003 2002 2004 2005 2006 2005
1994 No HIA 2012 No HIA; Review 1983 No HIA
No HIA No HIA; Review No HIA No HIA; No OR/RR of CA; Review No HIA; No OR/RR of CA No HIA; No OR/RR of CA No HIA; No OR/RR of CA; Review
TE D
1999 2001 1992 2008 2014 2012 2008
EP
Triantafillidis et al Ullman Ullman et al Ullman et al Ullman et al Usaj et al van Staa et al Velayos et al Velayos et al Venkataraman et al Venkatesh et al Viennot et al Vilien et al von Herbay et al Vu et al Waye Werner M et al Wong et al Woolrich et al Xie J et al Yashiro Zisman et al Zisman et al
AC C
Supplementary Table 2. Quality measures of included studies16. 1 2 3 4 5a 5b 6 7 Quality Measures Gupta Y N Y Y N N Y Y Korelitz Y Y Y Y Y Y Y Y Nieminen Y Y Y Y Y Y Y Y Rubin Y Y Y Y Y Y Y Y 10 Rutter Y Y Y Y Y Y Y Y Rutter11 Y Y Y Y Y Y Y Y
8
Total
N N N N N N
5 8 8 8 8 8
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Impact of Mucosal Inflammation on Risk of Colorectal Neoplasia in Patients with Ulcerative Colitis: A Systematic Review and Meta-Analysis
1. IBD: inflammatory bowel disease 2. UC: ulcerative colitis 3. CRN: colorectal neoplasia
SC
4. CRC: colorectal cancer
RI PT
List of Acronyms
guidelines 6. OR: odds ratio 7. CI: confidence interval
M AN U
5. PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses
8. HIA: histologic inflammatory activity
10. CA: cancer
TE D
9. RR: relative risk
AC C
EP
11. CD: Crohn’s disease