Surgical Endoscopy

, Volume 28, Issue 7, pp 2039–2047

Routine colonoscopy is not required in uncomplicated diverticulitis: a systematic review

Authors

    • Department of SurgerySt Antonius Hospital
    • Department of SurgeryGelderse Vallei Hospital
  • D. Boerma
    • Department of SurgerySt Antonius Hospital
  • R. Timmer
    • Department of GastroenterologySt Antonius Hospital
  • B. van Ramshorst
    • Department of SurgerySt Antonius Hospital
  • L. A. Dieleman
    • Department of GastroenterologyUniversity of Alberta
    • Centre of Excellence for Gastrointestinal Inflammation and Immunity ResearchUniversity of Alberta
  • H. L. van Westreenen
    • Department of SurgerySt Antonius Hospital
    • Isala Clinics
Review

DOI: 10.1007/s00464-014-3447-4

Cite this article as:
de Vries, H.S., Boerma, D., Timmer, R. et al. Surg Endosc (2014) 28: 2039. doi:10.1007/s00464-014-3447-4

Abstract

Background

It is generally accepted that patients following an episode of diverticulitis should have additional colonoscopy screening to rule out a colorectal malignancy. We aimed to investigate the rate of CRC found by colonoscopy after an attack of uncomplicated diverticulitis.

Methods

MEDLINE, Embase, and Cochrane databases were searched systematically for clinical trials or observational studies on colonic evaluation by colonoscopy after the initial diagnosis of acute uncomplicated diverticulitis, followed by hand-searching of reference lists.

Results

Nine studies met the inclusion criteria and included a total number of 2,490 patients with uncomplicated diverticulitis. Subsequent colonoscopy after an episode of uncomplicated diverticulitis was performed in 1,468 patients (59 %). Seventeen patients were diagnosed with CRC, having a prevalence of 1.16 % (95 % confidence interval 0.72–1.9 % for CRC). Hyperplastic polyps were seen in 156 patients (10.6 %), low-grade adenoma in 90 patients (6.1 %), and advanced adenoma was reported in 32 patients (2.2 %).

Conclusion

Unless colonoscopy is regarded for screening in individuals aged 50 years and older, routine colonoscopy in the absence of other clinical signs of CRC is not required.

Keywords

ColonoscopyDiverticulitisColorectal cancerUncomplicated diverticulitisSystematic review

Diverticulosis of the colon is an acquired condition that forms typically at the site where the vasa recta penetrate the muscle layer to reach submucosa and mucosa [1]. Its prevalence increases with age, varying from approximately 5 % in those younger than 40 years and increasing to 50–66 % in patients older than 80 years[25]. Acute diverticulitis results from inflammation of a diverticulum due to micro- or macro perforation. It is thought that ~20 % of the patients with diverticulosis will experience an inflammatory complication of the disease, ranging in severity from uncomplicated diverticulitis to a complicated course characterized by abscess formation and perforation, although a recent study suggests a rate as low as 1 % over 11 years [6, 7].

Clinically, an episode of acute diverticulitis is typically characterized by left lower quadrant abdominal pain, fever, and leukocytosis. Imaging is not always necessary in patients with mild clinical symptoms or recurrent disease. However, it is generally accepted that patients following an episode of diverticulitis should have additional imaging to rule out a colorectal malignancy [8]. It is most useful to confirm the diagnosis and to assess the severity and extent of disease, because there can be discrepancy between clinical presentation and the extent of disease [9]. Computed tomography (CT) of the abdomen often is the diagnostic test of choice, although ultrasound also can be useful to diagnose early, uncomplicated diverticulitis [10, 11].

Because colorectal cancer (CRC) may present itself with similar symptoms as diverticulitis, professional medical organizations, such as the American College of Gastroenterology, the American Society of Colon and Rectal Surgeons, and the European Association for Endoscopic Surgery as well as the Association of Coloproctology of Great Britain and Ireland, recommend colonic evaluation by colonoscopy to exclude CRC after an episode of acute diverticulitis [1215]. Data supporting this recommendation are sparse and largely are based on small cohorts and/or case reports [16].

The purpose of this systematic review was to evaluate the necessity of performing routine colonoscopy in patients following an episode of uncomplicated acute diverticulitis, i.e., inflammation without abscess formation or perforation, to rule out a colorectal malignancy. The prevalence of CRC in patients with uncomplicated acute diverticulitis was established on the basis of the available literature.

Methods

Literature search

Data collection and analysis were performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [17]. A systematic literature search with predefined search terms was performed using the databases PubMed, EMBASE, and the Cochrane Library for articles and abstracts published from 1970 to December 2012. A librarian assisted with the electronic search. The keywords diverticulitis, colorectal neoplasms or CRC, and colonoscopy were used. The search strategy included medical subject heading (MeSH) terms and free text. Additionally, a manual cross-reference search was performed of the bibliographies for relevant articles covering CRC and diverticulitis not found in the first search. The day of the last search was December 6, 2012.

Study selection

Eligible studies for inclusion were randomized, clinical trials or observational studies in which routine colonic evaluation by colonoscopy was performed after the initial diagnosis of acute uncomplicated diverticulitis to rule out CRC. Acute uncomplicated diverticulitis was defined as the presence of colonic diverticular disease with localized colonic wall thickening and/or stranding of pericolonic fat, whereas a complicated course of diverticulitis was defined by the presence of abscesses, fistula, purulent or faecal peritonitis, or perforation [18]. Exclusion criteria were: patients <18 years, case reports, narrative reviews, animal studies, and studies in patients with complicated diverticulitis. To separate high- and medium-quality studies from low-quality studies, the system for assigning level of evidence from the Centre for Evidence-Based Medicine (CEBM) in Oxford, UK was used [19].

Data abstraction

Titles and abstracts of all retrieved records and subsequent full-text articles were examined independently by two investigators (HSdV, HLvW) to identify studies meeting the inclusion criteria. If there were discrepancies in selection, consensus was reached through discussion among the reviewers. If there was suggestion of overlap of the cohorts among the included studies, only data from the most comprehensive cohort were included.

Data were extracted using a standardised data collection form that included the following characteristics: year of publication, study design, country where the study was performed, basic characteristics (age, number of patients with uncomplicated diverticulitis), diagnostic modality used to confirm the diagnosis of acute uncomplicated diverticulitis, number of CRC detected, number of polyps detected, follow-up period, and stated endpoint. In case these data could not be retrieved from the published studies, additional patient data were obtained were requested and obtained from several corresponding authors (see Acknowledgments).

Endpoints of interest

The primary outcome of interest of our systematic review was the number of CRC detected (by colonoscopy) in patients with uncomplicated diverticulitis. Because most CRCs develop from adenomas, a secondary endpoint was the number of polyps (which were divided in hyperplastic polyps, low-grade adenoma, and advanced adenoma) detected by colonoscopy in patients with uncomplicated diverticulitis. Advanced adenoma was defined as either an adenoma of 10 mm or greater in diameter and/or more than 25 % villous components and/or severe dysplasia [20].

Statistical analysis

The prevalence of CRC with 95 % confidence intervals was calculated using GraphPad Prism version 4.00 (GraphPad Software, San Diego, CA).

Results

Literature search and study selection

As depicted in Fig. 1, the search strategy identified a total number of 418 potential studies after removal of duplicates. After review on title and abstracts, a total number of 380 studies did not meet the inclusion criteria and could be excluded. The remaining studies were subsequently assessed, full text if available; ten studies were only available as conference abstracts. Review of the bibliographies of the full-text articles revealed five additional relevant articles that were not covered by our literature search.
https://static-content.springer.com/image/art%3A10.1007%2Fs00464-014-3447-4/MediaObjects/464_2014_3447_Fig1_HTML.gif
Fig. 1

Prisma flow diagram of the systematic literature review. *Ten studies were assessed of which only the abstract was available. #Of these ten studies, three were included. **This category included studies on colonoscopies for colorectal cancer in asymptomatic individuals, management of diverticulosis/diverticulitis, imaging studies, fistulae, and diverticulitis, etc. ***These studies included patients with both complicated and uncomplicated diverticulitis. Distinct data on the subgroup of patients with uncomplicated diverticulitis were not given

The final selection included nine studies that matched our inclusion criteria; six studies were published as full papers, and three studies were only published as conference abstracts.

Study characteristics

Study characteristics of the nine included studies are given in Table 1. The included studies represented a total number of 2,490 patients with uncomplicated diverticulitis. Of these patients, 1,468 patients (59.0 %) underwent subsequent colonoscopy. Follow-up by barium enema or CT colonography was performed in 75 patients (3.0 %). Reasons for not performing colonoscopy were only mentioned in one study [23], which included a total number of 284 patients with uncomplicated diverticulitis. These reasons were severe comorbidity in combination with high age (n = 5), treating physician deemed colonoscopy not to be warranted (n = 44), and colonoscopy being planned outside the study period or in another hospital (n = 5). Of the 1,022 patients (41 %) with no record of colonic evaluation, two studies evaluated 751 patients (73 %) by searching national cancer registries for the occurrence of CRC up to 1 year after an episode of uncomplicated diverticulitis [26, 27]; Table 2.
Table 1

Basic characteristics of the studies included

Author, year, country [reference]

Study design

No. of patients included

No. of patients with uncomplicated diverticulitis

Age (yr) total cohort

Diagnostic modality

CEBM level

Duration of study

Comments

Ahmeidat et al., 2012, UK [21]

Retrospective cohort study

92

56

Mean age 61 (range 19–92) Uncomplicated diverticulitis: Mean age 60 (range 31–88)a

Abdominal CT scan

Level 3/4

20 months

Conference abstract

Alexandersson et al., 2012, Iceland [22]

Cohort study

118

118

Median age 57 (IQR 50–67) years

Abdominal CT scan

Level 3

15 months

Conference abstract

Van de Wall et al., 2012, The Netherlands [23]

Retrospective single centre cohort study

307

284

Mean age 57.4 (SD 13.2) in patients who underwent endoscopy, 59.6 (SD 14.5) in patients without endoscopy

Abdominal CT scan or ultrasonography

Level 3

3-year study interval

 

Schout et al., 2012, The Netherlands [24]

Retrospective single centre cohort study

422

422

Mean age 63 (range 31–95 yr)a

Abdominal CT scan or ultrasonography

Level 3

10-year study interval

 

Schmilovitz-Weiss et al., 2012, Israel [25]

Retrospective single centre cohort study

200

186

Mean age 61.97 ± 11.56 years (in patients with uncomplicated diverticulitis)

Total cohort:

Mean 61.8 (SD 14.3)

Abdominal CT scan in 84.5 % of all patients

Level 3

7-year study interval

 

Lau et al., 2011, Australia [26]

Retrospective multi-centre cohort study

1088

931a

Uncomplicated diverticulitis: Mean 59 (range 20-100)

Total cohort:

59.8 (SD 15.2)

Abdominal CT scan

Level 3

7-year study interval

 

Westwood et al., 2011, New Zealand [27]

Single-centre retrospective longitudinal study

292

292

Median age 60 (range 23–95)

Abdominal CT scan

Level 3

5-year study interval

Colonic evaluation was also performed by CT colonography in patients in which colonoscopy was incomplete

Lahat et al., 2007, Israel [28]

Prospective cohort study

86

86

Mean age 60.4 (SD 14.4)

Abdominal CT scan

Level 3

30 months

This study was a randomized controlled trial comparing early and late endoscopy in patients with uncomplicated diverticulitis

Elramah et al., 2010, USA [29]

Retrospective multi-centre cohort study

130

115

Not reported

Abdominal CT scan

Level 3

9-year study interval

Conference abstract

aInformation obtained from corresponding author of the study

Table 2

Number of malignancies and polyps in patients with uncomplicated diverticulitis

Author, year, country [reference]

No. of patients with uncomplicated diverticulitis

No. of patients with uncomplicated diverticulitis who underwent (full) colonoscopy

No. of colon carcinomas detected by colonoscopy

Colonic polyps in patients who underwent colonoscopy

No. of patients without colonic evaluation included in follow up

Follow-up

Comments

Hyperplastic polyps

Low-grade adenoma

Advanced adenoma

Ahmeidat et al., UK [21]

56

44

0

3

2

Colonoscopy

 

Alexandersson et al., 2012, Iceland [22]

118

108

0

9

9

1

Colonoscopy after 6–8 weeks

 

Van de Wall et al., 2012, The Netherlands [23]

284

202

2

15

.

18

7

Endoscopy after a mean interval of 8.9 weeks (SD 10.6).

 

Schout et al., 2012, The Netherlands [24]

422

378

8

39

44

Sigmoidoscopy,

Colonoscopy, barium enema colon x-ray (n = 42), CT-colography (n = 3) 6–10 weeks after discharge

 

Schmilovitz-Weiss et al., 2012, Israel [25]

186

93

0

5

26

1

93

Colonoscopy 4 to 6 weeks after hospital discharge

This cohort included a comparator, a group in which no colonoscopy was performed. The follow up of this cohort however is not clearly stated by the authors.

Lau et al., 2011, Australia [26]

931

267

5

33

39

10

664

Colonoscopy within an average time of 70 days after the initial CT scan or registration in the WA cancer registry.

For the 664 patients with no record of colonic evaluation, the Western Australian

(WA) Cancer Registry was searched for registrations of CRC within 1 year after CT scan. A total of 6 patients were found to have CRC.

Westwood et al., 2011, New Zealand [27]

292

175

1

20

19

11

87

Colonoscopy, CT colonography (n = 30) or registration in the New Zealand Cancer registry at least one year after the initial CT scan for acute uncomplicated diverticulitis

For patients with no record of colonic evaluation, the New Zealand Cancer Registry was searched for registrations of CRC. No patients were found to have CRC.

Lahat et al., 2007, Israel [28]

86

86

0

0

3

2

-

3-11 days in the early endoscopy group, 6 weeks after CT scan in the late endoscopy group

 

Elramah et al., 2010, USA [29]

115

115

1

N.R.

N.R.

N.R.

-

Follow-up colonoscopy within 6 months of the acute attack

 

CRC colorectal cancer, N.R. not reported

Endpoint of interest

Of the total number of 1,468 patients with uncomplicated diverticulitis who underwent colonoscopy, 17 patients were diagnosed with CRC. The prevalence of CRC detected by colonoscopy was 1.16 % (95 % confidence interval (CI) 0.72–1.9 %). A total number of 6 patients with CRC were identified among the 751 patients without colonic evaluation by searching the national cancer registries, bringing the total prevalence to 1.0 % (95 % CI 0.69–1.6 %). Hyperplastic polyps were seen in 156 patients (10.6 %) during colonoscopy, low-grade adenoma in 90 patients (6.1 %), and advanced adenoma was reported in 32 patients (2.2 %).

Discussion

Principal findings

The key finding of this systematic review is that the number of CRC detected by subsequent colonoscopy after a period of uncomplicated diverticulitis is ~1.0 %. This number is comparable to the number of CRC found during screening colonoscopies in asymptomatic individuals. On a total number of 2,881,763 screening colonoscopies for CRC performed in several western countries in asymptomatic individuals aged 40 years and older, the incidence of CRC ranges from 0.5–1.1 % [3039]. Notably, the largest study performed included 2,821,392 screening colonoscopies and found a CRC prevalence of 0.9 % [33]. Thus, the occurrence of CRC in patients with uncomplicated diverticulitis is not higher than in the population at large.

Most CRCs are thought to develop from adenomatous polyps by way of the adenoma-carcinoma sequence [40], of which advanced adenomas are considered to be the clinically relevant precursors of CRC [41]. In our systematic review, advanced adenomas were found in 2.2 % of the patients with uncomplicated diverticulitis who underwent colonoscopy, whereas the incidence of advanced adenoma ranges from 4.8–11.7 % in screening colonoscopies [3039]. Suggesting that precursors or CRC are comparable or even lower in patients with uncomplicated diverticulitis compared with asymptomatic individuals aged 40 years and older.

A recent systematic review by Sai et al. estimated the prevalence of underlying adenocarcinoma of the colon in patients in whom acute diverticulitis was diagnosed with CT at 2.1 versus 0.68 % in the general population [16]. However, most studies included in their systematic review were not designed to answer the question whether or not follow-up colonoscopy after an episode of acute diverticulitis is necessary. In six of ten studies included, follow-up was done by surgery and both patients with complicated and uncomplicated diverticulitis were included. In our study, we focused on patients with uncomplicated diverticulitis, because a complicated course of disease with diverticular abscesses and/or peritonitis generally requires resuscitation and stabilization of the patient followed by surgery [42, 43], and a subsequent colonoscopy is performed less often. Moreover, complicated diverticulitis is more difficult to differentiate from neoplasia by abdominal CT or ultrasound, and therefore our finding of a low rate of CRC in patients with uncomplicated diverticulitis cannot be extrapolated to patients with a complicated course of diverticulitis. In a study by Bahadursingh, a 6.8 % cancer rate was found during pathology examination of surgical specimens from patients with complicated diverticulitis [44]. Nevertheless, a study by Krones and coworkers, focusing on the advanced neoplasia rates in surgical specimens of patients operated for diverticulitis, found a statistically significant reduced rate of advanced colonic neoplastic lesion in nearly all age categories for patients with diverticulitis compared with the published prevalence of advanced colonic neoplastic lesion in people at average risk [45].

Additionally, Sai and coworkers calculated the prevalence of colon cancer in patients older than 55 years in the general population to be 0.68 % by using the data from the U.S. National Cancer Institute’s Surveillance Epidemiology and End Results (SEER) program [16]. We did not perform such an analysis, because in our opinion it is more useful to compare the prevalence of CRC and advanced adenoma in patients with uncomplicated diverticulitis to the prevalence found in asymptomatic individuals (during screening colonoscopies). As stated above, these reports indicate that the incidence of CRC found in uncomplicated diverticulitis is comparable to the number found in asymptomatic individuals, whereas the rate of advanced adenoma is lower in patients with uncomplicated diverticulitis.

Strengths and limitations of this review

To our knowledge, this is the first systematic review on the rate of CRC and advanced adenoma in patients with uncomplicated diverticulitis. The rate of follow-up colonoscopies was 59 % and the effect of incomplete follow-up in both studies with the lowest follow-up rate was reduced by cross-checking their national cancer registry database for CRC at least 1 year after the initial CT scan for diagnosing diverticulitis [26, 27]. In the one study that mentioned reasons for not performing colonoscopy, the major reason was that the treating physician deemed colonoscopy not to be warranted due to the fact that these patients were asymptomatic at the outpatient clinic in combination with a clinically mild initial episode of diverticulitis [23]. Because most of the studies are retrospective in nature, this relatively low rate of colonoscopies might therefore be due to the low clinical suspicion for CRC by the treating physician.

Although our study adds to the understanding of the rate of CRC and advanced adenoma in patients with uncomplicated diverticulitis, there also are some weaknesses that we have to address. First of all, most studies included in our review were retrospective and showed methodological weaknesses, including heterogeneity in patients and the lack of a control population, reflected in a CEBM level of evidence of medium- to low-quality studies. Additional patient data therefore were requested from the respective corresponding authors to minimize this bias. Furthermore, none of the studies included used the ideal randomised, controlled trial methodology and most of the studies were of small sample size, thereby reducing statistical power. Although three of the included studies were only available as (conference) abstracts, we decided to include these studies in order to avoid publication bias, as pointed out by the Centre for Reviews and Dissemination guidance [46]. There was no discrepancy between the results of these abstracts and the results from the others studies, and further study details were obtained by contacting the corresponding authors, as recommended by Dundar and colleagues [47].

As pointed out by Westwood and coworkers, all patients included had a definite CT or US diagnosis of acute uncomplicated diverticulitis and the findings cannot be generalized to patients with a purely clinical diagnosis of diverticulitis [27]. As there is lower diagnostic certainty in this setting, colonoscopy, CT, or US is useful to confirm the presence of diverticular disease and exclude other pathology.

Implications

Our systematic review strongly suggests that the rate of CRC and advanced adenomas in patients with uncomplicated diverticulitis are equal or even less than the rates encountered in asymptomatic individuals. Unless colonoscopy is regarded for screening in individuals aged 50 years and older, subsequent evaluation of the colon is not required. We therefore agree with the Dutch CBO Guideline on Acute Diverticulitis, which states that there is no indication for routine endoscopic evaluation after an episode of acute uncomplicated diverticulitis, unless the patient has persistent complaints or alarming symptoms for CRC [48].

Acknowledgments

The authors thank the following authors for providing additional data on their studies: Prof Frizelle, Dr. Wallace, Dr. Schout, Dr. Ahmeidat, Dr. Alexandersson, and Dr. Schmilovitz.

Disclosures

All authors declare no conflicts of interest.

Funding

None

Copyright information

© Springer Science+Business Media New York 2014