The incidence of diverticular disease and hospitalization for diverticulitis has been increasing over the past 3 decades in North America and Europe.1 Up to 25% of patients with diverticulosis will present to a hospital with an episode of acute diverticulitis in their lifetime.2 Computerized tomography has been shown to be a valuable diagnostic tool in acute diverticulitis and can be used to accurately grade the severity of disease at presentation and predict the need for surgery.3,4 Approximately 70% to 90% of admitted patients with acute diverticulitis are successfully treated nonoperatively to hospital discharge.5,6 After the resolution of diverticulitis, current practice guidelines recommend routine colonoscopic evaluation to exclude other diagnoses, namely invasive malignancy or IBD.7–9 However, these recommendations are largely based on expert opinion alone. In addition, with the increased fidelity of modern CT for the diagnosis of acute diverticulitis, the diagnostic yield of routine colonoscopy has been questioned.10 A number of case series exist, and the reported incidence of invasive malignancy on endoscopy in patients with diverticulitis has ranged from 0% to 2.8%.10–17 In addition, the incidence of advanced adenoma varies greatly among series with a reported incidence of 0% to 8.6%.10–17 The variation in these findings is probably due to the heterogeneous methodology used in these series and has led to inconsistent conclusions regarding the use of routine colonoscopy following an episode of acute diverticulitis.10,17
The purpose of this study was to determine the diagnostic yield of colonoscopy for clinically significant neoplasia (invasive malignancy or advanced adenoma) following the successful nonoperative management of acute diverticulitis in a large urban health region.
MATERIALS AND METHODS
A retrospective chart review was performed of all adult patients (>18 years) who were admitted in the Calgary Health Region between January 21, 2007 and March 31, 2010 with a diagnosis of acute left-sided diverticulitis (International Classification of Diseases, 10th Revision codes K572, K573, K578, and K579). Our health region is the sole provider of in-patient care for our large urban population (1.1 million), and, therefore, this series includes all admitted patients within this population. All patients had the diagnosis of acute diverticulitis confirmed either by CT of the abdomen or at surgery. Demographic, clinical, radiological, treatment, and outcome variables were abstracted for the index admission. Patient comorbidities were classified with the use of the age-adjusted Charlson Comorbidity Index,18 anemia was defined as a hemoglobin on admission of less than 137 g/L for men and 120 g/L for women, and diverticulitis severity was classified by CT by using the modified Hinchey classification.19 Patients who were successfully treated nonoperatively were selected for inclusion in this study. A regional endoscopy database was cross-referenced, and all endoscopy reports before index admission and within 1 year after admission were reviewed. Endoscopies performed more than 1 year after admission were not included in the analysis, because we presumed these were not performed for either the diagnosis of diverticulitis or the exclusion of other diagnoses. Patients who underwent complete colonoscopy within the 2 years before admission were excluded, because these patients were unlikely to have significant neoplasia on repeat endoscopy and are therefore not routinely referred for colonoscopy at our center. All pathology reports from polypectomy specimens and endoscopic biopsies were also reviewed. Advanced adenomas were classified as adenomas >1 cm in largest dimension, adenomas with >25% villous component, serrated adenomas, or adenomas with high-grade dysplasia. The primary end point in this study was the incidence of advanced adenoma or invasive malignancy detected endoscopically within 1 year of admission for acute diverticulitis managed nonoperatively to hospital discharge. In addition, the provincial cancer registry (which includes data on all patients diagnosed with invasive malignancies within the province) was reviewed to identify patients from this cohort who were diagnosed with a colorectal malignancy between January 1, 2007 and December 31, 2010. Based on a recent publication demonstrating a low incidence of cancer or advanced adenoma in a series of uncomplicated diverticulitis,10 an a priori subgroup analysis of patients who presented with uncomplicated and complicated diverticulitis was performed. Post hoc subgroup analyses of patients less than 50 years of age and patients 50 years of age or older at admission were also performed.
Statistical analysis was performed by using STATA version 12.0 (STATA Corp, College Station, TX). Descriptive statistical analysis was performed by using the Student t test for continuous variables, and the χ2 test or Fisher exact test, as appropriate, was performed for nonparametric variables. Univariate analysis and multivariate logistic regression was performed by using forced entry of predefined variables (age, presence of anemia, presence of abscess, and history of previous attacks) to identify risk factors for advanced adenoma or invasive malignancy. Odds ratios with 95% CIs were calculated. Statistical significance was defined as p < 0.05 for univariate and multivariate analysis.
This study was approved by the University of Calgary Conjoint Health Research Ethics Board.
During the study period, 645 patients were admitted with a diagnosis of acute diverticulitis, and 506 (78.4%) patients were successfully treated nonoperatively to hospital discharge. Of these, 48 patients had a previous complete colonoscopy within the 2 years before admission and were excluded, leaving 458 patients for analysis (Fig. 1).
Of the 458 patients, 249 patients (54%) underwent endoscopy within 1 year of admission; 245 patients (98.4%) underwent colonoscopy (with or without sigmoidoscopy), and 4 patients (1.6%) underwent sigmoidoscopy alone. The median time to endoscopy was 90 days (interquartile range, 85 days), and 214 patients (86%) had complete colonoscopic evaluation to the cecum. Patients undergoing endoscopy were significantly younger (mean age 55 years vs 59 years, p = 0.006), had fewer comorbidities (mean Charlson Comorbidity Index, 0.7 vs 1.5, p < 0.001), presented with abscess more frequently (29.7% vs 17.7%, p = 0.003), and were less likely to have an endoscopic evaluation within 2 to 5 years before admission (5.6% vs 13.9%, p = 0.003) in comparison with patients who did not undergo endoscopy (Table 1). There was no difference between patients who underwent endoscopy versus those who did not with respect to the presence of anemia or history of a previous attack of diverticulitis.
Of those patients who underwent endoscopy within 1 year of admission, 77 (30.9%) patients were found to have polyps, 19 (7.6%) patients had advanced adenomas, and 4 (1.6%) patients were found to have an invasive malignancy. Together, 23 patients (9.2%) were found to have an advanced adenoma or invasive malignancy at endoscopy performed within 1 year of admission for acute diverticulitis.
With respect to the location of pathological findings on endoscopy, polyps and advanced adenomas were discovered throughout the colon and rectum; 33.8% of polyps and 42.1% of advanced adenomas were located in the colon proximal to the splenic flexure; 42.9% of polyps and 31.6% of advanced adenomas were found in the descending or sigmoid colon; and 20.8% of polyps and 26.3% of advanced adenomas were found in the rectum (Table 2). In contrast, all invasive malignancies were found in the sigmoid colon. On retrospective review of CT images, all 4 of these patients presented with inflammation involving their sigmoid colon and abscess. There was no evidence of a discrete mass on imaging.
We performed an a priori subgroup analysis of patients who presented with complicated and uncomplicated diverticulitis. Because this cohort included only patients managed nonoperatively, there were no patients who presented with fistula or obstruction. Therefore, all patients included in the complicated diverticulitis subgroup presented with either a pericolic or pelvic abscess (n = 74). Patients presenting with abscess had a significantly higher rate of advanced adenoma or invasive malignancy (18.9% vs 5%, p = 0.001) and invasive malignancy (5.4% vs 0%, p = 0.007) in comparison with patients who presented with uncomplicated diverticulitis (Table 3).
We performed an additional subgroup analyses on patients less than 50 years of age and those 50 years of age and older at the time of presentation. For patients less than 50 years of age (n = 91), only 3 patients (3.3%) had an advanced adenoma on endoscopy. All 3 patients had serrated adenomas, and all presented with an abscess. Moreover, no patients in this subgroup were found to have an invasive malignancy on endoscopic examination. For patients 50 years of age or older (n = 158), 16 patients (10.1%) had an advanced adenoma, and 4 patients (2.5%) had an invasive malignancy on endoscopy.
Univariate analysis and multivariate logistic regression was performed for predefined risk factors for advanced adenoma or invasive malignancy (Table 4). On univariate analysis, significant risks factors for advanced adenoma or invasive malignancy were age (OR 1.04 (1.01–1.07), p = 0.02), and the presence of abscess (OR 4.30 (1.77–10.5), p = 0.001). On multivariate logistic regression, age (OR 1.04 (1.01–1.08), p = 0.02) and the presence of abscess (OR 4.15 (1.68–10.3), p = 0.002) remained significant risk factors for advanced adenoma or malignancy.
The provincial cancer registry was reviewed to identify any patients who were managed nonoperatively and who were diagnosed with a colorectal malignancy from the date of their admission to December 31, 2010 (median follow-up, 31 months; range, 11–48 months). Apart from the 4 patients diagnosed endoscopically within 1 year, no additional patients from this cohort were diagnosed with colorectal cancer.
Current guidelines recommend routine endoscopic evaluation following an episode of acute diverticulitis; however, evidence supporting these recommendations is lacking.7–9 Recent advances in CT technology, combined with increased use of imaging studies in the evaluation of patients presenting with abdominal pain, may obviate the need for routine endoscopic evaluation. Indeed, recent evidence suggests that CT can accurately diagnose acute diverticulitis with a sensitivity of 97% and a specificity of 98%.4
In this study, routine endoscopic follow-up of patients treated nonoperatively for CT-confirmed acute diverticulitis demonstrated a 9.2% incidence of clinically significant neoplasia; the incidence of advanced adenoma and invasive malignancy on endoscopic follow-up was 7.6% and 1.6%. All nonoperatively managed patients with complicated acute diverticulitis presented with either a pericolic or pelvic abscess (modified Hinchey grade 1b or 2). No patients with obstruction or fistula were included in this series, because these patients typically undergo surgical resection at our institution. When patients were stratified into complicated vs uncomplicated diverticulitis, the incidence of clinically significant neoplasia was significantly greater in those who presented with complicated disease (18.9% vs 5.1%). In multivariate analysis, presentation with abscess was a significant risk factor for the presence of clinically significant neoplasia on endoscopy. Interestingly, all 4 patients in this cohort diagnosed with an invasive malignancy on endoscopic follow-up presented with abscess.
This series includes all patients admitted with a CT-confirmed diagnosis of acute diverticulitis within the Calgary Health Region. Because this health region is the sole provider of in-patient care in metropolitan Calgary, it is a population-based assessment of the utility of diagnostic colonoscopy following the nonoperative management of diverticulitis. All endoscopic procedures in our region are recorded in a regionwide institutional database and our provincial cancer registry requires mandatory reporting of all invasive malignancies diagnosed in our province. Therefore, we are confident that our series includes all patients who underwent diagnostic endoscopy and all patients diagnosed with invasive malignancy following an admission for acute diverticulitis within this time period.
There are several limitations in our study. Although this series constitutes the second largest in the literature, the uncomplicated diverticulitis subgroup (n = 175) is relatively small and may preclude a definitive conclusion regarding the risk of clinically significant neoplasia in this group. The rate of endoscopic follow-up was lower than expected (54%). This may be due in part to the retrospective nature of the study. Patients who underwent preadmission colonoscopy more than 2 years before admission (13.9% of patients) may not have been referred for colonoscopy after admission because the surgeon did not feel that a repeat colonoscopy was warranted. In addition, we included only those patients who had a follow-up colonoscopy within 1 year of admission, because we assumed that a diagnostic colonoscopy performed beyond 1 year was unlikely to be related to their index admission. It is difficult to ascertain the indication for a procedure with certainty in a retrospective review. Given this relatively modest rate of endoscopic follow-up, we cross-referenced the provincial cancer database, and no invasive malignancies were detected at median follow-up of 31 months from presentation. Although mandatory reporting of invasive malignancies exists in our province, the database may have missed patients who migrated outside of the province in the follow-up period and patients who did not receive a biopsy-confirmed diagnosis, both of which would be relatively rare events. Another possible limitation of this study is the lack of obstructing or fistulizing disease in this series. These complicated cases are typically managed operatively and therefore were excluded because our series was confined to nonoperative patients only. This may limit the applicability of our findings to the broader group of patients presenting with complicated diverticulitis.
Several other published series examine endoscopic follow-up of diverticulitis.10–17 The reported incidence of endoscopically diagnosed advanced adenoma and invasive malignancy in these series vary from 0% to 2.8% and 0% to 8.6%.10–17 This variation can be accounted for by differences in study design: the inclusion of only uncomplicated10 vs both complicated and uncomplicated diverticulitis11–17; mandatory10,11,14–17 vs selective12,13 imaging to confirm the diagnosis; the inclusion of operatively10–12,17 vs nonoperatively13–16 managed patients; the inclusion10–13,16 vs exclusion14,15,17 of patients with preadmission endoscopy; the variable rate of endoscopic follow-up of eligible patients (ranging from 29%17 to 87%15); the variable timing of endoscopic evaluation (from the time of admission14,15 to any time after the date of diagnosis)10; and the inclusion11,12,14–17 vs exclusion10,13 of incomplete colonoscopy or poor bowel preparation. Overall, despite the methodological variations, our findings are largely consistent with these previous reports.
There are several possible explanations for our observation of an increased risk of clinically significant neoplasia in patients presenting with abscess. Although it is unlikely that there is a biological difference between these groups, patients with complicated disease may present with more inflammatory changes, making it more difficult to identify an invasive malignancy or mass on CT and a higher rate of misdiagnosis. There may also be other environmental or inherited factors that may predispose patients to complicated diverticulitis and neoplasia, such as a low-fiber diet; however, this is entirely speculative and cannot be supported or refuted with our data. Although a conclusive explanation for our findings is unclear at this time, the low rate of clinically significant neoplasia in uncomplicated diverticulitis has been observed in another large series.10
In our series, an a priori subgroup analysis of patients with CT-confirmed uncomplicated diverticulitis demonstrated an incidence of advanced adenoma of 5.1%, with no cases of invasive malignancy on endoscopy or in review of the provincial cancer registry. The incidence of advanced adenoma and invasive malignancy in this group is not statistically different than that in average-risk individuals referred to our regional screening center (6.6%, p = 0.54; 0.38%, p = 0.85),20 as well as those included in a meta-analysis of screening studies (5.7%, p = 0.88; 0.3%, p = 0.98).21 Taken together, our data suggest that patients with CT-confirmed uncomplicated diverticulitis are not at increased risk of clinically significant neoplasia in comparison with average-risk individuals. Therefore, patients may not require routine diagnostic colonoscopy following acute uncomplicated diverticulitis, unless other indications exist. However, given the sample size of this subgroup analysis, further studies are required to support this conclusion. Only 1 other series specifically examined the yield of endoscopic follow-up in patients with imaging-confirmed uncomplicated diverticulitis.10 Westwood et al10 report an incidence of invasive malignancy and advanced adenoma of 0.49% and 4.9%. However, this report has some methodological concerns. Although their reported endoscopic follow-up rate was 70%, the authors included patients who underwent colonoscopy up to 2 years before the diagnosis of diverticulitis. By virtue of a recent endoscopy, these patients would have a lower risk of invasive malignancy following a diagnosis of diverticulitis, which may have biased their findings. In addition, patients who underwent poor or incomplete colonoscopy were excluded from their analysis, which may have also biased their findings. Three of the 4 cancers in our series were found in patients whose initial colonoscopy was incomplete. Finally, patients were included if they underwent colonoscopy at any time after their CT, with no disclosure of the mean time or range of time from diagnosis to endoscopy, which raises the possibility of interval development of invasive malignancy or advanced adenoma if the follow-up endoscopy was performed well after the index admission.
Following nonoperative management of CT-confirmed diverticulitis, the incidence of clinically significant neoplasia on diagnostic endoscopy was 9.2%. However, in patients with uncomplicated diverticulitis, the incidence was only 5.1%, which is comparable to the incidence in average-risk individuals without diverticulitis. Based on our series, early diagnostic colonoscopy following the nonoperative management of acute uncomplicated diverticulitis may not be necessary, although this finding requires further study for confirmation.
1. Weizman AV, Nguyen GC. Diverticular disease: epidemiology and management. Can J Gastroenterol. 2011;25:385–389
2. Stollman N, Raskin JB. Diverticular disease of the colon. Lancet. 2004;363:631–639
3. Ambrosetti P. Acute diverticulitis
of the left colon: value of the initial CT and timing of elective colectomy. J Gastrointest Surg. 2008;12:1318–1320
4. Werner A, Diehl SJ, Farag-Soliman M, Düber C. Multi-slice spiral CT in routine diagnosis of suspected acute left-sided colonic diverticulitis
: a prospective study of 120 patients. Eur Radiol. 2003;13:2596–2603
5. Mäkelä JT, Kiviniemi HO, Laitinen ST. Spectrum of disease and outcome among patients with acute diverticulitis
. Dig Surg. 2010;27:190–196
6. Binda GA, Arezzo A, Serventi A, et al.Italian Study Group on Complicated Diverticulosis (GISDIC). Multicentre observational study of the natural history of left-sided acute diverticulitis
. Br J Surg. 2012;99:276–285
7. Stollman NH, Raskin JB. Diagnosis and management of diverticular disease of the colon in adults. Ad Hoc Practice Parameters Committee of the American College of Gastroenterology. Am J Gastroenterol. 1999;94:3110–3121
8. Rafferty J, Shellito P, Hyman NH, Buie WDStandards Committee of American Society of Colon and Rectal Surgeons. . Practice parameters for sigmoid diverticulitis
. Dis Colon Rectum. 2006;49:939–944
9. Jacobs DO. Clinical practice. Diverticulitis
. N Engl J Med. 2007;357:2057–2066
10. Westwood DA, Eglinton TW, Frizelle FA. Routine colonoscopy
following acute uncomplicated diverticulitis
. Br J Surg. 2011;98:1630–1634
11. Elramah MM, Horwitz JP, Einstein MM, et al. Colonoscopic evaluation after acute diverticulitis
: is there an association with colonic malignancy? Gastrointestinal Endosc. 2010;71:AB188
12. Schmilovitz-Weiss H, Yalunin E, Boaz M, et al. Does a colonoscopy
after acute diverticulitis
affect its management?: a single center experience. J Clin Gastroenterol. 2012;46:317–320
13. Hjern F, Jonas E, Holmström B, Josephson T, Mellgren A, Johansson C. CT colonography versus colonoscopy
in the follow-up of patients after diverticulitis
: a prospective, comparative study. Clin Radiol. 2007;62:645–650
14. Lahat A, Yanai H, Menachem Y, Avidan B, Bar-Meir S. The feasibility and risk of early colonoscopy
in acute diverticulitis
: a prospective controlled study. Endoscopy
15. Sakhnini E, Lahat A, Melzer E, et al. Early colonoscopy
in patients with acute diverticulitis
: results of a prospective pilot study. Endoscopy
16. van de Wall BJ, Reuling EM, Consten EC, et al. Endoscopic evaluation of the colon after an episode of diverticulitis
: a call for a more selective approach. Int J Colorectal Dis. 2012
17. Lau KC, Spilsbury K, Farooque Y, et al. Is colonoscopy
still mandatory after a CT diagnosis of left-sided diverticulitis
: can colorectal cancer
be confidently excluded? Dis Colon Rectum. 2011;54:1265–1270
18. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–383
19. Wasvary H, Turfah F, Kadro O, Beauregard W. Same hospitalization resection for acute diverticulitis
. Am Surg. 1999;65:632–636
20. Rostom AM, Dube C, Pontifex D, Bridges R, Hildsen R. Results of the first 10,000 colonoscopies at the Forzani & MacPhail Colon Cancer Screening Centre. Can J Gastroenterol. 2010;24:A54
21. Heitman SJ, Ronksley PE, Hilsden RJ, Manns BJ, Rostom A, Hemmelgarn BR. Prevalence of adenomas and colorectal cancer
in average risk individuals: a systematic review and meta-analysis. Clin Gastroenterol Hepatol. 2009;7:1272–1278