Aliment Pharmacol Ther 2011; 33: 358–365
Background Ulcerative colitis with diverticulosis (UCD), segmental colitis associated with diverticulosis (SCAD) and acute uncomplicated diverticulitis (AUD) may affect the same colonic regions, but the real incidence of these entities in clinical practice is unknown.
Aim To assess the incidence and the endoscopic findings of UCD, SCAD and AUD.
Methods From January 2004 to June 2009, 8525 consecutive colonoscopies were performed. Diagnosis of the diseases was based on specific endoscopic and histological (UCD and SCAD), and on endoscopic and radiological (AUD) patterns.
Results Ulcerative colitis with diverticulosis was diagnosed in 25 patients (0.3%), SCAD was diagnosed in 129 patients (2%) and AUD was diagnosed in 130 patients (2%). In UCD, the inflammation in colonic area harbouring diverticula always affects the overall colonic mucosa in all cases, involving also diverticular orifices. The endoscopic characteristic of SCAD is that inflammation is mainly detected within the inter-diverticular mucosa without involvement of the diverticular orifices. In AUD, the inflammation affects primarily diverticular orifice and peri-diverticular mucosa.
Conclusions In clinical practice, the incidence of mucosal inflammation in the presence of colonic diverticular disease is low and endoscopy is the mainstay of differential diagnosis.
Ulcerative colitis (UC), segmental colitis associated with diverticulosis (SCAD) and acute uncomplicated diverticulitis (AUD) are characterised by inflammation of the colonic mucosa that may affect the same colonic regions. In particular, left-sided colon is primarily affected in UC,1 and sigmoid and descending colon are primarily involved in SCAD and AUD.2, 3
Ulcerative colitis may be sometimes associated with diverticulosis, although the prevalence of diverticulosis in patients with UC seems to be lower than in general population.4, 5
The incidence of SCAD and, in particular, AUD is now increasing.2, 3 The endoscopic findings of these diseases may be similar. To distinguish among these three diseases may be extremely difficult, even when the results of endoscopy, surgery and histological evaluation are available. In past years, distinguishing between UC and SCAD was not a problem because the incidence of SCAD was low and most of patients were erroneously considered as suffering from UC with diverticulosis (UCD).6 New endoscopic and histological patterns were proposed for SCAD, with particular attention to the differential diagnosis with inflammatory bowel disease (IBD).2, 6 At the same time, distinguishing between SCAD and AUD may be difficult as it is still unclear whether SCAD is a complication of diverticular disease or, actually, a clinical autonomous entity.7
Endoscopy may be useful to differentiate these three entities. The endoscopic findings in UC have been well described,8 as well as new endoscopic criteria in diagnosing SCAD have been recently described.2 However, there has been no attempt for endoscopic classification of AUD. Moreover, it may be sometimes difficult to differentiate between SCAD and UCD.
We conducted a prospective study to evaluate the incidence of these three entities in clinical practice according to endoscopic classification of these three diseases.
Patients and methods
From January 2004 to June 2009, a prospective study was performed involving four primary Endoscopic Centres in central-south Italy [Servizio di Gastroenterologia Territoriale, ASL BAT, Andria (BAT); U.O.C. di Gastroenterologia, ASL Roma H, Albano Laziale (Roma); U.O.D. di Endoscopia Digestiva e Nutrizionale, Ospedale ‘S. Eugenio’, Roma; U.O.D. di Gastroenterologia, Ospedale ‘Cristo Re’, Roma].
During the study period, 8525 consecutive colonoscopies were performed: 2238 colonoscopies were performed in in-patients, 6287 were performed in out-patients. Colonoscopy was performed due to intestinal symptoms (abdominal pain, diarrhoea, constipation, rectal bleeding) in 6041 patients, whereas it was performed as screening in 1484 patients. Exclusion criteria were Crohn’s disease, large bowel resection and a previous endoscopic or radiological diagnosis of colon cancer. Moreover, other forms of chronic colitis (e.g. lymphocytic, collagenous or ischaemic), as well as complicated diverticulitis, were excluded by final assessment.
The study was approved by the Institutional review Board, and informed consent was obtained from all patients.
Diagnostic criteria for, UCD, SCAD, and AUD are summarised in Table 1.
|UCD||1. Macroscopic and microscopic inflammation involving the colonic mucosa coming from the rectum to the proximal segments|
2. Presence of diverticula in the sigmoid and descending colon, with macroscopic and microscopic inflammation of the diverticular ostia
|SCAD||1. Macroscopic and microscopic inflammation of the sigmoid-descending colon mucosa without involvement of the diverticular ostia|
2. Normal aspect of the rectum and proximal colon
|AUD||1. Macroscopic and microscopic inflammation of the colon harbouring diverticula|
2. Stage 0-Ia according to modified Hinchey classification
Diagnosis of UCD was made according to endoscopic and histological criteria: macroscopic and microscopic inflammation involving the colonic mucosa coming from the rectum to the proximal segments with presence of diverticula in the sigmoid and descending colon (Figure 1). Endoscopic severity was graded according to the ‘Mayo Subscore for endoscopy’:9
- 0: normal mucosal appearance;
- 1: Erythema, mild mucosal friability, reduction of vascular pattern;
- 2: Absence of vascular pattern, severe erythema, mucosal friability, erosions;
- 3 Spontaneous bleeding, ulcerations.
The criteria for SCAD included endoscopic lesions within the sigmoid-descending colon harbouring diverticula, combined with endoscopic and histological rectal and proximal colon sparing.10 The endoscopic classification of SCAD was based on a recent score formulated prospectively by the authors.2 This score is not based on the progressive severity of the disease, but it is based on the character of the endoscopic damage:
- (i) ‘Crescentic fold disease’. The colonic mucosa shows some swollen red patches ranging from 0.5 to 1.5 cm in diameter, without haemorrhage or ulceration, confined to the crescentic mucosal fold. Diverticular orifices are always spared.
- (ii) ‘Mild-to moderate ulcerative colitis-like’ pattern (Figure 2). This endoscopic pattern includes diffuse loss of vascular pattern, mucosal oedema and hyperaemia and diffuse erosions. Moreover, in these cases diverticular orifices are always spared.
- (iii) ‘Crohn’s disease colitis-like’ pattern. This endoscopic pattern shows scattered aphtous ulcers, which may be found within a normal colonic mucosa. Diverticular orifices are always spared.
- (iv) ‘Severe ulcerative colitis-like’ pattern. This endoscopic pattern shows diffuse loss of vascular pattern, diffuse hyperaemia, and easy bleeding at contact with colonoscope. Oedema in the mucosa with ulcerations was marked, sometimes with reduction of the lumen. The diverticular orifices are not always easy to recognise, but they may be visible and spared by inflammation at maximal air inflation. Sometimes, inflammation also involves diverticular orifices. In these cases, differential diagnosis with UCD is based on SCAD rectal sparing.
The endoscopic characteristic of SCAD is that inflammation is mainly detected within the inter-diverticular mucosa without involvement of the diverticular orifices.
Diagnosis of AUD was made according to endoscopic and radiological criteria: macroscopic and microscopic inflammation involving the colon harbouring diverticula, and affecting bowel wall without complications, assessed by computerised tomography (CT) scan.11 In particular, complicated diverticulits was defined as stage Ib-III or the presence of fistula or obstruction on CT according to a modified Hinchey classification.12
After exclusion of complicated acute diverticulitis on CT scan, patients underwent colonoscopy. As there is no endoscopic classification of severity of AUD, endoscopic severity was arbitrarily graded as follows:
- (i) Mild (Figure 3): diverticular inflammation localised within and around diverticula;
- (ii) Moderate: diverticular inflammation that affects both peri- and inter-diverticular mucosa;
- (iii) Severe: inflammation that affects the overall colonic segment harbouring diverticula.
All patients underwent the same standard bowel preparation for colonoscopy prescribed in our centres consisting in an oral polyethylene glycol solution (Selg-Esse, Promefarm, Milano, Italy) to be taken during the evening. The following morning pancolonoscopy (clean colon colonoscopy) was performed. Using standard biopsy forceps, six bioptic samples of colonic mucosa were collected in the colonic tract affected by lesions for histological assessment. Moreover, six bioptic samples of the colonic mucosa were obtained from every other colonic region (ascending, transverse colon and rectum) to confirm the histological diagnosis of SCAD.
Descriptive analysis was performed, and continuous data were expressed as the mean (range) and categorical data as numbers (percentage).
Ulcerative colitis was diagnosed in 242 patients (3%), and diverticulosis was found in 1104 patients (13%). Among them, SCAD was diagnosed in 129 patients (11%, 2% of the overall studied population), UCD was diagnosed in 25 patients (2%, 0.30% of the overall studied population) and AUD was diagnosed in 130 patients (12%, 2% of the overall studied population). The results are summarised in Table 2.
|Endoscopic spectrum||UCD (25 pts)||SCAD (129 pts)||AUD (130 pts)|
|Mild (8 pts)||Moderate (10 pts)||Severe (7 pts)||A (72 pts)||B (36 pts)||C (13 pts)||D (8 pts)||Mild (67 pts)||Moderate (42 pts)||Severe (21 pts)|
|Diffuse inflammation||7 (87.5)||9 (90.0)||7 (100)||–||–||–||8 (100)||1 (1.5)||1 (2.4)||7 (33.3)|
|Inflammation of inter-diverticular mucosa||1 (12.5)||1 (10.0)||–||71 (98.6)||34 (94.4)||11 (84.6)||–||2 (3.0)||12 (28.6)||13 (61.9)|
|Inflammation of peri-diverticular mucosa||–||–||–||1 (1.4)||2 (5.6)||2 (15.4)||–||64 (95.5)||29 (69.0)||1 (4.8)|
Ulcerative colitis with diverticulosis
According to the above mentioned criteria, UCD was diagnosed in 25 patients (15 Men, 10 Women, mean age 63.68 years, range 63–82 years). The most frequent form of disease diagnosed was procto-sigmoiditis in 13 patients (52%), followed by left-sided colitis in seven subjects (28%), and pancolitis in 5 (20%). Diverticula were localised in sigmoid region in 14 cases (56%) and in sigmoid-descending colon in 11 (44%), respectively.
In this disease, the inflammation in colonic area harbouring diverticula always affects the overall colonic mucosa in all cases, involving also diverticular orifices (see Figure 2). Rectal bleeding was the most recorded isolated symptom (11 patients, 44%), followed by diarrhoea (4 patients, 16%), and by abdominal pain (2 patients, 8%). Eight patients (32%) experienced more than one symptom.
Segmental colitis associated with diverticulosis
According to the above mentioned criteria, SCAD was diagnosed in 129 patients (77 Men, 52 Women, mean age 63.68 years, range 43–87 years). The endoscopic lesions affected primarily inter-diverticular mucosa in 121 patients (94%). Only in the severe form of the disease (‘Severe ulcerative colitis-like’ pattern SCAD) the inflammation involved the overall colonic mucosa in eight subjects (6%). Chronic diarrhoea was the most recorded isolated symptom (58 patients, 45%), followed by recurrent abdominal pain (15 patients, 12%), and by recurrent rectal bleeding (19 patients, 15%). Thirty-seven patients (29%) experienced more than one symptom.
Acute uncomplicated diverticulitis
According to the above mentioned criteria, AUD was diagnosed in 130 patients (81 Men, 49 Women, mean age 64.71 years, range 40–85 years). Eighty-three patients (64%) referred a previous history of diverticulosis or diverticular disease, whereas diverticular disease was diagnosed for the first time in 47 patients (36%). In the 66 in-patients (51% of the overall population affected by AUD), we performed an ‘early colonoscopy’ (within 7 days since admission, and after a 7 days course of antibiotic therapy) in patients affected by AUD13 in order to exclude other diseases (in particular colorectal cancer or IBD) and in order to confirm the AUD diagnosis as soon as possible, whereas in the 64 out-patients (49% of the overall population affected by AUD) colonoscopic examination was the first diagnostic tool in diagnosing AUD. No complication was recorded after colonoscopy, and it was able to confirm the radiological diagnosis in all patients undergoing to the intervention. In AUD, the inflammation affects primarily diverticular orifice and peri-diverticular mucosa. Essentially, we found inflammation restricted to mucosa surrounding diverticular orifices in 64 (49%) patients (see Figure 3). When the inflammation increases, it begins to involve also inter-diverticular mucosa, as we found in 12 of 42 patients (29%) and in 13 of 21 (62%) with moderate and severe form of AUD, respectively. Abdominal pain was the most recorded isolated symptom (79 patients, 61%), followed by diarrhoea (15 patients, 12%), constipation (11 patients, 8%) and by rectal bleeding (9 patients, 7%). Sixteen patients (12%) experienced more than one symptom.
Current literature data showed that incidence of diverticulosis is increasing.14 Most of these patients are suffering from diverticular disease (DD), but often the endoscopists are faced up to make a correct differential diagnosis between DD and other diseases in which diverticulosis can be found. This occurrence is particularly true regarding patients affected by UCD and SCAD.
This study showed that the incidence of diverticulosis in UC is low in clinical practice. Two recent studies reported that prevalence of diverticulosis in UC ranges from 2% to 8%, affecting in particular patients older than 50 years.4, 5 This study found a similar incidence (3%), with a mean age of 63.68 years. The first consideration about this study is that incidence of UCD, SCAD and AUD is low in clinical practice. These results may be related to the overall number of colonoscopies performed in the observed period (8525 colonoscopies performed in four centres during a 4.5-year period). This endoscopic volume would appear quite low, suggesting that a referral bias is possible. However, we think that this may be also a strength of this study because it describes what is the real incidence of these diseases in clinical practice.
An interesting finding arising from this study is that SCAD and AUD showed surprising similar incidence. The literature incidence of SCAD ranges from 0.3% to 4%.15, 16 Two recent prospective studies confirmed that the incidence of SCAD ranged from 0.3%9 to 2%.2 In particular, this last prospective study found that about 12% of patients with diverticulosis are affected by SCAD.2 Our prospective study found a SCAD incidence of 1% (11% of the population harbouring diverticula) similar to the highest values reported in literature. Although the incidence of this disease is low, considering these results, an endoscopic diagnosis of SCAD may be expected to occur once every 300–400 colonoscopies. This study confirms a low incidence of the disease, but an increasing occurrence of new cases may be registered given the high number of colonoscopies currently carried out at our endoscopic centres.
The incidence of AUD is increasing. A recent nationwide in-patient sample survey, performed in United States, found that the incidence of patients admitted with acute diverticulitis raised from 120 500 in 1998 to 151 900 in 2005 (26% increase).3 Another recent study from United Kingdom found that annual admissions for DD (including both symptomatic DD and acute diverticulitis) increased from 71 to 263. There was a significant trend of decreasing mean age from 71.2 years in 1995 to 68.1 years in 2004, and admissions younger than 50 years raised from 8% in 1995 to 42% in 2003,17 with poor outcome in the elderly.18 We found a low AUD incidence of 2% (12% of the population harbouring diverticula). As this study only assessed AUD, it is probably that the real incidence of acute diverticulitis is higher than recorded. Actually, we should consider that complicated diverticulitis is managed surgically, the acute complicated diverticulitis incidence is increasing particularly in younger people,3 and in most of patients with CT findings compatible with diverticulitis do not undergo colonoscopy. Performing colonoscopy in acute diverticulitis is still controversial, because acute inflammation of the diverticula may be at risk of perforation or bleeding.19, 20 But recent literature data found that early colonoscopy (within 3 and 11 days after the admission) is as safe and effective as late colonoscopy (within 6 and 19 weeks following admission), without any complication in both approaches. We performed colonoscopy within 7 days after exclusion of complicated diverticulitis [which is considered the only parameter avoiding early colonoscopy13] and after a course of antibiotic therapy, and we did not record any complication. Somebody may argue that computerised tomography colonography (CTC) may have better diagnostic potential for imaging of diverticular disease-specific findings, when compared with colonoscopy. Moreover, CTC is less uncomfortable and may be preferred by a majority of patients.21 However, we cannot forget that diverticular disease often shows circular muscle hypertrophy that leads to a thickened colonic wall. This condition needs a correct differential diagnosis as colorectal adenocarcinoma or new diagnosis of IBD (namely SCAD or UCD) can be established.22 On these bases, colonoscopy is advisable in these patients to make a correct diagnosis. The safety of colonoscopy is an important issue in these patients. Our opinion is that an adequate course of antibiotic therapy before colonoscopy is a good option to reduce the risk of perforation. We performed early colonoscopy after a 7-day course of antibiotic therapy, after careful clinical evaluation and after exclusion of free perforation by abdominal CT. No perforation was recorded during or after colonoscopy. Therefore, our opinion is that gentle colonoscopy with minimal air insufflation can be carried out safely in that patients. If a diagnosis of diverticulitis is confirmed, the procedure may terminate because the risk of perforation cannot be completely avoided.23
As the incidence of these three diseases is quite low in clinical practice, what is the role of colonoscopy in managing these three diseases in clinical practice?
First of all, our study found that endoscopy is mainstay to discriminate these three clinical entities. In fact, colonoscopy found inflammation localised in different fashions. In UCD, inflammation affects the overall mucosa of the region harbouring diverticula, also involving colonic mucosa surrounding diverticular orifices. In SCAD, inflammation affects primarily inter-diverticular mucosa, and only in severe cases can we see the overall involvement of the colonic mucosa harbouring diverticula. In these last cases, the correct differential diagnosis between SCAD and UCD is based on endoscopic and histological rectal assessment, which shows sparing by inflammation, and which, by definition, characterised SCAD instead of UC. On the contrary, in AUD the inflammation affects primarily the colonic mucosa surrounding diverticula, and only in more severe forms of the disease can we see the overall involvement of the colonic mucosa surrounding diverticula. In these last cases, differential diagnosis between SCAD and AUD is based on three criteria:
- (i) CT scan finding of thickening of the colonic wall, sometimes associated with fat stranding, in AUD instead of SCAD;24
- (ii) Endoscopic findings in AUD are characterised by aspecific signs, such as diffuse mucosal oedema and fragility, and sometimes bleeding coming from diverticula,13, 25 whereas SCAD is characterised by diffuse erosions and, sometimes, by ulcers;2
- (iii) Histological findings of SCAD are characterised by IBD-like pictures in most of cases,2, 9 whereas AUD mainly shows an active, unstructured inflammatory infiltrate.26, 27
But more important to the clinician is assistance regarding the appropriate clinical circumstances and timing of endoscopy. This detail is particularly true in managing the disease. Knowing the prevalence and the natural history of each disease, endoscopic appearance may be a key point to programme adequate clinical and endoscopic surveillance. For example, the treatment of UCD is similar to that of a standard UC that may evolve to a more severe form of the disease.28 SCAD has usually a more benign course,29, 30 but some cases of severe disease may require immunosuppressors or surgery. Moreover, AUD may be managed with conservative therapy,31 but it often recurs with complication.32 In light of this information, it is probable that less aggressive forms of SCAD (types A and C) required less aggressive clinical and endoscopic follow-up, whereas UCD or more aggressive forms of SCAD (types B and D) required more aggressive clinical and endoscopic follow-up. An opening question remains the role of endoscopy in the follow-up of AUD as there are no data about the role of colonoscopy in long-lasting management of this disease.
In conclusion, this study showed that incidence of UCD, SCAD and AUD is low in clinical practice. Although endoscopy is the mainstay of diagnosis, further studies are needed to solve the questions opening by this study.
Declaration of personal and funding interests: None.