Introduction
In recent years, with the westernization of eating habits and aging population in
Japan, the prevalence of colonic diverticulum has increased, and the opportunity to
treat colonic diverticulosis has increased worldwide. Colonic diverticulitis, which
is one of the complications of colonic diverticulosis, is diagnosed by clinical symptoms
and imaging, such as computed tomography (CT), and colorectal cancer (CRC) is one
of the differential diagnoses. The American Gastroenterological Association (AGA)
guidelines recommend that patients undergo a colonoscopy 4 to 8 weeks after an event
of diverticulitis if they have not recently undergone colonoscopy [1]. This recommendation is based on some observational studies of patients with diverticulitis
diagnosed by imaging and subsequent colonoscopy. According to the AGA technical review,
1 in 67 patients (1.5 %) initially diagnosed with diverticulitis are estimated to
be misdiagnosed with CRC on follow-up colonoscopy [2]. However, it is unclear whether there is a relationship between diverticulitis and
CRC in the Japanese population [3]. In a previous study, it was reportedly uncertain whether patients with colon diverticulitis
have a higher rate of CRC than the general population [4]
[5]. The 2017 Japan Gastroenterological Association (JGA) guidelines also recommend
performing a colonoscopy to exclude malignancies other than colonic diverticulosis
[3]. However, whether it is appropriate to have a similar strategy in Japan is unclear
because diverticular disease in the right colon is more prevalent in Japan than in
Western countries [3]. In addition, the current guidelines do not specifically indicate the risk of CRC
in patients with acute diverticulitis. In the present study, we investigated the impact
of initial colonoscopy after an acute colonic diverticulitis event on the detection
of colorectal neoplasms in comparison to the impact of colonoscopy due to positive
fecal immunochemical tests (FITs).
Patients and methods
There were 297 consecutive patients with suspected acute colonic diverticulitis diagnosed
by computed tomography (CT) who were hospitalized in the Toyonaka Municipal Hospital
from February 2011 to December 2019. Acute diverticulitis was diagnosed based on CT
images showing bowel wall thickening and fat stranding consistent with diverticulitis,
abdominal pain and elevated C-reactive protein levels and white blood cell counts.
We enrolled diverticulitis patients with an initial colonoscopy or no colonoscopy
history within 5 years as the diverticulitis group, and all of these patients underwent
colonoscopy 24 weeks after an event of diverticulitis except for those who underwent
surgery. During the same period, we enrolled asymptomatic patients who were referred
to our hospital due to positive FIT results from a CRC screening program as a control
group. Japan introduced a CRC screening program with FIT in 1992. The patients in
the control group showed at least one positive result with a cut-off value of 100 ng/mL
in a 1-day test or 2-day test from the screening program and visited our hospital
for colonoscopy. Then, we retrospectively evaluated and compared patients with acute
diverticulitis and FIT-positive patients in terms of their CRC neoplasia rates.
The present study was conducted in accordance with the Declaration of Helsinki, and
approval was obtained from the Institutional Review Board of Toyonaka Municipal Hospital
(27 January 2020, No. 2019-03-03). The requirement for informed consent was waived
via the opt-out method on our hospital website.
Colonoscopy procedure
We routinely used an Olympus CF-Q260AI or CF-H290I series device (Olympus Optical
Co., Tokyo, Japan) under conventional white light and narrow-band imaging. Carbon
dioxide insufflation was used for all colonoscopies. If a patient preferred sedation,
we used intravenous midazolam under oxygen saturation and electrocardiographic monitoring.
Anticonvulsants, including scopolamine butyl bromide and glucagon, were used unless
contraindicated. We detected colorectal polyps during colonoscopy, the indication
for endoscopic resection (ER) was a polyp size > 5 mm, and diminutive polyps were
allowed to be followed up without resection based on the judgment of each endoscopist
and in adherence to the Japan Gastroenterological Endoscopy Society (JGES) guideline
[6]. However, diminutive flat and depressed lesions that were difficult to distinguish
from adenoma or carcinoma were resected. The final decision to perform ER was made
by each physician based on the bowel preparation and patient’s condition. All removed
lesions were diagnosed by full-time pathologists at our hospital.
Definitions
Cecum intubation was defined as the passage of the colonoscope tip from the rectum
to the ileocecal valve. Each endoscopist evaluated the quality of bowel preparation
using the Boston Bowel Preparation Scale (BBPS) [7].
In the present study, CRC was defined as submucosal and highly invasive cancer without
ER indications. Advanced neoplasia was defined as intramural or slightly submucosal
invasive cancer or adenoma with a diameter of at least 10 mm and high-grade dysplasia,
villous or tubulovillous histologic characteristics, or any combination of these features
[8]. The adenoma detection rate (ADR) and advanced neoplasia detection rate (ANDR) were
defined as the proportion of colonoscopies that led to the removal of at least one
histologically confirmed colorectal adenoma per colonoscopy. The polyp detection rate
(PDR) was defined as the proportion of colonoscopies that led to the detection of
at least one polyp without histological confirmation.
Endpoints
The primary endpoint was the incidence of colorectal neoplastic lesions (CRC and advanced
neoplasia) in the diverticulitis group and the FIT-positive group. The secondary endpoints
were the ADR and PDR in the two groups.
Statistical analysis
All continuous variables are presented as the mean ± standard deviation (SD). Categorical
variables are expressed as frequencies and percentages. The differences in categorical
variables were evaluated for statistical significance by t-tests, and differences
in continuous variables were evaluated with the Fisher’s exact tests.
We stratified patients in both groups by age: under 40 years, 40 to 59 years, and
60 years or older. Then, we performed propensity score matching to minimize the potential
confounding variables and selection biases. Age and sex were used as matching variables
between the two groups, and we created a 1:1 matched cohort with an alpha of 0.05
to minimize the impact of potential selection bias. All calculated P values were two-sided, and P < 0.05 was considered statistically significant. All statistical analyses were performed
with JMP statistical software (ver. 14.3, SAS Institute Inc., Cary, North Carolina,
United States).
Results
Patients
During the study period, 282 patients were finally diagnosed with colonic diverticulitis,
except for 15 patients diagnosed with other diseases ([Fig. 1]). Acute diverticulitis was treated with bowel rest and antibiotics, but 36 patients
required surgery. Of the patients with colonic diverticulitis, we excluded 139 patients
with a previous colonoscopy within 5 years, unexamined colonoscopy results because
of diverticulitis recurrence or other reasons, or unwillingness to undergo colonoscopy.
Finally, 143 patients (60 %) underwent colonoscopy. Of them, four surgical patients
underwent colonoscopy over 24 weeks. During the same study period, 4232 patients underwent
colonoscopy because of positive FIT results. We excluded 399 patients due to a history
of previous colonoscopy and 2014 patients due to unavailable data ([Fig. 1]). Finally, we enrolled 143 patients in the diverticulitis group and 1891 patients
in the FIT-positive group as controls.
Fig. 1 Study flowchart for the diverticulitis and FIT-positive groups.
Patient characteristics and colonoscopy quality indicators in patients with acute
diverticulitis and positive FIT results
[Table 1] shows the patient characteristics and quality indicators of the colonoscopies. The
diverticulitis group was significantly younger than the FIT-positive group (53 ± 15
vs. 67 ± 12 years, P < 0.0001), but there was no significant sex difference between groups (male, 56 %
vs. 59 %). The sedative midazolam was more commonly used in the diverticulitis group. The
average BBPS score, caecum intubation rate, intubation time, and total procedure time
did not significantly differ between the two groups. Compared to the diverticulitis
group, the FIT-positive group had a significantly higher number of detected polyps
(1.0 vs 2.7, P < 0.0001), a higher CRC rate (0 vs 2.7 %, P = 0.0061), ANDR (5.6 vs. 14.0 %, P = 0.0017), ADR (19.6 vs. 53.2 %, P < .0001), and PDR (44.1 vs. 91.0 %, P < .0001). There were no CRC patients in the diverticulitis group ([Table 1]).
Table 1
Baseline patient characteristics of the diverticulitis and FIT-positive groups and
their quality indicators.
|
Diverticulitis
|
FIT-positive
|
P value
|
|
N
|
143
|
1819
|
|
|
Age mean ± SD
|
53.1 ± 15.3
|
67.3 ± 11.7
|
< 0.0001
|
|
Sex, male (%)
|
56
|
59
|
0.5354
|
|
Sedative drug use[1], N (%)
|
88 (76.5)[2]
|
924 (55.1)[2]
|
< 0.0001
|
|
BBPS score, mean ± SD
|
7.8 ± 1.3
|
7.5 ± 1.5
|
0.0528
|
|
Caecum intubation rate, N (%)
|
143 (100)
|
1683 (98.5)[2]
|
0.1388
|
|
Intubation time, min
|
9.3
|
9.8
|
0.4222
|
|
Total procedure time, min
|
22.2
|
27.9
|
< 0.0001
|
|
Total number of detected polyps
|
1.0 ± 1.6
|
2.7 ± 2.6
|
< 0.0001
|
|
Total number of resected polyps
|
0.4 ± 0.9
|
1.2 ± 1.6
|
< 0.0001
|
|
CRC, N
(%, 95 % CI)
|
0
(0)
|
49
(2.7, 2.0–3.5)
|
0.0061
|
|
ANDR, N
(%, 95 % CI)
|
8
(5.6, 2.9–10.7)
|
254
(14.0, 12.4–15.6)
|
0.0017
|
|
ADR, N
(%, 95 % CI)
|
28
(19.6, 13.9–26.8)
|
968
(53.2, 50.9–55.5)
|
< 0.0001
|
|
PDR, N
(%, 95 % CI)
|
63
(44.1, 36.2–52.2)
|
1655
(91.0, 89.6–92.2)
|
< 0.0001
|
FIT, fecal immunochemical test; BBPS, Boston Bowel Preparation Scale; CRC, colorectal
cancer; ANDR, advanced neoplasia detection rate; ADR, adenoma detection rate; PDR,
polyp detection rate.
1 Midazolam.
2 We calculated the frequencies and percentages and evaluated the differences for statistical
significance, excluding cases with missing data.
Detection rate of colorectal neoplasia according to age group
Because CRC rates increase with age, we divided the patients into three groups according
to age: under the age of 40 years, 40 to 59 years, and 60 years or older. [Table 2] shows the number of patients and colorectal neoplasia detection rate of each age
group. There were no significant sex differences between the diverticulitis and FIT-positive
groups at any age. The CRC detection rates and ANDRs did not differ between the two
groups at any age, but there were higher rates in the FIT-positive group than in the
diverticulitis group, although there were no CRC patients under the age of 40 in either
group. The ADR and PDR were significantly higher in the FIT-positive group than in
the diverticulitis group at all ages.
Table 2
Number of patients and colorectal tumor detection rate for each age group.
|
|
Diverticulitis
|
FIT-positive
|
P value
|
|
Age, years
|
Colorectal neoplasia
|
N = 143
|
N = 1819
|
|
|
< 40
|
N
|
32 (22.3)
|
30 (1.6)
|
|
|
Male sex (%)
|
15 (46.9)
|
18 (60.0)
|
0.8017
|
|
CRC, N (%)
|
0 (0)
|
0 (0)
|
0.0000
|
|
ANDR, N (%)
|
0 (0)
|
3 (10.0)
|
0.0667
|
|
ADR, N (%)
|
2 (6.3)
|
13 (43.3)
|
0.0007
|
|
PDR, N (%)
|
6 (18.8)
|
24 (80.0)
|
< 0.0001
|
|
40 to 59
|
N
|
65 (45.5)
|
382 (21.0)
|
|
|
Male sex (%)
|
38 (58.5)
|
243 (63.6)
|
0.4877
|
|
CRC, N (%)
|
0 (0)
|
6 (1.6)
|
0.3090
|
|
ANDR, N (%)
|
5 (7.7)
|
38 (10.0)
|
0.5686
|
|
ADR, N (%)
|
16 (24.6)
|
192 (50.3)
|
0.0001
|
|
PDR, N (%)
|
28 (43.1)
|
335 (87.7)
|
< 0.0001
|
|
≥ 60
|
N
|
46 (32.2)
|
1407 (77.4)
|
|
|
Male sex (%)
|
24 (52.2)
|
805 (57.2)
|
0.5460
|
|
CRC, N (%)
|
0 (0)
|
43 (3.1)
|
0.2287
|
|
ANDR, N (%)
|
3 (6.5)
|
213 (15.1)
|
0.1060
|
|
ADR, N (%)
|
10 (21.7)
|
763 (54.2)
|
< 0.0001
|
|
PDR, N (%)
|
29 (63.0)
|
1296 (92.1)
|
< 0.0001
|
FIT, fecal immunochemical test; CRC, colorectal cancer; ANDR, advanced neoplasia detection
rate; ADR, adenoma detection rate; PDR, polyp detection rate.
Propensity score matching
We performed propensity score matching analysis using age and sex because male sex
as well as old age are strongly correlated with colorectal neoplasia [9]. After propensity score matching, we analyzed a total of 276 age- and sex-matched
patients in both groups ([Table 3]). There was no significant difference in the detection rates for CRC (0 % vs. 0.7 %,
P = 0.3164) or advanced neoplasia detection (5.8 % vs 7.3 %, P = 0.6259) between the diverticulitis group and the FIT-positive group. The ADR and
PDR were significantly higher in the FIT-positive group than in the diverticulitis
group (20.3 % vs 43.5 %, P < 0.0001; 45.7 % vs 86.2 %, P < 0.0001).
Table 3
Patient characteristics and colorectal neoplasm detection rate after propensity score
matching.
|
Diverticulitis
|
FIT-positive
|
P value
|
|
N
|
138
|
138
|
|
|
Age
|
54
|
54
|
0.99
|
|
Sex, male (%)
|
56
|
57
|
0.81
|
|
Sedative drug use[1], N (%)
|
84 (60.9)
|
75 (54.3)
|
0.0023
|
|
BBPS sore, mean ± SD
|
7.8 ± 1.3
|
7.6 ± 1.5
|
0.2337
|
|
Caecum intubation rate, N (%)
|
138 (100)
|
138 (100)
|
|
|
Intubation time, min
|
9.3 ± 6.6
|
11.0 ± 9.7
|
0.4409
|
|
Total procedure time, min
|
22.2 ± 8.9
|
27.9 ± 14.4
|
0.0041
|
|
Total number of detected polyps
|
1 ± 1.6
|
1.9 ± 1.8
|
< 0.0001
|
|
Total number of resected polyps
|
0.4 ± 0.9
|
1 ± 1.5
|
< 0.0001
|
|
Colorectal cancer, N
(%, 95 % CI)
|
0
(0)
|
1
(0.7, 0.1–4.0)
|
0.3164
|
|
Advanced neoplasia, N
(%, 95 % CI)
|
8
(5.8, 3.0–11.0)
|
10
(7.3, 4.0–12.8)
|
0.6259
|
|
ADR, N
(%, 95 % CI)
|
28
(20.3, 14.4–27.8)
|
60
(43.5, 35.5–51.8)
|
< 0.0001
|
|
PDR, N
(%, 95 % CI)
|
63
(45.7, 37.6–54.0)
|
119
(86.2, 79.5–91.0)
|
< 0.0001
|
FIT, fecal immunochemical test; BBPS, Boston Bowel Preparation Scale; CRC, colorectal
cancer; ANDR, advanced neoplasia detection rate; ADR, adenoma detection rate; PDR,
polyp detection rate.
1 Midazolam.
Discussion
We found that the CRC rate after acute diverticulitis was not as high as that in asymptomatic
patients with positive FIT results detected by a national CRC screening program in
Japan. Propensity matching analysis after matching for sex and age also revealed that
there was no significant difference in the CRC rate and ANDR between groups, but the
ADR and PDR were significantly higher in the FIT-positive group. These results may
suggest that colonoscopy can be recommended after colonic diverticulitis for CRC detection,
as it is after positive FIT results, but does not seem to have as much impact after
colonic diverticulitis as after positive FIT results regarding adenoma or polyp detection.
There have been reports on the necessity of colonoscopy to exclude CRC after colonic
diverticulitis. In a systematic review reported by Sai et al., 2.1 % of colonic diverticulitis
patients diagnosed by CT who underwent surgery (14/771) were found to have CRC by
barium enema or colonoscopy within 24 weeks. The authors concluded that CRC is more
common in patients with colon diverticulitis than in the general population [10]. The highest incidence of CRC after diverticulitis was reported to be 7.4 % (11/147)
by Choi et al. [11]. The lowest value was 0.25 % (1/404) by Lecleire et al. [12]. They concluded that the adenoma and CRC rates for colonoscopy after acute diverticulitis
were similar to those for the screening program [12]. The incidence rates differ among reports. It has been questioned how essential
colonoscopy is after diverticulitis.
A recent study recommended selecting patients with CRC risk factors for colonoscopy
because the risk for CRC increases with older age, male sex, or family history [13]
[14]
[15]
[16]. Chan DKH et al. reported that among patients with diverticulitis under 50 years
old who underwent colonoscopy, CRC or advanced adenoma were not observed in the younger
patients [17]. Therefore, the authors concluded that follow-up colonoscopy after diverticulitis
might not be necessary for younger patients [17]. In the present study, we evaluated the incidence of CRC among sex-matched groups
stratified by age. There were no patients under 40 years of age with CRC or advanced
neoplasia in the diverticulitis group, and the ANDR tended to be higher in the FIT-positive
group, but the difference was not significant. For patients aged 40 years or older,
the CRC detection rate and ANDR were not significantly different between the two groups.
Colonic diverticulosis is an age-dependent disease, but diverticulitis was more common
in younger patients than diverticular bleeding [18]. In the present study, patients under 40 years of age were more common in the diverticulitis
group than in the FIT-positive group. Therefore, this result could support that colonoscopy
may not always be necessary after diverticulitis for patients aged under 40 years
if they do not have CRC risk factors.
Colorectal screening with a chemical fecal occult blood test (FOBT) can suppress CRC-related
mortality, which has been demonstrated by several randomized controlled trials on
screening [13]
[14]
[15]
[16]. Compared with FOBT, FIT has the advantage of not being affected by diet [19]. Therefore, FIT is currently used for CRC screening in Japan. The rate of CRC detection
by colonoscopy after a positive FIT result is 3.9 % to 4.2 % [13]
[14]
[15]
[16]. In the present study, we defined patients with positive FIT results as the control
group, because it is difficult to use screening colonoscopy data for asymptomatic
populations in Japan who are not covered by the insurance system. Colonoscopy for
FIT-positive patients aged 40 years or over is common and acceptable in Japan because
the CRC screening program is offered to the population aged 40 years or over. Therefore,
we think these patients are reasonable as a control group.
There is a strength in the present study. We evaluated the ADR in addition to the
CRC and advanced neoplasia rates. Colorectal adenoma is a well-established precursor
lesion for the majority of CRCs. Until now, there has been only one report analyzing
the incidence of adenoma in patients with acute diverticulitis [12] that showed that there was no significant difference in ADR between acute diverticulitis
patients with a mean of 61 years and control patients with a mean of 61 years who
underwent screening colonoscopy (12.1 vs 14.6 %, P = 0.35). Although the mean age of our diverticulitis group was younger (53 years)
than that of this cohort, the ADR in the diverticulitis group was significantly lower
than that in the FIT-positive group (19.6 vs 53.2 %, P < 0.0001). After the propensity score matching analysis using age and sex, the ADRs
and PDRs, which were surrogate markers, in the diverticulitis group were approximately
half those in the FIT-positive group.
There are several limitations to the current study. First, this study has a single-center
retrospective design. Our cohorts could represent real-world data in Japan because
our hospital is a regional core municipal hospital with 613 beds, and we analyzed
consecutive patients with diverticulitis. Second, the diverticulitis group had a small
sample size, which could have resulted in the absence of CRC patients in this group
in the present study. Consequently, this study was underpowered to show a difference
in CRC incidences between patients with positive FITs and those with diverticulitis.
Therefore, we evaluated adenoma, which can be a surrogate marker for CRC. Third, almost
half of the patients with positive FIT results during the study period were excluded
due to missing or unavailable data, which may cause bias. However, we were able to
enroll 1819 FIT-positive patients, which we considered a sufficient number for the
control group to compare with the diverticulitis group because of the very narrow
95 % confidential intervals in CRC rate, ANDR, ADR, and PDR. Finally, we did not consider
the patient history of other cancers, family history of CRC, or disease sidedness,
all of which could have affected the results.
Conclusion
We found that patients with acute diverticulitis and those with positive FIT results
had similar CRC rates and ANDRs, but those with acute diverticulitis had significantly
lower ADRs and PDRs. It may not be necessary to perform subsequent colonoscopy to
exclude malignancy for patients with colonic diverticulitis diagnosed by cross-sectional
imaging methods such as CT.
