Keywords
serrated epithelial change - inflammatory bowel disease - dysplasia - endoscopy
Introduction
Patients with inflammatory bowel disease (IBD) have an increased risk of developing
colorectal cancer (CRC), which rises further with longer disease duration, greater
inflammatory activity, and more extensive colonic involvement.[1]
[2] The relative risk is at least twofold higher in patients with colonic Crohn's disease
or ulcerative colitis compared to the general population.[3]
[4]
Serrated epithelial change (SEC) in colonic biopsies is an increasingly recognized
histopathological finding in patients with IBD. However, there is a lack of robust
data on its prognostic significance. Several studies have suggested an association
between SEC and an increased risk of colorectal dysplasia.[5]
[6]
[7] Consequently, the identification of SEC may have implications for dysplasia surveillance
intervals in this population. Nonetheless, the clinical significance of SEC remains
uncertain.
SEC is histologically distinct from hyperplastic polyps, which exhibit evenly spaced,
straight funnel-shaped crypts with a basally located proliferation zone, and from
sessile-serrated lesions, which are characterized by serrated architecture extending
to the base of irregularly distorted crypts. While hyperplastic polyps and sessile-serrated
lesions are typically visible endoscopically, SEC may appear as polypoid or nodular
lesions but can also be invisible and only detectable through random biopsies of flat
mucosa. Currently, No widely accepted histopathological criteria exist for diagnosing
SEC. However, it has been described as colonic mucosa with goblet cell–rich epithelium
and serrated crypt architecture without basal crypt distortion or dysplasia.[6]
[8]
[9]
Dysplasia detection in IBD has traditionally relied on mucosal inspection with targeted
and random biopsies to identify both visible and invisible lesions.[10] In the past decade, advances such as chromoendoscopy and high-definition imaging,
endorsed by the 2015 SCENIC international consensus[11] and the 2021 American Gastroenterology Association update,[12] have significantly improved dysplasia detection rates.[13]
[14] Dye-based and virtual chromoendoscopy may also contribute to the declining incidence
of IBD-related CRC.[14]
[15] However, invisible dysplasia, which may include SEC, remains a concern due to inherent
limitations of random biopsies, complicating surveillance strategies in affected patients.
The potential association between SEC and dysplasia presents a challenge for safe
and effective dysplasia surveillance for IBD patients.
With the clinical significance of SEC remaining unclear, our aims were to expand the
understanding of the clinical, endoscopic, and histological features of IBD patients
with SEC and determine the overall rate and incidence of dysplasia.
Materials and Methods
We performed a single-center, retrospective study at the Gold Coast University Hospital,
a large tertiary referral center in Queensland, Australia.
Study Population
Patients with a confirmed diagnosis of IBD (Crohn's disease or ulcerative colitis)
with SEC were identified by searching the Gold Coast University Hospital pathology
database for the terms “serrated,” “epithelial,” and “change” in pathology reports.
The first colonic biopsy demonstrating SEC between January 1, 2017 and December 31,
2024 was designated as the index SEC. Colonic biopsies obtained from colonoscopy or
flexible sigmoidoscopy were included and classified as targeted or nontargeted based
on the presence of a visible lesion. Patients were included if they (1) were aged
18 years or older and (2) had a confirmed diagnosis of IBD. Patients were excluded
if (1) there was no follow-up colonoscopy or flexible sigmoidoscopy, (2) SEC was identified
in a surgical specimen, or (3) the diagnosis was IBD-unidentified. Two expert gastrointestinal
pathologists reviewed relevant pathology slides to confirm the diagnosis of SEC.
Patient and Endoscopic Data
Clinical, endoscopic, and histological data were extracted from electronic medical
records and endoscopy reports. Clinical data included year of IBD diagnosis, patient
age, IBD treatment at the time of index SEC diagnosis, disease duration, smoking status,
co-existence of primary sclerosing cholangitis, prior history of conventional colonic
dysplasia, and family history of CRC. Endoscopic data included the method of SEC detection,
use of high-definition white light endoscopy or chromoendoscopy, SEC morphology, SEC
location, presence of multifocal SEC, pseudopolyps, endoscopic inflammation, and histological
inflammation. Colonic segment concordance between SEC and dysplasia was also recorded.
Follow-up time was defined as the interval from the index SEC diagnosis to the most
recent repeat endoscopic assessment.
IBD Characteristics
The Montreal classification system was used to define disease extent and location
for UC (E1: proctitis, E2: left-sided colitis, E3: pancolitis) and CD (L1: ileal,
L2: colonic, L3: ileocolonic; B1: non-stricturing and nonpenetrating, B2: stricturing,
B3: penetrating). Endoscopy reports and photo-documentation were reviewed to identify
the presence of endoscopic inflammation. Previous biologic treatment exposure and
disease treatment at the time of index SEC diagnosis were captured. Treatment categories
included 5-aminosalicylates, immunomodulators (thiopurine, methotrexate), small molecules
(tofacitinib, upadacitinib), biologic therapy (infliximab, adalimumab, golimumab,
vedolizumab, ustekinumab), corticosteroids and combination therapy (immunomodulator
with a biologic).
Histology
In this study, SEC was defined as nondysplastic colonic mucosa with goblet cell–rich
epithelium and distorted serrated architecture involving the upper crypts, without
distortion of the basal crypt architecture (see [Fig. 1]).
Fig. 1 Serrated epithelial change. Nondysplastic colonic mucosa with goblet cell-rich epithelium
and distorted, serrated architecture involving the upper crypts, without distortion
of the basal crypt architecture.
Serrated Epithelial Change Characteristics
The location of SEC was defined by the colonic segment in which it was identified.
Endoscopy and pathology reports were used to determine whether SEC was detected on
random biopsies of flat mucosa or on targeted biopsies of nodular or polypoid mucosa.
Multifocal SEC was defined as SEC identified in multiple colonic segments during the
same endoscopic assessment. Consecutive SEC was defined as SEC identified in a different
colonic segment from the previous endoscopic assessment. The locations of SEC and
dysplasia were also reviewed.
Statistical Analysis
Descriptive statistics were used to summarize key data pertaining to patient demographics,
endoscopy, and histology. To account for the small sample size and minimize the influence
of outliers in the data distributions, median values with interquartile range (IQR)
were used to report the patient age (years) and endoscopic follow-up time (months).
Mean values with standard deviation (SD) were used to report disease duration (years),
consistent with methodologies used in comparable studies. The estimated prevalence
of SEC and the overall rate and incidence of dysplasia are also calculated.
Results
Population
The Gold Coast University Hospital pathology database was used to identify 113 patients
with endoscopic histopathology reports containing the words “serrated”, “epithelial”
and “change”. Of these, 78 had no IBD diagnosis and were excluded. The remaining 35
reports, each corresponding to a unique patient with confirmed IBD and SEC, were included
for further review. Histopathology slides were re-examined by two gastrointestinal
pathologists to confirm the diagnosis of SEC. Nine patients had no follow-up endoscopic
assessment, leaving a total of 26 patients for the final analysis (see [Fig. 2]). Over the study period, 1,621 colonoscopies were performed on individual patients
with IBD.
Fig. 2 Flowchart of patient inclusion. A total of 113 histopathology reports from colonic
biopsies containing the terms “serrated,” “epithelial,” and “change.” Of these, 35
corresponded to a patient with confirmed IBD and their pathology slides were reviewed
by two gastrointestinal pathologists to confirm SEC. Twenty-six patients had at least
one follow-up endoscopic assessment and were included in the final analysis. IBD,
inflammatory bowel disease; SEC, serrated epithelial change.
Patient Characteristics
Key patient demographics and relevant endoscopic and histological data are presented
in [Table 1]. The cohort included a total of 26 patients, 14 (53.8%) of whom were males. The
median age at the time of the index diagnosis of SEC was 58 years (IQR: 51–68). There
were 16 (61.5%) patients with either colonic or ileocolonic Crohn's disease, and 14
(87.5%) had an inflammatory phenotype of disease. There were 10 (38.5%) patients with
ulcerative colitis, and 6 (60%) had a pan-colonic phenotype. The mean disease duration
was 15 years (SD: 10), and the median follow-up time was 24.5 months (range: 5–60).
Table 1
Patient characteristics and key clinical, histological, and endoscopic data
|
KEY DATA
|
VALUE
|
|
Number of patients
|
26
|
|
Number of SEC specimens
|
30
|
|
Median age (IQR), y
|
58 (51–68)
|
|
Male gender
|
14 (53.8%)
|
|
Diagnosis
|
|
Crohn's disease
|
16 (61.5%)
|
|
L1
|
0 (0%)
|
|
L2
|
11 (68.8%)
|
|
L3
|
4 (31.2%)
|
|
L4
|
0 (0%)
|
|
B1
|
14 (87.5%)
|
|
B2
|
1 (6.25%)
|
|
B3
|
1 (6.25%)
|
|
Ulcerative colitis
|
10 (38.5%)
|
|
E1
|
1 (10%)
|
|
E2
|
3 (30%)
|
|
E3
|
6 (60%)
|
|
Mean duration of IBD, y (SD)
|
15 (10)
|
|
Median follow-up, mo (range)
|
24.5 (5–60)
|
|
Endoscopic method of SEC diagnosis
|
|
Colonoscopy
|
24 (92.3%)
|
|
Flexible sigmoidoscopy
|
2 (7.7%)
|
|
HDWL endoscopy
|
16 (61.5%)
|
|
Chromoendoscopy (dye-based or virtual)
|
10 (38.5%)
|
|
SEC morphology
|
|
Nodular/polypoid mucosa
|
18 (60%)
|
|
Size of lesion ≥10 mm
|
12 (66.7%)
|
|
Size of lesion <10 mm
|
6 (33.3%)
|
|
On flat mucosa
|
12 (40%)
|
|
SEC location
|
|
Cecum
|
9 (30%)
|
|
Ascending colon
|
3 (10%)
|
|
Transverse colon
|
5 (16.7%)
|
|
Descending colon
|
3 (10%)
|
|
Sigmoid colon
|
3 (10%)
|
|
Rectum
|
7 (23.3%)
|
|
Multifocal SEC
|
5 (19.2%)
|
|
SEC on consecutive endoscopy (different colonic segment)
|
7 (26.9%)
|
|
Active medications for IBD
|
|
Oral/rectal 5-ASA
|
9 (34.6%)
|
|
Immunomodulators
|
2 (7.7%)
|
|
Small molecules
|
2 (7.7%)
|
|
Biologics
|
6 (23.1%)
|
|
Combination therapy (biologic and immunomodulator)
|
4 (15.4%)
|
|
None
|
3 (11.5%)
|
|
Previous biologic exposure
|
|
Yes
|
12 (46.2%)
|
|
No
|
14 (53.8%)
|
|
Patient factors
|
|
PSC
|
3 (11.5%)
|
|
Smoker
|
3 (11.5%)
|
|
Family history of CRC
|
4 (15.4%)
|
|
Previous adenomas with dysplasia
|
6 (23.1%)
|
|
Endoscopic and histologic factors
|
|
Colonic strictures
|
0 (0%)
|
|
Tubular colon
|
3 (11.5%)
|
|
Post-inflammatory polyps
|
5 (19.2%)
|
|
Histologic inflammation
|
12 (46.2%)
|
|
Endoscopic inflammation
|
13 (50%)
|
|
Overall rate of dysplasia
|
1/26 (3.8%)
|
|
Incidence rate of dysplasia
|
1.65/100 person-years
|
Abbreviations: 5-ASA, 5-aminosalicylic acid; CRC, colorectal cancer; HDWL, high-definition
white light; IBD, inflammatory bowel disease; IQR, interquartile range; PSC, primary
sclerosing cholangitis; SD, standard deviation; SEC, serrated epithelial change.
Serrated Epithelial Change Characteristics
A total of 30 specimens with SEC were identified. Of these, 18 (60%) were visible
endoscopically as polypoid or nodular mucosa, with 12 (66.7%) measuring ≥10 mm. Colonoscopy
was the primary endoscopic method of diagnosis in most cases (92.3%). High-definition
white light endoscopy and chromoendoscopy were used in 16 (61.5%) and 10 (38.5%) cases,
respectively. SEC was primarily located in the caecum (30%), rectum (23.3%), and transverse
colon (16.7%), and exhibited multifocal distribution in 5 (19.2%) patients. Endoscopic
and histological inflammation were demonstrated in 50 and 46.2% of cases, respectively.
Dysplasia
The overall dysplasia rate in this cohort was 3.8%, represented by one case detected
among 26 patients. The incidence rate of dysplasia was 1.65 cases per 100 person-years,
calculated by dividing this single dysplasia event by the total person-years of follow-up.
Discussion
This study describes a cohort of IBD patients with SEC consisting of an almost equal
male-to-female distribution, with a median age of 58 years (IQR: 51–68) and a mean
disease duration of 15 years (SD: 10). Gender distribution, age at index diagnosis
of SEC, and disease duration were all comparable to those reported in previous studies.[5]
[6]
[7] Approximately half of the patients in this study were being treated with advanced
therapies, including biologics and small molecules, which were not widely available
at the time of earlier studies.[5]
[6]
[7] Additionally, nearly half of the patients had evidence of endoscopic and histological
inflammation at the time of SEC diagnosis. These data reflect a population with long-standing
active disease and treatment complexity, both of which may contribute to the development
of colorectal dysplasia.[5]
[6]
[7]
Our findings indicate that the highest occurrence of SEC was observed proximal to
the splenic flexure (56.7%) and in the rectum (23.3%). This pattern contrasts with
previous studies by Parian and colleagues, where SEC predominantly occurred in the
left colon and rectum in approximately 75% of cases. This difference may be attributed
to the small sample size of our analysis. However, multifocal SEC was observed in
19.2% of patients, which is a proportion closely aligning with the 22.4% previously
reported. We observed that about one-quarter of patients had SEC detected in a different
colonic segment on successive endoscopic assessments, a phenotype associated with
an increased risk of dysplasia.[6]
[7]
Notably, we observed 40% of SEC was not endoscopically visible and detected only through
random biopsies of flat mucosa, despite chromoendoscopy being employed in 38.5% of
cases. Prior studies indicate that rates of invisible SEC are as high as 66.3 to 73.4%,[6]
[7] although the use of chromoendoscopy was not specified. Johnson and colleagues reported
a lower rate of 62%, where chromoendoscopy was used in 20% of cases.[5] These differences may be explained by the application of high-definition endoscopy
and higher rates of chromoendoscopy within our cohort. These findings raise concerns
regarding the effectiveness of white-light colonoscopy in detecting SEC and underscore
the limitations of random biopsies. It is plausible that the actual prevalence of
SEC is underestimated.[5]
[6]
[7]
In our study, endoscopic and histological inflammations were frequently associated
with SEC, occurring in 50 and 46.2% of cases, respectively. Previous studies reported
similar figures, with moderate to severe endoscopic inflammation observed in 56.1%
and histological inflammation in 49.2% of cases.[6] A subsequent study reported rates of 41.8 and 57.1%, respectively.[7] These data suggest that inflammation and ongoing disease activity may drive specific
molecular alterations resulting in SEC. Moreover, given the known correlation between
disease activity and CRC, the development of SEC may potentially represent a morphological
indicator or precursor of dysplasia.
In our cohort, we observed a dysplasia rate of 3.8%, represented by a single case
of nonconventional low-grade dysplasia associated with SEC. This rate is considerably
lower than the reported 11.2 to 21%,[5]
[6]
[7]
[16] likely due to variable study designs and heterogenous populations. When adjusting
for the differences in follow-up time, the incidence rate of dysplasia was 1.65 cases
per 100 person-years, which closely aligns with the findings reported by Parian and
colleagues.[6] Interpretation of this data, however, is challenging given the small sample size
and presence of several potential confounders in the affected patient, including previous
conventional dysplasia, a family history of CRC, and a tubular colon. Notably, all
procedures in our analysis were performed using high-definition endoscopes, with chromoendoscopy
employed at a relatively high frequency of 38.5%. Consistent with the SCENIC consensus
recommendations, these methods are known to improve dysplasia detection in this population,[11] but their utility in detecting SEC specifically remains unclear and warrants further
investigation.
Several studies have reviewed SEC and colorectal dysplasia, but current data are inconsistent.
SEC was first described by Kilgore and colleagues in 2000, who identified “hyperplastic-like
mucosal change” in surgical specimens of adenocarcinoma from patients with Crohn's
disease. They suggested a potential link between this histological finding and colorectal
adenocarcinoma,[17] but the study was likely underpowered. Parian and colleagues published a large case
series of 187 IBD patients, reporting that 21% of patients with SEC had synchronous
or metachronous colorectal dysplasia after adjusting for prior dysplasia. Longer disease
duration increased the likelihood of dysplasia,[6] a factor also validated in our cohort. The study was possibly subject to detection
and selection bias, as patients with SEC underwent more endoscopic assessments and
had longer follow-up times. In contrast, other studies have demonstrated no significant
associations between SEC and colorectal dysplasia after adjusting for confounding
factors.[5]
[16] The variability in findings stresses the need for high-quality research, as compelling
evidence of a causal relationship between SEC and dysplasia in the IBD cohort remains
elusive.
Overall, SEC emerges as a histopathological diagnosis of increasing interest within
the IBD population, though its clinical significance remains to be fully elucidated.
The current literature on its association with colorectal dysplasia lacks consensus
and is primarily constrained to low-quality analyses. Considering the growing attention
on SEC, we propose that this unique descriptive analysis, set within the era of advanced
endoscopic imaging and techniques, can stimulate further much-needed research in this
area.
Strengths
This study represents the largest Australian cohort of IBD patients with SEC. The
integrity of histological data is supported by the expertise of gastrointestinal histopathologists
at the Gold Coast University Hospital. Additionally, this study included only patients
with follow-up endoscopic assessment using high-definition endoscopes and chromoendoscopy,
both of which have been demonstrated to enhance dysplasia detection.
Limitations
The retrospective design inherently limits control over data collection and may introduce
bias through incomplete or inaccurate documentation in electronic records. The small
sample size, which resulted from a stringent patient selection process, may limit
data acquisition and the generalizability of our findings. The absence of a control
group precludes comparison of dysplasia rates between patients with and without SEC,
which is necessary to more definitively ascertain the risk SEC confers for colorectal
dysplasia. While this was not a primary aim, it limits the ability to fully contextualize
the clinical significance of SEC. Lastly, the indications for endoscopic assessment
were not provided, which may introduce a recall and detection bias, as the frequency
of procedures varies significantly depending on disease severity, as does the need
to perform segmental colonic biopsies. This also appeared to be a common challenge
in related studies. These factors emphasize the need for larger, prospective studies
to overcome these limitations and to validate our findings.
Conclusion
In this uncontrolled study, SEC was predominantly found in patients with long-standing
colitis as polypoid or nodular lesions situated proximal to the splenic flexure and
associated with both endoscopic and histological inflammation. We observed a solitary
case of nonconventional dysplasia, occurring in a patient with several other risk
factors for dysplasia, which confounds interpretation. Given the small sample size,
robust conclusions about the role of SEC in dysplasia development cannot be drawn.
Overall, SEC remains an understudied entity, necessitating further research into any
potential association with dysplasia, particularly in the context of advanced endoscopic
imaging techniques such as chromoendoscopy.
