Introduction
With advances in technicality skills and preference of sedation by patients, terminal
ileal (TI) intubation during colonoscopy has improved from 60%–75% to 95%–98% in recent
times.[1],[2],[3],[4] In clinical practice, TI abnormalities such as nodularity, ulcers, and erythema
are often noted during colonoscopy evaluation for a variety of indications. The decision
to take a biopsy is usually based on the endoscopist's discretion. We also encounter
at times situations where the clinical symptoms and TI findings are conflicting or
inconclusive. Biopsy for histopathological diagnosis, most often, adds to the conundrum.
Thus, the present study was planned as an audit of TI biopsies at our center in the
past 5 years. The primary aim of the study was to retrospectively determine the clinical
presentation and ileal mucosal changes during colonoscopy for which TI biopsies were
taken. The secondary endpoint of the study was to determine the specific histopathology
which had the best yield for specific colonoscopy findings.
Materials and Methods
The study was done at the Institute of Gastrointestinal Sciences, Gleneagles Global
Health City, Chennai. All patients who underwent ileocolonoscopy between 2012 and
2016 were included. Patient data included age, gender, and indication for the procedure.
The ileal mucosal changes were documented from patient records and static images.
Indications for colonoscopy were classified into the following distinct prototype
groups. These were the working diagnosis made by the consultants at the consultation
clinic following standard guidelines for diagnosis and documented in the patient records.
-
Irritable bowel syndrome (IBS) – As per Rome III criteria
-
Suspected Colorectal cancers (CRC) – This included patients presenting as unexplained
anemia, bleeding per rectum, cancer surveillance in patients with family history of
gastrointestinal cancer
-
Inflammatory bowel disease (IBD) – Naïve patients presenting as chronic diarrhea,
bloody or blood admixed with mucus diarrhea, CECT imaging showing long segment TI
thickening
-
Chronic colonic infection, for example, tuberculosis: In suspected ileocecal tuberculosis
with systemic symptoms, symptoms of partial small bowel obstruction, ileocecal mass
-
Others such as diverticulosis/itis, nonspecific lower quadrant abdominal pain, postradiation
enteritis, postorgan transplant patients with pain abdomen/diarrhea.
Ileal mucosal changes were classified as [Figure 1].
Figure 1 Flowchart for the study
-
Ileal ulcers – Aphthous or large ulcers
-
Ileal mucosal nodularity with or without surface ulceration
-
Nonspecific – Effaced mucosa, loss of villi, patchy erythema.
Histopathology changes were classified based on the standard protocol of reporting
[5] as:
Acute ileitis
The presence of mucosal aphthous ulcers with scattered eosinophils, neutrophil infiltration,
and alternate bands of fibrosis in the lamina propria extending up to the mucosal
surface.
Chronic ileitis
Characterized by distortion of the mucosal crypt architecture, the predominance of
lymphomononuclear cells, plasma cells, eosinophils, pyloric metaplasia, goblet cell
rich crypts (hypercrinia), mucosal basal plasmacytosis, and broadening of the ileal
villous tips. Among these, however, topographical changes such as crypt branching,
crypt shortening, and the crypt loss were considered as hallmarks of chronicity.
Chronic ileitis with activity and granulomas
Crohn's disease is typified by mucosal pericryptal microgranuloma, rail track/deep
burrowing ulcers, basal plasmacytosis, dense lymphocytic infiltrations, and epithelioid
granulomas in mucosa and submucosa, necrosis as definitive of tuberculosis.
Others
Lymphoid aggregation, patchy inflammatory infiltrate with no crypt distortion.
Rarely specific features
Eosinophilic ileitis, ischemic ileitis, radiation enteritis, graft versus host disease,
and tumors such as lymphoma.
The methodology of the study is seen in the consort chart [Figure 1].
Ileocolonoscopy was performed under sedation by senior consultants until 2014. Later,
senior registrars were involved who worked under supervision and guidance of senior
consultants. Being a corporate hospital, biopsies were taken only after consensus
of the treating consultant. All reports were seen and approved by the incharge consultant,
as per the hospital policy.
Since this is a retrospective audit, no fixed number of biopsies was set for inclusion.
The protocol followed is four biopsies from the suspicious areas in ileum. All samples
were fixed in 10% formalin. Standard processing techniques were followed and sections
4–6 microns were stained routinely with hematoxylin and eosin and special stains were
used when indicated. The inadequate sample reported as no mucosal glands and predominantly
fibrocollagenous tissue were excluded from analysis.
Statistical analysis
The results were interpreted using rows and columns contingency tables for determining
statistical significance using Chi-square test. P < 0.05 was considered statistically
significant. Sensitivity, specificity, positive, and negative predictive values were
calculated.
The retrospective study was approved by the Ethics Committee of the Institution.
Results
The indications for ileocolonoscopy and the corresponding ileal findings are shown
in [Table 1]. One hundred and nine patients had isolated ileal lesions. The median age was 44.1
years (range 8–80 years). Men outnumbered women in a ratio of 82:27.
Table 1
: Clinical indications and ileal mucosal changes
|
Indication
|
Ulcers (large/small) (%)
|
Stricture (%)
|
Mucosal nodularity (%)
|
Nonspecific (erythema, denuded/effaced mucosa, loss of villi) (%)
|
P
|
|
CRC=Colorectal cancer, IBD=Inflammatory bowel disease, IBS=Irritable bowel syndrome
|
|
CRC (12)
|
6 (50)
|
0
|
4 (33.3)
|
2 (16.7)
|
0.06
|
|
IBD (24)
|
14 (58.3)
|
3 (12.5)
|
3 (12.5)
|
4 (16.7)
|
|
|
IBS (70)
|
32 (45.7)
|
1 (1.5)
|
31 (44.3)
|
6 (8.5)
|
|
|
Others (3)
|
2 (66.6)
|
-
|
1 (33.4)
|
-
|
|
|
Total (109)
|
54 (49.5)
|
4 (3.7)
|
39 (35.8)
|
12 (11)
|
|
The major clinical indications for ileocolonoscopy were IBS (64.2%), followed by IBD
(22%).
Ileal mucosal changes in specific clinical situations
As shown in [Table 1], ulcers (aphthoid) were the most frequent finding in suspected IBD cases, during
CRC screening and those with an IBS. Mucosal nodularity was frequent in IBS and commonly
noted in cases with suspected CRC. Nonspecific findings ranged from 8.5% to 16.7%
[Figure 2].
Figure 2 Representative ileal mucosal changes (clockwise) - (a) Normal ileum, (b) ileal erythema,
(c) mass lesion and ulcers in ileum, (d) aphthous ulcers, (e) Large ulcer with slough,
(f) Nodularity and ulcerations at IC region
Histological correlation with ileal mucosal changes
Ulcers in ileum [Table 2] was most often reported as chronic ileitis (46.2%), followed by nonspecific changes
(35.2%) that included patchy inflammation and/or lymphoid aggregates. One specimen
each had specific findings of cytomegalovirus infection and eosinophilic enteritis.
Biopsies from strictures revealed nonspecific findings in 50% of cases and acute or
chronic ileitis in one case each. Biopsy from nodular ileal lesions, were predominantly
nonspecific (74.4%), followed by acute (15.4%) and chronic ileitis (10.2%). About
50% of specimens with nonspecific ileal changes had nonspecific histological changes
as well. Thus, significant histological findings of chronic ileitis with activity/granuloma
and acute ileitis were more common in cases with ileal ulcers (P = 0.002) rather than
other findings. Ileal ulcers had the highest sensitivity, PPV, and NPV for significant
histological findings of acute or chronic ileitis [Table 3]. Ileal nodularity and nonspecific mucosal changes had the least sensitivity, PPV,
and NPV. Nonspecific mucosal changes such as patchy inflammation, effaced mucosa,
and loss of villi (on NBI) had very low sensitivity but high specificity (89.3%),
suggesting that these findings were unlikely to predict any significant histopathological
changes.
Table 2
: Endoscopic and histological correlation
|
Ileal mucosal changes
|
Histology
|
P
|
|
Rare causes (CMV, eosinophilic) (%)
|
Others (patchy inflammation, lymphoid aggregates) (%)
|
Acute ileitis (%)
|
Chronic ileitis
|
|
With activity (%)
|
With granulomas (%)
|
|
CMV=Cytomegalovirus
|
|
Ulcers (54)
|
2 (3.7)
|
19 (35.2)
|
8 (14.8)
|
24 (44.4)
|
1 (1.8) 0.002
|
|
Stricture (4)
|
-
|
2 (50)
|
1 (25)
|
1 (25)
|
-
|
|
Nodularity (39)
|
-
|
29 (74.4)
|
6 (15.4)
|
4 (10.2)
|
-
|
|
Nonspecific changes (12)
|
-
|
6 (50)
|
3 (25)
|
3 (25)
|
|
|
Total (109)
|
2 (1.8)
|
56 (51.4)
|
18 (16.5)
|
32 (29.4)
|
1 (0.9)
|
Table 3
Predicting outcome of isolated ileal findings at histology
|
Percentage (95% CI)
|
|
|
hSensitivityh
|
Specificityh
|
PPVh
|
NPVh
|
|
CI=Confidence interval, PPV=Positive predictive value, NPV=Negative predictive value
|
|
Ulcers
|
64.7
|
66.1
|
63.5
|
67.3
|
|
|
(50-77.6)
|
(52.2-78.2)
|
(53.4-72.5)
|
(57.6-75.7)
|
|
Nodularity
|
19.6
|
48.2
|
25.6
|
39.7
|
|
|
(9.8-33.1)
|
(34.7-62)
|
(15.8-38.8)
|
(32.7-47.1)
|
|
Nonspecific changes
|
11.8
|
89.3
|
50
|
52.6
|
|
|
(4.5-23.8)
|
(78.1-95.9)
|
(25.6-74.4)
|
(49.3-56)
|
Discussion
TI biopsies for varied mucosal changes are frequently recommended for histopathological
correlation. At present, there are no specific guidelines to suggest the role of such
biopsies.
There have been a few studies on TI biopsies and its impact on clinical decision-making.
In 1985, Borsch and Schmidt prospectively evaluated 400 consecutive patients successfully
undergoing TI endoscopy with biopsy.[6] Although pathological abnormalities were identified in only 5% of these biopsies,
diagnostic information was obtained in 30% after excluding suspicious abnormalities.
Zwas et al.[3] evaluated 144 patients undergoing colonoscopy. Biopsy of the TI was successful in
130 patients. About 12.5% of symptomatic patients and 2.7% of asymptomatic patients
had histological abnormalities. Geboes et al.[7] observed ileal mucosal abnormality in 48% and histological changes in 49% of patients
undergoing colonoscopy for enterocolitis. The authors concluded that ileoscopy was
beneficial in carefully selected patients with IBD-related symptoms.
A recent study from a center in north India [8] on 1632 colonoscopy reported ulcers in the ileocaecal region (ileum: 40%; cecum:
33%; and ileocecum in remaining) in 104 patients. The predominant presentation in
this series was lower gastrointestinal bleed.
Common indications for ileocolonoscopy in our series were suspected IBS, IBD, and
CRC in that order. As is recommended in the guidelines, colonoscopy was considered
complete only when terminal ileum was seen. Most frequent indications for ileal mucosal
biopsy were ileal ulcers in 54 patients (49.5%), nodularity in 39 patients (35.8%),
and nonspecific findings in 12 (11%). Three patients had postsolid organ transplant
status and underwent ileocolonoscopy for pain abdomen and/or diarrhea. Of the 109
biopsies sent, nonspecific findings of patchy inflammation with/without lymphoid aggregates
was the most common finding (51.4%), followed by chronic ileitis (30.3%) and acute
ileitis (16.5%). The yield of detecting significant histological abnormality (acute
and chronic ileitis, specific etiology) was highest for ileal ulcers (35/54, 64.8%)
followed by isolated ileal stricture and nonspecific mucosal changes (50% each). Isolated
ileal nodularity was associated with a significant histological abnormality in only
a quarter of cases. The positive likelihood ratio of detecting histological abnormality
for various ileal mucosal changes indicated that TI biopsy had the best yield for
ileal ulcers. Based on our observational retrospective study, we reiterate that nonspecific
ileal mucosal changes and nodularity do not require a tissue biopsy. This observation
is similar to McHugh et al. who reported that diagnostic yield of TI biopsy varied with indication and endoscopic
findings. Biopsy was of greatest value in patients undergoing endoscopy for known
or strongly suspected Crohn's disease, or with an abnormal imaging study of the TI.
Biopsy of endoscopically normal mucosa was unlikely to yield a diagnostic useful information.[9]
Conclusion
Ileal ulcers are the significant colonoscopy findings where tissue biopsy is likely
to yield a definitive diagnosis and justify specific management. Biopsies from nonspecific
ileal changes and nodularity should be discouraged as it is unlikely to pick up any
major abnormality. Thus, “take a bite only when you see an ileal ulcer!”
Financial support and sponsorship
Nil.
