Introduction
Extrapulmonary tuberculosis (TB) accounts for approximately 20% of patients having
TB and 7%–10% of them have abdominal TB.[1],[2],[3] While abdominal TB continues to be a common problem in developing countries including
India, increasing population migration and epidemic of acquired immunodeficiency syndrome
have led to resurgence of this disease in regions where TB had previously been largely
controlled.[1],[2],[3] Abdominal TB remains an important disease in countries where TB is still endemic.[1],[2],[3],[4],[5],[6],[7] While TB is endemic in India, there has been a rise in number of patients with Crohn's
disease (CD) from every part of India.[8],[9] In a study by the Indian Society of Gastroenterology Task Force on inflammatory
bowel disease (IBD) including 1159 patients with IBD from secondary and tertiary care
centers in India, 35.3% of them had CD.[10]
The clinical, morphological, and histological features of gastrointestinal (GI) TB
and CD are so similar that it becomes difficult to differentiate between these two
entities.[11],[12],[13],[14] In geographical regions such as India where both GI TB and CD are prevalent, differential
diagnosis between the two is challenging.[11],[12],[13],[14] The natural history of CD is quite different from that of GI TB. While GI TB gets
cured with appropriate anti-TB treatment, CD has a remitting/relapsing or persistent
course and stays life-long usually. Although treatment with an immunosuppressant with
a mistaken diagnosis of CD may worsen TB, an empiric treatment with anti-TB therapy
on the other hand may delay the diagnosis of CD in such patients. Furthermore, use
of biologics, which have become one of the important treatment modalities for CD,
can lead to flare of TB if the mistaken diagnosis of CD is made in a patient having
GI TB.[15],[16],[17] Because of similarity in the clinical presentation, endoscopic appearance, and histological
characteristics between these two diseases, many of them are treated empirically with
anti-TB drugs at times.[9],[18]
Diagnostic Criteria
Diagnostic criteria for gastrointestinal tuberculosis
A definitive diagnosis of GI TB can be made based on the presence of any of the following
three criteria:
-
Culture of tissue (colonic biopsy, lymph nodes) resulting in growth of Mycobacterium tuberculosis
-
Histological demonstration of typical acid-fast bacilli (AFB)
-
Histological evidence of necrotizing granuloma.
While a positive culture, demonstration of AFB on smear, and histological evidence
of a caseating granuloma are gold standard tests for the definitive diagnosis of TB,
the sensitivity of the combination of the three or any one test is approximately 30%–50%.[19] In rest of the patients, the diagnosis is considered presumptive based on clinical,
endoscopic, radiological, and histological evidence. The unequivocal response to anti-TB
therapy for 6 months confirms the diagnosis of TB in these patients. It should be
emphasized that adherence to treatment is extremely important in those with presumptive
diagnosis of TB who are receiving anti-TB therapy empirically.
Diagnostic criteria for Crohn's disease
There is no single gold standard test, and the diagnosis of CD is based on a combination
of clinical features, endoscopic characteristics, and histological characteristics.
Both the Asian Pacific Associations of Gastroenterology and the Indian Society of
Gastroenterology guidelines on CD suggest exclusion of infective causes, especially
TB before making a diagnosis of CD.[20],[21]
Sites of Involvement in Gastrointestinal Tuberculosis and Crohn's Disease
TB may involve any part of GI tract (GIT), from the esophagus to the anal canal. There
is a striking tendency for intestinal TB to affect the ileum and ileocecal (IC) area
preferentially.[22] Other common sites of involvement in TB are colon, appendix, and jejunum [Table 1]. In the colon, the frequency of involvement decreases segmentally from the ascending
colon to the rectum. The involvement of rectum and anal canal is uncommon in patients
with GI TB. Patients with TB may have involvement of lymph nodes, peritoneum, and
solid viscera either alone or in combination with GIT.
Table 1
Sites of involvement in patients with intestinal tuberculosis and Crohn's disease
|
Site
|
Amarapurkar et al.[8]
|
Makharia et al.[23]
|
Li et al.[24])
|
|
CD (n=26) (%)
|
ITB (n=26) (%)
|
CD (n=53) (%)
|
ITB (n=53) (%)
|
CD (n=130) (%)
|
ITB (n=122) (%))
|
|
CD=Crohn's disease, ITB=Intestinal tuberculosis
|
|
Stomach
|
0 2 (7.7)
|
3 (5.6)
|
1 (1.8))
|
|
|
|
Duodenum
|
1 (3.8)
|
2 (7.7)
|
3 (5.6)
|
2 (3.7))
|
|
|
|
Jejunum
|
2 (7.7)
|
1 (3.8)
|
9 (16.9)
|
0
|
|
|
|
Ileum
|
16 (61.5)
|
20 (76.9)
|
13 (24.5)
|
8 (15)
|
86 (66.2)
|
73 (59.8))
|
|
Ileocecal region
|
20 (76.9)
|
21 (80.7)
|
47 (88.6)
|
49 (92.4)
|
72 (55.4)
|
92 (75.4))
|
|
Ascending colon
|
18 (69.2)
|
21 (80.7)
|
34 (64.1)
|
23 (43.4)
|
60 (46.2)
|
76 (62.3))
|
|
Descending colon
|
10 (38.4)
|
5 (19.2)
|
32 (60.3)
|
7 (13.2)
|
44 (33.8)
|
30 (24.6))
|
|
Sigmoid colon
|
|
|
35 (66.3)
|
6 (11.3)
|
58 (44.6)
|
28 (23))
|
|
Rectum
|
8 (30.7)
|
5 (19.2)
|
33 (62.2)
|
10 (18.8)
|
50 (38.5)
|
20 (16.4))
|
|
Anal canal
|
2 (7.7)
|
1 (3.8)
|
8 (15)
|
2 (3.7)
|
|
|
Types of Morphological Lesions
The changes caused by acute inflammation of the colonic mucosa, namely, edematous
mucosal folds, mucosal ulcerations and nodularity, luminal narrowing, strictures,
and pseudopolyps, can occur in both intestinal TB and CD. The morphological features
of intestinal TB (ITB) and CD depend on the stage of the disease. The lesions may
vary from mild lesions such as loss of vascular pattern, erythema, and superficial
ulcerations to more advanced lesions including deep ulcerations, nodularity, and strictures.
One should realize that the evolution of disease takes from months to years both for
GI TB and CD. The classical morphological features, as described in most textbooks,
are seen in advanced stages of the disease. Some of the patients may present in early
stages of evolution of the disease, thereby one may not find classical lesions and
instead finds minute lesions. The wisdom lies in identifying these minute lesions
so that the progression of the disease can be prevented with timely institution of
appropriate treatment. [Table 2] shows the types of lesions and t
Table 2
Terminology of endoscopic lesions in Intestinal tuberculosis and Crohn's disease
|
Mucosal damage
|
Description
|
Grading
|
|
Loss of vascular pattern
|
Loss of normal mucosal appearance without well.demarcated, arborizing capillaries
|
From patchy or blurred to complete loss
|
|
Erythema
|
Unnaturally reddened mucosa
|
From discrete or punctiform to diffuse erythema
|
|
Granularity
|
Mucosal pattern produced by a reticular network of radiolucent foci of 0.5.1 mm of
diameter with a sharp light reflex
|
From fine to coarse or nodular due to abnormal light reflection
|
|
Friability/bleeding
|
Bleeding or intramucosal hemorrhage before or after the passage of the endoscope
|
From contact bleeding (bleeding with light touch) to spontaneous bleeding
|
|
Erosion
|
A definite discontinuation of mucosa <3 mm in size. Also
described as pinpoint ulceration
|
Isolated, diffuse
|
|
Aphthous ulcer
|
White depressed center surrounded by a halo of erythema
|
Isolated, multiple
|
|
Ulcer
|
Any lesion of the mucosa of unequivocal depth with or without reddish halo
|
Isolated or multiple based on morphology: Circular, linear, stellar, serpiginous,
irregular shape, superficial or deep
|
|
Ulcer size
|
Defined in mm or classified as: ≤5 mm; 5.20 mm; >20 mm
|
|
|
Stenosis
|
Narrowing of the lumen
|
Single, multiple, passable (by standard adult endoscope), un-passable, passable after
dilation, ulcerated, nonulcerated
|
|
Postinflammatory polyps
|
Polypoid lesion, usually small, glistering, isolated or multiple, scattered throughout
the colon Sometimes cylindrical or giant (>2 cm) in size
|
Isolated, diffuse, occluding (“gaint”)
|
|
Cobblestone
|
Mucosal pattern with raised nodules, resembling the paving of a ÂgromanÂh roa
|
With or without ulceration
|
Pathological and Clinical Correlation
The diameter of a normal intestine decreases from the duodenum to jejunum and then
to the ileum. The diameter of the jejunum is 3 cm and that of the ileum is 2.5 cm.[25],[26] The content, which flows through the intestine, is mostly chyme (liquid). The diameter
of the intestinal lumen, at which the obstruction to the flow of chyme occurs, is
not well established. In a rat model, it was observed that the pressure in the intestine
started rising and the flow of the contents started decreasing when the luminal diameter
decreased to 60% of the original diameter.[27] Therefore, in the early stage of diseases, when there is no cicatrization of the
lumen, the flow of chyme is not affected and thus patients may not have any intestinal
symptoms. With progression of the disease, the effective diameter of the intestinal
lumen decreases, resulting in obstruction to the flow of the intestinal contents.
Therefore, the severity of symptoms increases progressively as the intestinal luminal
diameter decreases. The initial symptoms of intestinal colic are usually episodic
and are precipitated by the episodes of GI infection resulting in transient mild intestinal
wall edema which further compromise the diameter of the already compromised intestinal
lumen because of the disease process. While inflammation caused by acute GI infection
in the normal intestine is not critical and does not lead to critical decrease in
the effective diameter of the intestine, even a mild decrease in the diameter of the
intestine can lead to a critically low diameter in already compromised lumen due to
the primary disease process. Such episodic symptoms may also be precipitated by a
high fiber diet or the ingestion of seeds. As disease progresses, the interval between
the episodes of symptoms also decreases. Ultimately, the symptoms of intestinal colic
become more frequent and patients develop partial intestinal obstruction. The ulcerations
and nodularity caused by the tissue destruction can lead to oozing of blood from the
ulcers, resulting in anemia.
After effective treatment of the disease, there is resolution of active inflammation
and healing of ulcers, which results in the improvement in symptoms. As it is well
known that chronic inflammatory process heals with fibrosis, the healing of ulcers
and inflammation of the intestine leads to scarring of the intestinal lumen and thus
compromise the normal intestinal diameter.[28] There is a fundamental difference in the diseases affecting solid/spongy organs
and luminal organs. In spite of residual changes in a part of the solid/spongy organ,
there still may be sufficient organ volume to maintain the functions of the organ.
On the contrary, healing of the intestinal lesions may result in the fibrotic stricture
and patients may continue to be symptomatic. Most of the times, however, despite the
residual intestinal stricture/narrowing, the luminal diameter may remain sufficient
to allow uninterrupted the passage of intestinal contents. On occasions, healing of
lesions can result in a critical stricture resulting in frequent symptoms and such
strictures may require resection or stricturoplasty.
Endoscopy in Intestinal Tuberculosis and Crohn's Disease
Endoscopic examination of the GIT is essential not only for the diagnosis but also
for evaluation the extent of the disease, assessment of the activity of the disease,
and monitoring the response of the therapy in most patients with GI TB and CD.
Ileocolonoscopy is the first recommended test for the diagnosis of both ITB and CD.[20],[21],[29]
Endoscopic features in patients with gastrointestinal tuberculosis
IC region is most often involved area in patients having GI TB, possibly because of
the increased physiological stasis, increased rate of fluid and electrolyte absorption,
minimal digestive activity, and an abundance of lymphoid tissue at this site.[30],[31],[32] Retrograde ileoscopy should be done in all patients suspected to have ITB and biopsies
should be obtained even if the mucosa in the terminal ileum appears normal. The biopsies
of normal appearing terminal ileum may reveal granulomas in additional 4% of patients
with clinical suspicion of ITB.[33],[34] Terminal ileal involvement alone with relative cecal sparing is uncommon in ITB.[35] In 10%–20% of patients, segmental colonic involvement may occur in the absence of
IC involvement.[36],[37] Approximately 5% of patients may present with pancolitis, which may mimic ulcerative
colitis.[38]
The endoscopic findings in ITB may vary from mild lesion(s), such as loss of vascular
pattern, erythema, small ulcerations, and superficial ulcerations, to more advanced
lesions including deep ulcerations, nodularity, and strictures [Figure 1]. The endoscopic lesions in ITB are classified as ulcerative (60%), ulcerohypertrophic
(30%) characterized by inflammatory mass around IC valve with ulcerated thick wall,
and hypertrophic (10%) characterized by fibrosis and pseudotumor lesions.[39] The ulcers in ITB are typically transverse, often circumferential with ill-defined,
sloping or overhanging edges, and are usually surrounded by inflamed mucosa.[19],[40],[41],[42] The strictures in patients with TB are generally shorter in length. IC valve is
often patulous in them [Figure 1]. A stricture in the intestine has also been described due to extraluminal cause
such as lymph nodes causing compression even without apparent mucosal lesion.[6]
Figure 1: Endoscopic lesions in a patient with intestinal tuberculosis. (a) Ulcers over the
ileocecal valve in a patient with intestinal tuberculosis. (b) Multiple nodules and
deep ulcers (with neoplasm-like appearance) in cecum with patulous ileocecal valve.
(c) Fixed patulous ileocecal valve with nodules in cecum and ascending colon in a
patient with intestinal tuberculosis. (d) Circumferential ulcerative lesion in a patient
with tuberculosis
In tuberculous colitis, the colonic mucosa surrounding an ulcer exhibits features
of inflammation, such as erythema, nodularity, or edema. The rectum is rarely involved.[43] Traction diverticula and sinus tracts due to adjacent lymphadenitis may be seen
in tubercular colitis.[39]
In gastroduodenal TB, comprising 1% of abdominal TB, the most common endoscopic findings
observed are ulceration and nodularity in the pyloroduodenal area along with strictures,
and the disease could be multifocal, involving pyloroduodenal area and the second
part of the duodenum (15%) or even present as antral hypertrophic nodular mass.[40]
Endoscopic features in Crohn's disease
CD can affect any part of the GIT, from the oropharynx to the anorectum. Once CD has
settled in, the location of the disease remains stable generally, although exceptions
do occur, especially following surgical resections [41] [Table 1]. Similar to TB, CD also affects IC area commonly.[42] In a multicenter study conducted by the Indian Society of Gastroenterology, of 397
patients with CD, 39.2% had involvement of both small and large intestine (ileocolonic),
31.2% had isolated colonic, and 22.9% had involvement of only small intestine. The
involvement of upper GI (UGI) was reported in 5.8% of patients. While 18.8% had stricturing
disease, 4.4% had fistulizing disease. Nearly 6.9% patients had perianal disease as
a disease modifier.[10]
Isolated perianal disease is rare (2%–3%) although perianal abnormalities, in conjunction
with other sites of intestinal involvement, are common.[43],[44],[45] UGI involvement, when occurs in patients with CD, it almost always occurs in association
with small intestinal and/or colonic involvement.
The earliest and most characteristic endoscopic finding in CD is the aphthous ulcer.[46],[47] It can be found throughout the GIT. An aphthous ulcer represents a small (≤5 mm)
superficial ulcer surrounded by a characteristic tiny rim of erythema [Figure 2]. They have clear margins and are often surrounded by normal colonic mucosa with
very little reactive change. Aphthous ulcers can be localized or multifocal.[48] They are often seen in groups and lie in longitudinal fashion in the GIT, tend to
enlarge concentrically, become nodular, and give rise to larger and deeper ulcerations.[46] The best model to study the earliest lesions in CD is the postoperative recurrence
condition, where these early lesions recur within weeks to months after surgery.[46]
Figure 2: Endoscopic features in a patient with Crohn's disease. (a) Erythema and loss of
vascular pattern in a patient with Crohn's disease. (b) A few aphthous ulcers in the
colon in a patient with Crohn's disease. (c) Discrete ulcers in the colon in a patient
with Crohn's disease. (d) Multiple deep longitudinal ulcers in colon in a patient
with Crohn's disease. (e) Longitudinal ulcers with cobblestone appearance in the colon.
(f) Cobblestone appearance in a patient with Crohn's disease
The ulcers in CD can be of various sizes and shapes. The mucosa lying between long
linear ulcerations can be normal or very edematous, reddish, and hyperplastic, giving
an appearance of cobblestone [Figure 2].[49] The involvement of the colon and small intestine in patients with CD is characteristically
patchy. The lesions may involve only one side of the colonic mucosa in the same segment
[Figure 3]. Rectum is involved in approximately 50% [Figure 2].
Figure 3: Skip lesions in a patient with Crohn's disease. (a) Discrete ulcerated lesion in
the rectosigmoid region. (b) Normal looking mucosa in the descending colon. (c) Multiple
ulcers with loss of vascular pattern in transverse colon. (d) Discrete ulcers in the
ascending colon. (e and f) Discrete ulcers in the cecum
Where the inflammation is deep and extensive, luminal narrowing or strictures can
occur.[48],[49] Strictures virtually always arise in areas having severe ulceration. Both length
and width of the strictures can vary considerably, 3–10 cm in length.[42] The strictures in patients with CD may be longer than that seen in patients with
TB. Strictures with surrounding nodularity and eccentric lumen may also be due to
malignant transformation.[50] The inflammation and ulcerations in CD are often deep, transmural, and can lead
to perforation, inflammatory mass, and/or fistula formation. Fistulas are most often
seen proximal to strictures or within the strictures and are frequently surrounded
by extensive inflammatory changes.[51]
When CD becomes quiescent, vascular pattern may remain diminished. In patients with
more extensive CD, irregular healing may lead to hypertrophic zones alternating with
areas of atrophy (pseudopolyps) and mucosal bridging.
Upper gastrointestinal involvement in Crohn's disease
UGI endoscopy is routinely performed in the assessment of children with suspected
IBD to accurately classify IBD.[52] While UGI endoscopy and biopsies may be useful in all patients for evaluation of
the extent/location of the disease at the time of diagnosis, whether it should be
performed routinely in all adult patients remains unclear.[52] The UGI endoscopy, however, should be done in at least in those who have UGI symptoms.
Isolated involvement of UGI tract is rare in patients with CD, and whenever there
is an involvement of UGI tract, it occurs concomitantly with the involvement of either
small intestine or the colon or both.[53] Gastroduodenal involvement often presents with additional symptoms such as epigastric
pain and features of high-gut obstruction. The ulcers in the gastroduodenal CD may
be linear or serpiginous, unlike circular or transverse ulcers seen in peptic ulcer
disease, which involves gastroduodenal area far more often than CD.[54] In isolated duodenal disease, any part of the duodenum can be involved, but the
second part is most frequently affected. Deep ulceration and inflammation can lead
to stricture in gastroduodenal area and high-gut obstruction.
Diaz et al. in a review of 20 articles reporting 2511 patients with CD (CD group)
reported involvement of UGI tract in 815 (34%) patients (UGI group).[55] The endoscopic features in the CD group were gastric erythema in 5.9%, gastric erosions
in 3.7%, duodenal ulcers in 5.3%, and erythema in the duodenum in 3% of patients.[55] The most common histopathological features in the CD group was nonspecific gastric
inflammation in 32%, gastric granuloma in 7.9%, and focal gastritis in 30.9% of patients.[55] In the UGI CD group, gastric inflammation was present in 84% of patients, followed
by duodenal inflammation in 28.2% and gastric granuloma in 23.2%. The detection of
UGI tract involvement is higher (17%–75% for upper endoscopy) when biopsies of gastric
mucosa are done routinely in patients with CD compared to the practice when UGI endoscopy
and biopsies are done only if one observes any abnormality in symptomatic patients
(0.5%–13%).[53],[55]
Endoscopy during Follow-Up of Patients
In patients treated for intestinal TB, it is expected that there will be complete
resolution of the lesions, detectable by endoscopy or by imaging, following completion
of therapy. In a study of 49 patients with IC or colonic TB treated with anti-TB treatment,
complete healing of mucosal lesions was observed in 73.5% of the participants at the
end of 2 months and in 100% of participants at the end of 6 months.[56],[57] It is therefore appropriate to re-endoscope patients who were treated empirically
for ITB and to establish complete mucosal healing at the end of 6 months. The Indian
Society of Gastroenterology Task Force on IBD was in agreement that this should be
done, in view of the possibility that recurrent symptoms would make it difficult to
judge whether the patient was in fact having ITB or CD during the first episode.
There is now emerging evidence that mucosal healing should be one of the goals of
the treatment of CD. Therefore, it appears appropriate that the status of healing
process is assessed during the treatment period. While many agree to do that, the
protocol and timing of assessment is yet to be defined.
When diagnosis remains in doubt, both endoscopic and histologic assessments are appropriate
as demonstrated by a large proportion of patients initially classified as indeterminate
colitis or possible IBD were diagnosed as non-IBD after 5 years (22.5% vs. 50%) in
the IBSEN study.[58] In patients of definite ileocolonic CD, routine endoscopic evaluation for patients
in clinical remission may not always be necessary, unless clinicians suspect that
re-evaluation may have influence on the optimization of the management strategy. Fecal
calprotectin or lactoferrin may be used as surrogate markers of mucosal healing and
may reduce the need of repeated endoscopies.[59]
Ileocolonoscopy is the gold standard in the diagnosis of postoperative recurrence.
It is recommended to be done at 6–12 months after surgery. After resection of the
IC region in patients with CD, the postoperative recurrence rate is around 65%–90%
within 12 months and 80%–100% within 3 years of surgery in the absence of postoperative
prophylaxis.[60],[61]
Mucosal biopsies
With increasing use of endoscopic procedures to visualize the intestinal lumen and
obtaining targeted biopsies from the diseased segments of the intestine, endoscopic
mucosal biopsies have mostly replaced the surgical biopsies for microscopic examination.
For a reliable diagnosis of CD, multiple biopsies from different segments of the colon
(including the rectum) and the ileum should be obtained.[62] Multiple biopsies imply a minimum of two samples from each site.[62] Biopsies should be representative from areas of minor and major inflammation to
mirror correctly the intensity and spectrum of inflammation. In addition, biopsies
must also be taken from “normal appearing” mucosa.[62] Targeted biopsies from areas of stenosis, any unusual polypoid lesions, or any other
lesion that may attract endoscopists attention should be obtained in separate bottles.
Biopsies should also be obtained for microbiological tests including molecular diagnosis
of TB (polymerase chain reaction [PCR], GenExpert), acid-fast stain, and culture.
Isolation of M. tuberculosis, using culture or by staining of the smear, is the most
specific method for the diagnosis of active TB. The sensitivity of both these techniques,
however, remains low.[63],[64],[65],[66] While culture can detect 10–100 bacilli/mL of the sample, M. tuberculosis bacilli
can be cultured from 10% to 30% of mucosal biopsies in patients with colonic TB.[63],[64],[65],[66] A presumptive diagnosis of TB may be considered by detection of AFB on intestinal
biopsies which have been reported with a variable frequency (25%–36%) in patients
with intestinal TB.[19],[67]
PCR using M. tuberculosis-specific primer shows a positivity varying from 20% to 64%.
Detection of M. tuberculosis by PCR has a fair sensitivity, but the results of a positive
PCR should be interpreted cautiously.[19],[68],[69],[70] PCR may be reported positive because of contamination in the laboratories. Passage
of saprophytic mycobacteria may also test positive in colonic biopsies. In one study,
in situ PCR for TB in mucosal biopsies was positive in 6 of 20 intestinal TB patients.[63]
Small Bowel Endoscopy
In patients suspected of having intestinal TB or CD, especially when ileocolonoscopy
does not show definite evidence of the disease and contrast-enhanced computed tomography
(CT) or CT-enterography (CTE) or magnetic resonance-enterography (MRE) shows small
bowel wall thickening or stricture (with or without proximal bowel dilation), small
bowel endoscopy may be needed to look for mucosal abnormalities suggestive of the
disease and to obtain biopsies for histological assessment. Balloon-assisted and spiral
enteroscopy are preferred modalities for evaluating the small bowel today because
of biopsy and therapeutic capabilities. Histological assessment of the small intestinal
lesion(s) is essential as the etiological diagnosis of the ulcerating lesions cannot
be made based on endoscopic appearances alone.[71] In patients with suspected CD and negative ileocolonoscopy, small bowel capsule
endoscopy (SBCE) may be the diagnostic modality for the evaluation of the small bowel,
in the absence of obstructive symptoms or known stenosis. The sensitivity and specificity
for the diagnosis of CD have been reported to be 100% and 91% by SBCE, 81% and 86%
by MRE, and 76% and 85% by CTE, respectively.[72]
In patients having intestinal obstructive features or known to have small intestinal
stricture, a cross-sectional imaging modality such as MR-enterography or CTE is a
preferable initial investigation.[52]
The data regarding capsule endoscopy in ITB are limited. Reddy et al. have described
multiple scattered short, oblique, or transverse mucosal ulcers with a necrotic base
in the jejunum and ileum in patients with ITB.[73] Cello described the characteristics of small intestinal tubercular ulcers to be
shallow with irregular geographic borders, and transverse in their lie, rather than
longitudinal.[74]
SBCE is a safe noninvasive modality for the diagnosis of CD. As per the results of
a meta-analysis, SBCE is superior to all other modalities for diagnosing nonstricturing
small bowel CD, with a number needed to test (NNT) of 3 to yield one additional diagnosis of CD over small bowel barium radiography
and NNT of 7 over colonoscopy with ileoscopy.[75]
Fidder et al. defined a positive SBCE result for CD as the presence of four or more
ulcers, erosions, or a region with clear exudate and mucosal hyperemia and edema.[76] They reported that the greatest yield of SBCE in diagnosing CD is achieved in young
patients who present with symptoms of abdominal pain plus diarrhea.[76] SBCE has a high negative predictive value for small bowel CD.[52]
Endoscopic differentiation of small bowel CD from drug-induced lesions or other diseases
is unreliable. Nonsteroidal anti-inflammatory drugs should be withdrawn at least 4
weeks before SBCE if there is a diagnostic possibility of CD.[52] The small lesions detected by SBCE may not be specific for CD and may be found in
healthy people and patients with drug-induced enteropathy and vasculitis. In a study,
among 102 patients with suspected CD, 37% were initially found to have small bowel
ulcerations on SBCE, but CD was diagnosed definitively in 13% only at 1-year follow-up.[77]
The Capsule Endoscopy Crohn's Disease Activity Index (CECDAI) or Niv score has been
recently validated in a multicenter prospective trial.[78],[79] This scoring index evaluates three parameters: inflammation (A), extent of disease
(B), and presence of strictures (C), both for the proximal and distal segments of
the small bowel [Table 3]. The final score is calculated by adding the two segmental scores: CECDAI = proximal
([A1 × B1] + C1) + distal ([A2 × B2] + C2.
Table 3
Capsule endoscopy Crohn's disease activity index[78]
|
Score
|
Point
|
Features
|
|
Inflammation
|
0
|
None
|
|
score
|
1
|
Mild to moderate edema/hyperemia/denudation
|
|
|
2
|
Severe edema/hyperemia/denudation
|
|
|
3
|
Bleeding, exudate, aphthae, erosion, small ulcer (<0.5 cm)
|
|
|
4
|
Moderate ulcer (0.5.2 cm), pseudopolyps
|
|
|
5
|
Large ulcer (<2 cm)
|
|
Extent of
|
0
|
No disease
|
|
disease
|
1
|
Focal disease (single segment)
|
|
score
|
2
|
Patchy disease (2.3 segments)
|
|
|
3
|
Diffuse disease (>3 segments)
|
|
Stricture
|
0
|
None
|
|
score
|
1
|
Single.passed
|
|
|
2
|
Multiple.passed
|
|
|
3
|
Obstruction (nonpassage)
|
CE may play an important role in the monitoring of patients with CD and may have a
unique role in assessing mucosal healing after medical therapy, for assessing early
postoperative recurrence. The role of SBCE in patients with established CD should
focus on patients with unexplained iron deficiency or obscure GI bleeding or in those
with unexplained symptoms, when other investigations are inconclusive.[52],[80]
After a negative ileocolonoscopy and negative CTE or small bowel follow through, SBCE
has not been found to be a cost-effective third test, even in patients with high pretest
probability of CD.[81] The International Conference on Capsule Endoscopy recommended that patients with
suspected CD should be selected to undergo SBCE if they present with typical symptoms
(chronic abdominal pain, chronic diarrhea, weight loss, or growth failure) plus either
extraintestinal manifestations (fever, arthritis or arthralgia, pyoderma gangrenosum,
perianal disease, or primary sclerosing cholangitis), inflammatory markers (iron deficiency,
erythrocyte sedimentation rate, C-reactive protein, leukocytosis, or serology), or
abnormal small bowel imaging (small bowel series or CT scan).[82]
The ability of CE in distinguishing between ITB and CD is not clearly established.
Endoscopic Dilatation of the Strictures
Endoscopic dilatation of strictures in CD is a safe and effective alternative to surgery
in experienced hands and should be considered before surgery in selected patients.[52] Endoscopic balloon dilatation has been used to treat uncomplicated strictures, with
a maximal length of 4 cm, in patients with CD.[83] It has often been combined with local injection of steroids at the site of the stricture.
Endoscopic dilatation should be done only after control of the disease activity. A
systematic review of 13 studies including 347 patients with CD showed that endoscopic
dilatation was mainly used in postsurgical strictures, with 86% technical success,
2% complication rate, and 58% long-term efficacy.[84] A stricture length equal to or less than 4 cm was associated with a surgery-free
outcome. Another study including 776 dilatations in 178 patients with CD further confirmed
that the endoscopic dilatations had been done mostly for anastomotic strictures (80%).[85] Technical success rate was 89%. In a subset of patients who were followed up, 80%,
57%, and 52% required no further intervention or one additional dilation only at the
end of 1, 3, and 5 years, respectively. Cumulative proportions of patients undergoing
surgery at 1, 3, and 5 years were 13%, 28%, and 36%, respectively. Complication rate
per procedure was 5.3%, and included bowel perforation (1.4%), major bleeding requiring
blood transfusion (1%), minor bleeding (1.3%), and abdominal pain or fever (1.5%).
Endoscopic Differences between Intestinal Tuberculosis and Crohn's Disease
Zhou and Luo [86] compared the records of 30 patients of CD and 30 patients with ITB. The comparison
of endoscopic findings among various studies between ITB and CD is presented in [Table 4].
Table 4
Endoscopic differences between gastrointestinal tuberculosis and Crohn's disease
|
|
Zhou and Luo[86]
|
Lee et al.[87]
|
Amarapurkar[8]
|
Makharia[23]
|
Li et al.[24]
|
Dutta et al.[88]
|
Zhang et al.[89]
|
|
CD (n=30) (%)
|
ITB (n=30) (%)
|
CD (n=44) (%)
|
ITB (n=44) (%)
|
CD (n=26) (%)
|
ITB (n=26) (%)
|
CD (n=53) (%)
|
ITB (n=53) (%)
|
CD (n=130) (%)
|
ITB (n=122) (%)
|
CD (n=30) (%)
|
ITB (n=24) (%)
|
CD (n=92) (%)
|
ITB (n=31) (%)
|
|
CD=Crohn's disease, ITB=Intestinal tuberculosis, IC=Ileocecal
|
|
Skip lesions
|
23.1
|
|
|
7.7
|
66
|
16.9
|
|
|
36.7
|
8.3
|
|
|
|
Aphthous ulcers
|
81.8
|
20
|
34.6
|
19.2
|
54.7
|
13.2
|
53.8
|
44.3
|
36.7
|
16.7
|
44.6
|
29
|
|
Linear ulcers
|
|
|
40.9
|
2.3
|
57.7
|
23.1
|
30.1
|
7.5
|
54.7
|
8.2
|
33.3
|
4.2
|
80.4
|
58.1
|
|
Transverse ulcers
|
|
|
25
|
65.9
|
|
|
|
|
4.6
|
41
|
6.7
|
25
|
18.5
|
64.5
|
|
Nodularity
|
|
|
|
|
|
|
24.5
|
49
|
48.5
|
32.8
|
|
|
|
|
Cobblestoning
|
20
|
0
|
34.1
|
6.8
|
57.7
|
23.1
|
16.9
|
0
|
27.7
|
1.6
|
10 0
|
33.7
|
32.3
|
|
Strictures
|
|
|
9
|
8 30.7
|
19.2
|
|
|
|
|
26.7
|
16.7
|
18.5
|
12.9
|
|
Pseudopolyps
|
46.7
|
40
|
27.3
|
52.3
|
34.6
|
19.2
|
|
|
|
|
30
|
8.3
|
|
Patulous IC valve
|
|
|
9.3
|
40
|
|
|
|
|
6.7
|
8.3
|
7.6
|
38.7
|
|
IC lesions
|
|
|
90.7
|
100
|
|
|
|
|
|
|
|
|
|
|
|
Anorectal lesions
|
|
|
84.1
|
9.1
|
|
|
|
|
|
|
20
|
4.2
|
17.4
|
6.5
|
|
<4 segments involved
|
|
|
18.2
|
81.8
|
|
|
|
|
|
|
|
|
|
|
Lee et al.[87] evaluated the diagnostic value of colonoscopy in differentiating ITB (n = 44) from CD (n = 44) and observed that four parameters (anorectal lesions, longitudinal ulcers,
aphthous ulcers, and cobblestone appearance) were significantly more common in patients
with CD than in patients with ITB. Four other parameters (involvement of fewer than
four segments, a patulous IC valve, transverse ulcers, and scars or pseudopolyps)
were observed more frequently in patients with ITB than in patients with CD.[87] A patulous valve with surrounding heaped-up folds or a destroyed valve with a fish
mouth opening was more likely to be caused by TB than that by CD. In their study,
if CD was diagnosed based on the presence of higher number of parameters characteristic
of CD than those characteristic of ITB, and vice versa, the diagnosis was correct
in 87.5% and incorrect in 8%.[87]
Amarapurkar et al. studied 26 patients with ITB and 26 patients with CD and found
that the accuracy of endoscopic features (deep linear/serpiginous ulcers and cobblestone
appearance) was 67.3% compared to the 84.6% for clinical score and 82.6% for positive
TB-PCR.[8]
We, in a study of 53 patients with intestinal TB and 53 patients with CD, observed
that while skip lesions, aphthous ulcers, linear ulcers, and cobblestoning were more
common in patients of CD than that in intestinal TB; nodularity was however more common
in those with intestinal TB.[23] We also observed that the involvement of rectum (62.2% vs. 18.8%, P < 0.001), sigmoid
colon (66% vs. 11.3%, P < 0.001), descending colon (60.3% vs. 13.2%, P < 0.001), ascending
colon (64.1% vs. 43.4%, P = 0.03), and jejunum (16.9% vs. 0%, P < 0.001) was significantly
more common in patients with CD than in patients with ITB.[23] There was no significant difference in the involvement of the IC region (88.6% vs.
92.4%, P = 0.5), ileum (24.5% vs. 15%, P = 0.2), stomach, and duodenum in patients
with CD and intestinal TB.[23]
Li et al.[24] in their study including 122 patients with ITB and 130 patients with CD observed
that that rectum and sigmoid colon involved lesions were more common in patients with
CD. On the other hand, cecum and IC valve involved lesions and fixed-open IC valve
were more often seen in patients with ITB. Ring-like ulcers and rodent-like ulcers
were more common in ITB, while longitudinal ulcers, grid-like ulcers, cobblestone
appearance, and nodular hyperplasia were more commonly found in CD [Table 4].
Dutta et al. followed up 30 patients of ITB and 30 patients of CD and reported that
patients with CD had longer duration of symptoms (P < 0.001), blood mixed stool (P
= 0.006), longitudinal ulcers (P = 0.005), skip lesions (P = 0.008), and involvement
of higher number of colonic segments (P = 0.004).[88] Yu et al.[90] studied 43 patients with ITB and 53 patients with CD and observed that transverse
ulcers and patulous IC valve were more common in ITB. Longitudinal ulcer was a significant
predictor in differentiating CD from ITB (odds ratio 35.5, confidence interval 1.8–683.2).
Zhang et al.[89] also reported that transverse ulcers, rodent-like ulcers, and patulous IC valve
were more common in those with ITB and longitudinal ulcers in patients with CD.
The reasons for a greater frequency of involvement of left side of the colon in patients
with CD and occurrence of aphthous ulcers, linear ulcers, cobblestoning of mucosa
in higher frequency in CD and nodularity, patulous IC valve in higher number in those
with ITB, are not well known and need more research.
Approach in a Patient Where a Differentiation between Intestinal Tuberculosis and
Crohn's Disease Cannot Be Made
The clinical, endoscopic, and histological features that help in differentiation between
ITB and CD are summarized in [Table 5]. From the [Table 5], it is obvious that almost all the features are often present in these two conditions
and none of the features is pathognomonic of either of the diseases. However, with
the use of a combination of these features, one may be able to make a diagnosis of
either TB or CD in almost half of the patients.
Table 5
Type of foreign body with the endoscopic removal accessory used in the two age groups
|
Variables
|
Intestinal Tuberculosis (%)
|
Crohnfs Disease (%)
|
|
Chronic diarrhea
|
20-40
|
60-80
|
|
Blood in the stools
|
10-20
|
50-70
|
|
Abdominal pain
|
90
|
60-80
|
|
Constipation
|
50
|
10-30
|
|
Symptoms Partial intestinal obstruction
|
50-60
|
20-30
|
|
Peri-anal disease 5
|
30-80
|
|
Fever
|
40-70
|
30
|
|
Loss of appetite
|
40-80
|
40-60
|
|
Weight loss
|
60-90
|
50-60
|
|
Extra-intestinal manifestations
|
10
|
20-50
|
|
Anal canal
|
<5
|
15-50
|
|
Rectum
|
10-20
|
40-60
|
|
Sigmoid colon
|
5-10
|
10-50
|
|
Involvement of Intestine Descending colon
|
5-10
|
10-50
|
|
Ascending colon
|
40-60
|
50-70
|
|
Ileocaecal area
|
70-90
|
60
|
|
Ileum
|
10-20
|
20-40
|
|
Aphthous ulcers
|
5-10
|
30-50
|
|
Endoscopic features Linear ulcers
|
5
|
20-30
|
|
Deep ulcers
|
50-70
|
30-40
|
|
Nodularity
|
50
|
20
|
|
Cobblestoning
|
None
|
15-20
|
|
Histological features Granuloma
|
30-60
|
30
|
|
Necrosis in the granuloma
|
10-30
|
None
|
There are situations where differentiation between ITB and CD is not possible despite
extensive investigation. Under these circumstances, both the Asia Pacific Association
of Gastroenterology and the Indian Society of Gastroenterology consensus on CD suggest
a trial of anti-TB treatment.[20],[21] This is based on the rationale that the treatment of TB is finite, whereas treatment
for CD continues indefinitely.[20],[21] Furthermore, there is a risk in treating patients with corticosteroids if the diagnosis
actually is TB. Obviously the decision to treat for one disease or the other would
take into account other clinical considerations including the nature of presentation
of the patient, whether acutely ill and requiring immediate definitive therapy or
not. It would also be accompanied by a complete discussion with the patient of the
possibilities and therapeutic alternatives. Compounding the problem of differentiation
between intestinal TB and CD and a therapeutic trial of anti-TB drugs to resolve the
issue, a proportion of patients with CD also respond to anti-TB therapy at least symptomatically.[91]
In patients treated for intestinal TB, it is expected that there will be a complete
resolution of the lesions, detectable by endoscopy or by imaging, following therapy.
It is, therefore, appropriate to re-endoscope the patients who were treated empirically
for ITB and to establish complete mucosal healing at the end of 6 months.[20],[21]
In summary, while many of endoscopic features overlap in patients with ITB and CD,
certain endoscopic features have been identified by most of the investigators as suggestive
of CD. These features include involvement of left side of the colon, involvement of
multiple segments of the intestine, and presence of longitudinal ulcers and cobblestoning
of the mucosa. Healing of the lesions should be demonstrated in patients suspected
to have ITB and being treated empirically with anti-TB therapy.
Financial support and sponsorship
Nil.
