Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31

https://doi.org/10.1007/s12664-023-01343-x

REVIEW ARTICLE

Evidence‑based approach to diagnosis and management

of abdominal tuberculosis
Daya Krishna Jha1 · Mythili Menon Pathiyil2 · Vishal Sharma3

Received: 21 December 2022 / Accepted: 20 January 2023 / Published online: 11 March 2023
© Indian Society of Gastroenterology 2023

Abstract
Abdominal tuberculosis is an ancient problem with modern nuances in diagnosis and management. The two major forms are
tuberculous peritonitis and gastrointestinal tuberculosis (GITB), while the less frequent forms are esophageal, gastroduodenal,
pancreatic, hepatic, gallbladder and biliary tuberculosis. The clinicians need to discriminate the disease from the close mimics:
peritoneal carcinomatosis closely mimics peritoneal tuberculosis, while Crohn’s disease closely mimics intestinal tuberculosis.
Imaging modalities (ultrasound, computed tomography, magnetic resonance imaging and occasionally positron emission tomog-
raphy) guide the line of evaluation. Research in diagnostics (imaging and endoscopy) has helped in the better acquisition of tissue
for histological and microbiological tests. Although point-of-care polymerase chain reaction–based tests (e.g. Xpert Mtb/Rif)
may provide a quick diagnosis, these have low sensitivity. In such situations, ancillary investigations such as ascitic adenosine
deaminase and histological clues (granulomas, caseating necrosis, ulcers lined by histiocytes) may provide some specificity to
the diagnosis. A diagnostic trial of antitubercular therapy (ATT) may be considered if all diagnostic armamentaria fail to clinch
the diagnosis, especially in TB-endemic regions. Objective evaluation with clear endpoints of response is mandatory in such
situations. Early mucosal response (healing of ulcers at two months) and resolution of ascites are objective criteria for early
response assessment and should be sought at two months. Biomarkers, especially fecal calprotectin for intestinal tuberculosis,
have also shown promise. For most forms of abdominal tuberculosis, six months of ATT is sufficient. Sequelae of GITB may
require endoscopic balloon dilatation for intestinal strictures or surgical intervention for recurrent intestinal obstruction, perfora-
tion or massive bleeding.

Keywords Abdominal tuberculosis · AFB · Ascites · Colonoscopy · Culture · Extrapulmonary tuberculosis ·

Gastrointestinal tuberculosis · Intestinal tuberculosis · Mycobacterium tuberculosis · PCR · Pancreatic tuberculosis ·
Peritoneal tuberculosis · Surgery · Tuberculous peritonitis · Xpert

Introduction the incidence of extrapulmonary tuberculosis (EPTB) is around

15% globally. Abdominal TB is among the common sites of
Infection with Mycobacterium tuberculosis lasts a lifetime and extrapulmonary involvement, where it tends to involve the gas-
the organism infects almost a quarter of the world population trointestinal tract, peritoneum, lymph nodes and solid organs
that remains at the risk of advancing to active disease [1, 2]. in that order. The diagnosis and management of abdominal TB
Tuberculosis affects nearly 10 million people and leads to death are challenging: the disease is usually paucibacillary with a low
in more than a million people annually, despite being a prevent- yield of microbiological tests and it mimics many conditions
able and curable disease [2]. It primarily involves the lung, but closely, resulting in diagnostic confusion [3, 4]. In certain cases,
relatively non-specific parameters such as ascitic fluid adenosine
deaminase levels are utilized for diagnosis [5]. When the diag-
* Vishal Sharma
nosis remains unclear even after all these modalities, a thera-
docvishalsharma@gmail.com
peutic trial with antitubercular therapy (ATT) is often started in
1
Army Hospital (Research and Referral), 110 010 Delhi, India TB endemic regions and the response to the therapy is assessed.
2
Saint Vincent Hospital, Worcester, MA, USA Since the disease is primarily a concern in the less developed
3 world, the development of evidence-based diagnosis and treat-
Department of Gastroenterology, Postgraduate Institute
of Medical Education and Research, Chandigarh 160 012, India ment has lagged.

13
Vol.:(0123456789)
18 Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31

A recent survey of clinicians suggested significant varia- commonest form of abdominal TB in most reports from
tions in clinical practice with respect to the diagnostic modali- tertiary care centers [25].
ties, treatment duration and follow-up modalities in abdomi- Certain comorbidities, especially chronic liver disease,
nal TB [6]. Therefore, the present review aims at providing specifically increase the predisposition to abdominal/PTB
an evidence-based summary to inform clinical practice for [30]. Additional risk factors for developing abdominal TB
abdominal TB. Because gastrointestinal tuberculosis (GITB, include younger age, female gender, Asian ethnicity, human
also termed intestinal TB) and tuberculous peritonitis (TBP, immunodeficiency virus (HIV) coinfection, immunosuppres-
also termed peritoneal tuberculosis [PTB]) are the most fre- sion, diabetes mellitus and peritoneal dialysis [30–32].
quent clinical problems, the review is largely focused on these
two entities.
Clinical presentation
Methods
In a systematic review on tuberculous peritonitis, abdomi-
Although the present review is a narrative review, a search nal pain (65%), fever (59%), weight loss (61%), diarrhea
of two databases, Pubmed and Embase, was performed on (21%) and constipation (11%) were the most frequently
December 1, 2022, to inform the review. We used MESH reported symptoms [32]. The presence of ascites (73%),
words and free terms to search for Abdominal Tuberculosis abdominal tenderness (48%), hepatomegaly (28%) and
OR Peritoneal Tuberculosis OR Tuberculous peritonitis OR splenomegaly (14%) were the most frequent clinical find-
Gastrointestinal Tuberculosis OR Intestinal Tuberculosis. We ings (16,197,489). Constitutional symptoms occur in half
aimed at providing evidence-based recommendations regard- of the patients. Because of the non-specific nature of the
ing several contentious areas in the diagnosis and management symptoms, the diagnosis can be delayed (7–24 weeks from
of abdominal TB, including evaluation pathways, diagnostic the onset of symptoms) [33]. Although less frequent than
armamentarium to be used (standard culture or liquid culture, GITB, TBP may also have features of intestinal obstruction,
number/amount of tissue or fluid, which polymerase chain especially in the presence of adhesions, peritoneal fibro-
reaction (PCR)–based tests, whether to use adenosine deami- sis or sclerosing encapsulating peritonitis (i.e. abdominal
nase and the cut-off for PTB) and treatment strategies (duration cocoon) [34].
of ATT, additional role of ancillary therapies such as steroids, The clinical presentation of intestinal TB is dominated by
use of diagnostic trial and the duration of such a trial, appro- abdominal pain (30% to 88%), fever (21% to 73%), diarrhea
priate methods of assessment after a diagnostic trial of ATT). (5% to 47%), loss of appetite (30% to 90%), loss of weight
Wherever systematic reviews (Tables 1 and 2) or randomized (8% to 80%), constipation (7% to 24%), and hematochezia
studies were available, we used these to summarize the man- (5% to 15%). Some patients may present with intestinal
agement recommendations [7–24]. Where systematic reviews obstruction (3% to 36%) [35]. Concomitant or past pulmo-
or randomized clinical trials (RCTs) were not available, we nary TB could be present in up to 25% of cases [36]. The
used observational studies to suggest an appropriate clinical rising incidence of inflammatory bowel disease (IBD) in
approach. We also identify advances in the field that are likely India does not seem to have reduced the numbers of GITB.
to be useful in clinical practice in the coming times, lacunae in Hence distinguishing the two continues to be a major chal-
the current literature and avenues for future research. lenge [37].

Epidemiology of abdominal tuberculosis Case definitions

The number of cases of abdominal TB as a fraction of all A large number of case definitions and classifications for
EPTB cases has been reported to vary from 2.7% to 21% abdominal TB are in vogue and are based on clinical pres-
[25, 26]. In a study from three states in India and based entation, morphological patterns, basis of diagnosis and
on the national tuberculosis program, abdominal TB con- response to ATT (Fig. 1). Traditionally, Paustian criteria
stituted 12.8% of all EPTB cases [26]. Lower treatment were proposed but are impractical in the modern era, as they
completion rates and worse outcomes have been reported rely on surgical specimens and animal inoculation. Logan
in abdominal TB [26, 27]. Among patients with abdomi- proposed a response assessment that continues to be used to
nal TB, both GITB and tuberculous peritonitis have been date for diagnosing GITB [38]. The cases of abdominal TB
reported as common sites [28, 29]. Because of a possi- could be defined on the basis of clinical presentation—pain
ble selection bias, PTB being easier to diagnose based (strictures, hypertrophic lesions) or diarrhea (diffuse ulcers)
on abdominal paracentesis, GITB is often reported as the for intestinal TB and pain predominant (peritoneal adhesions

13
 Table 1  Summary of available evidence based on the systematic reviews on tuberculous peritonitis
Research query Reference Evidence base Comparison Results Interpretation

What are the imaging features in Chen et al. 4 studies, 483 participants Smooth peritoneal thickening The specificity and sensitivity of No single finding seems ade-
CT to differentiate TBP from 2020 [7] 84% and 59%, respectively quately sensitive and specific
peritoneal carcinomatosis? Chen et al. 5 studies, 605 participants Ascites The sensitivity and specificity of
2020 [7] 50% and 58%, respectively
Chen et al. 3 studies, 313 participants Lymph node necrosis The sensitivity and specificity
2020 [7] and 21% and 100%, respec-
tively
What is the diagnostic accuracy Su et al. 6 studies IFN-γ level in diagnosing TB The sensitivity was 0.93 (95% Similar performance as ascitic
of (IFN-γ level testing for 2013 [8] peritonitis CI: 0.87–0.97) and specificity, adenosine deaminase
the diagnosis of tuberculous 0.99 (95% CI: 0.97–1.00) Unclear if any incremental benefit
peritonitis?
What is the diagnostic accuracy Riquelme et al. 4 studies, 264 participants Using ADA cut-off 36–40 IU/L The pooled sensitivity and Variable cut-off and methods used
of adenosine deaminase in 2006 [9] in peritoneal fluid to diagnose specificity are 100% and 97%, in participating studies
peritoneal fluid for the diagno- peritoneal TB respectively Good sensitivity and specificity
sis of tuberculous peritonitis? Shen et al. 2013 16 studies, 1574 participants ADA in peritoneal fluid The pooled sensitivity was for the diagnosis
[10] 0.93 (95% CI: 0.89–0.95),
specificity was 0.96 (95% CI:
0.94–0.97)
Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31

Tao et al. 17 studies, 1797 subjects Using ADA 35 IU/L to diagnose The sensitivity and specificity
2014 [11] peritoneal TB against CRS are 93% and 94%, respectively
Zhou et al. 24 studies, 3044 participants Variable usually composite Pooled sensitivity of 93% (95%
2022 [12] reference CI: 0.89–0.95) and 95% (95%
CI: 0.93–0.96), respectively
What is the diagnostic accu- Luo et al. 8 studies, 786 participants T-SPOT in peripheral blood Pooled sensitivity and specificity Poor specificity limits use as
racy of T-SPOT testing for 2020 [13] compared to composite refer- of PB T-SPOT in diagnosing many positive cases would have
the diagnosis of tuberculous ence standard peritoneal TB were 0.91 (95% alternative diagnosis
peritonitis? CI: 0.88–0.94) and 0.78 (95%
CI: 0.73–0.81)
Luo et al. 7 studies, 423 participants T-SPOT in peritoneal fluid com- The pooled sensitivity and speci-
2020 [13] pared to composite reference ficity of PF T-SPOT are 0.90
standard (95% CI: 0.85–0.94) and 0.78
(95% CI: 0.72–0.83)
What is the diagnostic accuracy Kohli et al. 2021 13 studies with 580 participants Xpert Mtb Rif against mycobac- Pooled sensitivity and specificity Xpert Mtb/Rif provides a quick
of Xpert MTB/RIF testing in [14] terial culture from ascitic fluid 59.1% (42.1–76.2) and 97.6% diagnosis with excellent speci-
ascitic fluid for the diagnosis of (95.4–98.9), respectively ficity. However, low sensitivity
tuberculous peritonitis? Sharma et al. 18 studies, 1099 samples Xpert Mtb Rif against mycobac- Pooled sensitivity and specificity means a significant number of
2021 [15] terial culture from ascitic fluid 64% (95% CI: 49–76%) and patients will be missed
97% (95–99), respectively
8 studies, 634 samples Xpert Mtb Rif in ascitic fluid Pooled sensitivity and specific-
compared to a composite refer- ity: 30% (22–40%) and 100%
ence standard (98–100%)
19

13
 20 Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31

or fibrosis) or distension predominant (ascites in PTB) for

RR relative risk, CT computed tomography, TBP tuberculous peritonitis, AFB acid fast bacilli, ADA adenosine deaminase, ATT​Antitubercular therapy, PB Peripheral blood, PF Peritoneal fluid,
6 months of therapy is adequate
observational data suggests that
only in the setting of peritoneal
tuberculous peritonitis [39]. The morphological patterns

Steroids should not be routinely

Poor quality of included studies

patients with TBP. However,

combination of this trial and
systematic review included
have also been described for intestinal TB and tuberculous

Only one RCT included in

peritonitis, but an overlap among these patterns is well rec-

Limited generalizability
ognized. A systematic review of definitions of PTB identi-
fied that the classification into wet, dry and fibrotic forms is
tuberculosis

dogged by overlapping features and is better avoided [39].

Interpretation

Therefore, clinicians should largely follow a case defini-

tion that provides information regarding the confidence
used

of diagnosis—a hierarchical strategy may be helpful and

CI Confidence interval, RCT​randomized controlled trial, CI confidence interval, TB tuberculosis, IFN‐γ Interferon‐gamma, CRS composite reference standard
would largely be consistent with the definitions of INDEX-
intestinal obstruction (RR: 0.18
compared with using ATT for

RR: 1.02, 95% CI: 0.97 to 1.08

ATT was more effective than

TB guidelines [40]. A microbiologically positive case would

Adjunctive steroids used with

the prevention of composite

[0.04–0.62] p = 0.008), and

end point (RR: 0.15 [0.04,
0.62], p = 0.008), sympto-

have the highest confidence in the diagnosis, while in the

matic stricture (RR: 0.15

absence of microbiological positivity, a diagnosis of a clini-

[0.03–0.99] p = 0.05)

cally diagnosed case is made. It is important to recognize

that all clinically diagnosed cases of abdominal TB are not
equal: certain findings such as caseating granulomas or
high ascitic adenosine deaminase levels may provide more
Results

certainty, while a diagnosis based on consistent clinical-

radiological findings and exclusion of alternative diagnosis
cal cure on 6 months compared

is much less secure. The confidence in the diagnosis has

Participants that achieved clini-
Adjunctive steroids with ATT

clinical relevance because the clinicians would need to fol-

compared to ATT alone

low patients with possible abdominal TB more closely and

to 9 months of ATT​

evaluate them for objective evidence of response to ATT.

Figure 1 suggests a hierarchical approach to the diagnosis of
abdominal TB based on confidence in the diagnosis.
Comparison

Evaluation

Serum markers/IGRA​
3 studies, 108 participants

3 trials, 294 participants

Interferon-gamma release assays (IGRAs) have emerged

as an important tool for the diagnosis of latent TB infec-
tion. These have also been used for possible discrimina-
Evidence base

tion of abdominal TB from mimics like Crohn’s disease or

other causes of ascites. These are believed to be helpful as
these are not impacted by Bacille Calmette-Guérin (BCG)
vaccination.
In a meta-analysis of 12 studies, the diagnostic accuracy
of peripheral blood (PB) T-SPOT and peritoneal fluid (PF)
Jullien et al.
2016 [17]
steroids for treating tuberculous 2019 [16]
Soni et al.
Reference

T-SPOT for TBP was evaluated. The pooled sensitivity and

specificity of PB T-SPOT for diagnosing peritoneal TB were
91% and 78%, respectively, while the pooled positive likeli-
hood ratio (PLR) and negative likelihood ratio (NLR) were
ATT must be used to achieve
What is the duration for which
What is the role of adjunctive

4.05 and 0.13, respectively. On the other hand, the pooled

sensitivity, specificity, PLR and NLR of PF T-SPOT for TBP
were 90%, 78%, 6.35, and 0.14, respectively. The results
Table 1  (continued)

summarized that both PB T-SPOT and PF T-SPOT are sensi-

Research query

clinical cure?

tive for diagnosing TBP. Still, the unsatisfactory specificity

peritonitis?

of these two methods limits application as rule-in tests for

peritoneal TB diagnosis [13] (Table 1). In another systematic

13
 Table 2  Summary of available evidence based on the systematic reviews on gastrointestinal tuberculosis
Research Query Reference Evidence base Comparison Results Interpretation

Accuracy of CT in differentiating Kedia et al. 6 studies, 638 patients Necrotic lymph node Pooled sensitivity of 23% (95% Necrotic lymph nodes very
ITB from CD 2017 [18] CI: 17–29%) and specificity of specific, but not sensitive for
100% (95% CI: 99–100%) GITB. Important to recognize
Kedia et al. 2 studies, 209 patients Ileocecal involvement The pooled sensitivity and that necrotic lymph nodes could
2017 [18] specificity were 64% (95% CI: occur in other conditions such
53–74%) and 77% (95% CI: as malignancy
68–84%), respectively
What is the clinical usefulness of Chen et al. 5 studies, 610 patients Composite reference using multi- Pooled sensitivity and specificity: Inability to differentiate latent
using IGRA in differentiating 2013 [19] ple parameters 74% (53–100%) and 87% from active TB
GITB from CD? (63–98%), respectively
Diagnostic odds ratio (DOR) in
differentiating ITB vs CD was
26.2
Ng et al. 8 studies, 705 patients Composite reference using multi- Pooled sensitivity and specific-
2014 [20] ple parameters ity 81% (75–86%) and 85%
(81–89%), respectively
Xu et al. 12 studies from Asia (Chinese) Performance of IGRAs Pooled sensitivity and specificity
2016 [21] (QuantiFERON-TB Gold or 82.8% (78.4–86.6%) and 86.7%
Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31

T-SPOT.TB) (83.2–89.6%), respectively

What is the usefulness of ASCA Ng et al. 6 studies, 603 patients Composite reference Pooled sensitivity and specificity Indian studies do not suggest
in differentiating CD from 2014 [20] were 33% (27–38%) and 83% ( much benefit
ITB? 77–88%), respectively
What is the diagnostic accuracy Sharma et al. 5 studies, 460 samples Composite reference standard Pooled sensitivity and specific- Poor sensitivity but excellent
of Xpert Mtb/Rif testing in 2021 [15] ity: 23% (16–32%) and 100% specificity. Good rule in test but
intestinal tissue for the diagno- (52–100%), respectively a poor rule out test
sis intestinal tuberculosis?
What is the clinical usefulness of Jin et al. 9 studies, 709 patients Composite reference Pooled sensitivity and specific- Fair sensitivity and good specific-
detecting MTB PCR (IS6110) 2017 [22] ity: 47% (42–51%) and 95% ity
in biopsy or fecal samples in (93–97%), respectively
differentiating GITB from CD? DOR of 21.92
What is the diagnostic accuracy Du et al. 9 studies, 692 patients Caseation versus composite Sensitivity: 21% (15–28%) Some of the features are specific
of histological features in dif- 2014 [23] reference Specificity: 100% (98–100%) but have a low sensitivity
ferentiating GITB and CD? DOR: 13.74
Du et al. 5 studies Confluent granulomas Specificity: 38% (30–47%)
2014 [23] Sp: 99% (95–100%)
DOR: 26.52
Du et al. 3 studies Ulcers lined by epithelioid Specificity: 41% (32–51%)
2014 [23] histiocytes Sp: 94% (88–98%)
DOR: 13.17

CT Computed tomography, ITB intestinal tuberculosis, CD Crohn’s disease, ASCA anti-Saccharomyces cerevisiae antibody, GITB Gastrointestinal tuberculosis, IGRA​ Interferon gamma release
assay; MTB PCR Mycobacterium tuberculosis-polymerase chain reaction test, DOR diagnostic odds ratio, CI confidence interval

13
21
22 Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31

Gastrointes Tuberculosis Peritoneal Tuberculosis

Clinical presenta Pain dominant Pain dominant

Diarrhea dominant Distension dominant

Morphological Ulcera ve Wet – asci c

Hypertrophic Dry plas
Stricturing/ Steno c Fibr fixed
Combina Cocoon

Diagnosis Clinical diagnosed (Probable or Possible)

Microbiologically diagnosed (Confirmed)

Confirmed Abdominal
Tuberculosis
Probable Abdominal
Tuberculosis Secure microbiological diagnosis
Possible Abdominal - Posi ve culture

Tuberculosis Largely certain diagnosis with exclusion of

- Posi ve validated PCR (Xpert MTB/RIF)
alterna ve causes - AFB pos ve (may ve in non-tubercular
pathologies also)
- c ADA, nega ve malignant cells
Diagnosis likely but no clear supp ng evidence - Casea granuloma on histology
- Clinically and radiologically consistent - Granuloma on histology, other causes excluded
- Tests not contributory - Necr lymph nodes on CT with non-
- Borderline ADA contributory cytology / biopsy
- Histology- Chronic inflamma
- Alterna ve diagnosis excluded

Increased certainty of diagnosis

Increased need for monitoring and follow-up if ATT started

Fig. 1  Summary of case definitions and a hierarchical approach to defining an abdominal tuberculosis case. ADA adenosine deaminase, CT com-
puted tomography, PCR polymerase chain reaction, AFB acid-fast bacilli

review, interferon-gamma levels in the ascitic fluid were abdominal TB is unclear and should not be routinely done in
reported to have an excellent sensitivity and specificity TB-endemic regions to diagnose active TB. However, it may
(93% and 99%, respectively), but the incremental values have a role in excluding TB if a diagnosis of CD is considered
over ascitic ADA are uncertain [8]. in TB-endemic settings, with the caveats already mentioned.
IGRAs have also been evaluated in multiple studies as
a diagnostic modality to discriminate GITB from Crohn’s Imaging modalities
disease (CD). Three meta-analyses have been published
(Table 2) [19–21]. In the most recent meta-analysis by Xu The initial imaging used for evaluation is often abdominal
et al., 12 studies were included and a pooled sensitivity ultrasound (USG) which can help in the evaluation of lym-
of 82.8% (78.4–86.6%) and a pooled specificity of 86.7% phadenopathy, peritoneal or omental thickening, ascites, mes-
(83.2–89.6%) were reported [21]. The authors of these meta- enteric changes and bowel wall thickening. The use of bowel
analyses concluded that IGRA is a good supplementary ultrasound in the evaluation of strictures has been reported in
method to discriminate between intestinal tuberculosis (ITB) the setting of IBD, although this is yet to be reported system-
and CD. Some other studies, however, report poor sensitiv- atically in the setting of GITB [43]. Nevertheless, ultrasound
ity and specificity and question the utility of this test [41]. provides a good initial evaluation in suspected abdominal
There are certain caveats to the use of this test—in TB- TB cases and can also provide material for microbiological,
endemic regions, patients with IBD would also be exposed cytological or histopathological evaluation [44]. In a system-
to tuberculosis. They may have a positive IGRA, while many atic review on the utility of abdominal ultrasonography for
with disseminated TB or malnutrition may not demonstrate the diagnosis of TB in the setting of HIV, the sensitivity and
immune responsiveness on exposure to TB antigens. A recent specificity for the diagnosis of abdominal TB were 63% and
report highlighted the value of quantitative measurements 68%, respectively, against a microbiological standard. These
using enzyme-linked immunoassay (ELISA), suggesting that findings suggest that a negative ultrasound should not be used
the levels of > 100 pg/mL had a higher sensitivity and speci- to exclude abdominal TB in HIV-positive individuals [24].
ficity for discriminating ITB from CD [42]. With the avail- Computed tomography (CT) has emerged as an excellent
able evidence, the role of IGRA in the diagnosis of active tool for evaluating abdominal diseases because of its easy

13
Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31 23

availability and excellent resolution. It provides good visu- especially peritoneal carcinomatosis. Although typically
alization of both luminal and extraluminal findings [43]. A lymphocyte dominance is seen in tubercular ascites, neu-
systematic review of six studies compared the yield of 17 trophils could be predominant in early stages and peritoneal
CT findings in discriminating TBP from peritoneal carci- dialysis associated with TBP.
nomatosis. The highest diagnostic accuracy for TBP was
achieved by smooth peritoneal thickening (AUC: 0.83), Ascitic adenosine deaminase
with fairly good specificity (84%), but limited sensitivity
(59%). The location and presence of ascites showed poor Adenosine deaminase (ADA) in peritoneal fluid offers a
diagnostic accuracy (AUC: 0.63) because of both poor sen- feasible, sensitive and highly specific test for TBP. While
sitivity (50%) and specificity (58%). Lymph node necrosis the multiple available test methods and different cut-offs by
and calcification showed an impressive specificity (95% different studies can pose a challenge, as ascitic fluid tap-
and 100%, respectively) but poor sensitivity (10% and 12%, ping and analysis is easy to do and yields immediate results,
respectively) [7]. ADA remains a relevant tool for screening and diagnosing
Another systematic review of six studies with 612 patients PTB in India. Additionally, as PTB is a paucibacillary dis-
assessed the role of various CT findings in the discrimi- ease with low mycobacterial numbers, ADA has a special
nation of GITB and CD. Certain findings such as necrotic role given its sensitivity and specificity. Four meta-analyses
lymph nodes (sensitivity: 23% and specificity: 100%) and have addressed the issue of diagnostic yield of ADA [9–12].
ileocecal involvement (sensitivity: 64% and specificity: 77%) The most recent systematic review included 24 studies with
suggested the diagnosis of GITB. Other features including 3,044 samples and found that the test had an excellent pooled
comb sign, skip lesions, asymmetric bowel involvement, sensitivity and specificity (93% and 95%, respectively) for
mural stratification, long segment involvement, fibrofatty the diagnosis of TBP. The PLR and NLR were also sup-
proliferation and left-sided colonic involvement were asso- portive of the use of this test. INDEX TB Guidelines suggest
ciated with a diagnosis of CD. However, none of these find- that a cut-off of 39 U/L be used for the diagnosis. However,
ings (except for necrotic lymph nodes) were pathognomonic there are certain caveats to the use of this test; false positives
of a particular diagnosis [18]. Certain other CT findings, could occur in peritoneal carcinomatosis, lymphoma, hemor-
which could suggest a diagnosis of abdominal TB, include rhagic ascites and pus, while false negatives could occur in
the presence of pulmonary involvement (15% to 25% cases), early stages or underlying cirrhosis (conflicting literature) [5,
omental line or rim (thick uniformly enhancing outer rim of 51]. If the diagnosis of TBP is based solely on ascitic ADA,
thickened omentum) and low visceral fat to subcutaneous we usually perform paracentesis and cytological analyses
fat ratio (< 0.63) [36, 45, 46]. Concomitant genitourinary three times to exclude peritoneal carcinomatosis with some
involvement, especially in females with salpingitis, hydrosal- degree of certainty [4, 5].
pinx, adnexal lesions and tubo-ovarian masses, could suggest
underlying TB [47]. USG or CT-guided fine-needle aspira- Microbiological tests on ascitic fluid
tion cytology or biopsy provides an opportunity to clinch a
microbiological or cytological/histological diagnosis from The yield of acid-fast bacilli (AFB) smear testing is
extraintestinal lesions such as lymph nodes, peritoneal or extremely poor with a sensitivity of < 5%. We do not rou-
omental deposits or thickening [44, 48]. tinely perform AFB staining on ascitic fluid. Mycobacterial
Advances in MRI could further improve diagnostics and culture has a sensitivity of around 35%, but would be specific
follow-up. Magnetic resonance enterography (MRE) has been if positive [32]. There is a growing interest in PCR-based
shown to identify more strictures than barium studies [49]. diagnostics, especially those which are available as point-of-
Also, diffusion-weighted imaging and apparent diffusion care tools. Xpert Mtb/Rif has emerged as an important tool
coefficient (ADC) have been found to be helpful in assessing for microbiological diagnosis of pulmonary and some forms
response as ADC values increase in responders [49, 50]. of extrapulmonary TB. The platform provides a point-of-
care PCR-based diagnosis rapidly and safely. Additionally,
Ascitic tests it also provides information about rifampin resistance and
helps in early diagnosis of multidrug-resistant tuberculosis
Typically, tubercular ascites is a low-serum ascites albumin (MDR-TB). Two systematic reviews have assessed the utility
gradient (SAAG) and high-protein ascites with a predomi- of Xpert Mtb/Rif testing for TBP [14, 15]. In one systematic
nance of lymphocytes. However, this typical pattern may not review, 18 included studies and 1099 samples were included.
be seen in patients with chronic liver disease (high SAAG), The pooled sensitivity and specificity with respect to cul-
or malnutrition (low protein) [4]. Ascitic fluid cytology is ture as a gold standard were 64% and 97%, respectively.
routinely performed in patients with ascites and provides However, against a comprehensive reference standard (eight
an opportunity to exclude important differential diagnoses, studies, 643 samples), the pooled sensitivity and specificity

13
24 Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31

were 30% and 100%. There are reports of other PCR-based 52.8% [56]. In real life, it may sometimes be difficult to
tests, but these tests have not been validated. In a study of obtain adequate tissue because samples are needed for mul-
multiplex PCR (16SrRNA, IS6110, and devR-based prim- tiple tests (histopathology, culture, polymerase chain reac-
ers), the sensitivity was 75.7% with a specificity of 100%, tion/Xpert Mtb/Rif). A recent standard treatment workflow
but the validation of this approach is required [52]. by Indian Council of Medical Research (ICMR) suggested
that at least six biopsy pieces be obtained in sterile saline for
microbiological testing [54].
Morphological or visual appearance Microbiological positivity is the gold standard for the
diagnosis of GITB. The tissue obtained by surgery/colo-
Laparoscopy noscopy should be subjected to microbiological testing.
Although AFB staining is extremely important for pulmonary
Laparoscopy with peritoneal biopsy is a tool for rapid and samples, the positivity rates are extremely low for GITB. TB
accurate diagnosis of TBP. However, it is not commonly culture provides an important tool for the diagnosis of GITB
performed due to the invasive nature of the procedure, com- and also helps in testing for drug sensitivity in appropriate
plications such as bowel perforation and bleeding and lack clinical settings. Culture positivity has been reported to vary
of availability at peripheral setups. Classical patterns of from 7% to 79%, but is usually less than 50% [35]. MGIT
TBP on laparoscopy were described as peritoneal thicken- may be preferable to the traditional culture in the Lowenstein-
ing with yellow-white tubercles and the visual appearance Jensen medium [57]. This is because MGIT provides early
was reported to be diagnostic in a majority of cases [53]. detection and possibly has higher sensitivity.
Peritoneoscopy also provides an opportunity to sample In a systematic review of nine studies with 709 patients, the
the lesions and achieve a histological or microbiological sensitivity and specificity of IS6110-based PCR testing for the
confirmation. diagnosis of GITB were 47% and 95%, respectively [22]. In a
systematic review of five studies with 460 samples, the pooled
sensitivity and specificity of Xpert Mtb/Rif for the diagnosis of
Colonoscopy findings GITB using intestinal tissue were 23% and 100%, respectively,
as compared to a composite reference standard [15]. No reports
Because the ileocecal region and the ascending colon are the are available on the utility of newer PCR-based diagnostics
most frequent sites of involvement in GITB, ileocolonos- such as Xpert Ultra. One study has reported that the sensitiv-
copy is an important tool in diagnosing GITB. The proce- ity and specificity of Truenat MTB Plus (TruPlus) were 70%
dure provides an opportunity to evaluate the morphological and 100%, respectively [58]. As with peritoneal tuberculosis,
pattern of involvement and also obtain tissue for histologi- multiplex PCR based on multiple probes has been found to be
cal and microbiological evaluation. Although none of the sensitive and specific but needs validation [52, 58].
endoscopic findings are specific to GITB, some findings
are considered to be suggestive of GITB. In a systematic Histopathology
review published in abstract form by Du et al., 12 studies
with 1134 patients were included. While the presence of Histopathological evaluation plays an important role in the
transverse ulcers (sensitivity: 43% and specificity: 88%) and diagnosis of GITB and its discrimination from CD. The find-
a patulous ileocecal valve (sensitivity: 38% and specificity: ings on histology may include the presence of epithelioid
91%) was suggestive of GITB, the presence of aphthous cell granulomas, caseating necrosis, confluent granulomas
ulcers, cobblestone appearance, skip lesions and longitu- and changes of chronic inflammation [59]. Since GITB is
dinal ulcers was more suggestive of CD [54]. In a Bayesian usually a paucibacillary disease, clinicians may have to rely
meta-analysis that studied multiple parameters, these find- on histology for the diagnosis. In a systematic review of
ings were confirmed [55]. diagnostic accuracy, including 10 studies with 692 patients,
three histological features were found to have high specific-
Intestinal biopsy‑based tests ity for the diagnosis of GITB (as compared to CD). These
included caseating granuloma (sensitivity of 0.21 and speci-
Microbiology ficity of 100%), confluent granuloma (sensitivity of 0.38 and
specificity of 99%) and ulcers lined by histiocytes (sensitiv-
The diagnosis of ITB often relies on testing of the intesti- ity of 0.41 and specificity of 94%) [23]. These three features,
nal tissue. It is important to obtain adequate tissue samples although specific, have poor sensitivity and none of these is
for evaluation. A study suggested that taking eight pieces positive in half of the cases of GITB. Therefore, a conclusive
(instead of four) increased the positivity of Mycobacterium histopathological diagnosis is only possible in some cases.
Growth Indicator Tube-960 (MGIT) culture from 40% to Further, granulomas in GITB are predominantly submucosal

13
Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31 25

in location and are dense (> 5/high-power field [HPF]) and and pericardial TB. A meta-analysis evaluating the use of
relatively larger in size (macrogranuloma, > 200 µm) [55, steroids for PTB showed adjunctive steroids used with ATT
60]. The presence of lymph nodal granuloma in the absence were more effective when compared with using ATT for the
of intestinal inflammation is highly specific for TB. Granu- prevention of composite endpoint, symptomatic stricture and
lomas in CD are commonly smaller in size (microgranu- intestinal obstruction. However, the study noted that due to
loma, < 200 µm), discrete and fewer in number [60]. the poor quality of studies involved in the review and meta-
Table 3 provides the sensitivity of various modalities for analysis, the findings could not be generalized and there is a
the diagnosis of TBP and GITB. need for prospective well-controlled trials [16]. Therefore,
routine use of steroids is not recommended for abdominal
TB.
Treatment and response
Endoscopic interventions and surgery
Antitubercular therapy and duration
Stricturing disease is common in GITB and may occur in a
A Cochrane meta-analysis of three RCTs (328 participants) quarter of patients with GITB. While three-fourths of the
compared the six-month regimen with the nine-month regi- stricturing disease have a clinical response to ATT (unpub-
men of ATT to treat adults with abdominal TB. The relapse lished meta-analysis), many patients continue to be sympto-
rates were not higher in patients who received a shorter dura- matic even after ATT and require additional therapy (endo-
tion (six months) of therapy with isoniazid, rifampin, pyrazi- scopic dilatation or surgery) [61, 62]. Further, many patients
namide and ethambutol. Also, the clinical cure rates at the may present directly with intestinal obstruction and may
end of therapy were similar between the two groups [17]. need emergency surgery. In patients with ongoing symp-
Out of the three RCTs, only one included participants with toms in spite of ATT, endoscopic dilatation could be done
PTB. The participants in this trial were treated thrice weekly for endoscopically reachable strictures [62]. The endpoint
under a directly observed therapy program and followed up for of dilatation is not well defined, but symptom resolution is
12 months after completing ATT. No difference was reported aimed for. Usually, endoscopic dilatation is safe and effica-
between the six-month and nine-month regimen on per-proto- cious [62]. Indirect evidence from CD-related strictures sug-
col analysis (91.5% vs. 90.8%) or intent-to-treat analysis (75% gests a dilatation of 15–18 mm and passability of a standard
vs. 75.85%) [59]. Additional observational data suggest that colonoscope as additional criteria, but these have not been
six months of therapy is sufficient in most cases, although the validated in the setting of GITB [35]. Strictures in GITB
guidelines provide clinicians with an option of extending the are likely to behave differently from CD-related strictures
duration of therapy on a case-to-case basis [28]. because effective ATT would ensure the absence of ongoing
inflammation and therefore, the effects of dilatation may be
Role of steroids more lasting. Apart from through-the-scope (TTS) balloon
dilatation, for which data is available, there is no data for
Corticosteroids could potentially offer benefits when used additional modalities, including endoscopic stricturotomy,
as an adjunctive by reducing inflammation and preventing in GITB. Certain patients may require surgery, especially if
post-inflammatory fibrosis and are used in TB meningitis they are suffering from unrelenting intestinal obstruction,

Table 3  Sensitivity of Test Gastrointestinal tuberculosis Peritoneal tuberculosis

various tests for diagnosis of
gastrointestinal tuberculosis and AFB stain < 5% 3%
peritoneal tuberculosis
Xpert MTB/RIF 23% against a composite reference 30% against a composite reference
60% against culture
Adenosine deaminase Not applicable 93% to 100%
MTB PCR (IS6110) 47% 25% to 80% (usually around 50%)
Multiplex PCR 75% 89%
Cultures 7% to 80% (usually around 40) 35%
Histology NA
Confluent granuloma 38%
Caseation 21%
Ulcers lined by epithelioid 41%
histiocytes

AFB acid fast bacilli, MTB PCR Mycobacterium tuberculosis-polymerase chain reaction test, NA not applicable

13
26 Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31

recurrent abdominal pain or episodes of obstruction, massive Drug resistance

gastrointestinal bleeding or perforation peritonitis [63, 64].
Drug resistance is an important global concern with regard to
Response assessment TB. Drug resistance in abdominal TB is likely to be similar to
overall drug resistance in a particular region. Several studies
As mentioned earlier, the diagnosis of abdominal TB is dif- have reported drug resistance in GITB. In a culture-based study
ficult because of the low positivity of microbiological tests. from Western India, of the 43 patients with culture positivity,
Certain diseases may closely mimic abdominal TB—like peri- 23% had resistance to at least one first-line drug, while 14%
toneal carcinomatosis (mimics TBP) and CD (mimics GITB). had MDR-TB [77]. In a recent study on 30 ileocecal biopsies
Unfortunately, clinical responses to ATT can be misleading. with positive TB cultures, four each (13.3%) had isoniazid and
There is evidence to suggest that antimycobacterial therapy MDR, while two each (6.7%) were pre-extremely drug resistant
may result in clinical response in patients with CD [65, 66]. (XDR) and mono-fluoroquinolone resistant [78]. This finding
On the other hand, patients with GITB may continue to be needs to be interpreted with caution, as only a subset of patients
symptomatic due to underlying strictures. In cases where had positive cultures. Another study from the Western India
the initial diagnosis was uncertain, it is important to look for reported an MDR-TB rate of 11.5% [79]. The rates were much
objective evidence of response to ATT. Trial of ATT (variably lower (4%) in an Xpert Mtb/Rif-based study from Northern
termed as a diagnostic trial or a therapeutic trial) is the stand- India [80].
ard strategy to discriminate GITB and CD in TB-endemic Figure 2 summarizes an evidence-based approach to the
areas. For ITB, healing of ulcers with ATT is an objective evaluation and treatment of abdominal TB and is largely
criterion of response to ATT, while the resolution of ascites is consistent with ICMR standard treatment workflow.
the criterion in PTB [66, 67]. Because the delay in diagnosis
of CD due to ATT could potentially result in worse outcomes,
it is important to make the discrimination between the two
entities as early as possible. In a study of > 700 patients with Additional types of abdominal tuberculosis
CD, ATT was responsible for a diagnostic delay and could
result in clinical response even in CD. A diagnostic delay was Gastroduodenal tuberculosis
associated with more stenosing complications and the need for
surgery [68]. In another report, the progression of an inflam- The stomach is an uncommon site of TB but could pre-
matory pattern of CD to stricturing disease was reported with sent with symptoms such as epigastric pain, gastric outlet
the administration of ATT [69]. obstruction, hematemesis and failure to thrive (in children)
In this regard, multiple studies have suggested that along with constitutional symptoms. Endoscopic findings
objective evidence of mucosal healing can be seen early can include ulcers, mass or growth, nodularity, stricture or
(at two months) and could potentially reduce unduly pro- extrinsic compression [81]. Deeper biopsies using endo-
longed ATT [70, 71]. Also, a two-month colonoscopy to scopic mucosal resection or well technique (biopsy upon
look for early mucosal response provides an opportunity biopsy) may improve the diagnostic yield. Endoscopic ultra-
to identify and address the causes for lack of response, sound could help in targeting extraluminal lesions, including
including alternative diagnosis or drug resistance [70]. the lymph nodes. Gastric cancer is an important differential
In patients who are not willing to undergo colonoscopy, diagnosis. Endoscopic balloon dilatation (usually success-
biomarkers such as fecal calprotectin can be used as a ful) or surgery may be needed if obstructive symptoms do
surrogate marker of mucosal response [72, 73]. Fecal cal- not improve with ATT [81, 82].
protectin appears to be a better biomarker as compared
to serum C-reactive protein (CRP) measurements [72]. Pancreatic tuberculosis
The utility of these non-invasive parameters for follow-up
was demonstrated in a study performed during the coro- Pancreatic TB is another uncommon manifestation of TB
navirus disease - 19 (COVID-19) pandemic, when physi- that closely mimics pancreatic cancer. It usually presents
cal follow-ups were difficult and endoscopic procedures as a solid pancreatic mass (especially in the head of the
restricted [74]. pancreas), but may present as a cystic lesion or peripan-
Follow-up in other types of abdominal TB utilizes a com- creatic lymphadenopathy [83]. The clinical manifestations
bination of imaging and clinical follow-up [75]. Tubercular may include abdominal pain, jaundice and constitutional
abdominal cocoon could respond clinically to ATT, but a symptoms. Although it was usually diagnosed as a histo-
significant number will require surgery [76]. logical surprise for surgical resection of presumed pancreatic

13
Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31 27

Fig. 2  Algorithmic approach to diagnosis and evaluation of abdominal tuberculosis. HIV human immunodeficiency virus, ADA adenosine deam-
inase, CT computed tomography, PCR polymerase chain reaction, CD Crohn’s disease, TB tuberculosis

13
28 Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31

cancer, the advent of endoscopic ultrasound has resulted in cells > 31.3% in peripheral blood (to discriminate GITB
more cases being diagnosed as part of the evaluation of and CD), proteomic-based approaches, and nuclear medi-
pancreatic masses. It should always be considered in the cine, including fluorodeoxyglucose (FDG)-positron emis-
differential diagnosis of pancreatic masses in TB-endemic sion tomography (PET) CT [90–94]. While some of these
regions [84]. approaches have not been helpful, others, like the use of
T-regulatory cells, appear promising but need validation
at additional centers and assessment of feasibility before
Hepatobiliary tuberculosis routine application. Some reports have evaluated the use of
artificial intelligence, including its application on textual
Hepatic involvement in TB could occur as part of localized data (reports) or radiological images with fair discrimina-
disease (mass or tuberculoma or abscess) or systemic disease tive power, but these are, as yet, beyond the realm of clinical
(granulomatous hepatitis). The clinical presentation could be use. Multiparameter models have also been assessed, and a
due to systemic disease (fever, weight loss, organomegaly) model based on a Bayesian network meta-analysis seems to
or local disease (abdominal pain or tenderness) [85]. Liver perform better than other models and is available online to
biopsy has an excellent yield in granulomatous hepatitis, discriminate ITB and CD [55].
while image-guided fine-needle aspiration biopsy can help Abdominal TB continues to be an important concern in
achieve diagnosis in localized forms of the disease. Micro- many regions of the globe. The rising incidence of IBD in
biological yield from hepatic lesions is usually good [86]. these regions poses additional challenges for diagnosis and
Hepatic tuberculosis should be considered in non-resolving management. Low diagnostic yield of microbiological tests
liver abscess, hepatic space-occupying lesion(s) or infiltra- in ascitic fluid and tissue biopsies is a concern and neces-
tive pattern of liver function tests with predominant alkaline sitates a diagnostic trial of ATT in a subset of patients. Close
phosphatase elevations. Biliary involvement could also occur follow-up with objective evidence of response to ATT is
in the form of biliary strictures or lymph nodal compression essential in such cases.
of the biliary system [85]. Gallbladder TB is very rare and
mimics gallbladder cancer [86, 87].
Declarations
Conflict of interest DKJ, MMP, and VS declare no competing inter-
ests.
Future aspects
Disclaimer The authors are solely responsible for the data and the con-
The low sensitivity of microbiological tests is the Achilles heel tents of the paper. In no way is the Honorary Editor-in-Chief, Editorial
Board members, the Indian Society of Gastroenterology or printer/pub-
in the diagnosis of abdominal TB. It is unclear if increasing lishers responsible for the results/findings and content of this article.
the amount of tissue or ascitic fluid could increase the myco-
bacterial yield on culture or other microbiological tests. It also
remains to be seen if improvement in PCR-based diagnostics References
could improve the diagnostic yield. The use of newer diagnos-
tic tests such as Xpert Ultra, although reported to have better 1. Barberis I, Bragazzi NL, Galluzzo L, Martini M. The history of
sensitivity for the diagnosis of tuberculous pleural effusions, tuberculosis: from the first historical records to the isolation of
has not been reported for TBP [88]. Xpert XDR could also Koch’s bacillus. J Prev Med Hyg. 2017;58:E9–12.
2. Global tuberculosis report 2022. Geneva: World Health Organi-
help in the detection of drug resistance to second-line thera- zation; 2022. Licence: CC BY-NC-SA 3.0 IGO. Accessed from
pies but has not been tested in abdominal TB. Multiplex PCRs https://​www.​who.​int/​teams/​global-​tuber​culos​is-​progr​amme/​tb-​
which have shown to be of excellent sensitivity and specificity repor​ts/​global-​tuber​culos​is-​report-​2022 1st February, 2022
3. Sharma V. Differentiating intestinal tuberculosis and Crohn
need to be validated in multicenter prospective studies before
disease: quo vadis. Expert Rev Gastroenterol Hepatol.
the clinical application can become routine. 2020;14:647–50.
The evaluation of small-bowel TB has been difficult 4. Sharma V, Debi U, Mandavdhare HS, Prasad KK. Tuberculosis
because of the difficulty in accessing this region. Capsule and other mycobacterial infections of the abdomen. In: Kuipers
EJ, editors. Encyclopedia of Gastroenterology. 2nd ed. Academic
endoscopy could help visualize the lesions but is limited by
Press; 2020. p. 646–659.
the inability to sample the lesions [89]. Advancements in 5. Dahale AS, Dalal A. Evidence-based commentary: ascitic adeno-
small-bowel endoscopy including motorized spiral endos- sine deaminase in the diagnosis of peritoneal tuberculosis. J Gas-
copy could help improve the diagnosis of small-bowel TB. trointest Infect. 2022;12:57–60.
6. Dawra S, Sharma V, Sharma A, Dutta U, Chawla MPS, Kochhar
The other interest has been in using biomarkers including
R. Survey of practices in managing abdominal TB. Int J Tuberc
tumor markers (to discriminate TBP from peritoneal carcino- Lung Dis. 2020;24:1109–11.
matosis), cytokines, CD4 + CD25 + FOXP3 + T-regulatory

13
Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31 29

7. Chen J, Liu S, Tang Y, et al. Diagnostic performance of CT for 25. Nath P. Epidemiology of gastrointestinal tuberculosis. In: Sharma
differentiating peritoneal tuberculosis from peritoneal carcino- V, editor. Tuberculosis of the gastrointestinal system. Singapore:
matosis: a systematic review and meta-analysis. Clin Radiol. Springer; 2022. https://​doi.​org/​10.​1007/​978-​981-​16-​9053-2_2.
2020;75:396.e7-396. 26. Cherian JJ, Lobo I, Sukhlecha A, et al. Treatment outcome of
8. Su SB, Qin SY, Guo XY, Luo W, Jiang HX. Assessment by meta- extrapulmonary tuberculosis under Revised National Tuberculosis
analysis of interferon-gamma for the diagnosis of tuberculous Control Programme. Indian J Tuberc. 2017;64:104–8.
peritonitis. World J Gastroenterol. 2013;19:1645–51. 27. Ade S, Harries AD, Trébucq A, et al. National profile and treat-
9. Riquelme A, Calvo M, Salech F, et al. Value of adenosine deami- ment outcomes of patients with extrapulmonary tuberculosis in
nase (ADA) in ascitic fluid for the diagnosis of tuberculous peri- Bénin. PLoS One. 2014;9:e95603.
tonitis: a meta-analysis. J Clin Gastroenterol. 2006;40:705–10. 28. Mandavdhare HS, Singh H, Dutta U, Sharma V. A real-world
10. Shen YC, Wang T, Chen L, et al. Diagnostic accuracy of adeno- experience with 6 months of antitubercular therapy in abdominal
sine deaminase for tuberculous peritonitis: a meta-analysis. Arch tuberculosis. JGH Open. 2019;3:201–5.
Med Sci. 2013;9:601–7. 29. Singh A, Sahu MK, Panigrahi M, et al. Abdominal tuberculosis in
11. Tao L, Ning HJ, Nie HM, Guo XY, Qin SY, Jiang HX. Diagnostic Indians: Still very pertinent. J Clin Tuberc Other Mycobact Dis.
value of adenosine deaminase in ascites for tuberculosis ascites: a 2019;15:100097.
meta-analysis. Diagn Microbiol Infect Dis. 2014;79:102–7. 30. Jeong YJ, Kang JY, Kim HW, et al. Association of underlying
12. Zhou R, Qiu X, Ying J, et al. Diagnostic performance of adenosine comorbidities and sites of tuberculosis: an analysis using surveil-
deaminase for abdominal tuberculosis: A systematic review and lance data. BMC Pulm Med. 2022;22:417.
meta-analysis. Front Public Health. 2022;10:938544. 31. Al-Zanbagi AB, Shariff MK. Gastrointestinal tuberculosis: a
13. Luo Y, Xue Y, Mao L, et al. Diagnostic value of T-SPOT.TB assay systematic review of epidemiology, presentation, diagnosis and
for tuberculous peritonitis: A meta-analysis. Front Med (Laus- treatment. Saudi J Gastroenterol. 2021;27:261–74.
anne). 2020;7:585180. 32. Sanai FM, Bzeizi KI. Systematic review: tuberculous peritonitis–
14. Kohli M, Schiller I, Dendukuri N, et al. Xpert MTB/RIF Ultra presenting features, diagnostic strategies and treatment. Aliment
and Xpert MTB/RIF assays for extrapulmonary tuberculosis and Pharmacol Ther. 2005;22:685–700.
rifampicin resistance in adults. Cochrane Database Syst Rev. 33. Shariff MK, Alzanbagi A, Sanai FM. Peritoneal tuberculosis. In:
2021;1:CD012768. Sharma V, ed. Tuberculosis of the Gastrointestinal System. Singa-
15. Sharma V, Soni H, Kumar-M P, et al. Diagnostic accuracy of pore: Springer; 2022. https://d​ oi.o​ rg/1​ 0.1​ 007/9​ 78-9​ 81-1​ 6-9​ 053-2_​
the Xpert MTB/RIF assay for abdominal tuberculosis: a sys- 10.
tematic review and meta-analysis. Expert Rev Anti Infect Ther. 34. Sharma V, Singh H, Mandavdhare HS. Tubercular abdominal
2021;19:253–65. cocoon: systematic review of an uncommon form of tuberculosis.
16. Soni H, Bellam BL, Rao RK, et al. Use of steroids for abdominal Surg Infect (Larchmt). 2017;18:736–41.
tuberculosis: a systematic review and meta-analysis. Infection. 35. Kedia S, Ahuja V. Intestinal tuberculosis: an overview. In: Sharma
2019;47:387–94. V, ed. Tuberculosis of the Gastrointestinal System. Singapore:
17. Jullien S, Jain S, Ryan H, Ahuja V. Six-month therapy for abdominal Springer; 2022. https://​doi.​org/​10.​1007/​978-​981-​16-​9053-2_6.
tuberculosis. Cochrane Database Syst Rev. 2016;11:CD012163. 36. Kedia S, Sharma R, Vuyyuru SK, et al. Addition of computed
18. Kedia S, Sharma R, Sreenivas V, et al. Accuracy of computed tomography chest increases the diagnosis rate in patients with
tomographic features in differentiating intestinal tuberculosis from suspected intestinal tuberculosis. Intest Res. 2022;20:184–91.
Crohn’s disease: a systematic review with meta-analysis. Intest 37. Dhoble P, Desai D, Abraham P. Is the rise in Crohn’s disease in
Res. 2017;15:149–59. India accompanied by a fall in intestinal tuberculosis? A single-
19. Chen W, Fan JH, Luo W, Peng P, Su SB. Effectiveness of inter- center experience Indian J Tuberc. 2021;68:210–4.
feron-gamma release assays for differentiating intestinal tubercu- 38. Logan VS. Anorectal tuberculosis. Proc R Soc Med.
losis from Crohn’s disease: a meta-analysis. World J Gastroen- 1969;62:1227–30.
terol. 2013;19:8133–40. 39. Ahamed ZR, Shah J, Agarwala R, et al. Controversies in classifica-
20. Ng SC, Hirai HW, Tsoi KK, et al. Systematic review with meta- tion of peritoneal tuberculosis and a proposal for clinico-radiolog-
analysis: accuracy of interferon-gamma releasing assay and anti- ical classification. Expert Rev Anti Infect Ther. 2019;17:547–55.
Saccharomyces cerevisiae antibody in differentiating intestinal 40. Sharma SK, Ryan H, Khaparde S, et al. Index-TB guidelines:
tuberculosis from Crohn’s disease in Asians. J Gastroenterol guidelines on extrapulmonary tuberculosis for India. Indian J Med
Hepatol. 2014;29:1664–70. Res. 2017;145:448–63.
21. Xu H, Li Y, Qian JM. A meta-analysis of the accuracy of 41. Sachdeva K, Kumar P, Kante B, et al. Interferon-gamma release
interferon-γ release assays in differentiating intestinal tubercu- assay has poor diagnostic accuracy in differentiating intestinal
losis from Crohn’s disease in Asia. Zhonghua Nei Ke Za Zhi. tuberculosis from Crohn’s disease in tuberculosis endemic areas.
2016;55:535–40. Intest Res. 2022. https://​doi.​org/​10.​5217/​ir.​2022.​00010.
22. Jin T, Fei B, Zhang Y, He X. The diagnostic value of polymer- 42. Zhao Y, Xu M, Chen L, Liu Z, Sun X. Levels of TB-IGRA may
ase chain reaction for Mycobacterium tuberculosis to distinguish help to differentiate between intestinal tuberculosis and Crohn’s
intestinal tuberculosis from Crohn’s disease: a meta-analysis. disease in patients with positive results. Therap Adv Gastroen-
Saudi J Gastroenterol. 2017;23:3–10. terol. 2020;18:1756284820922003.
23. Du J, Ma YY, Xiang H, Li YM. Confluent granulomas and ulcers 43. Goyal P, Shah J, Gupta S, Gupta P, Sharma V. Imaging in dis-
lined by epithelioid histiocytes: new ideal method for differentia- criminating intestinal tuberculosis and Crohn’s disease: past,
tion of ITB and CD? A meta analysis. PLoS One. 2014;9:e103303. present and the future. Expert Rev Gastroenterol Hepatol.
24. Van Hoving DJ, Griesel R, Meintjes G, Takwoingi Y, Maartens 2019;13:995–1007.
G, Ochodo EA. Abdominal ultrasound for diagnosing abdominal 44. Kumar S, Gupta P, Sharma V, et al. Role of ultrasound-guided
tuberculosis or disseminated tuberculosis with abdominal involve- fine-needle aspiration cytology of omentum in diagnosis of
ment in HIV-positive individuals. Cochrane Database Syst Rev. abdominal tuberculosis. Surg Infect (Larchmt). 2019;20:91–4.
2019;9:CD012777. 45. Ramanan RV, Venu V. Differentiation of peritoneal tuberculosis
from peritoneal carcinomatosis by the omental rim sign. A new

13
30 Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31

sign on contrast enhanced multidetector computed tomography. 64. Limpin ET, Lopez MPJ, Maglangit SACA, et al. Surgical man-
Eur J Radiol. 2019;113:124–34. agement of patients with GI tuberculosis. Dis Colon Rectum.
46. Yadav DP, Madhusudhan KS, Kedia S, et al. Development and 2023;66:106–12.
validation of visceral fat quantification as a surrogate marker for 65. Agrawal G, Clancy A, Sharma R, Huynh R, Ramrakha S, Borody
differentiation of Crohn’s disease and intestinal tuberculosis. J T. Targeted combination antibiotic therapy induces remission in
Gastroenterol Hepatol. 2017;32:420–6. treatment-naïve Crohn’s disease: a case series. Microorganisms.
47. Naeem M, Zulfiqar M, Siddiqui MA, et al. Menias CO. Imag- 2020;8:371.
ing manifestations of genitourinary tuberculosis Radiographics. 66. Pratap Mouli V, Munot K, Ananthakrishnan A, et al. Endoscopic
2021;41:1123–43. and clinical responses to anti-tubercular therapy can differentiate
48. Gupta P, Kumar S, Sharma V, et al. Common and uncommon intestinal tuberculosis from Crohn’s disease. Aliment Pharmacol
imaging features of abdominal tuberculosis. J Med Imaging Ther. 2017;45:27–36.
Radiat Oncol. 2019;63:329–39. 67. Sharma V, Mandavdhare HS, Lamoria S, Singh H, Kumar A.
49. Krishna S, Kalra N, Singh P, et al. Small-bowel tuberculosis: a Serial C-reactive protein measurements in patients treated for sus-
comparative study of MR enterography and small-bowel follow- pected abdominal tuberculosis. Dig Liver Dis. 2018;50:559–62.
through. AJR Am J Roentgenol. 2016;207:571–7. 68. Banerjee R, Pal P, Girish BG, Reddy DN. Risk factors for diagnos-
50. Mathur P, Sharma R, Kandasamy D, Kedia S, Gamanagatti S, tic delay in Crohn's disease and their impact on long-term com-
Ahuja V. Can ADC be used as a surrogate marker of response plications: how do they differ in a tuberculosis endemic region?
to therapy in intestinal tuberculosis? Abdom Radiol (NY). Aliment Pharmacol Ther. 2018;47:1367–74.
2019;44:3006–18. 69. Gupta A, Pratap Mouli V, Mohta S, et al. Antitubercular ther-
51. Dahale AS, Puri AS, Sachdeva S, et al. Reappraisal of the role of apy given to differentiate Crohn’s disease from intestinal tuber-
ascitic fluid adenosine deaminase for the diagnosis of peritoneal culosis predisposes to stricture formation. J Crohns Colitis.
tuberculosis in cirrhosis. Korean J Gastroenterol. 2021;78:168–76. 2020;14:1611–8.
52. Hallur V, Sharma M, Sethi S, et al. Development and evaluation 70. Sharma V, Mandavdhare HS, Dutta U. Letter: mucosal response in
of multiplex PCR in rapid diagnosis of abdominal tuberculosis. discriminating intestinal tuberculosis from Crohn’s disease-when
Diagn Microbiol Infect Dis. 2013;76:51–5. to look for it? Aliment Pharmacol Ther. 2018;47:859–60.
53. Bhargava DK, Shriniwas, Chopra P, Nijhawan S, Dasarathy S, 71. Park YS, Jun DW, Kim SH, et al. Colonoscopy evaluation after
Kushwaha AK. Peritoneal tuberculosis: laparoscopic patterns and short-term anti-tuberculosis treatment in nonspecific ulcers on the
its diagnostic accuracy. Am J Gastroenterol. 1992;87:109–12. ileocecal area. World J Gastroenterol. 2008;14:5051–8.
54. Du J, Ma YY, Ma PP, Chen CX. Endoscopic features for differen- 72. Sharma V, Verma S, Kumar-M P, et al. Serial measurements of
tiation between Crohn’s disease (CD) and intestinal tuberculosis faecal calprotectin may discriminate intestinal tuberculosis and
(ITB): a meta-analysis. J Dig Dis 2014; 15 Suppl. 1: 147. Crohn’s disease in patients started on antitubercular therapy. Eur
55. Limsrivilai J, Shreiner AB, Pongpaibul A, et al. Meta-analytic J Gastroenterol Hepatol. 2021;33:334–8.
Bayesian model for differentiating intestinal tuberculosis from 73. Jo HH, Kim EY, Jung JT, et al. Value of fecal calprotectin meas-
Crohn’s disease. Am J Gastroenterol. 2017;112:415–27. urement during the initial period of therapeutic anti-tubercular
56. Mehta V, Desai D, Abraham P, et al. Do additional colonoscopic trial. Clin Endosc. 2022;55:256–62.
biopsies increase the yield of Mycobacterium tuberculosis cul- 74. Shukla J, Jena A, Singh H, Mandavdhare HS, Dutta U, Sharma
ture in suspected ileo-colonic tuberculosis? Indian J Gastroenterol. V. Management of gastrointestinal tuberculosis during COVID
2018;37:226–30. pandemic: lessons for posterity. Dig Liver Dis. 2021;53:394–6.
57. ADULT ABDOMINAL TUBERCULOSIS. Standard Treat- 75. Sharma V, Singh H, Mandavdhare HS. Defining ‘satisfactory
ment Workflows of India, 2022, Special Edition on Paediatric response’ to therapy in abdominal tuberculosis: a work in pro-
and Extrapulmonary Tuberculosis, Indian Council of Medical gress. Infect Disord Drug Targets. 2020;20:111–4.
Research, Department of Health Research, Ministry of Health 76. Sharma V, Mandavdhare HS, Rana SS, Singh H, Kumar A, Gupta
and Family Welfare, Government of India Accessed https://​stw.​ R. Role of conservative management in tubercular abdominal
icmr.​org.​in/​images/​Adult_​Extr_​Tuber​culos​is/1_​Adult_​Abdom​ cocoon: a case series. Infection. 2017;45:601–6.
inal_​TB_​18032​022.​pdf on 1st February 2023 77. Sonambekar A, Desai D, Abraham P, et al. Drug resist-
58. Sharma K, Sharma M, Sharma V, et al. Evaluating diagnostic per- ance in intestinal tuberculosis: a reason to worry? JGH Open.
formance of Truenat MTB Plus for gastrointestinal tuberculosis. 2017;1:22–4.
J Gastroenterol Hepatol. 2022;37(8):1571–8. 78. Sharma K, Sharma M, Sharma V, et al. MTBDRplus and MTB-
59. Makharia GK, Ghoshal UC, Ramakrishna BS, et al. Intermittent DRsl for simultaneous diagnosis of gastrointestinal tuberculosis
directly observed therapy for abdominal tuberculosis: a multi- and detection of first-line and second-line drug resistance. J Gas-
center randomized controlled trial comparing 6 months versus 9 troenterol Hepatol. 2023. https://​doi.​org/​10.​1111/​jgh.​16124.
months of therapy. Clin Infect Dis. 2015;61:750–7. 79. Udgirkar S, Jain S, Pawar S, Chandnani S, Contractor Q, Rathi P.
60. Pulimood AB, Ramakrishna BS, Kurian G, et al. Endoscopic Clinical profile, drug resistance pattern and treatment outcomes
mucosal biopsies are useful in distinguishing granulomatous coli- of abdominal tuberculosis patients in Western India. Arq Gastro-
tis due to Crohn’s disease from tuberculosis. Gut. 1999;45:537–41. enterol. 2019;56:178–83.
61. Aggarwal P, Kedia S, Sharma R, et al. Tubercular intestinal stric- 80. Bellam BL, Mandavdhare HS, Sharma K, et al. Utility of tis-
tures show a poor response to anti-tuberculous therapy. Dig Dis sue Xpert-Mtb/Rif for the diagnosis of intestinal tuberculo-
Sci. 2017;62:2847–56. sis in patients with ileocolonic ulcers. Ther Adv Infect Dis.
62. Kumar P, Jena A, Birda CL, et al. Safety and efficacy of non- 2019;6:2049936119863939.
fluoroscopic endoscopic dilatation of gastrointestinal tuberculosis 81. Shah J, Maity P, Kumar-M P, Jena A, Gupta P, Sharma V.
related strictures. BMC Gastroenterol. 2022;22:60. Gastroduodenal tuberculosis: a case series and a management
63. Singh H, Krishnamurthy G, Rajendran J, et al. Surgery for abdom- focused systematic review. Expert Rev Gastroenterol Hepatol.
inal tuberculosis in the present era: experience from a tertiary-care 2021;15:81–90.
center. Surg Infect (Larchmt). 2018;19:640–5. 82. Dalal A, Puri AS, Sachdeva S, Sakuja P. Nonsurgical manage-
ment of gastroduodenal tuberculosis: Nine-year experience from
a tertiary referral center. Endosc Int Open. 2019;7:E1248-52.

13
Indian Journal of Gastroenterology (January–February 2023) 42(1):17–31 31

83. Sharma V, Rana SS, Kumar A, Bhasin DK. Pancreatic tuberculo- 91. Gupta A, Sharma K, Sharma V, et al. Comparative evaluation
sis. J Gastroenterol Hepatol. 2016;31:310–8. of interleukin-10, transforming growth factor-β, and interleu-
84. Panic N, Maetzel H, Bulajic M, Radovanovic M, Löhr JM. Pan- kin-17 in gastrointestinal tuberculosis and Crohn’s disease. Int
creatic tuberculosis: a systematic review of symptoms, diagnosis J Mycobacteriol. 2022;11:384–8.
and treatment. United European Gastroenterol J. 2020;8:396–402. 92. Rampal R, Kedia S, Wari MN, et al. Prospective validation of
85. Amarapurkar DN, Patel ND, Amarapurkar AD. Hepatobil- CD4+CD25+FOXP3+ T-regulatory cells as an immunological
iary tuberculosis in western India. Indian J Pathol Microbiol. marker to differentiate intestinal tuberculosis from Crohn’s dis-
2008;51:175–81. ease. Intest Res. 2021;19:232–8.
86. Hickey AJ, Gounder L, Moosa MY, Drain PK. A systematic review 93. Rukmangadachar LA, Makharia GK, Mishra A, et al. Pro-
of hepatic tuberculosis with considerations in human immunode- teome analysis of the macroscopically affected colonic mucosa
ficiency virus co-infection. BMC Infect Dis. 2015;15:209. of Crohn's disease and intestinal tuberculosis. Sci Rep.
87. Krishnamurthy G, Singh H, Rajendran J, et al. Gallbladder tuber- 2016;6:23162.
culosis camouflaging as gallbladder cancer - case series and 94. Singh AK, Kumar R, Gupta P, et al. FDG-PET-CT enterography
review focusing on treatment. Ther Adv Infect Dis. 2016;3:152–7. helps determine clinical significance of suspected ileocecal thick-
88. Aggarwal AN, Agarwal R, Dhooria S, Prasad KT, Sehgal IS, ening: a prospective study. Dig Dis Sci. 2021;66:1620–30.
Muthu V. Xpert MTB/RIF Ultra versus Xpert MTB/RIF for diag-
nosis of tuberculous pleural effusion: a systematic review and Publisher's note Springer Nature remains neutral with regard to
comparative meta-analysis. PLoS One. 2022;17: e0268483. jurisdictional claims in published maps and institutional affiliations.
89. Rana SS, Sharma V, Sharma R, Nada R, Gupta R, Bhasin DK.
Capsule endoscopy in small bowel Crohn’s disease and tubercu- Springer Nature or its licensor (e.g. a society or other partner) holds
losis. Trop Doct. 2017;47:113–8. exclusive rights to this article under a publishing agreement with the
90. Jain T, Ram S, Kumar H, Saroch A, Sharma V, Singh H. Ascitic author(s) or other rightsholder(s); author self-archiving of the accepted
and serum levels of tumor biomarkers (CA 72–4, CA 19–9, manuscript version of this article is solely governed by the terms of
CEA and CA 125) in discrimination of cause of ascites: a pro- such publishing agreement and applicable law.
spective study. Arq Gastroenterol. 2022;59:198–203.

13