Você está na página 1de 9

Clinics and Research in Hepathology and Gastroenterology (2011) 35, 6069

SEMINAR

Peritoneal tuberculosis
La tuberculose pritonale
A. Guirat a,, M. Koubaa b, R. Mzali a, B. Abid a, S. Ellouz c, N.
Affes a, M. Ben Jemaa b, F. Frikha a, M. Ben Amar a, M.I. Beyrouti
a

Service de chirurgie gnrale, CHU Habib Bourguiba, 3029 Sfax, Tunisia


Service des maladies infectieuses, CHU Hdi Chaker, 3029 Sfax, Tunisia
c
Service danatomie pathologique, CHU Habib Bourguiba, 3029 Sfax, Tunisia
b

Available online 6 January 2011

Summary
The peritoneum is one of the locations outside the most common pulmonary
tuber- culosis. Peritoneal tuberculosis poses a public health problem in endemic regions of
the world. The phenomenon of migration, the increased use of immunosuppressive therapy
and the epi- demic of AIDS have contributed to a resurgence of this disease in regions where
it was previously controlled. The aim of this review is to expose the clinical, biologic end
radiologic futures of the peritoneal tuberculosis and to present the methods of diagnosis and
treatment. The diag- nosis of this disease is difcult and still remains a challenge because of
its insidious nature, the variability of presentation and limitations of available diagnostic
tests. The disease usually presents a picture of lymphocytic exudative ascites. There are
many complementary tests with variable sensitivities and specicities to conrm the diagnosis
of peritoneal tuberculosis. Isola- tion of mycobacteria by culture of ascitic uid or histological
examination of peritoneal biopsy ideally performed by laparoscopy remains the investigation
of choice. The role of PCR, ascitic adenosine deaminase, interferon gamma and the
radiometric BACTEC system can improve the diagnostic yield. An antituberculous treatment
with group 1 of the WHO for 6 months is sufcient in most cases.
2010 Elsevier Masson SAS. All rights reserved.
Rsum Le pritoine est lune des localisations extrapulmonaire les plus frquentes de
la tuberculose. La tuberculose pritonale pose un problme de sant publique dans certaines rgions endmiques du monde. Le phnomne de migration, lutilisation plus
frquente dimmunosuppresseurs et lpidmie du sida ont contribu une rapparition de
cette maladie dans les rgions o elle tait prcdemment contrle. Le but de cette mise
au point est de dgager les particularits cliniques, biologiques et radiologiques de la
tuberculose pritonale et de prsenter les moyens de diagnostic et les modalits
thrapeutiques. Le diagnostic de cette maladie est difcile et demeure un d en raison de
sa nature insidieuse, de la variabilit

Corresponding author. BP 12 bureau de poste, Jadida, 3027 Sfax, Tunisia.


E-mail address: ahmedguirat@yahoo.fr (A. Guirat).

0399-8320/$ see front matter 2010 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.gcb.2010.07.023

Peritoneal
61
tuberculosis

A. Guirat et61al.

de sa prsentation et des limites des examens de diagnostic disponibles. Cette pathologie se


manifeste habituellement par un tableau dascite exsudative prdominance lymphocytaire.
Il existe une panoplie dexamens complmentaires dont la sensibilit et la spcicit sont
variables pour conrmer le diagnostic de tuberculose pritonale. Lisolement dune
mycobac- trie par culture du liquide dascite ou de la biopsie pritonale idalement
ralise par voie laparoscopique reprsente lexamen de choix. Le rle de la PCR, ladnosine
dsaminase asc- itique, linterfron gamma et le systme radiomtrique de BACTEC
permettent damliorer le rendement diagnostique. Un traitement pendant six mois avec des
antituberculeux du groupe
1 de lOMS est sufsant dans la plupart des cas.
2010 Elsevier Masson SAS. Tous droits rservs.

Introduction
Described for the rst time in 1843 [1], peritoneal tuberculosis (PT) is due to the development of Kochs Bacillus
(KB) in the peritoneum. It is a disease that poses a public
health problem in endemic areas. The PT risks to evolve
into complications such as septicemia, acute intestinal
occlu- sion and infertility in women. The diagnosis of this
disease is a challenge; the absence of specic clinical
signs, the lack of paraclinical tests with high predictive
value and the pauci-bacillary nature of this disease make
the diagnosis of tuberculosis often based on histological
study of biopsies of peritoneum performed ideally by
laparoscopy.

Pathogenesis
Pathogenic agent
PT is caused by several species of Gram-positive bacteria
known as Mycobacterium tuberculosis complex (M. tuberculosis, M. africanum, M. bovis, M. caprae, M. microti, M.
pinnipedi. . .). They are united in the same unit for genetic
similarities reasons. These bacteria belong to the family of
Mycobacteriacea and the order of Actinomycetales: they
are characterized by a thick wall, which is rich in fat giving
them special dyeing properties and a relative resistance to
many antiseptics (soda, acid, detergent. . .). The
mycobacterium is 2 to 5 microns in length, very sensitive
to heat but resis- tant to cold and desiccation, colored red
by the fuchsin, not discolored by nitric acid or alcohol
from where the nomina- tion of acid-fast bacillus-resistant
(AFB). It grows in aerobic strictly between 35 and 37 C on
enriched media including that of Lowenstein-Jensen [2].
Atypical mycobacteria are bacteria present in the
environment and usually non-pathogenic, but which may
be
responsible
for
peritoneal
mycobacteriosis.
Mycobacteruim avium complex is the most common cause
of mycobacterio- sis in immunosuppressed patients (AIDS)
[3,4].

Modes of contamination
Mycobacteria can infect the peritoneum by different
mech- anisms:
hematogenous spread of bacilli from a primary
pulmonary site of tuberculosis infection to secondary
foci such as

the peritoneum. An immune response occurs usually and


the evolution is generally towards healing. However,
some bacteria remain quiescent in the peritoneum for
months or years. Several causes may reduce the
defenses of the organism and can cause the reactivation
of the bacilli and their multiplication [5,6];
by ruptured retroperitoneal and mesenteric lymph, which
nodes into the peritoneal cavity resulting in the
formation
of caseous foci partitioned by brosis in places adhering
to the intestinal loops. These secondary foci are formed
fol- lowing the ingestion of the bacteria crossing the
intestinal mucosa to the lymph of the peritoneal cavity
or following hematogenous spread as the rst
mechanism [79];
from lesions in adjacent organs such as intestine or the
fallopian tubes [1012];
by direct contamination of the peritoneum in patients
with chronic renal failure under peritoneal dialysis [13].

Mechanism of ascites formation


In PT, ascites usually results from peritoneal lymphatic
obstruction caused by a blocking of the peritoneal uid
reabsorption [14]. Rarely, ascites may result from portal
hypertension secondary to portal vein compression by
tuber- culous lymph nodes but the characteristics of
ascites are so different [10,15,16].

Epidemiology
In recent years, there has been an increase in the total
number of new cases of tuberculosis. The World Health
Orga- nization (WHO) has estimated that the number of
new cases of PT disease in 2007 was about 9.27 million.
This repre- sents an increase from 9.24 million cases in
2006, 8.3 million cases in 2000 and 6.6 million cases in
1990. Most of the esti- mated cases in 2007 were recorded
in Asia (55%) and Africa (31%), a small proportion of cases
in the Eastern Mediter- ranean Region (6%), the European
Region (5%) and Region of the Americas (3%). Of the 9.27
million new cases in 2007, an estimated 1.37 million (15%)
were suffering from AIDS, 79% of these patients came from
the African Region and 11% of the Southeast Asia Region
[17].
The PT remains a public health problem in endemic
areas. It represents 1% to 2% of all localizations of

Peritoneal
62
tuberculosis
tuberculosis [18] and
localizations [1921]. It

A. Guirat et62al.
31%

to

58%

of

abdominal

is associated with pulmonary tuberculosis in 3.5% of cases


[22]. As in pulmonary tuberculosis where the incidence is
higher among younger subjects between 25 and 44 years
in developing countries [23], the PT is typically observed in
young adults between the third or fourth decade [22,24].
For some authors, the average age is even younger than 28
years [25]. A frequency among women was found in most
published series. Neither pathogenic nor physiological
explanation is advanced for this frequency.

Table 1 Frequency of various clinical signs of peritoneal


tuberculosis according to the literature.

Contributing factors

Abdominal distension
Abdominal tenderness
Constipation
Diarrhea

The factors favoring overall tuberculosis disease are:


the pandemic infection of acquired immunodeciency
virus [26,27]: about 9% of new cases of PT (31% in
Africa) had occurred among AIDS patients [25];
the prolonged corcticosteroid [28];
the low socio-economic conditions of poor living, poor
hygiene and overcrowding. The literature data suggest
that the incidence of the disease is particularly high in
rural areas [22,29];
treatment with TNF (tumor necrosis factor) [3032].
Other predisposing factors for the occurrence of PT
have been reported in the literature such as:
chronic renal failure on peritoneal dialysis [33];
the BCG therapy: use in bladder instillation of Bacillus Calmette-Gurin in the treatment and prevention of
recurrent supercial bladder carcinoma has been implicated in outbreaks of PT [34];
alcoholism and alcoholic cirrhosis [35,36].

The clinical manifestations


PT is a subacute disease and its symptoms often evolve
over a period of several weeks to months. The average
consulta- tion time compared to the rst symptom varied
between 2 and 4 months [29].
This disease can be manifested by different clinical
forms. The presenting symptoms of PT are unspecic,
which makes the diagnosis a challenge. Some clinical signs
are frequently found in most series, such as ascites (most
com- mon clinical sign), abdominal pain, night sweats,
fever and abdominal distension (Table 1). In most cases,
the discovery of PT is done during the waning of ascites.
PT is diagnosed in 6.6% of the ascites cases [37]. This rate
can reach 65.5% during exploration of exudative ascites in
an endemic area [38].

The diagnostic approaches


Assessment standard blood
The standard blood test abnormalities are non-specic and
therefore they have a low diagnostic value. A mild normocytic normochromic anemia and thrombocytosis are the
main abnormalities [11,36,39]. The white blood cells rate
is generally normal, however lymphocytosis is often found

Clinical Signs
Ascites
Abdominal pains
Isolated fever

Hepatomegaly
Splenomegaly

Frequency (%)
35100
[2932,35,38,43,47,95,103]
49100
[6,27,30,31,35,38,43,47,103]
5276.1
[27,29,30,35,38,43,47,50,103,104]
62.573 [38]
47.7 [22]
7.131 [32,35,3840,44,99]
4.71.4
[22,29,30,39,40,44,99,103]
2.38.2
[22,29,36,40,44,72,94,99,105,106]
2.34.3
[22,29,36,47,94,99,105]

[39,40]. The sedimentation rate is always increased but


the values do not exceed 60 mm per hour in more than
half of the cases [39].

Analysis of ascetic uid


Macroscopic
In PT, ascites is usually yellow citrine. This characteristic is
found in 77% to 91% of the cases [39,41]. The ascites uid
may also take a cloudy appearance, chylous or hematic.
Protein content
In PT, the exudative ascites is a quasi-constant [42],
usually with protein content higher than 30 g/l.
Cellularity
The ascites uid of PT is often rich in cells (> 400 E/ml)
with lymphocyte predominance (> 60%) [22,43]. A recent
meta- analysis that included 39 English sets about PT has
found a lymphocytic predominance in 68.3% of cases [22].
The pre- dominance of neutrophils is sometimes found
especially in case of concomitant renal failure or in case of
infection of ascites with another common germ [22]. Thus,
an exudative ascites rich in lymphocytic cells is suggestive
but not specic of PT.
Cytological study
Cytological examination of the ascitic uid must always
be performed to look for neoplastic cells indicating the
existence of carcinomatous ascites, which is the main differential diagnosis of PT.
Biochemical of ascitic uid
The literature review shows that no biochemical marker is
useful for the diagnosis of PT.
LDH essay. Normally the rate of LDH in ascites is less than
half that of serum. In case of tuberculosis, a rate of LDH
greater than 90 IU/l has a sensitivity ranging between 77%
and 90%, but a low specicity (14%) [22].Serum amino-acid
gradient (SAAG)

This gradient is calculated by measuring the difference


between the concentrations of albumin in serum and in
ascites collected on the same day. It reects the characteristic of exudative ascites when its value is greater than
11 g/l and portal hypertension when its value is greater
than 11 g/l [44]. This specicity for portal hypertension
is estimated at 97% [45]. A low gradient less than 11 g/l
was found in 100% of tuberculosis patients [43,46,47],
but this sensitivity decreases in patients having an associated cirrhosis reaching values ranging from 29% to 88%
[26,43]. Besides specicity, it is low. Indeed, SAAG less
than 11 g/l is also found in patients having carcinomatosis
[43,47,48].
Glucose determination. A low concentration of glucose in
the ascites uid may be in favor of PT [22,40]. However,
data are insufcient to decide about the value of a
systematic assay.
Thus, assays of LDH, SAAG and glucose have a specicity too low to be recommended for the diagnosis of
PT.
Microbiological study
Microscopy: search of Mycobacterium tuberculosis (MT).
The microscopic diagnosis on a single smear made from the
crushed uses dye specic property of the wall of mycobacteria, which is acid-alcohol resistance. This is a quick
technique that is also accessible. Due to the paucibacillary nature of PT, microscopic examination has a
sensitivity of less than 10% [22,35,38,46,49].
Culture. The isolation of MT in the culture of ascitic uid
or in cell homogenates of biopsies conrm the diagnosis of
PT.
Solid culture medium. This test requires 10 to 50 ml of
ascites, ultra-centrifuged and then cultured on
Lowenstein- Jensen medium. This method has a sensitivity
of around 35% [35,50]. The major drawback of culture is
the time needed to obtain the result. Indeed, the average
time of bacterio- logical results is 45 days [22], which is
not conforming to the requirements and urgent diagnosis.
Liquid culture medium. The use of automated BACTEC
is a recent method allowing for more rapid detection of
MT in an average of 14 to 27 days [22,35]. It is a radiometric system allowing the acceleration of the discovery
of MT by reducing the culture time by half [51]. The
time of bacteriological results is shortened to 15 days but
requires one liter of ascites ultra-centrifuged. This method
has a large positivity rate, which varies from 66% to 83%.
A comparative study of its use, compared to LowensteinJensen medium on any type of sample showed a higher
sensitivity of detection of MT as atypical mycobacteria
(100% versus 51.7%), a better specicity (94.7% versus
78.9%) and a higher negative predictive value (100% versus 78.9%) [48]. However, this method is more expensive
than the solid culture medium, and it uses a radioactive
material, which poses a problem of manipulation. To overcome these drawbacks, other non-radioactive culture
media have been developed with the same advantages of
rapid growth [52]. These environments are used manually
(BBL tube, MGITTM , MB Redox system) or via automated
systems (BP BACTECTM , 9000MB, BACTECTM MGITTM 960 Bac
T/Alert
3DTM ).

CA-125 determination
Some authors have found that rates of CA-125 were high
in both serum and ascites in the majority of patients with
exudative ascites whatever the etiology [53,54]. Other
stud- ies have reported high levels of this marker in both
serum and ascites of patients with PT [55,56] and can
therefore be confused with ovarian carcinoma at an
advanced stage. Thus, a high rate of CA-125 should always
evoke PT in differ- ential diagnosis of ovarian carcinoma
mainly among women living in endemic areas. In addition,
the nding of a high CA125 in ascites rich in proteins in a woman may lead to a
false diagnosis of ovarian cancer and tuberculosis
underrate.
Moreover, several studies have noted that under
quadru- ple anti-tubercular drugs, after histological
conrmation of the PT diagnosis, the rate of CA-125
decreased gradually and returned to normal even after 1 to
2 months on average [22,53,54,56,57]. Thus, the
determination of CA-125 seems useful for evaluating the
therapeutic response of PT. This feature remains to be
conrmed by prospective studies.

The determination of the activity of adenosine


deaminase in ascites (ADA)
Several studies have reported an increase of ADA in some
body uids in disorders that involve an immune reaction
cell type such as tuberculosis, empyema, rheumatoid
arthritis and malignant lymphomas [58,59].
In PT, the increase of adenosine deaminase may be the
result of activation of T cells in response to mycobacterial
antigens, on the threshold of 30 U/l and in the absence of
immunosuppression or cirrhosis, this test has a sensitivity
of
96%, a specicity of 98% (Table 2), positive predictive
value of 95% and a negative predictive value of 98%
[22,60,61]. It is a cheap and reproducible test.
According to the recommendations of the Society of
Pneumology in the French language for the management of
tuberculosis in France published in 2004, the utility of the
peritoneal assay of ADA depends more on economic and
epi- demiological criteria than on its sensitivity or its
specicity: in endemic area where invasive techniques like
laparoscopy are not available or costly, the ADA is helpful.
Elsewhere, the ADA will be useful if positive, it conrms a
clinical suspicion of TB for inoperable patient [62].
In a recent meta-analysis that included four prospective
studies reporting 264 cases of ascites, Riquelme et al. have
shown that for a rate between 36 and 40 IU/L, the ADA has
a sensitivity of 100% and a specicity of 97% for the
diagnosis of PT [63]. The authors concluded that the
determination of the ADA is a rapid and accurate test for
the diagnosis of PT. In case of high rate, the introduction
of empirical antituberculosis treatment is justied while
waiting for the results of mycobacteria culture or biopsy
[63].

Immunological tests
Tuberculin skin test (TST)
The TST is testing cell memory response to mycobacterial
antigens. It explores the capacity of memory lymphocytes in the patient after stimulation with these antigens,
to secrete cytokines (IFN-gamma especially) responsible

Table 2

Sensitivity, specicity and threshold values of ADA in some studies reported in the literature.

Series
Voigt et al. [107]
Bhargava et al. [108]
Ribera et al. [109]
Soliman et al. [110]
Sathar et al. [68]
Brant et al. [111]
Burgess et al. [60]
El Abkari et al. [24]
Riquelme et al. [63]
Sharma et al. [112]
Dewivedi [113]

Year
1989
1990
1991
1994
1995
1995
2001
2006
2006
2006
2008

Number
41
87
86
50
93
44
178
123
264
119
49

Sensitivity (%)
100
100
100
94.4
96
100
94
96
100
97
100

Specicity (%)
96
97
97
100
96
92
92
98
97
94
96.6

Value (U/L)
112.6
36
40
28
30
31
30

39
58
33

for induration at the injection site, if the subject is


ill [64].
The positivity of the TST is not specic for active
tubercu- losis, but merely reects an awareness by prior
contact with the MT. Its interpretation varies depending
on the immune status of patients, history of vaccination,
contagion or pri- mary infection. Similarly, the sensitivity
of this test is low; false negatives are reported in 15% to
60% in the literature [65,66].

Gene amplication by ligase chain reaction (LCR)


It is a molecular amplication test recently introduced into
medical practice. According to Gamboa et al. [73], this
method has a sensitivity of 77.7%, specicity of 98.7% and
a negative predictive value of 95.2% for the diagnosis of
extra pulmonary tuberculosis; its diagnostic effectiveness
is then signicantly higher than that of PCR. Unfortunately,
this test is still very expensive and therefore difcult to
insert in practice especially in endemic countries.

Determination of gamma interferon


It is a quantitative test that assesses the cell-mediated
immune response against MT. It is based on the fact that
T cells secreting gamma interferon were necessarily aware
in advance [22]. This assay has a sensitivity of 89% for the
diagnosis of latent forms of PT [67]. Some authors
advocate the measurement of gamma interferon coupled
with that of adenosine deaminase, thereby increasing
their respec- tive sensitivity and specicity for the positive
diagnosis of tuberculosis [68]. According to Sathar et al.
[68], the gamma interferon assay has a sensitivity of 93%
and a specicity of
98% when the rate was 3.2 IU/ml, the sensitivity of adenosine deaminase is 93% and its specicity is 96%, when the
rate is 30 IU/l. Unfortunately this is not a common practice
test. Thus, the dosage of interferon gamma may be an
alter- native diagnosis compared to surgery in countries
with high endemic TB.

Radiological explorations

Determination of specic immunoglobulin


A recent review on the immunological diagnosis of tuberculosis has shown that the determination of IgG antibody
response to 43-kDa antigen of MT in ascites had a high sensitivity approaching 100% and a specicity of 95.7% [69]. It
is a very expensive test performed in specialized centers.

Abdominal ultrasounds
Abdominal ultrasound can evoke some signs in favor of PT
but cannot conrm diagnosis [74], and can follow the evolution under treatment. The ascites is the most frequent
ultrasound sign (45% to 100% of cases) [75,76]. It is free in
most cases and appears as trans-sonic images in which the
bowel loops oat. Ascites may take echogenicity related to
its exudative nature [77]. Sometimes, the ascites is partitioned, which is an argument in favor of its tubercular
origin and could predict technical difculties during diagnostic laparoscopy. Other signs suggestive of PT can be
highlighted, such as the thickening of the peritoneum, the
clumping of bowel loops side by side each other or to the
anterior abdominal wall [78], peritoneal nodules that are
the equivalent of granulations sitting at the visceral and
parietal peritoneum [75,78,79], adhesions visualized as
hypoechoic linear structures in thin strips and oating in
the tuberculosis peritoneal effusion [78,80], and the deep
lymph nodes often described as hypoechoic multiple
masses and sometimes conuent [75,76,80]. All these
echographic signs taken sep- arately have no diagnostic
value in the PT. On the contrary, their association should
suggest the diagnosis [75]. How- ever, the problem of
differential diagnosis with peritoneal carcinomatosis
remains.

The molecular genetics


Testing the polymerase chain reaction
(PCR)
The reaction of gene amplication by PCR is a rapid test
to isolate the MT between 24 to 48 hours [70]. The use in
clinical practice of this technique is limited by high cost
and low sensitivity (60% to 80%) [71,72].

Computed tomography (CT)


This examination is often done through the assessment of
exploration of exudative ascites isolated in order to nd a
primary tumor such as advanced ovarian cancer, the main
differential diagnosis of PT [56,81].
The major CT signs suggestive of PT are:

Figure 1
Free high-density ascites with thickening of the
pari- etal peritoneum (arrow).

the ascites, which typically has a high density (25 to 45


HU) [82,83], but at an early stage it may have a uid
density [84,85] (Fig. 1);
lymph nodes, which usually have hypodense center [84]
corresponding to the caseating with hyperdense rim but
can also be calcied [86];
thickening of the mesentery and omentum [8789];
regular and uniform thickening of the peritoneum [84,89]
(Fig. 1);
agglutination of the intestinal loops [85,89].

The surgical biopsies


Surgical approach
They remain the ultimate means to conrm the diagnosis
of PT-related histological proof. These biopsies are typically done by laparotomy or in recent years ideally by
laparoscopy. They must interest the peritoneum zones,
which contain nodular lesions. The feasibility of
laparoscopy in the exploration of ascites is well
established [9093]. Cumulative data from 402 patients in
11 studies, reported at a recent systematic review,
showed rates of sensitivity and specicity of 93% and 98%
respectively [22]. In addition, laparoscopy is a surgical
approach that gives less parietal prejudice and less
postoperative pain than laparotomy. It also allows the
reduction of hospital stay and quicker reha- bilitation.
The complications rate of laparoscopy biopsies is around
2.7%; the most common are intestinal perforations and
bleeding from wounds of large vessels [11,93] especially
in the bro-adhesive forms [24]. Therefore the diagnostic
laparoscopy in case of exudative ascites should be
entrusted to an experienced surgeon.
Macroscopic ndings
The macroscopic appearance of PT is typical when we nd
regular white granules with equivalent sizes (1 to 3 mm),
numerous nodules, dispersed and friable at biopsy, hyperemia of the peritoneum, lamentous (or rope) peritoneal
adhesions and clumping of bowel loops with brin
deposition [94].

Figure 2 Tuberculous granuloma consisting in giant cells of


Langhans-type, epithelioid cells, histiocytes and eosinophils.

However, face to atypical macroscopic PT, we should


always evoke the peritoneal carcinomatosis, which typically includes peritoneal nodules of varying sizes, irregular,
inverted, retractable, preferably located on the diaphragm
and/or in the pelvis with a non-inammatory peritoneum.
The pathological study of biopsies
Histologically, typical lesions of tuberculosis is represented
by the tuber, which corresponds to caseating epithelioid granuloma containing multi-nucleated giant cells
Langhans-type, epithelioid histiocytes and lymphocytes
forming a ring surrounding a central area of caseous
necrosis (Fig. 2).

Percutaneous biopsy
Liver biopsy
A granulomatous liver disease may be associated with peritoneal involvement in varying proportions ranging from 25%
to 48% [38,72,95]. For some authors, liver biopsy must be
systematic face to any suspicion of PT, it would provide,
in some cases, the only evidence of tuberculous etiology
without resorting to invasive procedures [96].
Peritoneal biopsies
This is a new minimally invasive method with little or no
complications and could allow to obtain very good results.
Vardareli et al. [47] reported 19 cases of PT with ascites
(bro-adhesive forms excluded) of which 18 were diagnosed by peritoneal biopsy under percutaneous radiological
guidance. Thus, this method could avoid the drawbacks
of surgery, and preliminary results seem very encouraging.
Studies on a larger scale trials will help to conrm the
contri- bution of this new diagnostic procedure.

Treatment
The treatment of PT is medical. The delay in initiating
treat- ment may increase mortality. Chow et al. [35]
reported a signicant deterioration in clinical status of
over 80% of

patients during the exploration period. The overall


mortality in this study was 35%, whereas in the subgroup
of patients with underlying cirrhosis mortality reached
73%. The aver- age mortality rate according to the
cumulative data from
18 studies, which included more than 800 patients was 19%
[14]. This underlines the importance of establishing appropriate treatment as soon as possible.
The antituberculosis drugs were classied into rst and
second line. Currently, WHO has divided them into ve
groups. The rst-line drugs are a group 1 of WHO: isoniazid
(INH), rifampicin (RIF), pyrazinamide (PZA), ethambutol
(EMB). Those of second-line bring together groups 2 to 5
of the WHO [97].
The currently recommended protocol combines four
drugs INH, RIF, PZA and EMB given daily for 2 months,
relayed by 4 months of combination therapy INH and RIF
[98]. Although the recommended duration of treatment of
PT is 6 months, some authors have reported treatment
dura- tion to 12 months [29]. There is however no
evidence to hold a treatment time beyond 6 months. Some
retrospec- tive series have used different regimens of 6 or
9 months, and found that the majority of their patients
has responded to treatment similarly [43,47]. One study
compared two dif- ferent durations of treatment (9
months and more than 12 months) and found no signicant
difference in terms of ther- apeutic response in either
group [99].
The favorable response to treatment results in the resolution of symptoms and the disappearance of ascites.
Laboratory abnormalities in favor of disease activity will
usu- ally normalize within 3 months after starting the
treatment [22].
The risk of hepatotoxicity of antituberculosis drug
increases in case of underlying liver disease and particularly cirrhosis. Stopping the treatment is recommended in
cases of elevated transaminases up to ve times normal in
asymptomatic forms, and three times normal in case of
signs of hepatotoxicity [98]. In this case it is important to
choose one of four alternatives:
RIF, PZA and EMB during 6 months;
INH, RIF, EMB during 2 months followed by INH and RIF
during 7 months;
RIF, EMB, uoroquinolone, and cycloserine (injection)
during 12 to 18 months;
EMB, streptomycin, uoroquinolones and other secondline anti-TB drugs.
In the serie of Sanai and Bzeizi [22], no cases of hepatotoxicity have been reported in 16 patients who had liver
disease and treated with INH, RIF, EMB and ooxacin
during
2 months followed by INH and EMB and ooxacin during 10
months.
The use of corticosteroids is still controversial. Some
authors have shown no benet in combining corticosteroids
with antituberculosis [100] while others have found the
opposite [101,102]. Prospective studies are needed to
better understand the contribution of corticosteroids.

Conclusion
PT remains a common public health problem in endemic
regions of the world. The diagnosis of this disease is
difcult

because of its clinical polymorphism. Assays of LDH, SAAG


and glucose have a specicity that is too low. PCR tests
are expensive and have low sensitivity. The research of MT
in the ascites uid is insensitive; culture in liquid media
such as BACTEC radiometric system improves the diagnostic sensitivity in a reasonable timeframe. The dosage of
the ADA is a quick test, reproducible with excellent sensitivity and specicity for the diagnosis of PT. It should
be used routinely in endemic countries. It is the same for
gamma interferon, which has a very high sensitivity and
specicity. The determination of CA-125 would allow the
assessment of therapeutic response. The ideal peritoneal
biopsies performed by laparoscopy can conrm the diagnosis while avoiding the disadvantages of laparotomy. The
interventional radiology seems promising for the
realization of peritoneal biopsies. Treatment involves
antituberculosis drug of group 1 according to the WHO
during 6 months.

Conict of interest statement


The authors have not declared any conict of interest.

Acknowledgments
The authors would like to thank Pr Ben Amar
Mohamed (benammed@yahoo.fr) and Mrs Ahlem Fendri
(ahlem fendri@yahoo.fr) for their precious help in the
translation of the manuscript and for her constant support.

References
[1] Dineen P, Homan WP, Grafe WR. Tuberculous peritonitis:
43 years experience in diagnosis and treatment. Ann Surg
1976;184:71722.
[2] Truffot-Pernot C. Mycobacterium tuberculosis. EMC biologie
clinique; 2005 [90050220].
[3] Huh JJ, Panther LA. Mycobacterium avium complex peritonitis in an AIDS patient. Scand J Infect Dis 2001;33:9368.
[4] Un-In Wu. Peritonitis due to Mycobacterium avium complex
in patients with AIDS: report of ve cases and review of the
literature. Int J Infect Dis 2009;13:28590.
[5] Mehta JB, Dutt A, Harvill L, Mathews KM. Epidemiology of
extrapulmonary tuberculosis. A comparative analysis with
pre-AIDS era. Chest 1991;99:11348.
[6] Singh MM, Bhargava AN, Jain KP. Tuberculous peritonitis. An evaluation of pathogenetic mechanisms, diagnostic
procedures and therapeutic measures. N Engl J Med
1969;281:10914.
[7] Lacut JY, Dupon M, Paty MC. Tuberculoses extrapulmonaires
: revue et possibilits de diminution des dlais
dintervention thrapeutique. Med Mal Infect 1995;25:304
20.
[8] Lazarus AA, Thilagar B. Abdominal tuberculosis. Dis Mon
2007;53:328.
[9] Rasheed S, Zinicola R, Watson D, Bajwa A, McDonald PJ.
Intra- abdominal
and
gastrointestinal tuberculosis.
Colorectal Dis
2007;9:77383.
[10] De Backer AI. Vascular involvement secondary to
tuberculosis of the abdomen. Abdom Imaging 2005;30:714
8.
[11] Marshall JB. Tuberculosis of the gastrointestinal tract and
peritoneum. Am J Gastroenterol 1993;88:98999.
[12] Tang LC, Cho HK, Wong Taam VC. Atypical presentation
of female genital tract tuberculosis. Eur J Obstet Gynecol
Reprod Biol 1984;17:3556

Você também pode gostar