Abdominopelvic tuberculosis in gynaecology: Laparoscopical and new laboratory findings

Authors


  • Pelvic TB diagnosis by LPS, ELIspot and histology.

  • This research was based on the Italian National Guidelines on Pelvic Infectious Disease and approved by local institutional ethics committee.

: Dr Andrea Tinelli, Department of Obstetrics and Gynecology, Vito Fazzi Hospital, Piazza Muratore, 73100 Lecce, Italy. Email: andtinelli@libero.it

Abstract

Introduction:  Tuberculosis (TB) is a rare curable infective disease, caused mainly by Mycobacterium tuberculosis, which in abdominopelvic (AP) localisation, can mimic a disseminated carcinomatosis. Symptoms of AP-TB are non-specific, so diagnosis is difficult and elusive as the affected patients have normal chest X-ray and elevated levels of CA125. Female ultrasonographic features of AP-TB mimic peritoneal carcinomatosis, and the computed tomography has also been suggested to be helpful, but the final diagnosis was reached by histology and serology.

Aim of the study:  To propose the validity of the combination of laparoscopy (LPS), histopathology and enzyme-linked immuno-spot (ELIspot) in the diagnosis of AP-TB.

Methods:  In the last two years, we had six women with suspect of AP-TB, who, after the routine exams, were referred for a diagnostic LPS that revealed turbid-free fluid in pelvis or ascites (collected for serology), multiple peritoneal and/or bowel tubercles (randomly sampled), fibrous bands, adhesions, hyperaemic and oedematous bowel loops.

Results:  LPS diagnosis was confirmed by intra-LPS biopsy of nodules and histological examination of specimens: epithelioid granulomas with central caseous necrosis in five patients (83.3%) and a non-caseating granulomatous inflammation in the last one. An outer layer of epithelioid histiocytes and Langhans cells was present in all patients. Using the ELIspot technique performed on free fluids, the final diagnosis of TB was made in all patients.

Conclusions:  Even if gynaecological LPS appearance of the peritoneum can mimic other conditions, ELIspot and histopathological exam can confirm the suspect of AP-TB.

Introduction

In developed countries abdominopelvic and peritoneal tuberculosis (TB) are rare and, in some women, they can mimic a peritoneal carcinosis;1 moreover, peritoneal TB is an uncommon but possible localisation of extrapulmonary TB (EPTB), especially in women.2

Mycobacterium tuberculosis (MTB) is the aetiological agent of TB with the highest incidence in not developed countries. Worldwide, the incidence of TB is growing, owing to the spread of human immunodeficiency virus (HIV), the most important cofactor of MTB infection. Recent reports suggest an increased incidence of abdominal TB in the USA and European countries, particularly in high-risk groups;3 in not well-developed countries, where sophisticated medical facilities are not easily available, laparoscopic examination of intra-abdominal organs should be considered as a useful tool of investigation.4 Immigrants from endemic areas and HIV women are the two population groups with higher risk for abdominal TB.

Because of its uncharacteristic presentation, TB should also be considered in the differential diagnosis of a patient with an abdominal/pelvic mass and ascites;5 signs and symptoms, imaging examinations and CA-125 serum levels in peritoneal TB may resemble that of ovarian carcinoma.6 Therefore, a precise diagnosis of abdominopelvic TB (AP-TB) requires different clinical evaluation.

For the diagnosis of AP-TB, the imaging findings can be misleading and the laboratory analyses are not, unfortunately, helpful.7 Besides, vague symptoms of low-grade, such as fever, mild abdominal pain, constipation, night sweats and bloating do not help clinicians, and sometimes, an abnormal chest X-ray, suggestive of previous TB, could be present.8

Laparoscopic observation can be used for the diagnosis of AP-TB but the uncharacteristic manifestation can lead to ovarian cancer misdiagnosis.

The computed topography (CT) scan of the abdomen can be helpful in suggesting the diagnosis.9 In all the suspected patients, the ultrasonographic features of AP-TB are: septated and particulate ascites, loculated fluid, thickened peritoneum and omentum and cul-de-sac nodularity,10 while the most common CT findings are: ascites, septation in the ascites, peritoneal thickening, mesenteric and omental involvement.11

In women, the routine imaging investigations by ultrasonographic or CT scan of the abdomen could occasionally reveal a cystic mass in the ovary or ascitis with features suggestive of secondary spreading over the omentum and peritoneal surface.12 For this reason, the initial clinical interpretation on the adnexal masses, cul-de-sac nodularity, ascites, omental and peritoneal thickening, weight loss and elevated CA-125 serum levels led to an erroneous preliminary diagnosis of disseminated ovarian cancer.13,14 The suspect of ovarian cancer should be confirmed by clinical investigations with CA125 serum levels in the normal range (0–35 U/mL) and only when there is no any analytical evidence of TB.15

TB can also be confirmed by DNA extraction from the frozen section specimens with subsequent analysis using polymerase chain reaction (PCR) but with some limitations.16 Serological tests such as tuberculin skin test (TST) is positive in TB patients and this can help clinicians for diagnosis.17

MTB is identified microscopically by Ziehl–Neelsen stain for acid-fast bacilli because these organisms resist the dilute acid decolorisation procedures common in many staining protocols. The technique allows the detection of MTB also in ascitic fluid but it is frequently negative.18

As acid-fast stains and special cultures of the ascitic fluid for MTB are frequently negative, another way to detect TB is the estimation of adenosine deaminase level in ascitic fluid.19

In conclusion, the diagnosis of TB is still no easy task, because its symptoms are few or non-specific and because the disease presents with signs similar to those of gynaecological malignancy.

Methods

In our experience, six patients were unwell at presentation and were hospitalised in Department of Obstetric and Gynaecology of Vito Fazzi Hospital during the period from 2004 to 2006; all records were reviewed with histology and bacteriological demonstration. The mean age of patients treated by diagnostic LPS was 44.5 years (range 30–67 years).

Presenting symptoms, at hospital admission, were as follows: fever > 38°C, weight loss, mild abdominal pain, constipation, ascites, abdominal pain and elevated CA-125 serum levels (100%), with the mean level of 54.3 U/mL (range of 80–39 U/mL). TST was positive in two patients (33.3%) and an abnormal chest radiography suggestive of previous TB was present in the elder patient (16.6%).

In the first clinical evaluation, three patients’ ultrasonographic evaluation showed cystic ovarian masses (50%) and the successive transvaginal pelvic ultrasonographic findings were: septated and particulate ascites, loculated fluid, thickened peritoneum and omentum and cul-de-sac nodularity in all women, three unilocular ovarian cysts (two simple adnexal cysts with size measured of 5 cm and one simple cyst of 3.5 cm of diameter) with negative value of resistive index Doppler analysis (cut-off of < 0.50).

The successive CT findings were the same in all patients: confirmation of ascites, peritoneal thickening, mesenteric and omental involvement.

The initial clinical interpretation based on the clinical signs, three cystic masses, ascites, omental and peritoneal thickening, weight loss, fever and elevated CA-125 serum levels, led to an erroneous preliminary diagnosis of a possible disseminated malignant cancer.

A diagnostic laparoscopy (LPS) was performed in all women, and the endoscopical appearance of abdominal TB was correlated with histological and bacteriological criteria for the diagnosis.

The abdominal LPS entering was by umbilicus zone under direct view, avoiding bowel or vessels perforation, which was according to the Direct Access by Optical Trocar (3 Endotip Trocars by Karl Storz, Germany and 3 Endopath Trocars by Ethicon, Jonhson & Johnson, New Brunswick, NJ, USA).

In all patients, free fluid culture was submitted to stain for acid-fast bacilli or analysed by enzyme-linked immuno-spot (ELIspot), in this last case to detect proteins of MTB. This technique is a powerful method for immunomonitoring purposes, including the detection and visualisation of proteins and cytokines released by single cells; ELIspot has a high sensitivity and specificity also in the early phase of TB infection.

Patients’ treatment procedure was based on the Italian National Guidelines on Pelvic Infectious Disease and approved by local institutional ethics committee.

Results

In all patients, the LPS diagnosis showed whitish yellowish–white nodules involving the entire abdominal cavity (Fig. 1), ascites together with ill-defined nodularities or thickening on the abdominal peritoneum surface (Fig. 2) and in the Douglas pouch and/or in the adnexal areas on pelvic examination, multiple fibrous bands and adhesions of the peritoneum to viscera (Fig. 3), thickening, hyperaemia and retraction of the greater omentum (100%), all signs compatible with peritoneal TB; the three cysts were removed and examined by histology (all serous simple cysts with inflammatory signs of the capsule).

Figure 1.

A nodule in the abdominal cavity of one patient.

Figure 2.

Nodularities or thickening on the abdominal peritoneum surface. This picture shows the nodularity behind the port site of trocar.

Figure 3.

Multiple fibrous bands and adhesions of the peritoneum to bowel and viscera. This picture shows a Fritz—Hug—Curtis Syndrome.

The ascitic fluid culture submitted to stain for acid-fast bacilli was positive for MTB in one patient (16.6%), consistent with TB infection.

Moreover, some biopsies performed on miliary nodules from the peritoneum, large intestine and ileocecal area resulted epithelioid granulomas with central caseous necrosis in five patients (83.3%) and a non-caseating granulomatous inflammation in the last one, both groups being characterised by an outer layer of epithelioid histiocytes and Langhans cells (100%).

Finally, ELIspot of free fluid culture was positive for MTB in all patients (100%).

For TB therapy all patients received a triple antitubercular treatment (ATT) corresponding to a combination of isoniazid, streptomycin (total dose 40 g) and pyrazinamide (for the first two months and then substituted with ethambutol), without rifampicin. In patients with abdominal TB the treatment was carried out up to six months with a follow up of one year.

After anti-TB therapy, all patients recovered as confirmed by clinical investigations and currently, none of them conceived or tried to conceive; all data are summarised in Table 1.

Table 1.  In our experience on female abdominal genital tuberculosis, we recovered and successfully treated six patients; all the dates showed our practice
 Patient 1Patient 2Patient 3Patient 4Patient 5Patient 6
Age (years)673930354551
Symptoms
(weight loss, mild abdominal pain, fever > 38°C, constipation, abdominal pain)
YesYesYesYesYesYes
Tuberculin skin testYesNotNotNotYesNot
AscitesYesYesYesYesYesYes
CA125 (0–35 U/mL)485540673980
Ovarian cystNotDiameter > 5 cmDiameter 3.5 cmDiameter > 5 cmNotNot
Computed tomography findings
(confirmation of ascites, peritoneal thickening, mesenteric and omental involvement)
YesYesYesYesYesYes
Diagnostic LPS by optical trocarsEndotipEndopathEndopathEndotipEndotipEndopath
Acid-resistant bacilli in ascitesNotYesNotNotNotNot
Peritoneal nodularityYesYesYesYesYesYes
Positivity to ELIspot testYesYesYesYesYesYes
Caseating granulomasYesYesYesYesYesNot
ATT for six months with recoveryYesYesYesYesYesYes

Discussion

TB is now undergoing a worrying recrudescence in developed countries, and in our country, it continues to pose a serious threat.

The 1995–2005 report on TB in Italy, performed by Italian Minister of Health, Government of Sanitary Prevention, Infective Disease and International Prophylaxis Office, confirmed an incidence of 7.1 cases per 100 000 in 2005, mainly men, and the AP-TB localisation was found in 14% of cases with extrapulmonary forms.

From all the patients who had been diagnosed with AP-TB, from 2004 to 2006, only six were successfully diagnosed by LPS at my department. During the same period, other patients were referred for suspected cases of EPTB to other clinical department at our hospital but diagnosed without laparoscopic biopsy.

Various investigations have been suggested as the gold standard of diagnosis of AP-TB, but despite their clinical utility these techniques often show poor sensitivity and sensibility.

Non-invasive tests, such as acid-fast stain and culture of the ascitic fluid are usually insufficient; in addition, AP-TB can arise with non-specific signs and symptoms mimicking other conditions: ultrasonographic features of the abdomen and pelvis can be less useful in this case as well as serum CA125 level is often > 35 U/mL, which is similar to that of ovarian malignancy. The CT is unspecific for final diagnosis of AP-TB and it is mandatory to quickly exclude a gynaecological malignancy. Consequently, there is an increasing need to use other diagnostic tools.

In our department, diagnostic LPS is currently utilised as main invasive examination tool for such patients: from 2004 to 2006, 182 diagnostic laparoscopies were performed, of these, 105 for fertility investigations; because of its safety, diagnostic LPS with histopathological findings plus ELIspot confirmation might be necessary, in our opinion, for the AP-TB diagnosis.

The laparoscopic approach is often very useful for the final diagnosis of AP-TB, because its symptoms vary greatly; diagnostic LPS may detect pathognomonic signs of rashes on the visceral peritoneum: enlarged mesenteric lymph nodes, a profound commissural process in the small pelvis and Douglas space, fibrous bands and adhesions, hyperaemic oedematous bowel loops or dense adhesions without ascites or free ascites with multiple yellowish–white nodules in the peritoneum, thickening and hyperaemia and retraction of the greater omentum.20

The LPS appearance of scattered yellowish–white nodules, approximately 1–5 mm in size, on the peritoneal surfaces and filmy adhesions was suggestive of TB peritonitis, miliary granulations, and inflammatory adherences on the visceral or parietal sheats.21

Diagnosis of pelvic–peritoneal TB is usually reached either by the miliary nodules on the peritoneum, large intestine and ileocecal area, or by laparoscopic-directed biopsy on multiple whitish tubercles in patients, because histopathological examination of organs revealed the presence of two types of granulomas: caseating and non-caseating.22

The multiple TB granulomas can be composed of inner caseous necrosis and outer layer of epithelioid histiocytes and Langhans cells, that lead to the diagnosis of abdominal peritoneal TB;23 but it is sometimes possible to diagnose AP-TB also by image-guided percutaneous peritoneal biopsy and by histological examination of specimens.19,24

Moreover, PCR analysis of the IS6110 sequence specific for the MT complex and of the 65 kDa antigen gene common to all mycobacteria resulted positive in all patients with EPTB in one study.2

In addition to histopathological investigation, AP-TB diagnosis was confirmed by ELIspot tests, an interesting technique with a great impact on the diagnosis of TB. ELIspot is an immunological assay where the immune cells and antigen-presenting cells are growth in a microtitre plate coated with an antibody against a particular cytokine. The detection of the cytokine of interest relies on a colorimetric reaction similar to an enzyme-linked immunosorbent assay. The high specificity of this method combined with its sensitivity makes ELIspot an attractive method for detection of specific protein-secreting cells, also in the early phase of TB infection.25

Concerning the possible worldwide usage of ELIspot in all the AP-TB diagnosis, the other case series of EPTB concern, mainly, the Asian reports; in studies involving immunocompetent adults, it has been observed that EPTB constituted about 15–20% of all cases of TB, while in HIV-positive patients, EPTB accounts for more than 50% of all cases of TB.26 Another epidemiological study, on distribution of EPTB cases by anatomical site in HIV-negative patients,27 reported 9% of genitourinary TB and 3% of abdominal TB (total of 12% of AP-TB), while an Indian research in 1997 reported 117 women with AP-TB as the cause of tubal blockage, on 300 infertile patients between the ages of 25 and 35 years and, finally, a Russian article in 1994 reported a genital TB of 11.9% of all the extrapulmonary cases.28

About the management of the disease, all patients are usually advised on ATT. They receive quadruple anti-TB chemotherapy (for one year) and follow-up visits every month for at least one year; on the other hand, the patient can be treated with streptomycin and rifampicin only for five months after diagnosis, with a monthly clinical follow up and a CA125 monitoring.12,17 In patients with AP-TB it has also been confirmed the efficacy of triple ATT with isoniazid, streptomycin (total dose 40 g) and pyrazinamide (for the first two months and then substituted with ethambutol), without rifampicin, for six months.5,8 Some patients can also receive methyl prednisolone, initially 20 mg/day, for one month, because ascites and abdominal pain abated earlier in patients on steroid therapy.3,4 Increased serum CA125 levels return to normal levels after one month of anti-TB therapy and follow up;14,15 currently, none of these patient conceived or tried to conceive, after AP-TB.

Conclusions

Diagnostic LPS can be essential for early diagnosis and management of AP-TB, because this disease requires only a conservative management. Patients with abdominopelvic suspect mass, with or without ascitis, with high serum CA125, should always raise a suspicion of TB and a diagnostic LPS combined either with peritoneal biopsy and histologic confirmation of epithelioid granulomas with central necrosis, either with ELIspot, should be performed to confirm the suspect of infective illness that mimics a gynaecological malignancy, especially in young and infertile women.29

Hence, because of its safety and capability to prevent unnecessary and painful laparotomies, diagnostic LPS seems to be a sufficient and safe method to provide AP-TB diagnosis.

However, if LPS is not feasible, laparotomy should be always performed and serum CA125 should be used to monitor the response of disease to ATT.30,31

Non-invasive tests such as TST, a chest RX or an acid-fast stain and culture of the ascitic fluid are usually insufficient and medical awareness of peritoneal TB is still lacking.

Free fluid culture by ELIspot can confirm a TB infection, can facilitate identification of TB and it could be an alternative tool for the early diagnosis of this infective disease, in order to avoid unnecessary extended surgery.

Ancillary