Peritoneal tuberculosis is a rare disease, often associated with a primary site of tuberculosis. Risk factors include HIV infection, diabetes mellitus, treatment with anti-tumor necrosis factor (TNF) agents, ongoing peritoneal dialysis and hepatic cirrhosis.

Bacilli can enter the peritoneal cavity in several ways, including transmural infection from diseased bowel or, more commonly, by hematogenous spread of infection from a pulmonary focus.1 Even though the lung is often the primary site of infection, there is clinical or radiologic evidence of pulmonary tuberculosis in only about a third of cases.2

We report the case of a 25-year-old woman who developed asthenia, anorexia, weight loss, metrorrhagia and pelvic pain over about 3 months. She was a secretary at a business office, former smoker, and had been appendicectomized one year earlier; there was no other relevant medical history.

Because of the symptoms described, an abdominal and pelvic computed tomography (CT) scan was performed, showing bilateral multiloculated ovarian cysts, with no other relevant findings. Later she underwent diagnostic laparoscopy, revealing multiple adhesions (frozen pelvis) and numerous white nodules/granulomas all over the peritoneum, which were biopsied.

Because of the suspicion of tuberculosis during the surgical procedure, a chest X-ray was performed, which showed pulmonary infiltrates in both upper lobes. Later, she underwent a thoracic CT scan, showing tree-in-bud pattern predominantly in the upper lobes of both lungs, with a cavitated image in the left upper lobe.

The tests for potentially immunosuppressive infections were negative, including HIV, Hepatitis B and C, Epstein Barr end Cytomegalovirus. Primary immunodeficiencies were also excluded, as immunoglobulin quantification was normal.

The patient underwent fiberoptic bronchoscopy. Bronchial washing culture and polymerase chain reaction (PCR) assay were positive for Mycobacterium tuberculosis, even though the test for acid-fast bacilli (AFB) was negative. Analysis of peritoneal fluid samples showed lymphocytosis, with negative AFB. Biochemical analysis of the fluid (including ADA levels) was not performed. Histological examination of the biopsies later confirmed granulomatous peritonitis and culture of the specimen was also positive for Mycobacterium tuberculosis.

Diagnosis of pulmonary and peritoneal tuberculosis was established and the patient started antituberculous therapy. The susceptibility testing showed sensitivity to all first-line drugs.

As stated before, immunosuppression plays a major role in the pathogenesis of peritoneal tuberculosis, but in this case the patient was immunocompetent. This diagnosis is frequently difficult, given the nonspecific signs and symptoms, which usually include ascites, abdominal pain and fever.3

Although the test for AFB in the peritoneal fluid is highly specific for the diagnosis, it lacks sensitivity,4 which often makes early diagnosis difficult.

New diagnostic procedures like PCR assay for M. tuberculosis could help to address this issue, since they can significantly decrease the time taken to achieve a correct diagnosis and are especially useful when AFB test is negative.5 In our case, PCR was only performed on the respiratory samples collected during bronchoscopy. Although our patient's condition remained relatively stable throughout the course of investigation, PCR assay can be of utmost value in patients with negative AFB with exclusively extrapulmonary tuberculosis, particularly if their condition is deteriorating, as it can provide a diagnosis much faster than mycobacterial culture of specimens, reducing morbidity and mortality.

Despite the fact that identification of M. tuberculosis in any material is the gold standard diagnostic method, negative result of culture cannot exclude the diagnosis of tuberculosis and, even if it is positive, it can be a slow process. In selected cases, laparoscopic evaluation may be appropriate when suspecting abdominal tuberculosis, since macroscopic findings can suggest the disease6 and it is an easier way of performing tissue biopsy. Since our patient did not have any respiratory complaints that would make one suspect tuberculosis, surgery in fact provided the first indications of tuberculosis and the tissue biopsies performed would eventually confirm the diagnosis.

In conclusion, this case emphasizes the need to consider tuberculosis as a differential diagnosis in young patients with constitutional symptoms, even if they have no respiratory complaints.