Purpose: Peritoneal tuberculosis and carcinomatosis are the most frequent etiologies of exsudative ascitis and require rapid diagnosis and treatment. The purpose of this study has been to evaluate the predictive value of clinical and complementary data for the etiologic diagnosis of exsudative ascites and to assess the results of laparoscopic surgery.
Material and methods: We report a prospective long-term study conducted over 10 years, having included all cases of exsudative ascites of unidentified etiology. We excluded patients with a history of anterior laparotomy and patients having a contraindication for laparoscopic surgery. Clinical signs, results of the cytochemical and bacteriological exam of the ascetic liquid and findings from radiological exams and endoscopic investigations were noted. Open laparoscopic exploration noted the macroscopic aspect and many peritoneal biopsies were obtained as well as a liver biopsy when possible. Statistical analysis was performed with SPSS10.0 software. The degree of statistical significance was set for P<0.05.
Results: We included 90 cases of isolated exsudative ascites. There were 59 cases of tuberculosis and 31 of carcinomatosis. Mean patient age was 47 years with a sex ratio of 0.5. Bowel transit disorders were significantly in favor of a carcinomatosis (P=0.04) while fever and nocturnal sweats were suggestive of tuberculosis (P=0.04) but in both instances, the positive predictive value (PPV) of these two signs was weak, respectively 29% and 43%. Relative to chemical and cytologic study of ascitic fluid, hemorrhagic fluid, low white cell count, low lymphocyte differential and presence of atypical cells were significantly linked with the carcinomatosis (P=0.01) but with a PPV<85%. Radiological exams were not very contributive for the etiologic diagnosis. The laparoscopic peritoneal aspect was typical of tuberculosis in 90% of proved cases and in 29% of carcinomatosis cases, underlying the importance of peritoneal biopsies for histological study. Hepatic tuberculosis was associated with peritoneal localisation in 48% of cases. We had no surgical mortality and the morbidity was about 1%.
Conclusion: Etiological diagnosis of exsudative ascitis remains difficult to establish. Peritoneal biopsies under videolaparoscopy are currently the "gold standard" for diagnosis.