Spontaneous bacterial peritonitis (SBP), an infection of ascitic fluid in the absence of localized intra-abdominal infection, is one of the main potentially fatal complications of cirrhosis. In the case of SBP, early diagnosis and rapid therapeutic care can improve patient survival (Garcia-Tsao, 2001).

The diagnosis of SBP is based on the detection of a polymorphonuclear neutrophils count equal to or greater than 250 /mm3 in the ascitic fluid (method of reference). However, obtaining an ascitic cell count is sometimes difficult because it can not always be performed in emergency especially outside the opening hours of the laboratory of Bacteriology. This raises the necessity of developing quick and easy alternative approaches of diagnosis.

Few groups have proposed the use of urinary reagent strip for rapid diagnosis of SBP. Nevertheless, the investigators clinical teams have shown that the sensitivity of this test was low in a large national multicenter prospective study involving more than a thousand patients (Nousbaum et al., 2007). The use of Multistix strips test is thus not recommended for the routine application of diagnosis of SBP due to its lack of sensitivity.

Although performed on small groups of patients, several studies have reported that IL-8 or IL-6 might be used as markers of ascitic fluid infections. Based on these data and confirmed by the investigators preliminary results the investigators propose to study on a broad recruitment of patients estimated to about 500 inclusions (about 45 infected patients) the interest of using IL-8 and IL-6 as predictive markers of SBP. The investigators propose to use an ELISA method, standardized, rapid and automated, applicable in the context of emergency (7 days a week and 24 hours a day) as previously described in the work conducted to exclude the urinary tract infection (Oregioni et al., 2005).

During the preliminary experiments conducted for this project, the investigators also found systematic variation of another marker, leptin. This is a protein hormone involved in the inflammatory and immune responses (Otero et al., 2005). It appears necessary to include the study of this marker in the analysis of differential protein response between patients suffering or not suffering from SBP.

The investigators therefore propose a diagnostic study, non-interventional, prospective, multicenter trial conducted over 2 years.

• The main objective is to establish the diagnostic performance (sensitivity, specificity, positive predictive value and negative) of IL-8 and IL-6, assayed in the ascites fluid by an automated ELISA in the early diagnosis of SBP.
• The secondary objectives are to confirm the interest of the measurement of leptin in the SBP and to establish the diagnostic performance of IL-8 and IL-6 or leptin according to different clinical features in patients (score Child-Pugh classification and history of SBP, ascitic fluid infection with positive bacterial culture).

Inclusion Criteria:

• Age > 18 years old
• Patient with social insurance
• Signature of informed consent
• Patient admitted for treatment of ascites or complications of cirrhosis. Diagnosis of cirrhosis relied on clinical, biological and morphological criteria (portal hypertention, hepatic biopsy…).

Exclusion Criteria:

• Patient who have received abdominal surgery within the last month.
• Patient with chylous ascites or ascites not related to portal hypertention (pancreatic ascites, hemoperitoneum, ascites observed during acute heart failure, peritoneal tuberculosis, hepatocellular carcinome…)
• Patient with obesity severe (IMC ≥ à 35 kg/m2)

All consecutive patients with cirrhosis admitted to the five participating center Patients were hospitalized or treated in an ambulatory setting for treatment of ascites or complications of cirrhosis.