BACKGROUND/AIMS: In cirrhotic patients, spontaneous bacterial peritonitis (SBP) may be difficult to distinguish from secondary peritonitis with occult intestinal perforation; Runyon's criteria (based on ascitic fluid glucose, protein and lactate dehydrogenase levels) are sensitive but not specific. Ascitic fluid carcinoembryonic antigen (CEA) and alkaline phosphatase (AP) are potential markers for secondary peritonitis. METHODS: Ascitic fluid CEA and AP levels were prospectively compared among three subject groups--cirrhotic patients with sterile ascites, cirrhotic patients with SBP, and patients (cirrhotic and non-cirrhotic) with perforation-related secondary peritonitis. RESULTS: The secondary peritonitis group (n = 38 including 11 cirrhotic patients) had significantly higher mean CEA and AP levels than the SBP (n = 34) and sterile ascites patients (n = 63). Of secondary peritonitis patients, 92% fulfilled predetermined criteria (either CEA >5 ng/ml or AP >240 units/l) versus only 12% of SBP patients; sensitivity was 92% and specificity 88% for differentiating secondary peritonitis from SBP. Runyon's criteria had a sensitivity of 97% and specificity of 56%. Stratification of secondary peritonitis patients by the presence or absence of cirrhosis did not alter our results. CONCLUSIONS: Ascitic fluid CEA or AP elevations appear to be sensitive and specific markers for perforation-related secondary peritonitis in cirrhotic as well as non-cirrhotic patients.
BACKGROUND/AIMS: In cirrhotic patients, spontaneous bacterial peritonitis (SBP) may be difficult to distinguish from secondary peritonitis with occult intestinal perforation; Runyon's criteria (based on ascitic fluid glucose, protein and lactate dehydrogenase levels) are sensitive but not specific. Ascitic fluid carcinoembryonic antigen (CEA) and alkaline phosphatase (AP) are potential markers for secondary peritonitis. METHODS: Ascitic fluid CEA and AP levels were prospectively compared among three subject groups--cirrhotic patients with sterile ascites, cirrhotic patients with SBP, and patients (cirrhotic and non-cirrhotic) with perforation-related secondary peritonitis. RESULTS: The secondary peritonitis group (n = 38 including 11 cirrhotic patients) had significantly higher mean CEA and AP levels than the SBP (n = 34) and sterile ascitespatients (n = 63). Of secondary peritonitispatients, 92% fulfilled predetermined criteria (either CEA >5 ng/ml or AP >240 units/l) versus only 12% of SBP patients; sensitivity was 92% and specificity 88% for differentiating secondary peritonitis from SBP. Runyon's criteria had a sensitivity of 97% and specificity of 56%. Stratification of secondary peritonitispatients by the presence or absence of cirrhosis did not alter our results. CONCLUSIONS: Ascitic fluid CEA or AP elevations appear to be sensitive and specific markers for perforation-related secondary peritonitis in cirrhotic as well as non-cirrhotic patients.