Clinical significance of serum procalcitonin levels in patients with acute or chronic liver disease

OBJECTIVE: To evaluate the diagnostic value of serum procalcitonin levels in patients with acute or chronic liver disease, with or without bacterial infections and to correlate the results with the clinical outcome and the laboratory findings for these patients. METHODS: One hundred and six consecutive hospitalized patients with liver disease were evaluated for procalcitonin levels on admission. Fifteen of them (14.2%) had acute alcoholic hepatitis on cirrhotic background (group A), 20 (18.9%) had alcoholic cirrhosis without hepatitis and/or bacterial infection (group B), 16 (15.1%) had decompensated cirrhosis with proved bacterial infection (group C), 42 (39.6%) had uncomplicated viral hepatitis-related cirrhosis (group D) and 13 (12.3%) had acute icteric viral hepatitis (group E). Serum procalcitonin levels were measured using an immunoluminometric assay. Statistical analysis was based on Student's t-test and the non-parametric Kruskall-Wallis test (P<0.05). RESULTS: Serum procalcitonin levels were significantly higher in cirrhotic patients with bacterial infection (9.80+/-16.80 ng/ml) than in those without bacterial infection (0.21+/-0.13 ng/ml, P=0.001), whereas they were within normal range (<0.5 ng/ml) in all patients with uncomplicated cirrhosis, irrespective of the cause of cirrhosis. Seven of 15 group A patients (46.2%) and 4/13 group E patients (30.8%), all of them cirrhotics, had procalcitonin levels higher than 0.5 ng/ml on admission, without established bacterial infection. CONCLUSION: Serum procalcitonin levels remain below the threshold of 0.5 ng/ml in all patients with uncomplicated cirrhosis, irrespective of the cause of the disease, while they are significantly elevated when bacterial infection complicates the course of the disease. A significant proportion of patients with acute alcoholic hepatitis on a cirrhotic background as well as of patients with acute on chronic viral hepatitis, without bacterial infection, exhibit serum procalcitonin levels above 0.5 ng/ml, suggesting that this cut-off value is probably not enough to discriminate between patients with or without bacterial infection within these subgroups of patients with liver disease.