Ultrasonographic and biochemical parameters in the non-invasive evaluation of liver fibrosis in hepatitis C virus chronic hepatitis.

Background: Prior studies suggest that platelet counts of <140 000/microL can discriminate patients with different stages of fibrosis.

Aim: To determine the added value of abdominal ultrasound analysis of morphological liver features in increasing the diagnostic accuracy of platelet counts for the prediction of liver fibrosis at histology.

Methods: In a retrospective study, clinical records of 1143 chronic hepatitis C patients at their first presentation, naives to both liver biopsy and anti-viral treatment, were reviewed. Sensitivity, specificity, positive and negative predictive values, positive and negative likelihood ratios of following indices were evaluated singularly or in combination: platelet counts <140 000/microL; nodular liver surface, spleen and portal vein size.

Results: All indices had specificity rate of > or =90% in excluding bridging fibrosis/cirrhosis, whereas sensitivity was acceptable (51%) for only platelet counts <140 000/microL. None of the ultrasonographic parameters singularly evaluated and reached an acceptable sensitivity rate. For ruling cirrhosis in or out, specificity rate was > or =82% for all tests, with the highest value reported by portal vein size. Low platelet counts plus nodular liver surface had the best sensitivity.

Conclusions: No additional significant predictive value was given by adding ultrasonographic parameters to low platelet counts, whereas only a mild non-significant improvement in sensitivity was obtained combining platelet counts <140 000/microL with the presence of nodular liver surface. The platelet counts <140 000/microL showed the best predictive value for including both significant fibrosis and cirrhosis.