Serum concentrations of the tissue polypeptide specific antigen in patients suffering from non-alcoholic...

Liver histology is the gold standard for diagnosis of non-alcoholic fatty liver disease. Ethical considerations and patient choice often preclude performing a liver biopsy, especially considering the rare but potential risk. Searching for a good serological marker substitute for the invasive procedure was the aim of our study. Keratins, mainly 8 and 18, play not only a mere structural role providing mechanical stability to hepatocytes, but also represent a target via toxic stress ultimately inducing apoptosis/necrosis. Tissue polypeptide-specific antigen (TPS), a serological mirror of keratin 18, is widely used as a marker for various cancers. This antigen was assessed in three different groups who were overweight or obese.

In this cross-sectional case-control study, 48 cancer-free patients with non-alcoholic steatohepatitis (NASH, Group 1), 48 patients with pure fatty liver (FL, Group 2), and 47 volunteers (Group 3) were studied. All of them were referred to our metabolic unit for routine evaluation.

The median (range) TPS levels were 123 (56-286) ng mL(-1) in NASH patients. FL patients and volunteers had significantly lower TPS levels, 76 (38-98) ng mL(-1) and 64 (28-87) ng mL(-1), respectively (P = 0.0001). A value of 88 ng mL(-1) in patients with underlying bright liver was associated with a high probability of NASH (sensitivity and specificity = 92% and 96%, respectively). One patient (2.1%) with FL had a TPS value > 88 ng mL(-1), but in the same group, 29 FL patients (60.4%) had an alanine aminotransferase value > 40 U L(-1). Based on a recent classification of liver fibrosis, the median (range) TPS values were significantly different among the stages: F1 (n = 23) = 100 (76-264) ng mL(-1); F2 (n = 21) = 134 (56-276) ng mL(-1); and F3 (n = 4) = 199.5 (123-286) ng mL(-1), respectively (P = 0.014).

Our study shows that TPS is a better marker than alanine aminotransferase activity, ultrasonography or the combination of both parameters in differentiating NASH from FL.