CYTOKERATIN INCLUSIONS IN ALCOHOLIC LIVER-DISEASE AND THEIR RELATION TO THE AMOUNT OF ALCOHOL INTAKE

In the present study, the frequency and the distribution pattern of im munoreactive hepatocytic cytokeratin (CK) inclusions was investigated using the monoclonal antibody (MAb) CAM 5.2 detecting CKs 8, 18 and 19 . The CK antigenicity of the inclusions was confirmed on frozen sectio ns with MAbs for the CKs 7, 8, 17, 18 and 19. The frequency of hepatoc ytic CK aggregates was compared to the presence of nonalcoholic and al coholic liver disease (ALD) as well as to the average all year daily e thanol intake of 195 consecutive males subjected to medicolegal autops y. Hepatocytic CK inclusions were infrequent in patients with normal l iver histology, portal fibrosis and/or portal inflammation. In ALD, ho wever, inclusions were observed in 35.6% of patients with fatty liver, in 63.2% of patients with alcoholic hepatitis, in 30.8% of patients w ith bridging fibrosis and in 60.0% of patients with liver cirrhosis. I n all specimens studied, the inclusions were reactive only for CKs 8 a nd 18, CKs 7, 17, and 19 being unreactive. The frequency of inclusion bodies increased, paralleling increasing average all-year daily alcoho l consumption. Compared to non-drinkers, a daily intake of between 40 and 80 g of absolute alcohol was associated with a statistically signi ficantly (relative risk, RR=6.6) increased risk of formation of aggreg ates. Similarly, daily consumption exceeding 80 g was related to incre ased (RR=6.0) frequency of CK aggregates. The frequency of full-blown ''classical'' Mallory bodies (MBs) was, however, increased (RR=8.9) on ly in patients with a daily intake exceeding 160 g. These results indi cate that CK inclusions are unequivocally related to the presence of A LD in longitudinal autopsy series. Their frequency in autopsy series w as, however, significantly lower than their frequency in previous stud ies comprising hospitalised patients diagnosed as having alcoholic liv er disease. Furthermore, our results suggest that a daily ethanol inta ke between 40 and 80 g can increase the risk of appearance of smaller CK inclusions, although the risk of occurrence of ''classical'' full-b lown MBs was associated only with heavy drinking exceeding 160 g of ab solute alcohol daily.