Send to

Choose Destination
J Hepatol. 2010 Oct;53(4):639-47. doi: 10.1016/j.jhep.2010.04.029. Epub 2010 Jun 22.

Cytokeratin 18-based modification of the MELD score improves prediction of spontaneous survival after acute liver injury.

Author information

Dept. of Gastroenterology and Hepatology, University Hospital, 45122 Essen, Germany.



Predicting the probability of patients with acute liver failure (ALF) to recover spontaneously is of major clinical importance. As apoptotic and necrotic cell death are crucial in the pathogenesis of ALF, we determined whether selected cell-death markers predict outcome of patients with ALF and/or discriminate between etiologies.


In a prospective study (11/2006-06/2009), 68 ALF patients were recruited consecutively. Data were collected over four weeks or until discharge, death or LTx, including CK18/M65 and M30 ELISA and glutathione S-transferase, subtype alpha. Data at date of admission and at the date of peak levels of M65 were individually analyzed and correlated with the patients' prognosis and etiology.


The predictive sensitivity of total serum M65 for lethal outcome was comparable to the Model for End-Stage Liver Disease (MELD) score at time of admission and at its peak value. In contrast, serum bilirubin levels had no prognostic value, neither at admission nor at later time points. In order to accurately predict the clinical prognosis of ALF patients, we tested a modified MELD score where CK18 M65 substituted bilirubin. This CK18/M65-based MELD score significantly better predicted the prognosis of ALF patients compared with the current MELD score or KCC. A combination of tested parameters contributed to improved discrimination of ALF etiologies by applying cell death and established laboratory parameters.


The CK18 M65-based MELD score has superior sensitivity and specifically predicts survival of ALF patients. Further prospective clinical studies could validate its potential role to predict requirement of LTx in ALF patients.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center