Inflammation-adapted Liver Stiffness Values for Improved Fibrosis Staging in Patients With Hepatitis C Virus and Alcoholic Liver Disease

Sebastian Mueller; Stefan Englert; Helmut K. Seitz; Radu I. Badea; Andreas Erhardt; Bita Bozaari; Michel Beaugrand; Monica Lupşor-Platon


Liver International. 2015;35(12):2514-2521. 

In This Article

Abstract and Introduction


Background & Aims: It is well known that inflammation increases liver stiffness (LS) in patients with chronic hepatitis C (HCV) and alcoholic liver disease (ALD) independent of fibrosis stage, but no inflammation-adapted cut-off values have been settled so far. An early identification of rapid fibrosers, however, is essential to decide whom to treat first with the novel but expensive antiviral drugs.

Methods: Liver stiffness, biopsy-proven fibrosis stages F0–F4 (METAVIR or Kleiner score) and routine laboratory parameters were studied in 2068 patients with HCV (n = 1391) and ALD (n = 677).

Results: Among the routine parameters for liver damage, AST correlated best with LS (HCV:r = 0.54, P < 0.0001 and ALD:r = 0.34, P < 0.0001). In the absence of elevated transaminases, cut-off values were almost identical between HCV and ALD for F1–2, F3 and F4 (HCV: 5.1, 9.0 and 11.9 kPa vs ALD: 4.9, 8.1 and 10.5 kPa). These cut-off values increased exponentially as a function of median AST level. The impact of AST on LS was higher in lobular-pronounced ALD as compared to portal tract-localized HCV. Most notably, Cohen's weighted Kappa displayed an improved agreement of the novel AST-dependent cut-off values with histological fibrosis stage both for HCV (0.68 vs 0.65) and ALD (0.80 vs 0.76).

Conclusions: The novel AST-adapted cut-off values improve non-invasive fibrosis staging in HCV and ALD and may be also applied to other liver diseases. Especially in HCV, they could help to decide whom to treat first with the novel but expensive antiviral drugs.


The assessment of liver fibrosis is essential to estimate prognosis, to predict complications, to develop treatment strategies and to monitor disease progression and response to treatment. This is particularly relevant for chronic hepatitis C infection as novel efficient but rather expensive treatment options are now available.[1,2] The introduction of elastography techniques such as transient elastography (FibroScan, TE) or acoustic radiation force imaging to measure liver stiffness (LS) has significantly improved the non-invasive and early diagnosis of fibrosis.[3–9] LS can be obtained in more than 95% of patients within minutes with high reproducibility.[10–12] Cut-off values have been established that allow the allocation to histological fibrosis stages with high accuracy.[5] For viral hepatitis, a cut-off value of 12.5 kPa has been widely accepted for the discrimination of liver fibrosis (F1–F3) from cirrhosis (F4), while LS values below 6 kPa are considered as normal.[6,7] However, it still remains a debate whether disease-specific cut-off values are required for all diseases or just for, e.g. cholestatic liver diseases.[9,13,14]

While LS has an excellent negative predictive value to exclude fibrosis, the interpretation of elevated LS is more complex and may be related to other conditions than fibrosis. Important factors that elevate LS include inflammation or liver damage,[15–17] congestion,[18] cholestasis[19] but also food intake[20–22] or amyloidosis.[23,24] In daily clinical practice, inflammation is the most common cause of non-fibrotic elevation of LS. However, the dissection of inflammation- from fibrosis-mediated LS elevation has not been settled so far.

The impact of elevated transaminase levels on fibrosis assessment via LS has long been recognized in patients with hepatitis B and C.[25,26] Even mildly elevated transaminases are known to increase LS[27] and normalization of bilirubin and transaminases has been postulated to improve fibrosis quantification.[27,28] Dramatically varying cut-off values were also reported for ALD.[8,29–31] However, in contrast to viral hepatitis, short interventions such as alcohol detoxification have demonstrated the dependence of elevated LS on histological alcoholic steatohepatitis and AST levels.[8,9,32] Moreover, relapse of alcohol consumption increased LS[33] and consideration of elevated AST levels significantly improved the diagnostic accuracy of LS in ALD patients.[8]

Short interventions such as alcohol withdrawal, however, cannot be applied to chronic viral hepatitis and other chronic self-propagating liver diseases. In addition, more accurate, inflammation-based cut-off values are urgently needed for patients with HCV as novel, efficient but expensive interferon-free treatment regimens are now available. As complication rate increases and treatment efficacy decreases in patients with manifest cirrhosis, an accurate and early identification of fibrosis independent of inflammation is needed. Therefore, the present multicentre study was set up primarily to identify clinically feasible inflammation-adapted cut-off values for improved fibrosis assessment in patients with HCV and ALD.