Liver fibrosis scores predict mortality in the general population

Background and aims:

Most studies evaluating non-invasive fibrosis measurement are limited to the hepatological setting. Studies focusing on the general population have been centered on methods such as transient elastography, which are not available to primary care givers. The aim was to evaluate the association of easily available non invasive markers of fibrosis in the general population with mortality.

Method:

This study is an analysis of the population-based CARLA study evaluating cardiovascular disease risk factors in a population of middle-elderly adults (45 – 83yrs) in Germany (Greiser BMC Cardiovascular Disorder 2005) between 2002 – 2006. Two follow-up examinations were performed 2007 – 2010 and 2013, respectively. Vital status of participants was regulary registered, last time in 2016. The cause of death was defined as specified in the official death certificate. Easily available non invasive fibrosis scores were calculated (NFS, FIB4, APRI, Lok, GUCI, Forns and BARD) and compared regarding overall mortality, CVD mortality and liver associated mortality with ROC curves.

Results:

1436 subjects were included in the analyses (mean age 67 (SD 9)years, male 55%, mean BMI 28.4 (SD 4.7)). During follow-up 220 people died, (100 CVD or liver associated). NFS and Forns had the greatest AUC for overall mortality and cardiovascular/liver related mortality [AUC = 0.78 (95%CI 0.73 – 0.82) and 0.75 (95%CI 0.70 – 0.80), respectively]. The NFS cut-off -0.43 was the best to predict mortality (Se 72%, Sp 72%). KM curves are shown on figure (log-rank 93.9, p < 0.001).

Fig. 1: KM curves

Conclusion:

The non-invasive fibrosis score, NFS, is an independent predictor of overall, cardiovascular and liver mortality in the general population. This easily available score (age, BMI, impaired fasting glucose or diabetes, AST, ALT, platelets and albumin) can be calculated in the primary care setting and select patients that should be referred for specialist care.