Abstract

Background:

FibroScan®-derived liver stiffness (LS) values above a threshold of 20 kilopascals (kPa) has been used to guide endoscopic screening for gastro-esophageal varices. However, there is limited data on the clinical significance of LS values above this threshold. We aimed to assess whether LS values ≥ 20 kPa are associated with death, decompensation, and hepatocellular carcinoma (HCC).

Methods:

We identified 20,776 patients who underwent Fibroscan LS measurement since 1/2014 at the Veterans Affairs Healthcare System and after excluding those with HCC, decompensation (ascites, hepatic encephalopathy, or gastro-esophageal variceal bleeding), or liver transplantation prior to LS measurement. Baseline characteristics were ascertained within 1 year prior to LS date. Patients were followed from LS date until 1/2022 for death, hepatic decompensation, variceal bleeding alone, and HCC. We used multivariable Cox proportional hazards regression to determine the association between LS measurements and these outcomes.

Results:

Among the 20,776 patients, 95.5% were male and mean age was 63.8 ± 9.6 years. LS values were <20 kPa in 17,175 (82.7%), 20 to <40 kPa in 2,724 (13.1%), 40 to <60 kPa in 470 (2.3%) and 60 to 75 kPa in 407 (2%). The etiology of liver disease was due to chronic hepatitis C infection in 50.2%, non-alcoholic fatty liver disease in 22.2%, and alcohol-related liver disease in 10.6%. Over a mean follow-up of 2.6 ± 1.8 years after LS date, 2,635 patients died, 1,106 had hepatic decompensation (of whom 211 had variceal bleeding), and 815 were diagnosed with HCC.

Compared to those with LS values 20 to <40 kPa, risk of mortality, decompensation, and variceal bleeding was lower for those with LS <20 kPa and higher across increasing LS groups (Table). With regards to HCC incidence, compared to those with LS 20 to <40 kPa (2.6 per 100 Person-Years [P-Ys]), the incidence was higher for those with 40 <60 kPa (3.9 per 100 P-Ys) but not 60 to 75 kPa (2.3 per 100 P-Ys).

On multivariable analysis, compared to the referent LS group 20 to <40 kPa, those in the 40 to <60 kPa and 60 to 75 kPa groups had significantly higher risks of death (aHRs 1.46 [95% CI 1.10-1.94] and 2.04 [95% CI 1.50-2.77]), decompensation (aHRs 2.04 [95% CI 1.46-2.87] and 3.80 [95% CI 2.82-5.11]), variceal bleeding (aHRs 3.43 [95% CI 1.96-5.97] and 2.94 [95% CI 1.66-5.21]), but not HCC (aHRs 1.23 [95% CI 0.82-1.84] and 0.87 [95% CI 0.51-1.47]). Compared to those with LS 20 to <40 kPa, those with LS <20 kPa had significantly lower risks of mortality (aHR 0.62 [95% CI 0.54-0.71]), decompensation (aHR 0.40 [95% CI 0.33-0.49]) and HCC (aHR 0.44 [95% CI 0.35-0.55]).

Conclusion:

Compared to a referent group of 20-40 kPa, higher LS values are associated with increased risk of death, decompensation, and variceal bleeding. Higher LS values beyond this referent group are not associated with an increased risk of HCC. Lower LS (<20 kPa) is associated with lower risk of all outcomes.