Ki-Chul Sung1, Michael P Johnston2, Mi Y Lee3, Christopher D Byrne4,5. 1. Division of Cardiology, Department of Internal Medicine, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea. 2. Department of Hepatology, University Hospital Southampton NHS Foundation Trust, Southampton, UK. 3. Division of Biostatistics, Department of R&D Management, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea. 4. Human Development and Health, Faculty of Medicine, University of Southampton, Southampton, UK. 5. National Institute for Health Research Southampton Biomedical Research Centre, University of Southampton and University Hospital Southampton National Health Service (NHS) Foundation Trust, Southampton, UK.
Abstract
BACKGROUND & AIMS: In a general population without known liver disease, we tested whether: (a) increased liver fibrosis scores (FIB-4 and APRI) are associated with liver cancer mortality and (b) the probability that a person with a higher score died of liver cancer. METHODS: In a retrospective occupational cohort who underwent annual/biennial health examinations (between 2002 and 2015), subjects were excluded with known chronic liver disease. Based on their baseline FIB-4 and APRI scores, subjects were categorised in low-/intermediate-/high-risk groups for advanced liver fibrosis. Using Cox proportional hazards regression analyses adjusted hazard ratios (aHR) were estimated for liver cancer mortality, with the low-risk FIB-4/APRI group as the reference. Harrell's C statistics were also calculated. RESULTS: In 200 479 participants, mean (SD) age was 36.4 (7.7) years. Median follow-up was 4.1 years (IQR 2.10-8.03) with 80 liver cancer deaths. High baseline FIB-4 or APRI scores occurred in 0.25% and 0.09% of subjects respectively. A high FIB-4 or APRI score was associated with a markedly increased risk of liver cancer mortality (aHRs 629.10 [95% CI 228.74-1730.20] and 80.42 [95% CI 34.37-188.18]) respectively. C statistics were FIB-4 = 0.841 (95% CI 0.735-0.946) and APRI = 0.933 (95% CI 0.864-0.999). CONCLUSIONS: In a general population without known liver disease, high FIB-4 or high APRI (in keeping with a high probability of advanced fibrosis) occurred in 0.25% (FIB-4) and 0.09% (APRI) of subjects. Both scores were associated with a markedly increased risk of liver cancer mortality and FIB-4 and APRI models both strongly predicted liver cancer mortality.
BACKGROUND & AIMS: In a general population without known liver disease, we tested whether: (a) increased liver fibrosis scores (FIB-4 and APRI) are associated with liver cancer mortality and (b) the probability that a person with a higher score died of liver cancer. METHODS: In a retrospective occupational cohort who underwent annual/biennial health examinations (between 2002 and 2015), subjects were excluded with known chronic liver disease. Based on their baseline FIB-4 and APRI scores, subjects were categorised in low-/intermediate-/high-risk groups for advanced liver fibrosis. Using Cox proportional hazards regression analyses adjusted hazard ratios (aHR) were estimated for liver cancer mortality, with the low-risk FIB-4/APRI group as the reference. Harrell's C statistics were also calculated. RESULTS: In 200 479 participants, mean (SD) age was 36.4 (7.7) years. Median follow-up was 4.1 years (IQR 2.10-8.03) with 80 liver cancer deaths. High baseline FIB-4 or APRI scores occurred in 0.25% and 0.09% of subjects respectively. A high FIB-4 or APRI score was associated with a markedly increased risk of liver cancer mortality (aHRs 629.10 [95% CI 228.74-1730.20] and 80.42 [95% CI 34.37-188.18]) respectively. C statistics were FIB-4 = 0.841 (95% CI 0.735-0.946) and APRI = 0.933 (95% CI 0.864-0.999). CONCLUSIONS: In a general population without known liver disease, high FIB-4 or high APRI (in keeping with a high probability of advanced fibrosis) occurred in 0.25% (FIB-4) and 0.09% (APRI) of subjects. Both scores were associated with a markedly increased risk of liver cancer mortality and FIB-4 and APRI models both strongly predicted liver cancer mortality.