Pretreatment gamma-glutamyl transferase predicts mortality in patients with chronic hepatitis B treated with nucleotide/nucleoside analogs.

ABSTRACT

Elevated serum gamma-glutamyl transferase (GGT) levels are associated with chronic hepatitis B (CHB)-related hepatocellular carcinoma. However, their role in predicting mortality in patients with CHB treated with nucleotide/nucleoside analogs (NAs) remains elusive. Altogether, 2843 patients with CHB treated with NAs were recruited from a multinational cohort. Serum GGT levels before and 6 months (Month-6) after initiating NAs were measured to explore their association with all-cause, liver-related, and non-liver-related mortality. The annual incidence of all-cause mortality was 0.9/100 person-years over a follow-up period of 17,436.3 person-years. Compared with patients who survived, those who died had a significantly higher pretreatment (89.3 vs. 67.4 U/L, p = 0.002) and Month-6-GGT levels (62.1 vs. 38.4 U/L, p < 0.001). The factors associated with all-cause mortality included cirrhosis (hazard ratio [HR]/95% confidence interval [CI] 2.66/1.92-3.70, p < 0.001), pretreatment GGT levels (HR/CI 1.004/1.003-1.006, p < 0.001), alanine aminotransferase level (HR/CI 0.996/0.994-0.998, p = 0.001), and age (HR/CI 1.06/1.04-1.07, p < 0.001). Regarding liver-related mortality, the independent factors included cirrhosis (HR/CI 4.36/2.79-6.89, p < 0.001), pretreatment GGT levels (HR/CI 1.006/1.004-1.008, p < 0.001), alanine aminotransferase level (HR/CI 0.993/0.990-0.997, p = 0.001), age (HR/CI 1.03/1.01-1.05, p < 0.001), and fatty liver disease (HR/CI 0.30/0.15-0.59, p = 0.001). Pretreatment GGT levels were also independently predictive of non-liver-related mortality (HR/CI 1.003/1.000-1.005, p = 0.03). The results remained consistent after excluding the patients with a history of alcohol use. A dose-dependent manner of <25, 25-75, and >75 percentile of pretreatment GGT levels was observed with respect to the all-cause mortality (trend p < 0.001). Pretreatment serum GGT levels predicted all-cause, liver-related, and non-liver-related mortality in patients with CHB treated with NAs.