Chronic hepatitis B (CHB) affects 257million individuals worldwide. In 2017, it caused around 39.7 million cases of cirrhosis and 0.4 million cirrhosis-related deaths in 2017. However, there is no specific treatment for liver fibrosis/cirrhosis. Although nucleos(t)ide analogues (NAs) profoundly suppress viral replication, fibrosis/cirrhosis progression can still occur in NA-treated patients.

Sodium-glucose cotransporter type-2 (SGLT2) inhibitors are antidiabetic drugs that may prevent fibrosis/cirrhosis progression by reducing hepatic steatosis/inflammation, dampening renin-angiotensin aldosterone system (RAAS) activation, and reducing fluid retention, effects of which are independent of glycemic control. Clinical studies in diabetic patients show SGLT2 inhibitors reduce hepatis steatosis/inflammation, regress ascites (a cirrhotic complication), and improve liver function parameters and survival prognosis in terms of model for end-stage liver disease (MELD) score. There are currently no randomized controlled trials (RCTs) on role of SGLT2 inhibitors in preventing fibrosis/cirrhosis progression in CHB patients. Magnetic resonance elastography (MRE) and transient elastography (TE) are non-invasive techniques for liver stiffness measurement (LSM), although MRE is more accurate than TE.

The investigators propose a double-blind, randomized, placebo-controlled trial to compare effect of empagliflozin (an SGLT2 inhibitor) with placebo (1:1 ratio) in preventing fibrosis progression in both diabetic and non-diabetic NA-treated CHB patients with significant/advanced fibrosis or compensated cirrhosis. 108 patients will be randomly sampled from our pre-existing TE database. Empagliflozin 10mg daily will be given to treatment arm. Placebo pills will be manufactured identical in appearance to empagliflozin. Subjects will receive active or placebo pills for three years, and undergo clinical, anthropometric and laboratory assessments (at baseline, weeks 8, 16, and every 4 months thereafter). They will undergo LSM by TE at baseline, end of first, second and third year, and by MRE at baseline and end of third year. Primary outcome is difference in change to liver stiffness (measured by MRE) from baseline between the two groups at the end of third year.

The study results will determine whether SGLT2 inhibitors can prevent hepatic fibrosis/cirrhosis progression in NA-treated CHB patients.

Chronic hepatitis B (CHB) affects 257million individuals worldwide. In 2017, it caused around 39.7 million cases of cirrhosis and 0.4 million cirrhosis-related deaths in 2017. However, there is no specific treatment for liver fibrosis/cirrhosis. Although nucleos(t)ide analogues (NAs) profoundly suppress viral replication, fibrosis/cirrhosis progression can still occur in NA-treated patients.

Sodium-glucose cotransporter type-2 (SGLT2) inhibitors are antidiabetic drugs that may prevent fibrosis/cirrhosis progression by reducing hepatic steatosis/inflammation, dampening renin-angiotensin aldosterone system (RAAS) activation, and reducing fluid retention, effects of which are independent of glycemic control. Clinical studies in diabetic patients show SGLT2 inhibitors reduce hepatis steatosis/inflammation, regress ascites (a cirrhotic complication), and improve liver function parameters and survival prognosis in terms of model for end-stage liver disease (MELD) score. Our preliminary data from a territory-wide electronic healthcare database shows SGLT2 inhibitors were associated with 58% reduction in risk of cirrhosis development over three years among CHB patients with diabetes (n=9,502). There are currently no randomized controlled trials (RCTs) on role of SGLT2 inhibitors in preventing fibrosis/cirrhosis progression in CHB patients. Magnetic resonance elastography (MRE) and transient elastography (TE) are non-invasive techniques for liver stiffness measurement (LSM), although MRE is more accurate than TE.

The novelty of utilizing "drug repositioning" by changing role of SGLT2 inhibitors in treating diabetes mellitus (DM) to preventing fibrosis/cirrhosis progression in CHB deserves exploration. We propose a double-blind, randomized, placebo-controlled trial to compare effect of empagliflozin (an SLGT2 inhibitor) with placebo (1:1 ratio) in reducing liver stiffness in both diabetic and non-diabetic NA-treated CHB patients with significant/advanced fibrosis or compensated cirrhosis. 108 patients will be randomly sampled from our pre-existing TE database. Empagliflozin 10mg daily will be given to treatment arm. Placebo pills will be manufactured identical in appearance to empagliflozin. Subjects will receive active or placebo pills for three years, and undergo clinical, anthropometric and laboratory assessments (at baseline, weeks 8, 16, and every 4 months thereafter). They will undergo LSM by TE at baseline, end of first, second and third year, and by MRE at baseline and end of third year. Primary outcome is difference in change to liver stiffness (measured by MRE) from baseline between the two groups at the end of third year. Secondary outcomes are remission of advanced fibrosis/cirrhosis, progression of advanced fibrosis to cirrhosis, improvement of laboratory results (liver transaminases, ferritin, glucose, lipid profiles) and anthropometric measurements.

The study results will determine whether SGLT2 inhibitors can prevent hepatic fibrosis/cirrhosis progression in NA-treated CHB patients.

Inclusion Criteria:

• Patients will be recruited if they have significant/advanced fibrosis or cirrhosis confirmed by MRE

Exclusion Criteria:

1. decompensated cirrhosis (variceal bleeding, ascites, hepatic hydrothorax, hepatic encephalopathy),
2. portal vein thrombosis,
3. alcohol intake >20g within last 2 years,
4. concurrent chronic liver disease (chronic hepatitis C infection, autoimmune hepatitis, Wilson's disease, hemochromatosis, primary biliary cholangitis, drug-induced),
5. history of malignancy including hepatocellular carcinoma (HCC),
6. pregnancy,
7. contraindications to empagliflozin (estimated glomerular filtration rate (eGFR) <45mL/min/1.73m2, recurrent genitourinary tract infections, gangrene, allergy),
8. contraindications to MRI (e.g., claustrophobia, implanted devices with ferromagnetic properties).

Individual participant data (IPD) will be made available in the form of excel files upon request by other researchers

The placebo pills will be manufactured in identical appearance to the study drug (empagliflozin)