Treatment cessation in chronic hepatitis B is associated with high rates of disease relapse. However patients who achieve the seroclearance of hepatitis B surface antigen (HBsAg) (<0.05 IU/mL) show good off-treatment durability after treatment cessation. Through the quantification of HBsAg, the study aims to investigate how low should quantitative HBsAg be before once can achieve successful disease control after treatment cessation.

The treatment paradigm of chronic hepatitis B (CHB) has been transformed by the introduction of nucleoside analogue (NA) therapy. Long-term NA therapy can suppress viral replication, improve liver histology, and reduce the risk of liver-related complications and mortality. For the large majority of CHB patients, NAs need to be taken long-term, since the occurrence of hepatitis B surface antigen (HBsAg) seroclearance, the established treatment endpoint of CHB, is a rare event. Nonetheless, once HBsAg seroclearance is achieved, virologic suppression and improvement in clinical outcomes remain sustained in almost all patients even after treatment discontinuation.

So far, attempts to discontinue NA therapy before HBsAg seroclearance had yielded variable results. Concerning hepatitis B e antigen (HBeAg)-negative CHB, prior studies found rates of disease relapse to range from 45% to 66% after treatment cessation. A recent multicenter prospective study involving three Hong Kong institutes (including our center) and 184 patients found the 48-week cumulative rate of virologic relapse to be 91.4%. While variations in relapse rates could be explained by the difference in study design and definition of endpoints, an important factor which could play a role is the serum HBsAg level of recruited subjects.

The quantification of serum HBsAg has recently been advocated as a marker of disease monitoring for CHB [10]. Low levels of serum HBsAg are associated with good immune control of the hepatitis B virus (HBV), with serum HBsAg levels of 100-200 IU/mL being predictive of eventual HBsAg seroclearance in treatment-naïve CHB. Nonetheless , in the abovementioned multicenter prospective study, only 4.8% and 14.1% of study subjects had HBsAg levels <100 IU/mL or <200 IU/mL at the point of treatment cessation - which could explain the study's high rate of disease relapse. In contrast, in another study of retrospective nature which found a lower relapse rate of 66%, 27.6% of patients had HBsAg levels <200 IU/mL at treatment cessation.

Although HBsAg seroclearance is the established treatment endpoint of CHB, patients would still have low amounts of intrahepatic HBV DNA. Moreover, HBsAg seroclearance via a conventional assay (HBsAg <0.05 IU/mL) does not mean absence of serum HBsAg, but merely serum HBsAg at undetectable levels. Using more sensitive assays, detectable HBsAg can be found in 29.1% to 53.2% of such patients. So despite the continued presence of HBV, the off-treatment durability of HBsAg seroclearance would suggest for treatment cessation, viral replication and viral protein production do not need to be totally absent, but at sufficiently low amounts to permit successful host immune control.

Hence, a relevant clinical question would be:

1. How low must serum HBsAg be for successful off-treatment durability? Must the "magic figure" of 0.05 IU/mL be reached? Or could an HBsAg level in the >0.05 - 200 IU/mL range be sufficient? This could have implications in drug budgeting. According to data retrieved from the Clinical Data and Analysis Reporting System (CDARS), Hospital Authority, NA prescription headcount in 2014 reached 36,287 prescriptions, with an estimated total drug expenditure of 381 million Hong Kong dollars. A finite duration of therapy for certain patients could help in the reduction of drug expenditure.

   Other clinical questions include:

2. Could a longer duration of NA therapy enhance the success of off-treatment durability? For many available therapy cessation studies, the mean NA therapy duration was approximately 1.5-3 years. There is emerging evidence demonstrating intrahepatic covalently closed circular DNA (cccDNA) could actually decrease profoundly after more than 5 years of NA therapy ; a low intrahepatic HBV cccDNA level could be the key to successful off-treatment durability.

3. Is there any clinical factor that can reliably predict off-treatment virologic relapse? A potential novel marker is the hepatitis B core-related antigen (HBcrAg), which detects an identical amino-acid sequence shared by HBeAg, hepatitis B core antigen and 22 kDa precore protein . Serum HBcrAg levels reflect the level of HBV disease activity [15], predict the occurrence of liver-related complications , and also correlate well with intrahepatic HBV cccDNA levels , the master intrahepatic reservoir for the production of HBV virions. Low levels of HBcrAg could reflect low levels of intrahepatic HBV cccDNA, meaning a better chance of maintaining off-treatment durability.

Aims and Hypotheses to be Tested:

Primary:

1. To determine the optimal HBsAg level for cessation of NA therapy in CHB

   Secondary:

2. To determine the rate of virologic relapse in after treatment cessation in CHB patients with serum HBsAg levels ≤200 IU/mL

3. To determine the role of serum HBcrAg in predicting virologic relapse after treatment cessation.

Study Design We propose a prospective observational study following the STROBE guidelines in which we will follow-up study subjects after cessation of NA for 48 weeks. HBsAg-positive patients from our clinic taking either entecavir or tenofovir therapy will be tested for HBsAg levels and screened for study eligibility. Those with serum HBsAg levels <200 IU/mL and fulfilling the inclusion and exclusion criteria listed above will be seen by one of the study investigators, who would explain the study objectives and offer the option of treatment discontinuation.

All enrolled study subjects, after cessation of therapy, will be evaluated at baseline, week 6, 12, 18, 24, 36 and 48, with liver biochemistry, alpha-fetoprotein, HBV DNA, HBsAg and HBcrAg levels determined at each clinic visit. A regular 6-week follow-up will be in place for the first 24 weeks, since approximately 75% of all virologic relapse occurs in the first 24 weeks - the regular monitoring will allow detection of relapse at an early stage. In addition, a dedicated research assistant will follow-up all study subjects and assist on all monitoring logistics. A telephone hotline will be provided to all study subjects if they have any additional queries.

For every HBV DNA measurement of >2,000 IU/mL, as second HBV DNA will be measured after 2 weeks. The primary outcome of the study will be virologic relapse, defined as serum HBV DNA >2,000 IU/mL in both 2 measurements of 2 weeks apart, regardless of serum ALT levels, in which the original NA (entecavir or tenofovir) will be recommenced. We chose HBV DNA >2,000 IU/mL as the threshold for restarting NA as this is the level in which CHB patients would need to be considered for treatment.

Secondary outcomes of the study include:

1. HBV DNA <200 IU/mL at week 48
2. Undetectable HBV DNA (<20 IU/mL) at week 48
3. HBsAg seroclearance (<0.05 IU/mL) at week 48

NA treatment could also be resumed at the discretion of the study investigators based on special clinical consideration, e.g. the sudden diagnosis of malignancy requiring chemotherapy. Patients will be always given the option of dropping out from the study; nonetheless we do not expect a high dropout rate (3.3% in our previous study).

Inclusion Criteria:

1. Age ≥18 years.

2. HBsAg-positive patients on entecavir or tenofovir therapy

3. Fulfill the European Association for the Study of the Liver of NA cessation:

   • HBeAg-positive at NA commencement: stable HBeAg seroconversion and undetectable HBV DNA and completed 12 months of consolidation therapy
   • HBeAg-negative: Virological suppression with undetectable HBV DNA (<10 IU/mL) for more than 3 years.

4. Normal levels of serum alanine aminotransferase (ALT) documented on two separate occasions 6 months apart.

5. Serum HBsAg between <200 IU/mL prior to NA cessation.

(At study commencement, in HBeAg-negative patients, we applied the APASL 2016 suggestion for at least 2 years with undetectable HBV DNA documented on three separate occasions 6 months apart. Upon publication of the EASL guidelines in 2017, we further scrutinized the inclusion criteria for HBeAg-negative patients to be viriological suppression with undetectable HBV DNA for more than 3 years. All recruited HBeAg-negative participants prior to the release of the EASL 2017 guidelines fulfilled the updated criteria).

Exclusion Criteria:

1. Concomitant liver diseases including chronic hepatitis C and D infection, Wilson's disease, autoimmune hepatitis, primary biliary cirrhosis and primary sclerosing cholangitis.

2. Significant alcohol intake (>30 grams per day).

3. Prior history of hepatocellular carcinoma (HCC) or any radiologic suspicion of HCC.

4. Decompensated liver disease (defined as Child's B or C cirrhosis), or presence of cirrhotic complications, including variceal disease, ascites, or history of hepatic encephalopathy.

5. Patient previously or currently on interferon therapy.

6. History of immunosuppressive therapy or organ transplantation.

7. Serious medical illness or malignancy.

   1. Patient previously or currently prescribed interferon therapy.
   2. Confirmed or radiologic suspicion of HCC.
   3. Serious medical illness or malignancy. -