Background: Acute-on-chronic liver failure (ACLF) is a syndrome characterized by acute deterioration in the setting of chronic liver disease associated with high short-term mortality. Currently, there is no specific treatment for patients with ACLF. Our previous results showed that Chinese herbal medicine (CHM) could reduce the mortality rate and the incidence of complications of ACLF effectively. In this study, we aim to conduct the multi-center randomized controlled trial to evaluate the effect of unified CHM formulas and provide propagable and high-level evidence for clinical practice.

Methods/design: This is a prospective, multicenter, centrally randomized controlled trial. Five hundred and ten patients diagnosed with HBV-related ACLF will be allocated in a 1:1 ratio to SMT group (standard medical therapy) and CHM group (CHM and SMT). The primary outcome is the transplant-free mortality rates at week 4, 8, 12, 24 and 48. Secondary outcomes include (1) the incidence of adverse reactions, (2) influence on liver function, (3) the incidence of serious complications and (4) the level of inflammatory cytokines.

Discussion: The effectiveness and safety of CHM formulas are assessed in the treatment of ACLF.

Study design This prospective, multi-center, parallel, centrally randomized controlled trial was designed in accordance with the Consolidated Standards of Reporting Trials (CONSORT) and Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guidelines, which is led by the principal investigator from 302 military hospital in Beijing, will be carried out from September 2018 and December 2020 among 16 tertiary academic hospitals in China. Patients with hepatitis B virus (HBV) related acute-on-chronic liver failure (ACLF) will be randomly allocated to the CHM group or the SMT group in a 1:1 ratio. The last patient will be enrolled before August 2019 to make sure the availability of 48 week's survival status by the end of this study. We will decide whether to expand additional institutions to guarantee adequate enrolments after interim assessment of effective participant numbers around December 2018. Patients in the SMT group will receive the standard medical treatment, and patients in the CHM group will receive rhubarb decoction enema for 1 week and CHM for 8 weeks additionally. All patients will be followed at pre-designed time points after 8 weeks' observational period until August 30, 2020 for the primary outcome and secondary outcomes.

The study is consistent with Good Clinical Practice (GCP) guidelines and has been approved by 302 military hospital ethics review committee(2018007D). Written informed consent is requisite to be signed by the patients or the authorized person under the willingness. The privacy and data of all participants will be protected carefully before, during and after the trial.

Patients will be consented and enrolled by co-principal investigators at each institution and managed by care providers including residents, attending physicians and nurses according to the protocol. Most of the patients will be treated in hospital for 8 weeks so clinicians can supervise the administration of CHM, in addition, the policy of reimbursement filed by the lab results at determined time points will boost the regular follow-up of this study.

Diagnostic criteria The diagnosis of HBV-ACLF refers to the guideline of liver failure issued by Chinese Group on the Study of Severe Hepatitis B (COSSH) in 2012.

The traditional Chinese medicine (TCM) syndrome was judged by two qualified senior TCM physicians simultaneously based on the published criteria to decide what CHM will be given, and the discrepancy will be solved by consulting another senior physician.

Criteria for discontinuing or modifying allocated interventions The study will end at once if severe adverse events associated with CHM occurs or sufficient evidence indicates that adding CHM is not superior to the SMT group. Besides, CHM will discontinue and the final status will be censored in the condition of liver transplantation or the onset of liver cancer. Allocated interventions can be changed upon participant's request. All these changes will be recorded and analyzed based on intention-to-treat (ITT).

Intervention Standard medical treatment The standard medical treatment (SMT) include a high-calorie diet; nucleoside analogues for HBV DNA-positive patients; sodium restriction, diuretics and paracentesis combined with albumin infusion for ascites; lactulose and L-ornithine aspartate and lactulose for hepatic encephalopathy (HE) and hyper-ammonia; hemostatic treatment for gastrointestinal hemorrhage; antibiotics for infections and renal replacement for hepatorenal syndrome and uremic symptoms. CHM formulas, whatever routes of administration, are prohibited in the SMT group.

Chinese herbal medicine formula There are two kinds of CHM based on TCM syndromes. The common components of the two CHM formulas include Artemisiacapillaris Thunb (30g), Salvia miltiorrhiza Bge (30g), Rhizoma Atractylodis Macroce- phalae (30g), Rubia cordifolia L (30g), Sieyesbeckiaorientalis L (30g). Additional four components of Paeoniae Radix Rubra (60g), Gardenia jasminoides Ellis (9g), Hedyotis diffusa Willd (30g), Bletilla striata (15g) are involved for excess syndrome, and five other components of Astragalus membranaceus (30g), Radix Pseudostellariae (15g), Radix Aconiti Lateralis Praeparata (10g), Galli Gigeriae Endothelium Corneum (20g), Polygonum cuspidatum Sieb.et Zucc (15g) are involved for deficiency syndrome. All ingredients are manufactured as a Chinese herbal granule in a specific weight ratio to the raw herb (Beijing Kang Rentang Pharmaceutical Co., Ltd, Beijing, China) in accordance with the quality criteria of Chinese pharmacopoeia in 2015. Patients will take CHM for 8 weeks. For patients who cannot take oral medicine can be changed to nasal feeding or colon route administration (colonic therapy system, IMS-100A,produced by the company of Sunny Medical,Beijing,China).

In addition, patients in the CHM group will also receive daily herbal enema of 100 microliter decoction composed by Radix et Rhizoma Rhei Palmata 30g and Fructus Mume 30g for 7 days expect the following contraindications: (1) severe hemorrhoid, (2) female patients during menstruation, (3) anal stenosis, (4) artificial anal and (5) cardiac disease.

Safety outcomes Primary vital signs, physical examinations, electrocardiogram and some laboratory tests such as routine blood test, urine test, renal function will be assessed before and after treatment to monitor the safety of this study. Furthermore, adverse events are also recorded in a predesigned case report form (CRF) through the whole study period. An independent safety committee will analyze the incidence and severity of adverse events to identify any unexpected adverse events or mortality in the CHM groups.

Other assessed parameters Epidemiological characteristics, the details of nucleoside/nucleotide analog therapy, family history of liver disease were collected from patient medical record, the level of HBV DNA, endotoxin, and lymphocyte subsets were tested as well as Child-Pugh score and the model of end-stage liver disease (MELD) score were computed at week 4, 8, 12, 24 and 48. The plasma level of HBV DNA, endotoxin, and cytokines were tested in a third-party clinical laboratories which is blind to the allocation information. Remnant plasma will be stored at -80˚C for ancillary studies. Women of childbearing age will undergo pregnancy test for the eligibility.

Sample size The proposed -12% superiority margin of 8-week mortality rate was chosen based on our previous results of 30.87% in the SMT group. Considering a power of 0.80 and 10% of attrition rate, 255 subjects are needed for each group to achieve the significance level of 0.05 in assessing the difference between the two groups.

Randomization The eligible patients will be randomly allocated to the SMT group and CHM group in a 1:1 ratio. The sealed numbers are generated by a computerized random number generator with statistical analysis system (SAS) software by an independent statistician from the center of clinical evaluation in China Academy of Chinese medical sciences (Beijing, China) and are concealed using opaque envelopes.

Trial monitoring Trial monitoring is assumed by an independent Contract Research Organization so as to verify that: (1) The rights and well-being of human subjects are protected. (2) The reported trial data are accurate, complete, and verifiable from source documents. (3) The conduct of the trial is in compliance with the currently approved protocol/amendment(s), with GCP, and with the applicable regulatory requirement(s). Monitors should be appropriately trained and have the scientific and/or clinical knowledge needed to monitor the trial adequately. A risk-based monitoring strategy is developed in this trial, which includes on-site and centralized monitoring, which is more likely than routine visits to all clinical sites and 100% data verification to ensure subject protection and overall study quality. On-site monitoring is an in-person evaluation carried out by the monitor at the sites at which the clinical investigation is being conducted. Centralized monitoring is a remote evaluation carried out at a location other than the sites at which the clinical investigation is being conducted. Key monitoring processes include subject inclusion process; implementation of the inclusion criteria; source data verification; data integrity, etc.

Data management and Statistical methods Data entry will be conducted by two independent administrators using online electronic CRF for accuracy and consistency. At the end of the study, the database will be locked and analyzed under the agreement of principal investigator and statistician, who are privileged to access the final trial dataset.

The randomized subjects who take at least one dose of CHM or SMT constitute the full analysis set (FAS) and those who complete all the visits constitute the per-protocol set (PPS). An efficacy assessment will be analyzed using FAS and PPS based on the principle of ITT. Subjects who have taken the CHM and been evaluated the safety at least once constitute the safety set (SS), which is used for safety evaluation. The statistical analysis will be performed using SAS 9.0 software by independent data management committee who are blind to the interventions.

Comparison of the two groups is conducted for the continuous variables with t-test and the categorical variables or rates with the chi-square test. The Kaplan-Meier overall survival curves will be generated and tested by log-rank. Cox proportional hazards regression model will be adopted to determine the prognostic factors associated with transplant-free survival. Multivariate logistic regression analyses will be adopted to explore factors associated with the incidence of liver-related complications.

The last observation carried forward approach was used for missing values. The statistical significance is defined as a two-sided P <0.05.

Inclusion Criteria:

• In-patients with acute-on-chronic liver failure;
• Patients with chronic hepatitis B or compensatory cirrhosis of hepatitis B;
• Patients who is willing to sign the informed consent;
• Patients aged among 16~65 years old.

Exclusion Criteria:

• Acute liver failure, or sub acute liver failure, or chronic liver failure;
• Acute-on-chronic liver failure not caused by hepatitis B but other disease such as autoimmune, drug, alcohol, toxic, parasites;
• Pregnant or lactating women;
• Primary liver cancer;
• Combined with other severe systematic disease and mental disease;
• Anti-HIV positive or combined with infection of hepatitis A, C, D, E virus or cytomegalovirus, Epstein-Barr virus;
• Enrolled in other clinical studies in last three months;
• Person unwilling to cooperate; 9. poor compliance, unable guarantee completing the protocol;
• Complicated with severe cerebral edema, severe infection, type I liver-kidney symptom, and gastrointestinal hemorrhage.