Study design and participants
This is a prospective, open-labeled and randomized controlled study registered at ClinicalTrial.gov (Number NCT01220492) and authorized by the General Logistic Ministry of Healthy, China. We carried out this study at a single center, Beijing 302 hospital, in the North of China. The study protocol was conformed strictly to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Ethics Committee of our hospital.
Decompensated liver cirrhosis is characterized by a series of clinical manifestations, including gastrointestinal bleeding, hepatic encephalopathy, jaundice and ascites, based on previously stable cirrhosis. Therefore, the criteria for enrollment of patients in this study included: manifestation of decompensated liver cirrhosis; 18-65 years of age, and positive testing for serum hepatitis B surface antigen (HBsAg) for more than 6 months, negative pregnancy test (female patients in fertile age), willing to participate in this clinical trial, able to understand and sign informed consent. The exclusion criteria were as follows: the presence of underlying neoplasms; evidence of significant extrahepatic biliary diseases; active thrombosis of the portal or hepatic veins; renal or respiratory failure; chronic HCV infection; other liver diseases including alcoholic liver disease, non-alcoholic fatty liver disease, drug-induced hepatitis, autoimmune hepatitis, Wilson disease, and hemochromatosis; diabetes mellitus; severe cardiovascular disease and drug-uncontrolled hypertension; the presence of severe comorbidities; pregnant woman; lack of a supportive family; and refusal to sign the informed consent form. All patients signed a written informed consent form in accordance with the Institutional Review Board guidelines.
Randomization
Patients were randomly allocated to the conventional treatment (control group) or UC-MSC infusion group (UC-MSC treatment group) at a ratio of 1:1 according to a computer-generated randomization list. Before treatment, all patients were assigned a unique randomization code. The randomization system allocated patients to receive either the conventional treatment or three UC-MSC infusions at 4-week intervals plus conventional treatment. This was an open-label study. Thus, both clinicians and participants were aware of the treatment allocation.
Procedures
UC-MSC were prepared in an approved good manufacturing practice (GMP)-compliant facility and identified as described previously 14 In brief, with the written consent of the maternity patients, fresh human umbilical cords were obtained after birth and collected in cold Xeno-free MesenCult-XF medium (STEMCELL Technologies Inc., Canada). The mesenchymal tissues in umbilical cord Wharton’s jelly were diced into cubes of approximately 0.5 cm3 and washed with Hanks’ balanced saline solution (Gibco Invitrogen). Then, the tissue pellets were seeded in a tissue culture flask (Corning Enterprises, Corning, NY) in α-MEM supplemented with 10% fetal bovine serum (STEMCELL Technologies, Vancouver, BC, Canada). The medium was replenished every 3 days. The UC-MSC were cultured and collected between the third and fourth passages for infusion.
For quality control of the UC-MSC, the cultured cells at the third passage were examined for the expression of their phenotypes by flow cytometric analysis, for example, high expression of CD44, CD73, CD90, and CD105, but no expression of CD31, CD34, CD45, or HLA-DR. Alternatively, the digested cells were cultured in conditioned medium (STEMCELL Technologies, Vancouver, BC, Canada), and were subsequently cultured for osteogenesis and adipogenesis differentiation assays. Briefly, the digested cells were cultured in osteogenesis-inducing medium (Gibco Invitrogen) consisting of MesenCult basal medium, osteogenic stimulatory supplements, dexamethasone, β-glycerophosphate and ascorbic acid ion for five weeks, or in adipogenic induction medium (Gibco Invitrogen) consisting of MesenCult basal medium and MesenCult adipogenic stimulatory supplements for two weeks. Subsequently, osteogenesis was assessed by alkaline phosphatase staining, and adipogenesis was assessed by oil red O staining. Moreover, the UC-MSC were negative for all the tested contaminants before infusion, including Mycoplasma spp., gram-positive and gram-negative bacteria, and fungi. The endotoxin levels were below 5 EU/kg and viability was >80%.
Patients in the control group received the conventional treatment which included on targeting abnormalities in gut-liver axis by antibiotic administration (i.e. rifaximin), improving the disturbed systemic circulatory function (i.e. longterm albumin administration), decreasing the inflammatory state (i.e. statins), and reducing portal hypertension (i.e. beta-blockers) be used to decrease cirrhosis progression in patients with decompensated cirrhosis. Management of specific complications of decompensated cirrhosis is followed by the practice guidelines for each complications of decompensated cirrhosis (such as variceal bleeding, ascites and encephalopathy).
In the UC-MSC treatment group, patients received UC-MSC therapy plus the conventional treatment. The UC-MSC therapy was administrated at a dose of approximately 0.5 × 106/kg body weight UC-MSC at the fourth passage, which were suspended in saline and were infused intravenously three times at 4-week intervals (Fig. 1). The patients in the UC-MSC treatment group were observed for 6 h after the UC-MSC therapy before discharged from the clinic.
After first UC-MSC infusion, the patients were followed up at weeks 2, 4, 8, 12, 24, 48, and at months 24, 48, 60, 75. The following tests were performed at weeks 0, 12, 24, 48: liver function tests: serum ALB, rothrombin activity (PTA), cholinesterase (CHE) and TBIL. The adverse events were collected during 48 weeks after the first infusion of UC-MSC. The adverse events and their severity were determined by means of vital signs and physical examinations, laboratory examinations, oral and telephone inquiries.
Outcome measures
The primary outcome measures were overall survival rate and the hepatocellular carcinoma (HCC)-free survival rate after UC-MSC infusions in each treatment group. The secondary outcome measures included the incidence of adverse events (e.g., fever, allergy, rash, and infection), effects of the treatment on liver function (including the levels of albumin [ALB], prothrombin activity [PTA], TBIL, and CHE), and the incidence of serious complications (including infection, gastrointestinal bleeding, encephalopathy, and hepatorenal syndrome).
Statistical analysis and sample size
The median overall survival was assumed to be 60 months in 60% of the patients who received the conventional medical treatment, and in 85% of those who received the UC-MSC therapy. We planned to recruit 100 patients in each group during a recruitment period of 40 months and follow for over 75 months, which would have more than 80% power to detect a significant risk difference between the two groups, at a two-sided a level of 0.05 (PASS 11.0). Based on the dropout or withdrawal rate, we determined a sample size of 125 patients per group.
All eligible participants who received the assigned treatment and were followed up were included in the analysis. We counted the number of endpoint events (death or HCC events) that occurred during the follow-up period and compared the event-free survival rate between the two groups by using Kaplan–Meier analysis. The post-treatment follow-up was continued until the date of an endpoint event, death, or the end of the study, whichever occurred first. Cox proportional-hazards models were developed to estimate the hazard ratio (HR) and 95% confidence interval (CI) for between-group comparisons with or without adjustments for stratification baseline model of end-stage liver disease (MELD) scores. Additionally, we performed landmark analyses to assess the endpoint events accordingly, with the hazard ratio calculated separately for events that occurred before or after the landmark points. Hypothesis testing was two-sided with an α value of 0.05. The χ² test or Fisher’s exact test was used to analyze categorical data, while the Student’s t test or paired Student’s t test was used to analyze continuous data.