Mesenchymal stem cell therapy in decompensated liver cirrhosis: a long-term follow-up analysis of the randomized controlled clinical trial

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.

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.

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.

UC-MSC was 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 was 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 5 weeks, or in adipogenic induction medium (Gibco Invitrogen) consisting of MesenCult basal medium and MesenCult adipogenic stimulatory supplements for 2 weeks. Subsequently, osteogenesis was assessed by alkaline phosphatase staining, and adipogenesis was assessed by oil red O staining. Moreover, the UC-MSC was 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., long-term 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, and 48, and at months 24, 48, 60, and 75. The following tests were performed at weeks 0, 12, 24, and 48: liver function tests: serum ALB, prothrombin 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.

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).

All eligible participants who received the assigned treatment and 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 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. In addition, 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.

A total of 252 decompensated liver cirrhosis patients who had chronic HBV infection were recruited and screened between October 15, 2010 and October 30, 2017. Among them, two patients were excluded because they refused to participate (n = 1), or failed to meet the inclusion criteria (n = 1). Therefore, 250 patients were randomly assigned to the control and UC-MSC treatment groups (n = 125 in each group). In the control group and UC-MSC treatment group, 10 and 10 patients withdrew the consent, and 4 and 7 patients were lost to follow-up, respectively. Therefore, 111 control patients and 108 UC-MSC treated patients were included in the primary analysis and followed up for 75 months (Fig. 2). The median follow-up of patients was 61 months in UC-MSC treatment group (range 1–85 months), and 57 months in control group (range from 2 to 82 months).

The baseline clinical parameters of the patients in the two groups are shown in Table 1. The baseline levels of total bilirubin (TBIL) (p < 0.001) and MELD score (p = 0.040) in the UC-MSC treatment group were higher than those in the control group; in addition, the baseline level of cholinesterase (CHE) in the UC-MSC treatment group was lower than that in the control group (p = 0.008). The other parameters at the baseline were comparable between the two groups. The patients in the UC-MSC treatment group were administered UC-MSC infusions, whereas the patients in the control group were administered the same volume of saline. All participants received the same conventional treatment throughout the study; for example, all the CHB patients received antiviral therapy with entecavir (0.5 mg daily), and the serum HBV DNA levels were below detection limit level (< 100 IU/mL) when enrolled in this study. In addition, patients with ascites in both the UC-MSC and control groups received the same dose of diuretics (spironolactone [40 mg daily] plus furosemide [20 mg daily]) until the ascites was treated. Other medications that might affect the outcome were not given in this study.

At the 75-month follow-up, the overall survival rate did not differ significantly between the UC-MSC treatment group and control group (Fig. 3a) (HR = 1.89, 95% CI 1.01–3.58, p = 0.051). While using Landmark analysis, the overall survival rate was significantly higher in the UC-MSC treatment group than in the control group during the 13- to 75-month follow-up (Fig. 3b) (HR = 3.68, 95% CI 1.38–9.85, p = 0.005), although the overall survival rate did not differ significantly between the UC-MSC treatment group and control group within the first 13-month follow-up (HR = 0.97, 95% CI 0.40–2.32, p = 0.930). These results indicate that although the UC-MSC treatment can significantly increase the survival rate in patients with decompensated liver cirrhosis, the superiority of the UC-MSC treatment to the conventional medical treatment may be evident only about 13 months after UC-MSC infusion.

To investigate the impact of UC-MSC infusion on liver function, the serum levels of ALB, PTA, CHE and TBIL were monitored within 48 weeks after the first UC-MSC infusion or conventional medical treatment. The levels of ALB had increased in comparison with the baseline in both groups during the 48-week follow-up, but it was significantly higher in the UC-MSC treatment group at the 24- and 48-week time points (Fig. 4a). The levels of PTA in the UC-MSC treatment group were markedly increased as compared with the baseline and control group during the 48-week follow-up (Fig. 4b). Astonishingly, the baseline level of CHE in the UC-MSC treatment group was obviously lower and the level of TBIL was obviously higher than that in the control group. After the UC-MSC infusions, the level of CHE was obviously increased and the level of TBIL was markedly decreased as compared with the baseline level during the 48-week follow-up; however, there was no significant difference in comparison with the control group (Fig. 4c, d). These data indicate that the UC-MSC therapy could alleviate hepatoinflammation and obviously improve liver function.

All the patients in the UC-MSC group tolerated the UC-MSC treatment well. Seven patients developed a self-limiting fever (body temperature, 37–38 °C) within 2–6 h after the UC-MSC transfusions but recovered within 12 h without any additional treatment. No other short-term clinical adverse effects, including allergy, rash, or infection were found. Long-term safety is a major concern in UC-MSC therapy for these patients. Therefore, we monitored the long-term adverse events, such as HCC events, during the 75-month follow-up after the first UC-MSC infusion. At the 75-month follow-up, cumulative incidences of HCC in control group and MSC treatment group were 20.7% (23/111) and 12.0% (13/108), respectively. However, the Kaplan–Meier analysis showed that there was no significant difference of HCC-free survival rate between the UC-MSC treatment group and control group (HR = 1.77, 95% CI 0.91–3.43, p = 0.104) (Fig. 5a). Interestingly, the HCC disease-free survival rate was seemingly higher in the control group than in the UC-MSC treatment group within the first 8 months of follow-up, but it was higher in the UC-MSC group between 8 and 75 months. However, landmark analysis did not reveal any significant difference in the HCC disease-free survival rate between the UC-MSC treatment group and control group within 8 months (HR = 1.09, 95% CI 0.24–4.98, p = 0.907) or between 8 and 75 months (HR = 2.02, 95% CI 0.91–4.47, p = 0.078) (Fig. 5b). These results indicate that the UC-MSC treatment did not pose a higher risk than the conventional treatment in terms of long-term safety for patients with decompensated liver cirrhosis.