Granulocyte colony-stimulating factor and autologous CD133-positive stem-cell therapy in liver cirrhosis (REALISTIC): an open-label, randomised, controlled phase 2 trial

Abstract

Background: Results of small-scale studies have suggested that stem-cell therapy is safe and effective in patients with liver cirrhosis, but no adequately powered randomised controlled trials have been done. We assessed the safety and efficacy of granulocyte colony-stimulating factor (G-CSF) and haemopoietic stem-cell infusions in patients with liver cirrhosis.

Methods: This multicentre, open-label, randomised, controlled phase 2 trial was done in three UK hospitals and recruited patients with compensated liver cirrhosis and MELD scores of 11·0-15·5. Patients were randomly assigned (1:1:1) to receive standard care (control), treatment with subcutaneous G-CSF (lenograstim) 15 μg/kg for 5 days, or treatment with G-CSF for 5 days followed by leukapheresis and intravenous infusion of three doses of CD133-positive haemopoietic stem cells (0·2 × 10⁶ cells per kg per infusion). Randomisation was done by Cancer Research UK Clinical Trials Unit staff with a minimisation algorithm that stratified by trial site and cause of liver disease. The coprimary outcomes were improvement in severity of liver disease (change in MELD) at 3 months and the trend of change in MELD score over time. Analyses were done in the modified intention-to-treat population, which included all patients who received at least one day of treatment. Safety was assessed on the basis of the treatment received. This trial was registered at Current Controlled Trials on Nov 18, 2009; ISRCTN, number 91288089; and the European Clinical Trials Database, number 2009-010335-41.

Findings: Between May 18, 2010, and Feb 26, 2015, 27 patients were randomly assigned to the standard care, 26 to the G-CSF group, and 28 to the G-CSF plus stem-cell infusion group. Median change in MELD from day 0 to 90 was -0·5 (IQR -1·5 to 1·1) in the standard care group, -0·5 (-1·7 to 0·5) in the G-CSF group, and -0·5 (-1·3 to 1·0) in the G-CSF plus stem-cell infusion group. We found no evidence of differences between the treatment groups and control group in the trends of MELD change over time (p=0·55 for the G-CSF group vs standard care and p=0·75 for the G-CSF plus stem-cell infusion group vs standard care). Serious adverse events were more frequent the in G-CSF and stem-cell infusion group (12 [43%] patients) than in the G-CSF (three [11%] patients) and standard care (three [12%] patients) groups. The most common serious adverse events were ascites (two patients in the G-CSF group and two patients in the G-CSF plus stem-cell infusion group, one of whom was admitted to hospital with ascites twice), sepsis (four patients in the G-CSF plus stem-cell infusion group), and encephalopathy (three patients in the G-CSF plus stem-cell infusion group, one of whom was admitted to hospital with encephalopathy twice). Three patients died, including one in the standard care group (variceal bleed) and two in the G-CSF and stem-cell infusion group (one myocardial infarction and one progressive liver disease).

Interpretation: G-CSF with or without haemopoietic stem-cell infusion did not improve liver dysfunction or fibrosis and might be associated with increased frequency of adverse events compared with standard care.

Funding: National Institute of Health Research, The Sir Jules Thorn Charitable Trust.

Figure 1

Trial profile We defined the mITT population as including any patient assigned to receive G-CSF who received a minimum of 1 day of G-CSF or any patient assigned to receive G-CSF plus stem-cell infusion who received at least one infusion of at least 0·17 × 10⁶ cells per kg. We defined the per-protocol population as including patients in the G-CSF group who attended all G-CSF visits and received an average daily dose of at least 12 μg/kg and patients in the G-CSF plus stem-cell infusion group who received 5 days of G-CSF and a cumulative cell infusion of at least 0·51 × 10⁶ cells per kg. All patients in the standard care group were included in the mITT and per-protocol populations. Because the second coprimary outcome allowed imputation of missing values, the mITT numbers are greater than the number of available day 90 values. mITT=modified intention to treat. G-CSF=granulocyte colony-stimulating factor. *Patient received no treatment so the event is attributed to the standard care group based on definition of the safety population.

Figure 2

Change in MELD and UKELD scores Observed median change in (A) MELD and (B) UKELD scores at baseline and days 30, 60, and 90. Bars show IQR. Model fit based on predictions from a linear mixed-effects model incorporating restricted cubic splines for (C) MELD and (D) UKELD respectively. G-CSF=granulocyte colony-stimulating factor. UKELD=UK End-Stage Liver Disease Score.