Abstract
BACKGROUND: Bone marrow mononuclear cells (BMMNCs) therapy should be effective for the improvement of liver function and short-term outcome in patients with liver cirrhosis, but few studies have explored the long-term prognosis of cirrhotic patients treated with BMMNCs. METHODS: In this retrospective study, eligible patients with liver cirrhosis were selected by using propensity score matching (PSM). Effect of BMMNCs on death was explored by Cox regression analysis, as well as competing risk analysis, where liver transplantation was a competing event. Hazard ratio (HR) and sub-distribution HR (sHR) were calculated. Subgroup analyses were performed based on the age, sex, Child-Pugh class, and model for end-stage liver disease (MELD) score. RESULTS: Overall, 260 patients were included, of whom 130 were treated with transhepatic arterial transplantation of BMMNCs. The median follow-up duration was 5.27 (range: 0.37-16.62) years. By adjusting by age, sex, and Child-Pugh score, multivariate Cox regression (HR = 0.707, P = 0.020) and competing risk analyses (sHR = 0.709, P = 0.026) demonstrated that BMMNCs were independently associated with a lower risk of death in cirrhotic patients in the overall analysis. Univariate Cox regression analyses demonstrated that BMMNCs were significantly associated with a decreased risk of death in the subgroup analyses of age ≤50 years (HR= 0.533, P = 0.016), male (HR = 0.626, P = 0.010), Child-Pugh class B (HR = 0.638, P = 0.026), and MELD score > 12 (HR = 0.483, P = 0.002), but not age > 50 years (P = 0.097), female (P = 0.170), Child-Pugh class A (P = 0.309), Child-Pugh class C (P = 0.369), or MELD score ≤12 (P = 0.096). CONCLUSION: BMMNCs can provide additional survival benefits in patients with liver cirrhosis.