The prognostic value of high-density lipoprotein cholesterol in patients with decompensated cirrhosis: a propensity score matching analysis

Abstract

Emerging evidence has implicated that high-density lipoprotein cholesterol (HDL-C) as a prognostic surrogate in the context of cirrhosis. However, an exact cutoff has not been fully elucidated. We aimed to clarify optimal cutoff of HDL-C for short-term mortality based on time-to-event analysis and validated this association by performing propensity score matching (PSM) analysis. A total of 238 patients with decompensated cirrhosis were enrolled. The optimal cutoff of HDL-C was initially determined by X-tile program. Independent risk factors for 180-day mortality were identified by multiple Cox regression. The Kaplan-Meier method was implemented to generate survival curves. A 1:2 ratio PSM was performed to diminish selection bias and potential confounders. The X-tile implied that the difference in survival was most significant for HDL <0.4mmol/L (<16mg/dL). Circulating HDL <0.4mmol/L exhibited an independent risk factor both in the entire cohort and PSM subset (HR 2.696, 95% CI 1.082-6.791, P = 0.033; HR 2.735, 95% CI 1.027-7.734, P = 0.048). Furthermore, HDL-C combined with conventional scoring systems had higher AUCs associated with poor prognosis than Child-Pugh classification or model for end-stage liver disease (MELD) in isolation (0.78 vs 0.66; 0.74 vs 0.54, P < 0.05 for both). HDL-C <0.4 mmol/L may serve as a readily available and robust cutoff for stratifying cirrhotic patients at high 180-day mortality risk. The incorporation of HDL-C to Child-Pugh classification or MELD has the potentials to provide substantially clinical relevance without extra economic cost.