2.2 Data collection
Patient medical data were extracted from the hospital information system as follows: demographics, alcoholism, smoking status, medical history (including cirrhosis etiology, diabetes, and hypertension), clinical presentation, imaging results, and laboratory results including alanine transaminase (ALT), aspartate transaminase (AST), bilirubin, serum albumin, lipid profile, international normalized ratio (INR), prothrombin time (PT), routine examination of blood, and serum ammonia.
2.3 Definitions
The cirrhosis diagnosis was based on pathologic findings or a combination of clinical presentation and imaging and laboratory results.13Chronic HBV infection was defined based on serum hepatitis B surface antigen positivity for >6 months. Alcohol-related cirrhosis was defined as cirrhosis together with alcoholism (alcohol consumption ≥20 g/d in women and ≥40 g/d in men for >5 years), in the absence of other liver diseases.14 ACLF was defined as INR ≥1.5 and serum bilirubin ≥5 mg/dL, complicated by encephalopathy and/or ascites within 4 weeks, in cirrhotic patients.15 HE was defined as abnormal neuropsychiatric manifestations and an abnormal ammonia level.16 Hypersplenism was defined as imaging results suggesting splenomegaly and platelet count <120 × 109/L 17 or a history of splenectomy owing to hypersplenism. Leukopenia, thrombocytopenia, and erythropenia were diagnosed based on white blood cell count <4.0 × 109/L, platelet count <100 × 109/L, and red blood cell count <3.5 × 1012/L for females or <4.0 × 1012/L for males, respectively. Smoking was defined as a history of smoking for more than 1 year. Moreover, the duration of liver disease was estimated using age at alcoholism or HBV infection onset.
2.4 Statistical analyses
Categorical variables are described as frequency (percentage) and were analyzed with chi-square tests. Normally distributed continuous variables are described as mean ± standard deviation and were analyzed with unpaired two-tailed Student’s t-tests, while other continuous variables are described as median (interquartile range) and were analyzed with Mann–Whitney U tests. Logistic regression with backward stepwise selection was used to determine the risks of complications by cirrhosis etiology, adjusting for sex, age, body mass index, hypertension, diabetes, disease duration, smoking, bilirubin, and albumin. The results are described as odds ratios (ORs) with 95% confidence intervals (CIs). P values less than 0.05 (two-tailed) indicated statistical significance. SPSS statistical software (version 22; IBM Corporation, Armonk, NY, USA) was used for all statistical analyses.
3 RESULTS
3.1 Patient characteristics
We retrospectively enrolled a total of 514 cirrhotic patients, comprising 445 with HBV-related cirrhosis and 69 with alcohol-related cirrhosis. Table 1 lists their demographic and clinical characteristics. Among the patients with HBV-related cirrhosis, 401 (90.11%) were taking antiviral therapy and 156 (60.94%) had a low HBV DNA level (defined as <2000 IU/mL).
3.2 Differences in hepatic function indexes by cirrhosis etiology
HBV-related cirrhotic patients had a higher rate of Child–Pugh grades B and C (81.50% vs 18.50%, P = 0.008), and alcohol-related cirrhotic patients had a lower serum albumin level (32.28 ± 7.02 vs 34.41 ± 6.38 mmol/L, P = 0.017). However, no differences were found in the levels of AST, ALT, bilirubin, PT, INR, or blood lipids between the two groups (Table 2).
3.3 Differences in complications by cirrhosis etiology
The rates of HE and ACLF were higher in alcohol-related cirrhotic patients than HBV-related cirrhotic patients (HE: 15.94% vs 4.49%, P = 0.001; ALCF: 7.25% vs 2.28%, P = 0.040). The serum ammonia level was also higher in alcohol-related cirrhotic patients with HE than HBV-related cirrhotic patients with HE (134.00 ± 62.99 vs 82.16 ± 17.82 μmol/L, P = 0.029). In contrast, the rates of HCC and hypersplenism were higher in HBV-related cirrhotic patients than alcohol-related cirrhotic patients (HCC: 39.55% vs 1.45%, P < 0.001; hypersplenism: 45.84% vs 28.99%, P = 0.009). No significant differences in the rates of jaundice, ascites, esophageal and gastric varices, or spontaneous peritonitis were observed between the two groups (Table 3).