Associations of gallbladder and gallstone parameters with clinical outcomes in patients with cirrhosis

Study design

This retrospective study was performed on the basis of our prospective database in which a total of 570 patients with cirrhosis were consecutively admitted to the Department of Gastroenterology of our hospital and underwent abdominal contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI) scans from December 2014 to November 2021. A diagnosis of liver cirrhosis was mainly based on clinical manifestations including gastrointestinal bleeding, ascites, and hepatic encephalopathy, laboratory tests including liver dysfunction, decreased serum albumin, and coagulation dysfunction, imaging features including liver morphology changes with splenomegaly on abdominal CT scans, and liver histological features including pseudolobule, if necessary.

Inclusion criteria were as follows: (1) patients who were diagnosed with liver cirrhosis and (2) patients who performed abdominal contrast-enhanced CT scans during hospitalization. Exclusion criteria were as follows: (1) a history of malignancy; (2) gallbladder and gallstone parameters could not be sufficiently evaluated on CT images; (3) a history of cholecystectomy; and (4) acute or chronic cholecystitis was diagnosed based on disease history, clinical presentations, and/or biochemical signs. Notably, patients with asymptomatic gallstones were not excluded.

Additionally, by reviewing both electronic medical records and imaging reports system, all patients without a history of chronic liver disease and hepatocellular carcinoma or other malignancies who were consecutively admitted to the Department of Gastroenterology of our hospital and underwent abdominal contrast-enhanced CT between January 2020 and October 2021 were selected as the control group.

Clinical data collection

All patients were subjected to complete clinical evaluation. Clinical data at admission were collected as follows: age, gender, red blood cell (RBC), white blood cell (WBC), platelets count (PLT), total bilirubin (TBIL), albumin (ALB), alanine aminotransferase (ALT), γ-glutamine transferase (GGT), serum creatinine (Scr), serum sodium (Na), prothrombin time (PT), and international normalized ratio (INR). The Child-Pugh score and the MELD score were calculated.

CT images

Two investigators independently reviewed all available CT images of each included patient, selected the specific layers, where the maximum length and width of gallbladder, the maximum thickness of the gallbladder wall, and the maximum length and width of gallstones were obtained, and then measured these parameters according to the standard methods as shown in Figure 1. A disagreement between them was resolved by discussing with another investigator to determine the most appropriate layer where the value measured should be maximal as the final value. Notably, all patients are routinely requested to be fasting before undergoing abdominal contrast-enhanced CT or MRI scan.

Figure 1

Gallbladder and gallstone parameters measured in CT images. (A) Measurement of gallbladder parameters in a 63-year-old woman with liver cirrhosis. The red line represents the maximum gallbladder length, the blue line represents the maximum gallbladder width, and the yellow line represents the maximum gallbladder wall thickness. The maximum gallbladder length was 77.6 mm, the maximum gallbladder width was 49.0 mm, and the maximum gallbladder wall thickness was 3.5 mm. (B) Measurement of gallstone parameters in a 65-year-old man with liver cirrhosis. The red line represents the maximum gallstone length, and the blue line represents the maximum gallstone width. The maximum gallstone length was 12.7 mm, and the maximum gallstone width was 12.4 mm. CT: computed tomography.

Follow-up

All enrolled patients were followed by reviewing inpatient and outpatient medical records and telephone visits. The last follow-up date was December 9, 2021. We recorded the dates of hepatic decompensation events and deaths during follow-up. Hepatic decompensation events evaluated in this study included gastrointestinal bleeding, ascites, hepatic encephalopathy, and jaundice.^{[15, 16, 17, 18]}

Statistical analyses

First, continuous variables were expressed as the mean ± standard deviation or median (range) and compared using nonparametric Mann-Whitney U test. Categorical variables were expressed as frequency (percentage) and compared using the χ² test. Second, the correlations of various gallbladder and gallstone parameters with the Child-Pugh score and the MELD score were analyzed by Spearman’s rank correlation tests, and the Spearman’s rank correlation coefficients (rs) were calculated. Third, X-tile software was utilized to calculate the optimal cutoff values of gallbladder and gallstone parameters for evaluating the long-term overall survival and hepatic decompensation events. Fourth, univariate Cox regression analyses were performed to explore the associations of gallbladder and gallstone parameters with survival and multivariate Cox regression analyses were performed by adjusting for age, sex, and the Child-Pugh score to identify which parameter was an independent predictor of survival. Hazard ratios (HRs) and their 95% confidence intervals (CIs) were calculated. Cumulative survival rates were evaluated by Kaplan-Meier curve analyses and compared by log-rank tests. Fifth, competing risk analyses were performed to analyze the associations of these parameters with hepatic decompensation events during follow-up, where death was considered a competing event. Sub-distribution hazard ratios (sHRs) and their 95% CIs were calculated. Cumulative rates of hepatic decompensation events were evaluated by Nelson-Aalen cumulative risk curve analyses and compared by Gray’s tests. All statistical analyses were performed on IBM Statistical Package for the Social Sciences (SPSS) version 25.0 (IBM Corp, Armonk, NY, USA), X-tile version 3.6.1 (Yale University, New Haven, CT, USA), and R version 4.1.2 (R Foundation for Statistical Computing, Vienna, Austria) with the packages ggplot2, survival, survminer, and cmprsk. A two-tailed P value < 0.05 was considered statistically significant.

Study population

Overall, 206 patients with cirrhosis were included (Figure 2). Median age was 56 years (28–89) and 142 (69%) patients were male (Table 1). During a median follow-up duration of 2.25 years (0.09–6.03), 54 patients developed gastrointestinal bleeding, 42 developed ascites, 10 developed hepatic encephalopathy, 29 died, and none underwent liver transplantation. Causes of death were related to liver diseases (n = 23), non-liver diseases (n = 1), and were unknown (n = 5).

Figure 2

Flowchart of cirrhotic patients’ selection. MRI: magnetic resonance imaging.

Cirrhosis versus non-cirrhosis

Overall, 104 patients with non-cirrhosis were selected as the control group (Supplementary Figure 1). Patients with cirrhosis had longer gallbladder and wider gallbladder, thicker gallbladder wall, and higher prevalence of gallstones than those with non-cirrhosis (Table 1).

Correlations of gallbladder and gallstone parameters with the Child-Pugh score and the MELD score in cirrhosis

Spearman’s rank correlation tests demonstrated that gallbladder width significantly correlated with the MELD score (P = 0.03, rs = 0.148) and gallbladder wall thickness significantly correlated with the Child-Pugh score (P = 0.005, rs = 0.196) and the MELD score (P = 0.002, rs = 0.212). However, gallbladder length, presence of gallstones, gallstones’ length, and gallstones’ width did not significantly correlate with the Child-Pugh score or the MELD score (Table 2).

Associations of gallbladder and gallstone parameters with survival in cirrhosis

The optimal cutoff values of gallbladder length, gallbladder width, gallbladder wall thickness, gallstones’ length, and gallstones’ width for predicting survival were 72.0, 38.6, 3.6, 1.9, and 2.0 mm, respectively.

Univariate Cox regression analyses demonstrated that gallbladder length < 72.0 mm, gallbladder width < 38.6 mm, gallbladder wall thickness < 3.6 mm, absence of gallstones, gallstones’ length < 1.9 mm, and gallstones’ width < 2.0 mm were not significant predictors of survival in cirrhosis. However, after adjusting for age, sex, and the Child-Pugh score, multivariate Cox regression analyses demonstrated that gallbladder width < 38.6 mm was an independent predictor of survival (HR = 3.01, 95% CI: 1.28–7.06; P = 0.01), but not gallbladder length < 72.0 mm, gallbladder wall thickness < 3.6 mm, absence of gallstones, gallstones’ length < 1.9 mm, or gallstones’ width < 2.0 mm (Table 3).

Kaplan-Meier curve analysis demonstrated that patients with gallbladder length < 72.0 mm had a significantly higher cumulative survival rate than those with a length of ≥ 72.0 mm (P = 0.049 by log-rank test) (Figure 3). But there was no significant difference in the cumulative survival rate between patients with gallbladder width < 38.6 mm versus those with a width of ≥ 38.6 mm (P = 0.10 by log-rank test), those with gallbladder wall thickness < 3.6 mm versus patients with a wall thickness of ≥ 3.6 mm (P = 0.14 by log-rank test), those without gallstones versus those with gallstones (P = 0.97 by log-rank test), those with gallstones’ length < 1.9 mm versus those with a length of ≥ 1.9 mm (P = 0.73 by log-rank test), or those with gallstones’ width < 2.0 mm versus those with a width of ≥ 2.0 mm (P = 0.73 by log-rank test).

Figure 3

Kaplan-Meier curve analysis demonstrating that patients with gallbladder length < 72 mm had a significantly higher cumulative survival rate than those with a length of ≥ 72 mm (P = 0.049 by log-rank test).

Associations of gallbladder and gallstone parameters with hepatic decompensation events in cirrhosis

The optimal cutoff values of gallbladder length, gallbladder width, gallbladder wall thickness, gallstones’ length, and gallstones’ width for predicting hepatic decompensation events were 55.8, 31.1, 3.4, 1.9, and 1.7 mm, respectively. Competing risk analyses demonstrated that gallbladder wall thickness < 3.4 mm was significantly associated with decreased risk of hepatic decompensation events (sHR = 1.62, 95% CI: 1.09–2.41; P = 0.02), but not gallbladder length < 55.8 mm, gallbladder width < 31.1 mm, absence of gallstones, gallstones’ length < 1.9 mm, or gallstones’ width < 1.7 mm (Table 4).

Nelson–Aalen cumulative risk curve analysis demonstrated that patients with gallbladder wall thickness < 3.4 mm had a significantly lower cumulative rate of hepatic decompensation events than those with a wall thickness of ≥ 3.4 mm (P = 0.02 by Gray’s test) (Figure 4). But there was no significant difference in the cumulative rate of hepatic decompensation events between patients with gallbladder length < 55.8 mm versus those with a length of ≥ 55.8 mm (P = 0.15 by Gray’s test), those with gallbladder width < 31.1 mm versus those with a width of ≥ 31.1 mm (P = 0.15 by Gray’s test), those without gallstones versus those with gallstones (P = 0.54 by Gray’s test), those with gallstones’ length < 1.9 mm versus those with a length of ≥ 1.9 mm (P = 0.76 by Gray’s test), or those with gallstones’ width < 1.7 mm versus those with a width of ≥ 1.7 mm (P = 0.54 by Gray’s test).

Figure 4

Nelson-Aalen cumulative risk curve analysis demonstrating that patients with gallbladder wall thickness < 3.4 mm had a significantly lower cumulative rate of hepatic decompensation events than those with a thickness of ≥ 3.4 mm (P = 0.02 by Gray’s test). Pts: patients.