The ability of the modified albumin-bilirubin-platelet score to predict high-risk varices in cirrhotic patients
摘要
The combination of a modified Albumin-bilirubin (mALBI) score, an objective indicator of liver function, and a platelet (PLT) count below 150 × 109/L is associated with an increased risk of developing high-risk varices (HRV) in patients with hepatocellular carcinoma (HCC). We aimed to determine optimal cut-off values for the mALBI-PLT score to predict HRV without the need for esophagogastroduodenoscopy (EGD) in cirrhotic patients without HCC.
MethodsA retrospective, single-center, observational cohort study included participants who were monitored for liver cirrhosis from December 2023 to September 2025. The mALBI-PLT score was calculated. Esophageal varices (EVs) of F2 and F3 were considered HRV.
ResultsThe study comprised 141 patients; 73 (51.8%) had F1 varices, 29 (20.6%) had F2 varices, and 39 (27.7%) had F3 varices. In univariate regression analysis, mALBI-PLT 3 (OR: 5.21, p < 0.001), mALBI-PLT > 2 (OR: 8.42, p < 0.001), and mALBI-PLT 4 (OR: 13.89, p < 0.001) scores were significantly more effective in predicting the presence of HRV compared to mALBI-PLT 2. In multivariate regression analysis, mALBI-PLT 3 (OR: 4.48, p = 0.002) and mALBI-PLT 4 (OR: 10.74, i < 0.001) scores were independently significantly more effective in predicting the presence of HRV than mALBI-PLT 2. With a mALBI-PLT score threshold of 2, the rate of avoiding unnecessary EGD was 49.3% (n = 36), and the false-negative rate was 5% (n = 7).
ConclusionIn predicting HRV in cirrhotic patients, a mALBI-PLT threshold of 2 demonstrated satisfactory sensitivity and negative predictive value, allowing for the avoidance of unnecessary EGD in almost half of the patients, while maintaining an acceptable false-negative rate.