A Sequential Combination of Baveno VI Criteria With Spleen Stiffness Measurement Improves the Screening of High-Risk Esophageal Varices in Compensated Advanced Chronic Liver Disease

Abstract

Background

Detecting high-risk esophageal varices (HRV) in cirrhotic patients is crucial for intervention and preventing variceal bleeding. While endoscopy is the gold standard for diagnosis, it is invasive and carries risks. Spleen stiffness measurement (SSM) may improve HRV screening and increase the spared endoscopy rate (SER).

Methods

We enrolled 129 patients with compensated advanced chronic liver disease (cACLD) who were candidates for variceal screening based on Baveno VI criteria (transient elastography ≥ 20 kPa or platelet count ≤ 150 × 10⁹/L). We collected SSM50Hz, SSM100Hz, liver stiffness measurement (LSM), and associated laboratory tests. Upper endoscopy was performed, and esophageal varices (EVs) types were defined as HRV according to Baveno VI.

Result

Ninety-two patients met the inclusion criteria. Alcohol was the most common etiology (32.6%). HRVs were found in 33 patients (35.9%). The AUROC values for SSM100Hz, SSM50Hz, Aspartate aminotransferase to Platelet Ratio Index (APRI), the combination of APRI and Fibrosis-4 (FIB-4), and FIB-4 were 0.807, 0.775, 0.753, 0.740, and 0.715, respectively. With a cutoff of 40 kPa, SSM100Hz showed a sensitivity of 97%, specificity of 45.8%, PPV of 50%, and NPV of 96.4%. The SER and missed HRV rate were 31.5% and 3%, respectively. The cutoff values for the greatest diagnostic accuracy in predicting HRV were 52.4 kPa for SSM50Hz, 0.98 for APRI, and 5.44 for FIB-4.

Conclusion

The sequential combination of Baveno VI with SSM100Hz improves HRV screening and SER. The SSM100Hz demonstrates the highest diagnostic accuracy with a low missed HRV rate of < 5%.