The accuracy of current criteria for ruling out large oesophageal varices (LEV) and other endoscopic lesions of portal hypertension (PH) may be compromised by obesity and MASLD/MASH. In the US multicentre SHUNT-V study, we evaluated the disease severity index (DSI) for detecting LEV and other lesions of PH at endoscopy. Subjects were adults with compensated cirrhosis scheduled for endoscopy to screen for varices. DSI was calculated from clearances of labelled cholates after oral and intravenous administration. DSI ≤ 18.3 was evaluated as a cut-off for ruling out LEV with acceptance criteria of negative likelihood ratio < 0.52 and sensitivity > 85%. SHUNT-V enrolled 306 subjects; 275 had both DSI and endoscopy, and 238 had Child-Pugh A cirrhosis (52.1% MASLD/MASH, 25.2% chronic hepatitis C and 15.6% alcoholic liver disease; 87% were overweight, 64% were obese and 54% had diabetes). AUROCs for DSI ranged from 0.81 to 0.82 for LEV and 0.79 to 0.80 for all significant PH lesions. DSI 18.3 had sensitivity 96.3%-100% for LEV and 97.3%-100% for all significant PH lesions. If DSI ≤ 18.3 were used as the sole determinant to defer EGD, 27%-35% of EGDs could have been avoided with 0%-3.7% of LEV and 0%-2.7% of all significant PH lesions missed. HepQuant DSI predicts the likelihood of LEV and significant PH lesions across a spectrum of patient characteristics and disease aetiologies. DSI, based on liver function and portal-systemic shunting, can aid in the decision to defer endoscopy for varices in patients with Child-Pugh A cirrhosis. The SHUNT-V study was registered at ClinicalTrials.gov (NCT03583996).
Read full abstract