When should I repeat the endoscopy in my patient with compensated cirrhosis, whom I just scoped and who had no or small varices?

Endoscopic surveillance for patients with compensated advanced chronic liver disease (cACLD) is currently guided by the presence or size of esophageal varices. Current guidelines recommend repeat endoscopy every 2-3 years for patients with no varices and every 1-2 years for those with small varices, but these intervals are based largely on expert opinion rather than objective risk stratification. We evaluated the Disease Severity Index (DSI) from the oral cholate challenge test as a predictor of clinical outcomes in 195 patients from the Hepatitis C Anti-viral Long-term Treatment against Cirrhosis (HALT-C) Trial (ClinicalTrials.gov, NCT00006164) with cACLD with no or small esophageal varices. Subjects were stratified into four clinical risk categories based on DSI and platelet count: DSI ≤18.3 (low risk), DSI ≤24 with platelet count >135 nL^-1 (low-to-moderate risk), DSI ≤24 with platelet count ≤135 nL^-1 (moderate-to-high risk), and DSI >24 (high risk). Kaplan-Meier analysis demonstrated a stepwise increase in adverse outcomes (variceal hemorrhage, ascites, encephalopathy, Child-Pugh progression, transplant, and death) across these categories. In multivariable logistic regression, DSI was a highly significant predictor of clinical outcome (P<0.001), while categories of no or small esophageal varices was not (P=0.07). Based on these findings, we recommend integrating DSI into surveillance strategies to guide the timing of repeat endoscopy: later endoscopy (2-3 years) for lower-risk patients and earlier endoscopy (1-2 years) for higher-risk patients. We recommend repeat DSI testing every 2-3 years for low risk, 1.5-2 years for low-to-moderate risk, 1-1.5 years for moderate-to-higher risk, and <1 year for highest risk. These results suggest that DSI is a robust, noninvasive tool for refining timing of repeat endoscopy and optimizing resource utilization in cACLD, regardless of disease etiology.