Abstract / Summary
Abstract Background Kidney dysfunction is associated with adverse outcomes in cirrhosis, but the prognostic significance of longitudinal creatinine-based eGFR decline in compensated cirrhosis remains uncertain. This study evaluated the association between longitudinal creatinine-based eGFR decline and long-term clinical outcomes in patients with compensated cirrhosis. Methods This retrospective cohort study included patients with compensated cirrhosis between 2016 and 2025. Creatinine-based eGFR trajectories were estimated from repeated serum creatinine measurements during routine follow-up, with rapid decline defined as an annual eGFR decline of more than 5 mL/min/1.73 m²/year. For the primary prognostic analysis, eGFR slope was estimated during the first 2 years after baseline, and mortality follow-up began at the 2-year landmark. Cox regression was used to evaluate all-cause mortality. Results The overall longitudinal cohort included 249 patients, of whom 37 (14.9%) had rapid creatinine-based eGFR decline. The primary 2-year landmark analysis included 191 patients, of whom 45 (23.6%) had rapid decline. During post-landmark follow-up, 29 deaths occurred, including 11 of 45 patients (24.4%) with rapid decline and 18 of 146 patients (12.3%) without rapid decline. Rapid decline was associated with subsequent mortality (crude hazard ratio [HR], 2.32; 95% CI, 1.10–4.93; P = 0.028) and remained associated after adjustment for MELD-Na and baseline ascites (adjusted HR, 2.55; 95% CI, 1.19–5.47; P = 0.016). A 1-year landmark sensitivity analysis yielded consistent results (adjusted HR, 3.42; 95% CI, 1.62–7.21; P = 0.001). Conclusions Rapid creatinine-based eGFR decline in patients with compensated cirrhosis was associated with a higher risk of mortality. Serial creatinine-based eGFR trajectories may provide additional prognostic information.