A model for Glomerular filtration Rate Assessment In Liver disease (GRAIL) in the presence of renal dysfunction.

Estimation of glomerular filtration rate (eGFR) in patients with liver disease is suboptimal in the presence of renal dysfunction. We developed a model for GFR assessment in liver disease (GRAIL) before and after liver transplantation (LT). GRAIL was derived using objective variables (creatinine, blood urea nitrogen, age, gender, race, and albumin) to estimate GFR based on timing of measurement relative to LT and degree of renal dysfunction (www.bswh.md/grail). The measured GFR (mGFR) by iothalamate clearance (n = 12,122, 1985-2015) at protocol time points before/after LT was used as reference. GRAIL was compared with the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) and Modification of Diet in Renal Disease (MDRD-4, MDRD-6) equations for mGFR < 30 mL/min/1.73 m(2). Prediction of development of chronic kidney disease (mGFR < 20 mL/min/1.73 m(2), initiation of chronic dialysis) and listing or receipt of kidney transplantation within 5 years was examined in internal cohort (n = 785) and external validation (n = 68,217, 2001-2015). GRAIL had less bias and was more accurate and precise as compared with CKD-EPI, MDRD-4, and MDRD-6 at time points before/after LT for low GFR. For mGFR < 30 mL/min/1.73 m(2), the median difference (eGFR-mGFR) was GRAIL: 5.24 (9.65) mL/min/1.73 m(2) as compared with CKD-EPI: 8.70 (18.24) mL/min/1.73 m(2), MDRD-4: 8.82 (17.38) mL/min/1.73 m(2), and MDRD-6: 6.53 (14.42) mL/min/1.73 m(2). Before LT, GRAIL correctly classified 75% as having mGFR < 30 mL/min/1.73 m(2) versus 36.1% (CKD-EPI), 36.1% (MDRD-4), and 52.8% (MDRD-6) (P < 0.01). An eGFR < 30 mL/min/1.73 m(2) by GRAIL predicted development of CKD (26.9% versus 4.6% CKD-EPI, 5.9% MDRD-4, and 10.5% MDRD-6) in center data and needing kidney after LT (48.3% versus 22.0% CKD-EPI versus 23.1% MDRD-4 versus 48.3% MDRD-6, P < 0.01) in national data within 5 years after LT. Conclusion: GRAIL may serve as an alternative model to estimate GFR among patients with liver disease before and after LT at low GFR.