Calculation of glomerular filtration rate based on Cystatin C in cirrhotic patients

Abstract

Plasma creatinine and creatinine clearance are of limited value for the estimation of renal function in cirrhotics. In these patients, cystatin C (Cys C) has been proposed as an alternative marker of glomerular filtration rate (GFR) and Cys C-based equations for calculation of GFR have been developed in non-cirrhotic patient cohorts. We retrospectively analyzed correlation, bias, precision and accuracy of two Cys C-based formulae (Larsson and Hoek) for GFR estimation in comparison with two creatinine-based equations (Cockroft & Gault and MDRD). The Cys C was determined by an immunonephelometric method. The GFR was measured by means of inulin clearance in 44 consecutive patients with liver cirrhosis. On average, inulin clearance was 28.3 (95% CI: 29.2-41.3 ml/min/1.73 m2). Creatinine as well as Cys C-based equations overestimated the true GFR by 105-154%. However, Cys C-based equations showed significantly lower bias and higher precision than the creatinine-based formulae. Correlation and accuracy tended to be better with the Hoek and Larsson equation than with the Cockroft & Gault or MDRD formulae. Hoek and Larsson equations showed a similar diagnostic performance in all statistical procedures. Our data suggest a significant improvement of GFR estimation in liver cirrhotics by means of the Cys C-based Hoek and Larsson formulae. However, all estimates remain a crude approximation of true GFR and thus cannot replace gold standard methods.