Performance of creatinine and cystatin C-based glomerular filtration rate estimating equations in a European HIV-positive cohort

Abstract

Objective: To validate glomerular filtration rate (GFR) estimating equations in white HIV-infected patients based on serum creatinine and/or serum cystatin C.

Design: Single-center, cross-sectional evaluation of the predictive performance of GFR estimators.

Methods: GFR was measured by iohexol plasma clearance. Serum creatinine (Scr) and serum cystatin C (Scyst) were measured by traceable and standardized methods. We evaluated the performance of the Modification of Diet in Renal Disease (MDRD) and the Chronic Kidney Disease-Epidemiology Collaboration (CKD-EPI) equations. We also studied the performance of the cystatin C-based equation (CKD-EPI Scyst) and the combined cystatin and creatinine-based equation (CKD-EPI combined), as recently proposed by the CKD-EPI group.

Results: Two hundred and three participants (18% of women) were included. Mean age was 49 ± 10 years. Mean measured GFR (mGFR) was 95 ± 24 ml/min per 1.73 m². CKD-EPI and CKD-EPI combined significantly outperformed the MDRD equation. The percentage of estimating results within 30% of mGFR was 75, 82 and 81% for the MDRD, CKD-EPI and CKD-EPI combined equation, respectively. Results favoring the CKD-EPI and CKD-EPI combined equation were especially observed for patients with mGFR over 90 ml/min per 1.73 m².

Conclusion: In our European HIV cohort, we confirmed that the creatinine-based CKD-EPI equation should replace the MDRD study equation. However, global performance of this equation remains worse than the performance observed in the general population. This lesser performance is particularly relevant in patients with measured GFR under and around 60 ml/min per 1.73 m². Moreover, the specific interest of Scyst-based equations is not confirmed in this population.