RESUMO
BACKGROUND: When estimating the glomerular filtration rate (GFR) in kidney transplant patients, significant differences have been found between MDRD and the 2009 CKD-EPI equations, and reference techniques. OBJECTIVE: To analyse and compare the performance of MDRD and the 2009 and 2012 CKD-EPI equations against 51Cr-EDTA plasma clearance in measuring GFR in 270 kidney transplant patients after one year. RESULTS: The mean measured GFR was 43.0±11.4 (18.2-79.4)ml/min/1.73m2, with creatinine levels of 1.42±0.46 (0.60-4.33)mg/dl and cystatin C levels of 1.45±0.53 (0.42-3.48)mg/l. This correlated moderately with creatinine (r=-0.61, P<.001) and cystatin C (r=-0.52, P<.001). Using linear regression techniques, it was found that creatinine, cystatin C, gender and age only explained 52% of GFR total variance. All equations overestimated GFR, with a mean bias of +11.1ml/min/1.73m2 for MDRD, +16.4ml/min/1.73m2 for 2009-CKD-EPI, +15ml/min/1.73m2 for CKD-EPI with cystatin C, and +14.1ml/min/1.73m2 for 2012-CKD-EPI with creatinine and cystatin C. eGFR by MDRD and the 2009 CKD-EPI equation correlated better with 51Cr-EDTA than CKD-EPI with creatinine and/or cystatin C. The overestimations were negatively correlated with creatinine and cystatin C levels, most significantly for CKD-EPI with creatinine and/or cystatin C when GFR was greater than 60ml/min/1.73m2. CONCLUSIONS: The 2012 CKD-EPI equations with creatinine and/or cystatin C significantly overestimate GFR in stage 1 and 2 chronic kidney disease. The MDRD equations is therefore recommended in these cases. The reference method used to measure GFR seems to heavily influence the bias of the equations.