Evaluating Potential Living Kidney Donors by Renal Function Equations Leads to Unnecessary Reduction of the Donor Pool

Abstract

Renal function equations are advocated for cheap and simple estimation of renal function. Many centres estimate glomerular fi ltration rate (GFR) by these equations to evaluate potential living kidney donors. However, these renal function equations were developed in populations with renal disease and generally perform poorly in subjects without renal function impairment. We therefore analysed MDRD and Cockcroft-Gault (CG) in comparison to true GFR, before as well as after donor nephrectomy in 250 consecutive donors who donated between 1996 and 2007. GFR was measured as the clearance of 125 I-iothalamate 4 months before and 2 months after kidney donation; MDRD and CG were calculated from data obtained on these same days. GFR and CG were normalized for body surface area to allow comparison with MDRD. Based on clinical practise, we applied a hypothetical lower limit of 80 ml/min/1. 73 m 2 by either MDRD or CG and analyzed true GFR as well as post-donation outcome in donors considered noteligible to donate by renal function equations. Mean donor age was 50±11 years, 56% female. For the entire population, predonation GFR was 103±15, CG 79±14 and MDRD 70±10 ml/min/1. 73 m 2 . After donation, mean GFR was 65±11, CG 61±11 and MDRD 52±7 ml/min/1. 73 m 2 . When the limit of eGFR<80 was applied, 210 donors would have been declined for donation by MDRD and 141 donors by CG. Donors with eGFR below the threshold were signifi cantly older and more often female. Donors with MDRD<80 were 51±10 years old and had a GFR of 102±15 ml/ min/1. 73 m 2 . After donation, GFR was 65±10 ml/min/1. 73 m 2 in this subgroup. Donors with CG<80 were 55±8 years old and had a GFR of 99±13 ml/min/1. 73 m 2 . Post-donation GFR was 62±10 ml/min/1. 73 m 2 in this subgroup. For comparison: donors with either MDRD 80 or CG 80 had a post-donation GFR of 70±10 ml/min/1. 73 m 2 (both p<0. 05). Thus, renal function equations considerably underestimated true GFR. Donors who would not have been eligible to donate based on MDRD or CG below 80 ml/min/1. 73 m 2 , displayed adequate and acceptable remnant renal function after donation. In conclusion, use of renal function equations for selection of kidney donors lead to an unnecessary reduction of the donor pool. MDRD in particular is not suitable for the use in donor screening.