To what extent estimated or measured GFR could predict subclinical graft fibrosis: a comparative prospective study with protocol biopsies.

Monitoring of allograft function entails methods more accurate than serum creatinine and creatinine-based GFR equations (eGFR). This prospective trial aimed at investigating the diagnostic accuracy of creatinine- and cystatin C-based eGFR with measured GFR (mGFR) and compared them with graft fibrosis detected by protocol biopsies (PBx). Forty-four kidney transplant recipients were enrolled. PBx were obtained postengraftment and at 6th and 12th months. GFR was measured by Tc-99m DTPA at 3th, 6th, and 12th months after transplantation. Significant correlation existed between eGFR and mGFR at 3, 6, and 12 months (P < 0.0001). Cystatin C-based Hoek and Larsson equations had the lowest bias and highest accuracy. The sum of interstitial fibrosis and tubular atrophy score increased from implantation to 6th and 12th months (0.52 ± 0.79, 0.84 ± 0.88, 1.50 ± 1.35). This was accompanied by reduction of mGFR from 54.1 ± 15.2 to 49.9 ± 15.2 and 46.8 ± 16.5 ml/min/1.73 m(2) , while serum creatinine, cystatin C, and eGFR remained stable. Neither creatinine- nor cystatin C-based GFR equations are reliable for detecting insidious graft fibrosis. In the first year after transplantation, mGFR, with its best proximity to histopathology, can be used to monitor allograft function and insidious graft fibrosis.