Résumé
A major problem in renal transplantation is identifying a grading system that can predict long-term graft survival. The present study determined the extent to which the two existing grading systems (Banff 97 and chronic allograft damage index, CADI) correlate with each other and with graft loss. A total of 161 transplant patient biopsies with chronic allograft nephropathy (CAN) were studied. The samples were coded and evaluated blindly by two pathologists using the two grading systems. Logistic regression analyses were used to evaluate the best predictor index for renal allograft loss. Patients with higher Banff 97 and CADI scores had higher rates of graft loss. Moreover, these measures also correlated with worse renal function and higher proteinuria levels at the time of CAN diagnosis. Logistic regression analyses showed that the use of angiotensin-converting enzyme inhibitor (ACEI), hepatitis C virus (HCV), tubular atrophy, and the use of mycophenolate mofetil (MMF) were associated with graft loss in the CADI, while the use of ACEI, HCV, moderate interstitial fibrosis and tubular atrophy and the use of MMF were associated in the Banff 97 index. Although Banff 97 and CADI analyze different parameters in different renal compartments, only some isolated parameters correlated with graft loss. This suggests that we need to review the CAN grading systems in order to devise a system that includes all parameters able to predict long-term graft survival, including chronic glomerulopathy, glomerular sclerosis, vascular changes, and severity of chronic interstitial fibrosis and tubular atrophy.