Interpretation of surveillance kidney allograft biopsies according to the Banff criteria

Ureteric lesions detected in normally functioning kidney allografts: risk factors and clinical implications

INTRODUCTION

We characterized the alterations in ureteral biopsies from normally functioning kidney allografts to study risk factors.

MATERIALS AND METHODS

We studied 55 ureteral fragments from kidney grafts obtained during cystoscopy for routine double-J stent extraction. We evaluated the type and severity of the lesions, the risk factors for their occurrence, and their relation to the evolution of the transplant, including the occurrence of renal rejection episodes or ureteral complications.

RESULTS

Borderline or rejection lesions were detected in 21 of the 55 fragments. Rejection lesions were more common among biopsies performed in the first 80 days (54.6%) than during the 120 days afterward (15.4%, P = .043). Similarly, urothelium reactivity was detected in 71.4% and 30.7% of the biopsies performed up to and after 4 months, respectively (P = .008). Urothelial atrophy was detected in 20% of the fragments, the age of the donors being higher in these cases (P = .026). There was a trend to the association of borderline or rejection lesions in the ureteral biopsies and a history of an acute renal rejection episode (P = .053). There were no detectable relations between those findings and the evolution of the transplant.

CONCLUSIONS

Thirty-eight percent of the biopsed ureters showed rejection or borderline lesions, these lesions were more common among biopsies done in the first months after transplantation. These findings are similar to the ones found in routine renal biopsies. We did not find any significant relation between the nature of the lesions encountered on the ureteral biopsies and the evolution of the transplant.

Prevalence and immunohistochemical findings of subclinical kidney allograft rejection and its association with graft outcome

Subclinical acute rejection (SAR) occurs in about 30% of stable renal transplant patients and may be a risk factor for a poor allograft outcome. In the present study, the prevalence and clinical features of subclinical rejection, and the expression of immune activation markers in surveillance graft biopsies were assessed and correlated with late graft outcomes. Protocol biopsies were obtained at 2 and 12 months post-transplant in 32 and 26 patients, respectively, with stable renal function. The Banff 1997 criteria were used for histological diagnosis. Graft function and survival and proteinuria were assessed during the 36 months of follow-up. Immunohistochemical evaluation of cell subpopulations and immunoactivation markers were performed on protocol biopsies. The prevalence of SAR at 2 months and of chronic allograft nephropathy (CAN) at 12 months in representative biopsies was 55 and 50%, respectively. Patients with SAR presented mononuclear cell infiltration with an increased expression of CD3, CD4, CD68, IL-2R and granzyme B. Kidney graft function was significantly worse in patients with SAR at 2 months who had chronic rejection on biopsy at 12 months, but SAR was not associated with a worse graft function, greater proteinuria or a lower graft survival in 3 yr of follow-up. In conclusion, we found an elevated prevalence of SAR at 2 months after transplantation with an increased expression of activation markers. Although an association of SAR with poor graft outcome was not observed, our results suggest that SAR is an immunologically active process and underscore the importance of protocol biopsies in the surveillance of transplanted kidneys.

Molecular markers in subclinical acute rejection of renal transplants

In this study, we evaluated the expression of molecular markers of acute rejection in protocol biopsies of patients with and without subclinical acute rejection (SAR). Protocol biopsies were performed at 2 months (n = 21) and 12 months (n = 14) after kidney transplantation in patients with stable allograft function. After biopsy tissue RNA isolation, reverse transcription and polymerase chain reaction (RT-PCR) for the glyceraldehyde 3-phospate dehydrogenase (GAPDH), perforin, granzyme B and Fas ligand genes were performed. The Banff 97 classification was used for histological diagnosis. Creatinine concentrations at 2 months were significantly higher in patients with SAR (1.46 +/- 0.27 x 1.18 +/- 0.24; p < 0.02). Perforin transcripts were found in 15 biopsy specimens, 10 of which had histological signs of SAR (p = 0.06). Granzyme B expression was found in 10 specimens, nine of which had SAR (p < 0.01). Fas ligand was expressed in seven specimens, and six of them were classified as SAR (p < 0.01). Perforin expression had the highest sensitivity (81%) for the diagnosis of SAR. Granzyme B and Fas ligand had specificity of 90%. At 12 months, there was no significant difference in creatinine concentrations for patients with and without previous SAR (1.63 +/- 0.57 x 1.28 +/- 0.31; p = 0.10). Molecular analysis revealed that there was no statistically significant difference in the expression of perforin and granzyme B in patients with and without SAR. Fas ligand expression was observed in five samples, four of which had histological signs of SAR (p = 0.03). At 12 months, perforin expression had the highest sensitivity (83%), and Fas ligand, the highest specificity (88%) for the diagnosis of SAR. We concluded that the expression of genes that encode proteins involved in the cytolytic attack against the allograft is increased in kidneys with SAR. These findings support the understanding that SAR is an active immune process potentially deleterious to renal allografts.