Histological analysis of pre-transplant deceased donor renal biopsies and its association with long-term graft survival and function

1Universidade Federal De Ciencias Da Saude De Porto Alegre (UFCSPA), Santa Casa De Porto Alegre Hospital, Nossa Senhora Da Conceicao Hospital, Porto Alegre, Brazil 2Santa Casa De Porto Alegre Hospital , Porto Alegre, Brazil 3Universidade Federal De Rio Grande Do Sul, Complexo Hospitalar Santa Casa, Hospital Nossa Senhora Da Conceicao, Porto Alegre, Universidade Luterana Do Brasil, Canoas, Brasil 4Universidade Federal De Ciencias Da Saude De Porto Alegre (UFCSPA), Santa Casa De Porto Alegre Hospital, Porto Alegre, Brazil

• Tubular atrophy; absent (0%), mild (<25%), moderate (26%-50%), severe (>50%).The glomerular filtration rate (GFR) was estimated by the CKD-EPI equation at 1, 3, and 5 years posttransplantation.Graft survival was calculated by the Kaplan-Meier method at 1, 3, and 5 years after kidney transplantation.The following variables were analyzed in the study as factors associated with graft loss: donor and recipient age and gender, standard or expanded criteria donor, presence of donor-specific antibodies (DSA) by the recipient, number of HLA-A/B/DR mismatches, class I and II panel reactive antibodies, occurrence of delayed graft function, occurrence of acute rejection episodes in the first year post-transplantation, cold ischemia time, and type of immunosuppressive therapy.Variables were analyzed by univariate and multivariate statistical methods using the Cox regression model.

Ethical issues
1) The research followed the tenets of the Declaration of Helsinki; 2) informed consent was obtained; and 3) This study was approved by the Ethics Committee of Universidade Federal de Ciências da Saúde de Porto Alegre (ethical code 495.047).

Statistical analysis
Statistical analysis was performed using the SPSS® (Porto Alegre, Brazil) version 20 software.Quantitative variables were presented as mean and standard deviation.Association of categorical variables was performed by the chi-square test, Fischer's exact test, and continuous variables by analysis of variance (ANOVA), adopting a significance level of P <0.05.ROC curve analysis was used to establish the cut-offs of GS >25% and GFR <30 mL/dL.Kaplan-Meier method was used to graft survival.Cox regression model was performed for univariate and multivariate analysis.

Results
Table 1 presents the clinical and demographic characteristics of donors and recipients.Approximately 60% of the donors and recipients were male, with a mean age of about 50 years.The ratio between SD and ECD donors was approximately 1:1 (SD = 49.3%,ECD = 50.7%).In the first year post-transplantation, 41.6% of the recipients suffered an acute rejection episode, 14.2% had DSA, and 64.9% had delayed graft function.Table 2 shows data on the lesions in renal compartments according to the Banff criteria.The number of glomeruli identified in the biopsies ranged from 6 to 145.Biopsies showing less than 10 glomeruli that could significantly represent the tubulointerstitial and artery compartments were included in the study but did not have the glomerular compartment evaluated or included in the results.Biopsies that had artifacts in the interstitial and tubular compartments as a result of the intraoperative frozen section examination were not evaluated in relation to these compartments.Twentysix biopsies did not show a good representation of the arteries.The predominant results in all compartments were the absence of lesions or mild chronic lesions.Only one biopsy showed severe GS (>50%), and this recipient lost the graft in the first year post-transplantation. Twelve biopsies showed severe arteriosclerosis .No transplanted kidney had severe interstitial fibrosis or tubular atrophy.Table 3 shows the results of GFR in relation to renal compartments at 1, 3, and 5 years post-transplantation.

Discussion
The present study evaluated 430 pre-implantation biopsies of renal compartments according to the Banff criteria.The scores of each compartment were not added since the classification does not include the sum of score per item.The semi-quantitative Banff grading systems were the most frequently used according to a systematic review that evaluated 47 studies correlating histopathological scores and transplantation prognosis (3), but the weight given to each component, combined into a composite score, varied among studies.Several scores have been proposed to evaluate renal lesions globally, and the most common are the Pirani-Remuzzi score (4) and the Maryland Aggregate Pathology Index (MAPI) (5).This variation in methodology may explain the conflicting results of the literature.
Our study showed a significant difference in graft survival at 1, 3, and 5 years post-transplantation between the different degrees of histopathological lesions in renal compartments.Regarding GS, graft survival was about 50% at one and three years and less than 45% at five years, when the percentage of sclerosed glomeruli was greater than 25% (moderate lesions).On the other hand, in the absence of GS or mild GS (less than 25%), the survival curves remained above 80% for all the analyzed periods.Gaber et al suggested in 1995 that a cut-off of 20% GS was associated with adverse prognosis (6).Bajwa et al (7) showed that a GS greater than 5% was associated with graft failure, whereas Cicciarelli et al (8) observed this same outcome associated with a GS greater than 10%.Other studies reported the association of different GS cut-off values with graft loss (9,10), whereas some studies found no association between them (11-15).Sung et al did not report GS as an independent risk factor for graft loss (16).Likewise, Edwards et al found no association between a GS greater than 20% and posttransplant graft loss (17).
The area under the ROC curves for graft loss and GFR below 30 mL/min with 25% GS were 0.613 and 0.628, respectively, showing a moderate discriminatory capacity, low sensitivity and high specificity.This result was slightly lower than the reported by a study that found an area under the curve of 0.7-0.8 for 20% GS (1).We observed a reduction in graft function as the chronic lesions in all renal compartments increased.At one year post-transplantation, the mean GFR of 44.43 mL/min for GS below 5% decreased to 36.72 ml/min when GS was in the range of 6-25% (P < 0.001) and to 31.64 ml/min when GS was in the range of 26%-50% (P = 0.75).At 3 years, the mean GFR for the absence of GS dropped from 52.23 mL/min to 39.62ml/min for mild GS (P < 0.001) and to 33.18 mL/min for moderate GS (P = 0.067).At five years, the reduction of GFR was not significant, and we attributed this result to the small number of patients who remained in the cohort.GFR significantly decreased as the degree of arteriosclerosis increased at 1, 3, and 5 years post-transplantation.The statistical power was small when comparing the group with severe arteriosclerosis due to the small number of cases in this category (8 patients).The interstitial and tubular compartments showed a significant inverse association of chronic alterations with graft function at 1 and 3 years post-transplantation, and, at 5 years, an inverse association between absence of lesions and mild lesions.
The present study is in agreement with the conclusions of Escofet et al and Randhawa et al (18,19), which showed a reduction in GFR with the increase in GS at 4 years and 1 year post-transplantation, respectively.On the other hand, Koppelstaetter et al (20), Pokorna et al (13), and Arias et al (10) did not find an independent association between GS and graft function at 1, 2, and 3 years post-transplantation, respectively.Lu et al (15), Oda et al (21), and Szanya et al (22) found an association between arteriosclerosis and GFR, but four other studies reported no association between these parameters (10,20,23,24).Few studies report an association between chronic tubulointerstitial damage and graft function.

Arias et al, Cockfield et al, and Koppelstaetter et al found
no association between chronic tubulointerstitial damage and graft function at 6 months, 1 year, and 3 years posttransplantation (10,11,20).
As limitations of our study, we highlight the lack of a global chronicity score; as previously mentioned, this study strictly adopted the Banff criteria.We adopted the assumption that chronicity in any anatomical compartment is a prognostic factor as a justification for the compartmentalized analysis (2).Biopsies with moderate and severe alterations accounted for a small number of the samples in our study, as they are generally Pre-transplant donor renal biopsies 253 refused for transplantation.There was no control of variability among pathologists.A small number of biopsies had one or more compartments that were not evaluated.The compartments that were not evaluated were excluded from the calculation of graft survival and function.

Conclusions
In summary, our study found a significant difference in the graft survival in the evaluated periods according to chronic lesions in all renal compartments, as well as an association with graft function.A donor age greater than 60 years was a risk factor for graft loss, but the distinction between SD and ECD donors showed no association with graft loss.Independent risk factors for graft loss were the presence of DSA, the occurrence of an acute rejection episode in the first year after transplantation, and the degree of GS.The authors suggest that chronic lesions in any renal compartment must be taken into account in the clinical decision to accept the organ, but the greater weight in this decision should be given to GS. Kidney recipients with more than 25% GS had a less favorable outcome in our study.

Limitations of the study
We highlight the lack of a global chronicity score, there was no control of variability among pathologists, and a small number of biopsies had one or more compartments that were not evaluated.

Figure 1
Figure1shows graft survival in relation to renal compartments at 1, 3, and 5 years after transplantation.

Table 1 .
Clinical and demographic data