The correlation of Banff scoring with reversibility of first and recurrent rejection episodes

Unsatisfactory reproducibility of interstitial inflammation scoring in allograft kidney biopsy

Interstitial inflammation scoring is incorporated into the Banff Classification of Renal Allograft Pathology and is essential for the diagnosis of T-cell mediated rejection. However, its reproducibility, including inter-rater and intra-rater reliabilities, has not been carefully investigated. In this study, eight renal pathologists from different hospitals independently scored 45 kidney allograft biopsies with varying extents of interstitial inflammation. Inter-rater reliabilities and intra-rater reliabilities were investigated by kappa statistics and conditional agreement probabilities. Individual pathologists' scoring patterns were examined by chi-squared tests and proportions tests. The mean pairwise kappa values for inter-rater reliability were 0.27, 0.30, and 0.26 for the Banff i score, ti score, and i-IFTA, respectively. No rater pair performed consistently better or worse than others on all three scorings. After dichotomizing the scores into two groups (none/mild and moderate/severe inflammation), the averaged conditional agreements ranged from 47.1% to 50.0%. The distributions of the scores differed, but some pathologists persistently scored higher or lower than others. Given the important role of interstitial inflammation scoring in the diagnosis of T-cell mediated rejection, transplant practitioners should be aware of the possible clinical implications of the far-from-optimal reproducibility.

Early Transplant Arteriopathy in Kidney Transplantation

BACKGROUND

Early dysfunction of renal allografts may be associated with vascular injury, which raises the specter of active rejection processes that require medical intervention. In our practice, we have encountered patients who present with delayed graft function and demonstrate a unique pattern of endothelial cell injury that raises concern for rejection in their biopsy. Therefore, we sought to systematically determine the biopsy characteristics and outcome of these patients.

METHODS

During a 17-year period at the University of Washington in Seattle, United States, we identified 24 cases of a distinct arterial vasculopathy presenting in the first year posttransplantation. This early transplant arteriopathy (ETA) is characterized by endothelial cell swelling and intimal edema but without the intimal arteritis that defines vascular rejection.

RESULTS

Approximately 1% of transplant biopsies during the study period showed ETA, almost all of which were in deceased donor organs (96%), and most presented with delayed graft function (54%) or increased serum creatinine (38%) soon after transplantation (median 13 days; range, 5-139). In this study, 77% of patients were managed expectantly, with only 2 patients (7.6%) subsequently developing acute vascular rejection. Except for 1 patient who died, all patients had functioning allografts at 1 year follow-up.

CONCLUSION

Recognizing ETA and distinguishing it from vascular rejection is important to prevent over-treatment because most patients appear to recover allograft function rapidly with expectant management.

Higher infiltration by Th17 cells compared with regulatory T cells is associated with severe acute T-cell-mediated graft rejection

The aim of this study was to evaluate whether the Th17 and Treg cell infiltration into allograft tissue is associated with the severity of allograft dysfunction and tissue injury in acute T cell-mediated rejection (ATCMR). Seventy-one allograft tissues with biopsy-proven ATCMR were included. The biopsy specimens were immunostained for FOXP3 and IL-17. The allograft function was assessed at biopsy by measuring serum creatinine (Scr) concentration, and by applying the modified diet in renal disease (MDRD) formula, which provides the estimated glomerular filtration rate (eGFR). The severity of allograft tissue injury was assessed by calculating tissue injury scores using the Banff classification. The average numbers of infiltrating Treg and Th17 cells were 11.6 ± 12.2 cells/mm² and 5.6 ± 8.0 cells/mm², respectively. The average Treg/Th17 ratio was 5.6 ± 8.2. The Treg/Th17 ratio was significantly associated with allograft function (Scr and MDRD eGFR) and with the severity of interstitial injury and tubular injury (P < 0.05, all parameters). In separate analyses of the number of infiltrating Treg and Th17 cells, Th17 cell infiltration was significantly associated with allograft function and the severity of tissue injury. By contrast, Treg cell infiltration was not significantly associated with allograft dysfunction or the severity of tissue injury. The results of this study show that higher infiltration of Th17 cell compared with Treg cell is significantly associated with the severity of allograft dysfunction and tissue injury.

Renal allograft biopsies with borderline changes: predictive factors of clinical outcome

The clinical outcome and appropriate management for patients showing 'borderline changes' on allograft biopsy after renal transplantation is still controversial. In an attempt to identify predictive factors of clinical outcome of patients with such lesions, we reviewed the clinical course of 91 patients with borderline changes. Multivariate analysis revealed significant and independent effects of histological stage (i + t < or = or > 2) and time to borderline changes (< or = or > 3 months after transplant) on serum creatinine levels at 1 year from borderline changes episodes (respectively, p = 0.04 and p = 0.02) and only a significant effect of time to borderline changes on serum creatinine levels at 2 years (p = 0.005). Renal function at 1 year and 2 years as 5- and 8-year graft survival were not significantly different in the group of patients treated with antirejection therapy (T group, n = 49) compared with the untreated group (UT group, n = 42). This study strongly suggests that borderline changes with histological score (i + t) > 2 and late episodes of borderline changes should be considered to be of poor prognosis.

Indicators of Acute Rejection on Tc-99m DTPA Renal Scintigraphy

INTRODUCTION

Specific criteria for diagnosing an acute rejection episode (ARE) are not present on renal scintigraphy. However, a deterioration in renal function observed on serial imaging is suggestive of an ARE during the early posttransplantation period. In this study, we evaluated Tc-99m DTPA renal scintigraphy findings among patients with renal allograft dysfunction. The aim was to define criteria for ARE on a single imaging study alone.

MATERIALS AND METHODS

This study included 82 patients with renal allograft dysfunction, each of whom had a specific biopsy-proven pathology. From the background-corrected Tc-99m DTPA renal time-activity curves, the following parameters were computed: the ratio of counts at peak perfusion to that at plateau (P:PL) and the ratio of counts at peak perfusion to that at peak uptake (P:U). The images were inspected visually for excretion patterns and other morphological changes.

RESULTS

A specific renogram pattern, that is an increase in perfusion to uptake ratio together with a nearly flat uptake curve and preserved peak/plateau pattern, was observed in 67% of low-grade AREs. In contrast, a flat uptake curve usually occurred together with a loss of peak/plateau pattern in a high-grade ARE. The findings that were highly specific for ARE on visual inspection were little or no collecting system activity, pelvic hypoactivity, and indistinctness of the corticomedullary junction.

CONCLUSIONS

Our results indicate that in low-grade ARE, there is decreased filtration relative to perfusion as evidenced by a decrease in the perfusion-to-uptake ratio with loss of a peak/plateau pattern. A hypoactive pelvis with little activity in the collecting systems may be due to edema in the renal interstitium and pelvicalyceal structures.

Lesions of T-cell-mediated kidney allograft rejection in mice do not require perforin or granzymes A and B

Organ allograft rejection is strongly associated with the presence of alloreactive cytotoxic T cells but the role of cytotoxicity in the pathologic lesions is unclear. Previous studies showed that the principal lesions of kidney rejection - interstitial infiltration, tubulitis, and endothelial arteritis - are T-cell-dependent and antibody-independent. We studied the role of cytotoxic granule components perforin and granzymes A and B in the evolution of the T-cell-mediated lesions of mouse kidney transplant rejection. By real-time RT-PCR, allografts rejecting in wild-type hosts at days 5, 7, 21, and 42 showed massively elevated and persistent expression of perforin and granzymes A and B, but evolution of tubulitis and arteritis did not correlate with increasing granzyme or perforin expression. Allografts transplanted into hosts with disrupted genes for perforin or granzymes A and B showed no change in tubulitis, arteritis, or MHC induction. Thus the development of the histologic lesions diagnostic of T-cell-mediated kidney transplant rejection are associated with but not mediated by perforin or granzyme A or B. Together with previous graft survival studies, these results indicate that the granule-associated cytotoxic mechanisms of T cells are not the effectors of T-cell-mediated allograft rejection.

Heterogeneity in the evolution and mechanisms of the lesions of kidney allograft rejection in mice

The natural history and pathogenesis of the pathologic lesions that define rejection of kidney transplants have not been well characterized. We studied the evolution of the pathology of rejection in mouse kidney allografts, using four strain combinations across full major histocompatibility complex (MHC) plus nonMHC disparities, to permit more general conclusions. Interstitial infiltrate, MHC induction, and venulitis appeared by day 5, peaked at day 7-10, then stabilized or regressed by day 21. In contrast, tubulitis, arteritis, and glomerulitis were absent or mild at days 5 and 7, but progressed through day 21, indicating separate regulation and homeostatic control of these lesions. Edema, hemorrhage, and necrosis also increased through day 21. All lesions were T-dependent, failing to develop in T-cell-deficient hosts. Allografts into immunoglobulin-deficient hosts manifested typical infiltration, MHC induction, and tubulitis at days 7 and 21, indicating that these lesions are alloantibody-independent. However at day 21 kidneys rejecting in immunoglobulin-deficient hosts showed decreased edema, arteritis, venulitis, and necrosis. Thus the three groups of lesions are: T-cell-mediated interstitial infiltration, MHC induction, and venulitis, which develops rapidly then stabilizes; slower but progressive T-cell-mediated tubulitis and arteritis; and late antibody-mediated endothelial injury, which contributes to late edema, arteritis, and venulitis.

Correlation between the Banff 97 classification of renal allograft biopsies and clinical outcome

The 1997 fourth Banff meeting revised the consensus for describing transplant biopsies. We have conducted a retrospective analysis of biopsies correlated between the Banff 97 classification and clinical outcome. The patients ( n=149), who had a total of 404 biopsy-proven rejections, were assessed and the biopsies taken from these patients were re-examined and classified according to the Banff 97 classification. Morphological changes in the glomeruli (g), interstitium (i), tubules(t), and arterial vessels (v) were scored. Severity of acute rejection was statistically associated with unresponsiveness to anti-rejection treatment ( P<0.0001) and predicted an increased risk of graft failure ( P<0.05). Each quantitative criterion (g, i, t, and v) was also statistically associated with unresponsiveness to anti-rejection treatment. Mean serum creatinine levels were significantly higher in the groups graded Banff 97 type I-III after 1 and 2 years of follow-up. The Banff 97 classification correlated with reversibility of rejection episodes and long-term graft survival.

International variation in the interpretation of renal transplant biopsies: report of the CERTPAP Project

BACKGROUND

The Banff working formulation of renal transplant pathology is intended to have international application. There remains a need to develop methods to harmonize the application of such grading systems between laboratories. Banff grades do not always permit precise management decisions to be made. Alternative schemes have been devised for the diagnosis of acute rejection, but there have been no independent tests of the different approaches.

METHODS

Sections from 55 renal transplant biopsies were circulated around the laboratories of 22 major transplant units for the Convergence of European Renal Transplant Pathology Assessment Procedures (CERTPAP) Project. Participating pathologists were asked to grade 32 different histological features, without any clinical information. After each circulation of five cases, feedback was provided to participants. Statistical evidence of improvement in interobserver variation was sought. At the end of the study, correlations with the original clinicopathological diagnosis were sought.

RESULTS

Interobserver variation was greater than has previously been reported. For every feature studied, some pathologists consistently under-grade or over-grade. There was relatively little evidence of improvement in interobserver variation as a result of the feedback system. No single feature permitted a reliable diagnosis of acute rejection. Applying the Banff and CCTT schemas to the histological grades showed no clear diagnostic advantage for either system, but a simple computer-based inference network, which combined data from 12 histological features, out performed either approach. Within the "protocol" biopsies studied, long-term survival correlated better with "acute" than with "chronic" histological features.

CONCLUSIONS

These results do not undermine the value of the Banff classification, but they demonstrate a need for caution when translating biopsy results between institutions. It is obvious that evaluation of biopsies in multicenter trials must be done in one center. In the management of individual patients, the need to interpret Banff grades in the light of local experience and clinical information is stressed.

Cohort study of the prognostic significance of acute transplant glomerulitis in acutely rejecting renal allografts

BACKGROUND

Acute transplant glomerulitis is a unique lesion in renal allografts, the prognostic significance of which is controversial. We conducted this retrospective cohort study to examine the independent prognostic significance of moderate-to-severe transplant glomerulitis in acute rejection.

METHODS

Renal allograft survival for patients with acute rejection were studied, comparing one group with significant glomerulitis (G, n=28) with those with no glomerulitis (NG, n=35). Clinical, biopsy, and demographic data and renal graft survival were compared, and the association of G with graft failure was examined.

RESULTS

In the G versus NG group, a greater percentage of patients were highly sensitized (peak panel reactive antibody value >80%; P=0.009), had had a previous renal transplant (40% vs. 11%; P=0.02), or had suffered from delayed graft function (P=0.03). The G group had a trend toward earlier rejection episodes (P=0.07), a significantly higher serum creatinine at the time of index biopsy (P=0.01), a higher prevalence of vascular rejection (P=0.02), and less improvement in mean reciprocal serum creatinine at 1-2 weeks after biopsy (P=0.02). Although there was a trend toward shorter allograft survival in the G group (P=0.09), the level of significance of which increased with adjustment for transplantation time period and the duration of the transplant-biopsy interval (P=0.06), the relative risk for graft loss was no longer significant when additionally adjusted for index biopsy Banff score (relative risk, 0.97; P=0.97).

CONCLUSION

In this study, G was significantly more common in highly sensitized patients and was strongly associated with vascular rejection biopsies but was not an independent predictor of graft survival.

RAD in de novo renal transplantation: comparison of three doses on the incidence and severity of acute rejection

BACKGROUND

The effects of three doses of RAD (40-O-[2-hydroxyethyl]-rapamycin), a novel macrolide with potent immunosuppressive and antiproliferative properties, on the incidence and severity of acute rejection episodes as well as its tolerability were evaluated in a dose-ranging study in de novo renal transplant recipients.

METHODS

In this double-blind, parallel group, multicenter study, recipients were randomized to receive 1 mg, 2 mg, or 4 mg/day of RAD in combination with Neoral (cyclosporine, USP MODIFIED) and corticosteroids. The incidence and severity of biopsy-proven acute rejection episodes, graft survival, patient survival, infection rates, laboratory measurements, and adverse events were compared across groups after 6 months of therapy.

RESULTS

Among the 103 recipients, patients receiving 1, 2, or 4 mg/day experienced a 32.4%, 14.7%, or 25.7% incidence of biopsy-proven acute rejection episodes within the first 6 months posttransplantation, respectively. Even though the study was not powered to demonstrate efficacy, the incidence of moderate and severe acute rejection episodes was found to be significantly lower among patients in the 2 mg and 4 mg/day groups than in the 1 mg/day group (P=0.002 and P=0.006, respectively). Overall graft and patient survival rates were excellent. RAD was generally well tolerated. Although blood lipid levels increased in all groups, changes were manageable with lipid-lowering agents and did not warrant discontinuation of study medication. The incidence of viral and fungal infections was low; however, it was higher among recipients treated with 4 mg/day.

CONCLUSIONS

In combination with Neoral and corticosteroids, RAD doses of 2 mg and 4 mg/day resulted in lower rates of acute rejection episodes and efficacy failure than the 1 mg/day dose and were significantly more effective in reducing the severity of rejection. Large-scale, controlled, follow-up studies are currently in progress to confirm these initial findings.

Ex vivo propagation and characterization of lymphocytes from rejecting rat-kidney allografts

Today, most clinically used methods for analysis of alloreactivity in organ transplantation are based on humoral immunity. In order to study the cellular alloresponse, a rat kidney transplantation model with culturing of graft infiltrating lymphocytes was developed. Kidney transplantations between inbred rat strains were performed with the animals initially immunosuppressed with cyclosporine. In order to initiate acute cellular rejection, immunosuppression was withdrawn after 10 days. Infiltrating lymphocytes were analysed using an in vitro culture system, allowing cells to propagate from the biopsies to culture medium. The propagated cells were counted and analysed for subtype activation markers and donor-specificity using flow cytometry and a proliferation assay. Syngeneically transplanted animals and animals given constant immunosuppression upon transplantation were used as controls. During rejection, significantly more T lymphocytes were propagating from the biopsies compared to controls. A higher percentage of the propagated T lymphocytes in the rejection group expressed activation markers [CD25 and major histocompatibility complex (MHC) class II antigen] compared to spleen- and peripheral blood T lymphocytes from the same individuals. Propagated mononuclear cells from biopsies in the rejection group were proliferating and showed donor-specific reactivity whereas mononuclear spleen cells from animals in the same group did not show this donor specificity. In conclusion, we have presented a rat kidney allotransplantation model with in vitro propagation of graft-infiltrating, activated and donor-specific T lymphocytes. This technique offers a possibility to study cellular reactivity in allotransplantation.

Quantitative assessment of the first acute rejection as a predictor of renal transplant outcome

BACKGROUND

Acute rejection (AR) of the transplanted kidney has been identified as the major risk factor for the development of chronic rejection and immunological graft loss. However, the clinical presentation and response to AR therapy can vary considerably between recipients.

METHODS

We studied the first AR episode in 201 kidney-only recipients transplanted between January 1987 and June 1998 who were biopsied between April 1993 and June 1998 and were graded using the Banff schema. All patients received cyclosporine-based immunosuppression. There were 134 cadaver donor (66.7%) and 67 live donor (33.3%) recipients followed for a mean of 46.2 (range 4-128) months. All Banff grade 1-3 and 40/78 borderline (BL) cases were treated for rejection after biopsy. These patients were compared with a contemporaneous control population who did not experience AR. Demographic risk factors associated with graft loss were identified in both univariate and multivariate analysis. Daily (0-18) serum creatinine (SCr) values during and after the AR were plotted for each patient to generate curves and calculate area under the serum creatinine versus time curve (mg/dl/day). Four response patterns to treatment were identified according to the velocity of % increase (V1) and decrease (V2) of serum creatinine. These were identified as rapid rise and fall (n=62); rapid rise and slow fall (n=43); slow rise and fall (n=55); and slow rise and rapid fall (n=41). Kaplan-Meier graft survivals were compared between the groups.

RESULTS

Any Banff grade was associated with increased risk for graft loss (P=0.0001). However, no significant differences were observed between the Banff BL and B1-3 groups, or among those BL patients who were treated or remained untreated for AR. Multivariate analysis identified a black recipient (P=0.03, risk ratio 2.0) and area under the serum creatinine versus time curve (P=0.0001, risk ratio 3.2) as significant risk factors for graft loss. The AR response pattern RS resulted in a significantly (P=0.0072) diminished 5-year graft survival (45%) compared with the other groups. Serum creatinine pattern, but not Banff grade, was also a significant (P=0.025) predictor of re-rejection.

CONCLUSIONS

These data suggest that all Banff grades, including BL, carry a significant risk for graft loss, and the initial response to antirejection therapy can predict long-term graft outcome. They support the practice of treating AR promptly and definitively and suggest that the RS subgroup of rejecting grafts could be targeted for additional antirejection therapy. This subgroup can be identified by 10 days of AR therapy, and should be the subject of further study.

A blinded retrospective analysis of renal allograft pathology using the Banff schema: implications for clinical management

BACKGROUND

We sought to determine whether diagnoses established through the Banff schema for evaluation of renal allograft pathology have implications for clinical management, compared with diagnoses established using descriptive terminology.

METHODS

All patients included in this study had mild to severe allograft rejection diagnosed, and, as part of a therapeutic protocol, they received OKT3 as primary anti-rejection therapy. We conducted a retrospective review of their renal allograft biopsy specimens and reclassified them, using the Banff schema, without knowledge of clinical information, laboratory data, or previous biopsy interpretation. Although there is no strict correspondence between descriptive diagnostic terminology and the criteria used in the Banff schema, for the purpose of comparisons, the following approximation was used: mild and mild to moderate rejection=Banff borderline and Banff grade 1, moderate and moderate to severe rejection=Banff grades 2A and 2B, and severe rejection=Banff grade 3. The diagnosis was considered concordant when the diagnosis by descriptive terminology and Banff grading were within the adopted approximation.

RESULTS

Of 96 biopsies specimens with mild to severe allograft rejection, 10 were insufficient for diagnosis, and three had changes of chronic allograft rejection. Of the remaining 83 biopsy specimens, 34 (41%) were concordant in interpretation of rejection grades, whereas 49 (59%) were discrepant. The greatest degree of concordance was in grades 2A (66.7%, 18 of 27) and 2B (64.7% 11 of 17), and the lowest was in the borderline category (11.8%, 2 of 17). The greatest degree of discrepancy was in normal and grade 3 (100%, 3 of 3 and 2 of 2, respectively), and the lowest was in grade 2A (33.3%, 9 of 27). Although primary anti-rejection therapy with OKT3 resulted in a high reversal rate of rejection (98%), there were 5 deaths, 12 graft loses, six episodes of serious infections, and three malignancies in this group of patients during a mean follow-up period of approximately 38 months.

CONCLUSIONS

Because patients with borderline changes and grades 1 and 2A rejection may be treated differently from patients with higher grades (2B and 3), the use of the Banff schema may allow for better adjustment of immunosuppressive therapy in response to specific grades of acute allograft rejection and may result in decreased complications of immunosuppressive therapy.

The US multicenter double-blind, randomized, phase III trial of thymoglobulin versus atgam in the treatment of acute graft rejection episodes following renal transplantation: rationale for study design

In this study intended to establish equivalence between two antibody therapies for acute rejection in kidney transplant recipients, it was important to develop a rigorous protocol. Assurance of the presence of acute rejection was imperative. Therefore, due to the lack of literature support for clinical assessment of renal dysfunction, histologic diagnosis of acute rejection was required for enrollment in the study. Likewise, supportive literature for a correlation between response to anti-rejection therapy and the severity of rejection lead to the decision that the study should be stratified by a measurement of rejection severity for which Banff criteria were used. Finally, quantification of the response to therapy was also measured against the available literature and a large, newly developed international database of kidney transplant rejection episodes (the Efficacy Endpoints database) where serum creatinine, expressed as a percentage of the baseline level at the time of rejection was shown to be the most effective, available clinical marker of rejection response. Therefore, the US Multicenter Phase III Trial for comparing Thymoglobulin to Atgam in the treatment of acute rejection exhibits a unique and detailed study design that could be implemented in future trials as well as in clinical practice to improve assessment of outcomes.

Histological outcome of acute cellular rejection in kidney transplantation after treatment with methylprednisolone

BACKGROUND

Several studies comparing the response of acute cellular rejection (ACR) episodes to different corticosteroid regimens have been conducted. However, in most of them, the histological evaluation of the infiltrate and its correlation with clinical response was not studied. The clinical and histological outcomes of 37 episodes of ACR treated with methylprednisolone (MP) were studied, with the aim to determine how long the infiltrate takes to be cleared after therapy.

METHODS

A total of 37 patients with biopsy-proven ACR were treated with 8 or 16 mg of MP/kg/day. Allograft biopsies were repeated at 5 and 10 days after the end of corticotherapy. Clinical and histological outcomes were compared.

RESULTS

Six patients were excluded; 15 (48.4%) patients responded to therapy; the mean serum creatinine of these patients reached normal levels in the 2 weeks that followed treatment. Nine patients (60%) of this group had signs of ACR on biopsies done 5 days after corticotherapy, and four (26.7%) maintained them on the 10th day. Among 16 patients with no clinical response, none reached normal serum creatinine levels; 15 (93.7%) had signs of rejection 5 days after treatment and maintained them on the 10th day. Histological signs of ACR disappeared in 73.3% of patients with clinical response 10 days after therapy, but in only 6.3% of patients with no response (P=0.001).

CONCLUSIONS

Biopsies performed 5 days after treatment show a high incidence of features of ACR; such features take on average 10 days to disappear in nearly 75% of cases with successful therapy with MP.

The Banff 97 working classification of renal allograft pathology

BACKGROUND

Standardization of renal allograft biopsy interpretation is necessary to guide therapy and to establish an objective end point for clinical trials. This manuscript describes a classification, Banff 97, developed by investigators using the Banff Schema and the Collaborative Clinical Trials in Transplantation (CCTT) modification for diagnosis of renal allograft pathology.

METHODS

Banff 97 grew from an international consensus discussion begun at Banff and continued via the Internet. This schema developed from (a) analysis of data using the Banff classification, (b) publication of and experience with the CCTT modification, (c) international conferences, and (d) data from recent studies on impact of vasculitis on transplant outcome.

RESULTS

Semiquantitative lesion scoring continues to focus on tubulitis and arteritis but includes a minimum threshold for interstitial inflammation. Banff 97 defines "types" of acute/active rejection. Type I is tubulointerstitial rejection without arteritis. Type II is vascular rejection with intimal arteritis, and type III is severe rejection with transmural arterial changes. Biopsies with only mild inflammation are graded as "borderline/suspicious for rejection." Chronic/sclerosing allograft changes are graded based on severity of tubular atrophy and interstitial fibrosis. Antibody-mediated rejection, hyperacute or accelerated acute in presentation, is also categorized, as are other significant allograft findings.

CONCLUSIONS

The Banff 97 working classification refines earlier schemas and represents input from two classifications most widely used in clinical rejection trials and in clinical practice worldwide. Major changes include the following: rejection with vasculitis is separated from tubulointerstitial rejection; severe rejection requires transmural changes in arteries; "borderline" rejection can only be interpreted in a clinical context; antibody-mediated rejection is further defined, and lesion scoring focuses on most severely involved structures. Criteria for specimen adequacy have also been modified. Banff 97 represents a significant refinement of allograft assessment, developed via international consensus discussions.

Utility of standardized histological classification in the management of acute rejection. 1995 Efficacy Endpoints Conference

BACKGROUND

Standardized histological grading of transplant kidney biopsies has become a primary criterion for diagnosis of rejection in immunosuppression clinical trials.

METHODS

A consortium of 19 transplant centers from North America, Europe, and Australia convened in 1995 to examine kidney transplant rejection. Data from the 1995 Efficacy Endpoints Conference were examined for frequency of adoption of Banff schema. Biopsy grading was correlated with clinical parameters of rejection and therapy response.

RESULTS

Histological confirmation of rejection episodes occurred in 73% of 953 cases, with Banff criteria adoption increasing in frequency between 1992 and 1995. Banff grading significantly correlated with clinical rejection severity (rejection creatinine: grade I, 2.8+/-0.2 mg/dl; grade II, 3.5+/-0.2 mg/dl; grade III, 4.1+/-0.3 mg/dl; P < 0.001), although nadir creatinines were similar. Response rates of Banff grades I and II to steroid therapy were not different, but only 42% of grade III rejections responded to steroids (P < 0.003. Banff grading also correlated with postrejection creatinine, day 15: grade I, 2.2+/-0.2 mg/dl; grade II, 3.0+/-0.2 mg/dl; grade III, 3.8+/-0.4 mg/dl (P < 0.001), and day 30: grade I, 2.1+/-0.1 mg/dl; grade II, 2.2+/-0.2 mg/dl; grade III, 2.7+/-0.2 mg/dl (P < 0.06). Banff grade III correlated with reduced graft survival at 1 year: grade I, 86%; grade II, 88%; grade III, 70% (P < 0.01).

CONCLUSIONS

This multicenter review of rejection severity confirms that standardized histologic classifications such as the Banff schema provide a reliable means for stratifying patient risk of treatment success or failure. These data support the use of Banff criteria in clinical trial design.

Diagnosis of early acute renal allograft rejection by evaluation of multiple histological features using a Bayesian belief network

BACKGROUND AND AIMS

The development of the Banff classification of renal transplant pathology has allowed the standardisation of approaches to transplant biopsy histology and reduced interobserver and interdepartmental variation. The usefulness of the Banff classification in the diagnosis of acute rejection has previously been tested by sending sections from 21 "difficult" biopsies to almost all of the renal transplant pathologists in the UK. Although the Banff classification improved reproducibility, the accuracy of diagnosis of early acute rejection was unchanged from the "conventional" approach. Perhaps this is because in making a diagnosis of acute rejection, the Banff classification uses only two features: tubulitis and intimal arteritis. To include more features on a systematic basis would be laborious for a human observer. Therefore, a Bayesian belief network was developed for this task.

METHODS

The network was initialised with observations from 110 transplant biopsies. Its performance was then tested on 21 biopsies that had been seen by 37 different renal transplant pathologists in an earlier study. These biopsies had been selected to represent histologically difficult problems but, in retrospect, they all had clear diagnoses of rejection or non-rejection on clinical grounds.

RESULTS

Using the Bayesian belief network, a relatively inexperienced pathologist made 19 of 21 correct diagnoses, better than had been achieved by any of the pathologists who had seen the same sections previously (17 of 21), and considerably better than the average proportion of correct diagnoses provided by all 37 renal transplant pathologists (65%). Application of the system by a second pathologist produced a tendency to overdiagnosis of acute rejection, illustrating the consequences of interobserver variation.

CONCLUSIONS

In the diagnosis of acute rejection, further useful information can be extracted from features that are currently not considered in the Banff classification. Integration of data by a computer can give a more reliable diagnosis of early acute rejection, but routine application will require the development of a more sophisticated system that can also accommodate clinical data, perhaps one that can continue to "learn" as more data are entered.

Correlation of clinical outcomes after tacrolimus conversion for resistant kidney rejection or cyclosporine toxicity with pathologic staging by the Banff criteria

BACKGROUND

Refractory rejection and cyclosporine (CsA)-induced nephropathy remain important causes of renal allograft loss. Previous studies demonstrated that 70-85% of the episodes of refractory acute rejection (AR) occurring in renal allograft recipients on a CsA-based immunosuppressive regimen could be salvaged by conversion to tacrolimus. No data are available regarding the correlation between allograft histology at the time of conversion and the response to tacrolimus. We examined the response to tacrolimus conversion in relation to preconversion biopsies stratified by the Banff criteria.

METHODS

Since May 1992, we have converted 22 patients from CsA to tacrolimus as part of a rescue protocol. We report on 18 patients in whom 6-month follow-up was available after conversion for biopsy-proven AR (n=13) or CsA toxicity (n=5). Sixteen patients were recipients of renal allografts, including three second transplants, and two were recipients of kidney-pancreas transplants. All patients with AR were treated with one or more courses of methylprednisolone and OKT3 before conversion. Renal allograft biopsies were interpreted by a transplant pathologist blinded to the clinical history, and graded according to the Banff criteria. Responses to tacrolimus were scored as improved (creatinine returned to within 150% of baseline), stabilized (creatinine rise arrested), or failed (returned to dialysis). RESULTS; Mean follow-up was 17.3+/-8 months. Fourteen of 18 patients (78%) showed improvement or stabilization in renal function as assessed by creatinine at 6 months or 1 year (when available). Of the 13 patients with histological AR, nine (69%) improved, including five of six with borderline AR, two of three with grade I AR, and two of four with grade II AR. Of the four other patients with AR, two stabilized and two failed. Three of five patients with severe clinical rejection requiring dialysis (range 2-16 weeks) recovered renal function after conversion. Of five patients with CsA toxicity, two (40%) improved. Seven of eight patients who were converted to tacrolimus less than 90 days after transplantation improved, compared with only 4 of 10 who were converted more than 90 days after transplantation. No grafts were lost in patients with a creatinine <3.0 mg/dl at the time of conversion versus two of seven grafts lost when the creatinine was 3.1-5.0 mg/dl and two of eight grafts lost when the creatinine was >5.0 mg/dl.

CONCLUSION

The majority of steroid and antilymphocyte antibody (OKT3 or ATGAM) unresponsive rejections in patients on CsA-based immunosuppression will improve or stabilize after conversion to tacrolimus. There was no correlation with allograft histology stratified by the Banff criteria and the response to tacrolimus. Although there was a trend toward a poorer response with more severe histological rejection, higher serum creatinine at the time of conversion, and longer time from transplantation to conversion, favorable responses were noted in all groups. This indicates that a trial of conversion is warranted, irrespective of the histological severity of injury.