Generating automated kidney transplant biopsy reports combining molecular measurements with ensembles of machine learning classifiers.

We previously reported a system for assessing rejection in kidney transplant biopsies using microarray-based gene expression data, the Molecular Microscope® Diagnostic System (MMDx). The present study was designed to optimize the accuracy and stability of MMDx diagnoses by replacing single machine learning classifiers with ensembles of diverse classifier methods. We also examined the use of automated report sign-outs and the agreement between multiple human interpreters of the molecular results. Ensembles generated diagnoses that were both more accurate than the best individual classifiers, and nearly as stable as the best, consistent with expectations from the machine learning literature. Human experts had ≈93% agreement (balanced accuracy) signing out the reports, and random forest-based automated sign-outs showed similar levels of agreement with the human experts (92% and 94% for predicting the expert MMDx sign-outs for T cell-mediated (TCMR) and antibody-mediated rejection (ABMR), respectively). In most cases disagreements, whether between experts or between experts and automated sign-outs, were in biopsies near diagnostic thresholds. Considerable disagreement with histology persisted. The balanced accuracies of MMDx sign-outs for histology diagnoses of TCMR and ABMR were 73% and 78%, respectively. Disagreement with histology is largely due to the known noise in histology assessments (ClinicalTrials.gov NCT01299168).

A Multicenter Application of the 2018 Banff Classification for BK Polyomavirus-associated Nephropathy in Renal Transplantation.

BACKGROUND: With current immunosuppressive regimens, BK polyomavirus-associated nephropathy (BKPyVAN) is still a matter of concern. Stratification of patients at risk for allograft loss is of uttermost importance to guide treatment choice and assess prognosis. In 2018, the Banff working group proposed a classification scheme for the prognosis of BKPyVAN, but external application on independent cohorts is yet to be performed. We investigated how the 2018 Banff classification would perform in a multicenter cohort comprising 50 cases of biopsy-proven BKPyVAN compared to previously published classification systems. METHODS: We analyzed consecutive BKPyVAN cases from two Dutch university hospitals between 2002 and 2013, retrieved clinical data, and scored all biopsies according to the Banff 2018 classification, and as a comparison, 4 previously proposed BKPyVAN classification systems. We used estimated glomerular filtration rate trajectories and death-censored graft survival as primary endpoints. RESULTS: The 2018 Banff classification did not associate with estimated glomerular filtration rate decline or graft failure and performed only slightly better than the 4 previously proposed classifiers. Anti-human leukocyte antigen donor-specific antibodies (DSAs), especially in combination with ongoing biopsy-proven BKPyVAN on follow-up, did correlate with graft function and survival. Patients who were DSA+/BKPyVAN+ on follow-up had more inflammation at the baseline biopsy, which by itself was not associated with graft outcomes. CONCLUSIONS: Neither the 2018 Banff BKPyVAN classification nor previously published stratification systems could be applied to our multicenter patient cohort. Our data suggest that there might be a prognostic value for follow-up biopsies and DSA measurements to improve risk stratification after BKPyVAN, although prospective multicenter efforts with protocol measurements are needed to confirm this.

Antibody-mediated rejection in the Banff classifications of 2007 and 2017: A comparison of renal graft loss prediction capability.

BACKGROUND: Antibody-mediated rejection (ABMR) is the leading cause of kidney graft loss worldwide. Criteria for acute humoral rejection (currently labeled active humoral rejection) established by the 2007 Banff classification are highly specific but lack sensitivity. Modifications to the Banff classification were introduced for its 2013 and 2017 versions in order to identify more cases of this entity. PURPOSE: We intend to demonstrate that, compared to its 2007 version, the 2017 Banff classification bears an improved capacity for graft loss prediction when histologic criteria for active ABMR are met. PATIENTS AND METHODS: Single-center retrospective cohort study. A random sample of 201 kidney recipients who underwent a graft biopsy since January 2004 was analyzed. Patients were classified as ever developing histologic characteristics of acute ABMR (2007 Banff) or not and renal survival between groups was compared. The same patients were then classified as ever developing histologic characteristics of active ABMR (2017 Banff) or not and renal survival was again compared. Presence of circulating donor-specific antibodies (DSA) was not taken into consideration. RESULTS: Patients were followed for a median 13.9 ±â€¯7.9 years, during which grafts were biopsied on 537 occasions (2.7 ±â€¯1.6 biopsies per graft). Baseline eGFR was 73.26 ±â€¯17.6 ml/min and baseline creatinine 1.14 ±â€¯0.25 mg/dl. Graft loss occurred in 38 recipients (18.9%) mainly due to ABMR (60.5%). Acute ABMR (2007 Banff) was identified in 11 recipients (5.5%) and graft survival did not differ between groups with and without active ABMR occurrence (log-rank p = 0.939). Active ABMR (2017 Banff) was found in 59 recipients (29%) and graft survival was better from the second post-transplant year onward in the group of patients without active ABMR occurrence (log-rank p = 0.001). Moderate microvascular inflammation was present in 89.6% of the 48 additional patients with active ABMR. CONCLUSION: The 2017 Banff classification identifies more patients who develop active ABMR and stratifies graft loss risk better than the 2007 version.

Classification of Acute Rejection Episodes in Kidney Transplantation: a Proposal Based on Factor Analysis.

INTRODUCTION: Kidney transplantation is considered the ideal treatment for end-stage renal disease. Acute rejection can influence graft survival. The aim of this study was to propose a classification system for acute rejection based on factor analysis. MATERIALS AND METHODS: Data were collected from kidney transplant recipients with acute rejection diagnosis based on standard histological variables, the presence of peritubular eosinophils, and immunolabeling for lysozyme and myeloperoxidase in kidney tissue. Factor analysis was employed for data reduction and generation of a new case classification, with orthogonal rotation as a strategy to simplify factors, and principal component analysis was used as an extraction method. RESULTS: Seventy-nine kidney biopsies were obtained from 74 patients. The total population was divided into humoral rejection (39.2%), cellular rejection (34.1%), and mixed acute rejection (26.7%). No significant differences were found between the three groups in clinical and biochemical variables. We extracted 4 factors using factor analysis. The 1st factor was characterized by the presence of capillaritis, plasma cells infiltration, tubulitis, and inflammation. The 2nd factor included positivity for lysozyme and myeloperoxidase, while the 3rd factor included the presence of eosinophils and glomerulitis. The 4th component consisted of the presence of C4d and endarteritis. The cases belonging to the 3rd factor showed the greatest increase in serum creatinine. The cases belonging to the 4th factor exhibited greater urinary excretion of proteins. CONCLUSIONS: This proposal of classification of acute rejection could contribute to evaluate the prognosis of kidney transplant recipients.

The bridge between transplantation and regenerative medicine: Beginning a new Banff classification of tissue engineering pathology.

The science of regenerative medicine is arguably older than transplantation-the first major textbook was published in 1901-and a major regenerative medicine meeting took place in 1988, three years before the first Banff transplant pathology meeting. However, the subject of regenerative medicine/tissue engineering pathology has never received focused attention. Defining and classifying tissue engineering pathology is long overdue. In the next decades, the field of transplantation will enlarge at least tenfold, through a hybrid of tissue engineering combined with existing approaches to lessening the organ shortage. Gradually, transplantation pathologists will become tissue-(re-) engineering pathologists with enhanced skill sets to address concerns involving the use of bioengineered organs. We outline ways of categorizing abnormalities in tissue-engineered organs through traditional light microscopy or other modalities including biomarkers. We propose creating a new Banff classification of tissue engineering pathology to standardize and assess de novo bioengineered solid organs transplantable success in vivo. We recommend constructing a framework for a classification of tissue engineering pathology now with interdisciplinary consensus discussions to further develop and finalize the classification at future Banff Transplant Pathology meetings, in collaboration with the human cell atlas project. A possible nosology of pathologic abnormalities in tissue-engineered organs is suggested.

The Banff 2015 Kidney Meeting Report: Current Challenges in Rejection Classification and Prospects for Adopting Molecular Pathology.

The XIII Banff meeting, held in conjunction the Canadian Society of Transplantation in Vancouver, Canada, reviewed the clinical impact of updates of C4d-negative antibody-mediated rejection (ABMR) from the 2013 meeting, reports from active Banff Working Groups, the relationships of donor-specific antibody tests (anti-HLA and non-HLA) with transplant histopathology, and questions of molecular transplant diagnostics. The use of transcriptome gene sets, their resultant diagnostic classifiers, or common key genes to supplement the diagnosis and classification of rejection requires further consensus agreement and validation in biopsies. Newly introduced concepts include the i-IFTA score, comprising inflammation within areas of fibrosis and atrophy and acceptance of transplant arteriolopathy within the descriptions of chronic active T cell-mediated rejection (TCMR) or chronic ABMR. The pattern of mixed TCMR and ABMR was increasingly recognized. This report also includes improved definitions of TCMR and ABMR in pancreas transplants with specification of vascular lesions and prospects for defining a vascularized composite allograft rejection classification. The goal of the Banff process is ongoing integration of advances in histologic, serologic, and molecular diagnostic techniques to produce a consensus-based reporting system that offers precise composite scores, accurate routine diagnostics, and applicability to next-generation clinical trials.

Optimizing rejection readouts in a corneal allograft transplantation model.

PURPOSE: To evaluate the feasibility of anterior segment spectral domain optic coherence tomography (ASOCT) as rejection readout in a keratoplasty mouse model and to compare ASOCT against the current standard (i.e., a clinical score system). Furthermore, to compare both approaches with respect to intra- and inter-individual observer variability and to calculate a critical point that distinguishes between rejection and non-rejection in ASOCT analysis. METHODS: Allogeneic penetrating keratoplasties (PKs) were performed using C3H/He donor mice and BALB/c recipient mice; syngeneic transplantations served as controls using BALB/c donors and recipients. Corneal graft rejection was determined with a clinical score. ASOCT was used to determine the central thickness of the corneal grafts in the same animals. The rejection status was corroborated with histopathological examination. RESULTS: The median survival time (MST) of the corneal allografts in the wild-type BALB/c mice was 12 days. Allogeneic transplantation led to a 100% rejection rate, whereas signs of rejection after syngeneic transplantation appeared in up to 20% of the mice. Central corneal thickness (CCT) determination via customized software revealed a direct correlation with the clinical score. Receiver operating curve (ROC) analysis confirmed CCT as a valid surrogate for rejection. Calculation of the area under the curve (AUC) revealed a value of 0.88 with an optimal cut-off at 267 pixels. CONCLUSIONS: An increase in the CCT during acute allogeneic corneal graft rejection significantly correlated with the clinical surrogate parameter "corneal opacity." ASOCT not only generates source data, but also analysis of the ASOCT data shows lower readout variability and fewer interpreter variations than the clinical score commonly used to define the time point of graft rejection in mice.

Impact of the Banff 2013 classification on the diagnosis of suspicious versus conclusive late antibody-mediated rejection in allografts without acute dysfunction.

BACKGROUND: The Banff classification is used worldwide to characterize pathological findings in renal allograft biopsies. During the 11th Banff meeting, relevant changes were introduced in the diagnostic criteria for Category 2 antibody-mediated rejection (ABMR). Here, we assess the effect of these changes on the diagnosis of late chronic ABMR. METHODS: Seventy-three indication renal graft biopsies (chronic dysfunction, proteinuria and/or the presence of de novo donor-specific antibodies) from 68 kidney transplant recipients initially classified following the Banff 2009 criteria were reviewed and reclassified as per the new Banff 2013 criteria. RESULTS: The diagnostic category changed in 18% of the study biopsies with Banff 2013. The reclassification mainly involved Category 2 cases, from which 23.5% of the biopsies from older patients with worse graft function were overlooked by Banff 2009. ABMR was ruled out in 13% of cases under the Banff 2009 criteria. A significant number of the study samples were conclusively diagnosed as ABMR (40% as per Banff 2009 and 74% as per Banff 2013; P = 0.006), because of the inclusion of microvascular inflammation and the acceptance of some ultrastructural diagnostic criteria. However, when following the criteria of the new classification, samples with histological signs of chronic ABMR, in which human leucocyte antigen donor-specific antibodies are not detected or ultrastructural studies are not performed, may be inadequately characterized. CONCLUSIONS: The Banff 2013 classification helps in making a diagnosis of late ABMR, identifying cases, decreasing the percentage of suspected ABMR and making more conclusive diagnoses.

Identifying Subphenotypes of Antibody-Mediated Rejection in Kidney Transplants.

The key lesions in antibody-mediated kidney transplant rejection (ABMR) are microcirculation inflammation (peritubular capillaritis and/or glomerulitis lesions, abbreviated "pg") and glomerular double contours (cg lesions). We used these features to explore subphenotypes in 164 indication biopsies with ABMR-related diagnoses: 137 ABMR (109 pure and 28 mixed with T cell-mediated rejection [TCMR]) and 27 transplant glomerulopathy (TG), identified from prospective multicenter studies. The lesions indicated three ABMR subphenotypes: pgABMR, cgABMR, and pgcgABMR. Principal component analysis confirmed these subphenotypes and showed that TG can be reclassified as pgcgABMR (n = 17) or cgABMR (n = 10). ABMR-related biopsies included 45 pgABMR, 90 pgcgABMR, and 25 cgABMR, with four unclassifiable. Dominating all time intervals was the subphenotype pgcgABMR. The pgABMR subphenotype presented earliest (median <2 years), frequently mixed with TCMR, and was most associated with nonadherence. The cgABMR subphenotype presented late (median 9 years). Subphenotypes differed in their molecular changes, with pgABMR having the most histologic-molecular discrepancies (i.e. potential errors). Donor-specific antibody (DSA) was not identified in 29% of pgcgABMR and 46% of cgABMR, but failure rates and molecular findings were similar to cases where DSA was known to be positive. Thus, ABMR presents distinct subphenotypes, early pg-dominant, late cg-dominant, and combined pgcg phenotype, differing in time, molecular features, accompanying TCMR, HLA antibody, and probability of nonadherence.

The Revised (2013) Banff Classification for Antibody-Mediated Rejection of Renal Allografts: Update, Difficulties, and Future Considerations.

The Banff 2013 classification (Banff 2013) for antibody-mediated rejection (ABMR) in renal allografts represents the first major revision of the original Banff classification for ABMR that was published in 2003. The main impetus for this revision was the need to include C4d-negative ABMR, although this revised classification contains a number of additional features based on findings reported from 2007 to 2013. Since its publication, several studies have examined the validity of different aspects of Banff 2013 and compared it to earlier (2003, 2007) versions of the Banff ABMR classification. Recent evidence, albeit limited, indicates that Banff 2013 represents an improvement over the previous versions, enhancing our ability to accurately diagnose cases of acute/active and chronic active ABMR on renal allograft biopsy. Molecular studies appear to justify the threshold value of glomerulitis plus peritubular capillaritis score ≥2 required by Banff 2013 for the diagnosis of C4d-negative ABMR; however, other aspects of the classification, including its overall interobserver reproducibility, the clinical significance of the category of C4d staining without evidence of rejection, and whether surrogate markers might potentially substitute for the requirement for the presence of donor-specific antibodies, require additional investigation.

Accuracy of Inpatient International Classification of Diseases, Ninth Revision, Clinical Modification Coding for Cytomegalovirus After Kidney Transplantation.

BACKGROUND: International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) coding for cytomegalovirus (CMV) has been used as a proxy for active CMV infection or disease occurring in the inpatient setting in retrospective studies of kidney transplant recipients using large amounts of administrative data. However, the accuracy of inpatient CMV coding has not been determined. METHODS: We identified 393 kidney transplant recipients who were readmitted to Barnes-Jewish Hospital in St. Louis, Missouri from January 1, 2007 to December 31, 2011 to determine the accuracy of the ICD-9-CM diagnosis code for CMV (078.5) in identifying active CMV infection or disease (asymptomatic viremia, CMV syndrome, or tissue-invasive CMV disease) in the inpatient setting, using microbiological, histopathologic, or ophthalmologic evidence for CMV as the gold standard. RESULTS: The sensitivity and positive predictive value of CMV coding in identifying active CMV infection or disease were 0.77 and 0.71, respectively. The specificity and negative predictive value were both 0.98. The sensitivity of CMV coding in identifying CMV syndrome or tissue-invasive CMV disease was 0.93. CONCLUSIONS: CMV coding had good accuracy in identifying active CMV infection or disease among readmitted kidney transplant recipients in our hospital. Further validation studies of CMV coding in other hospitals are needed to obtain more generalizable estimates of the accuracy of CMV coding.

Heart transplantation: review.

Heart transplantation is currently the definitive gold standard surgical approach in the treatment of refractory heart failure. However, the shortage of donors limits the achievement of a greater number of heart transplants, in which the use of mechanical circulatory support devices is increasing. With well-established indications and contraindications, as well as diagnosis and treatment of rejection through defined protocols of immunosuppression, the outcomes of heart transplantation are very favorable. Among early complications that can impact survival are primary graft failure, right ventricular dysfunction, rejection, and infections, whereas late complications include cardiac allograft vasculopathy and neoplasms. Despite the difficulties for heart transplantation, in particular, the shortage of donors and high mortality while on the waiting list, in Brazil, there is a great potential for both increasing effective donors and using circulatory assist devices, which can positively impact the number and outcomes of heart transplants.

Banff 2013 update: Pearls and pitfalls in transplant renal pathology.

The pathological classification of rejection in renal allografts (Banff classification) has undergone substantial evolution for more than 20 years, and has been the diagnostic gold standard in clinical practice. The 2013 updated Banff classification encompasses a revised scheme of antibody-mediated rejection (ABMR) that consists of donor-specific antibody (DSA) positivity, characteristic histological manifestations for both acute and chronic ABMR, and DSA-induced endothelial cell injury which is represented by either C4d positivity, microvascular inflammation or expression of activated endothelial gene transcripts. Other modified criteria include a C4d positivity threshold, and histological definition of transplant glomerulitis and transplant glomerulopathy. Morphologically, glomerulonephritis, either recurrent or de novo, can be challenging to differentiate from ABMR-mediated transplant glomerulitis. Endothelial arteritis by itself does not warrant the diagnosis of acute T-cell mediated rejection; ABMR should also be considered based on the DSA test results. With regard to polyomavirus BK-associated nephropathy, immunohistochemical examination using anti-simian virus (SV) 40 antibody can be a promising method to assess the quantitative viral load of polyomavirus BK and graft survival. In summary, the 2013 updated Banff classification strictly defines ABMR with histopathological and serological criteria irrespective of C4d positivity. Inclusion of gene expression data relevant to ABMR highlights that the Banff criteria have entered the era of 'Seeing the Unseen' schemes, reflecting recent advances in understanding the molecular events in allograft injury.

The impact of C4d-negative acute antibody-mediated rejection on short-term prognosis among kidney transplant recipients.

AIM: We aimed to investigate the clinical and pathological features of C4d-negative acute antibody-mediated rejection (aAMR), and examined the impact of C4d-negative aAMR on short-term prognosis. METHODS: From 2005 to 2011, 626 kidney transplantations were performed in our institution and related hospitals. We excluded 174 ABO-incompatible transplantations, and analysed clinical and pathological data from the remaining 452 until December 2013. RESULTS: During the follow-up period, 39 patients underwent aAMR. We divided them into two groups. According to C4d positivity in each patient's first AMR, we divided the cohort into a C4d-positive aAMR group and a C4d-negative aAMR group, using the new Banff 2013 classification. We compared each aAMR patient's features to controls. Clinical and pathological characteristics were similar in both groups and the short-term outcomes of the two groups were similar, but both were worse than control. CONCLUSION: C4d-negative aAMR resembles C4d-positive aAMR in terms of clinical and pathological features, and that C4d positivity has no influence on short-term outcome.

Heart transplantation: review / Transplante cardíaco: revisão

ABSTRACT Heart transplantation is currently the definitive gold standard surgical approach in the treatment of refractory heart failure. However, the shortage of donors limits the achievement of a greater number of heart transplants, in which the use of mechanical circulatory support devices is increasing. With well-established indications and contraindications, as well as diagnosis and treatment of rejection through defined protocols of immunosuppression, the outcomes of heart transplantation are very favorable. Among early complications that can impact survival are primary graft failure, right ventricular dysfunction, rejection, and infections, whereas late complications include cardiac allograft vasculopathy and neoplasms. Despite the difficulties for heart transplantation, in particular, the shortage of donors and high mortality while on the waiting list, in Brazil, there is a great potential for both increasing effective donors and using circulatory assist devices, which can positively impact the number and outcomes of heart transplants.

RESUMO O transplante cardíaco é atualmente a abordagem cirúrgica definitiva padrão-ouro no tratamento da insuficiência cardíaca refratária. No entanto, a escassez de doadores limita a realização de um número maior de transplantes cardíacos, situação em que vem aumentando a utilização de dispositivos de assistência circulatória mecânica. Com indicações e contraindicações bem estabelecidas, além de diagnóstico e tratamento de rejeição, por meio de protocolos definidos de imunossupressão, os resultados do transplante cardíaco são muito favoráveis. Dentre as complicações precoces que podem impactar a sobrevida, destacamos a disfunção primária do enxerto, a disfunção do ventrículo direito, rejeição e infecções; já as complicações tardias incluem a doença vascular do enxerto e as neoplasias. Apesar das dificuldades para realização do transplante cardíaco, em especial pela escassez de doadores e pela elevada mortalidade em fila de espera, no Brasil, existe um grande potencial, tanto no aumento de doadores efetivos, quanto na utilização de dispositivos de assistência circulatória, o que pode vir a impactar positivamente no número e nos resultados do transplante cardíaco.

Bronchoscopic procedures and lung biopsies in pediatric lung transplant recipients.

Bronchoscopy remains a pivotal diagnostic and therapeutic intervention in pediatric patients undergoing lung transplantation (LTx). Whether performed as part of a surveillance protocol or if clinically indicated, fibre-optic bronchoscopy allows direct visualization of the transplanted allograft, and in particular, an assessment of the patency of the bronchial anastomosis (or tracheal anastomosis following heart-lung transplantation). Additionally, bronchoscopy facilitates differentiation of infective processes from rejection episodes through collection and subsequent assessment of bronchoalveolar lavage (BAL) and transbronchial biopsy (TBBx) samples. Indeed, the diagnostic criteria for the grading of acute cellular rejection is dependent upon the histopathological assessment of biopsy samples collected at the time of bronchoscopy. Typically, performed in an out-patient setting, bronchoscopy is generally a safe procedure, although complications related to hemorrhage and pneumothorax are occasionally seen. Airway complications, including stenosis, malacia, and dehiscence are diagnosed at bronchoscopy, and subsequent management including balloon dilatation, laser therapy and stent insertion can also be performed bronchoscopically. Finally, bronchoscopy has been and continues to be an important research tool allowing a better understanding of the immuno-biology of the lung allograft through the collection and analysis of collected BAL and TBBx samples. Whilst new investigational tools continue to evolve, the simple visualization and collection of samples within the lung allograft by bronchoscopy remains the gold standard in the evaluation of the lung allograft. This review describes the use and experience of bronchoscopy following lung transplantation in the pediatric setting.

Current views on chronic rejection after lung transplantation.

Chronic lung allograft dysfunction (CLAD) was recently introduced as an overarching term mainly to classify patients with chronic rejection after lung transplantation, although other conditions may also qualify for CLAD. Initially, only the development of a persistent and obstructive pulmonary function defect, clinically identified as bronchiolitis obliterans syndrome (BOS), was considered as chronic rejection, if no other cause could be identified. It became clear in recent years that some patients do not qualify for this definition, although they developed a chronic and persistent decrease in FEV1 , without another identifiable cause. As the pulmonary function decline in these patients was rather restrictive, this was called restrictive allograft syndrome (RAS). In the present review, we will further elaborate on these two CLAD phenotypes, with specific attention to the diagnostic criteria, the role of pathology and imaging, the risk factors, outcome, and the possible treatment options.