An integrated molecular diagnostic report for heart transplant biopsies using an ensemble of diagnostic algorithms.

BACKGROUND: We previously reported a microarray-based diagnostic system for heart transplant endomyocardial biopsies (EMBs), using either 3-archetype (3AA) or 4-archetype (4AA) unsupervised algorithms to estimate rejection. In the present study we examined the stability of machine-learning algorithms in new biopsies, compared 3AA vs 4AA algorithms, assessed supervised binary classifiers trained on histologic or molecular diagnoses, created a report combining many scores into an ensemble of estimates, and examined possible automated sign-outs. METHODS: We studied 889 EMBs from 454 transplant recipients at 8 centers: the initial cohort (N = 331) and a new cohort (N = 558). Published 3AA algorithms derived in Cohort 331 were tested in Cohort 558, the 3AA and 4AA models were compared, and supervised binary classifiers were created. RESULTS: A`lgorithms derived in Cohort 331 performed similarly in new biopsies despite differences in case mix. In the combined cohort, the 4AA model, including a parenchymal injury score, retained correlations with histologic rejection and DSA similar to the 3AA model. Supervised molecular classifiers predicted molecular rejection (areas under the curve [AUCs] >0.87) better than histologic rejection (AUCs <0.78), even when trained on histology diagnoses. A report incorporating many AA and binary classifier scores interpreted by 1 expert showed highly significant agreement with histology (p < 0.001), but with many discrepancies, as expected from the known noise in histology. An automated random forest score closely predicted expert diagnoses, confirming potential for automated signouts. CONCLUSIONS: Molecular algorithms are stable in new populations and can be assembled into an ensemble that combines many supervised and unsupervised estimates of the molecular disease states.

Exploring the cardiac response to injury in heart transplant biopsies.

BACKGROUND: Because injury is universal in organ transplantation, heart transplant endomyocardial biopsies present an opportunity to explore response to injury in heart parenchyma. Histology has limited ability to assess injury, potentially confusing it with rejection, whereas molecular changes have potential to distinguish injury from rejection. Building on previous studies of transcripts associated with T cell-mediated rejection (TCMR) and antibody-mediated rejection (ABMR), we explored transcripts reflecting injury. METHODS: Microarray data from 889 prospectively collected endomyocardial biopsies from 454 transplant recipients at 14 centers were subjected to unsupervised principal component analysis and archetypal analysis to detect variation not explained by rejection. The resulting principal component and archetype scores were then examined for their transcript, transcript set, and pathway associations and compared to the histology diagnoses and left ventricular function. RESULTS: Rejection was reflected by principal components PC1 and PC2, and by archetype scores S2TCMR, and S3ABMR, with S1normal indicating normalness. PC3 and a new archetype score, S4injury, identified unexplained variation correlating with expression of transcripts inducible in injury models, many expressed in macrophages and associated with inflammation in pathway analysis. S4injury scores were high in recent transplants, reflecting donation-implantation injury, and both S4injury and S2TCMR were associated with reduced left ventricular ejection fraction. CONCLUSION: Assessment of injury is necessary for accurate estimates of rejection and for understanding heart transplant phenotypes. Biopsies with molecular injury but no molecular rejection were often misdiagnosed rejection by histology.TRAIL REGISTRATION. ClinicalTrials.gov NCT02670408FUNDING. Roche Organ Transplant Research Foundation, the University of Alberta Hospital Foundation, and Alberta Health Services.

Impact of Rabbit Antithymocyte Globulin Dose on Long-term Outcomes in Heart Transplant Patients.

BACKGROUND: Optimal dosing strategies have not been established for rabbit antithymocyte globulin (rATG) after heart transplantation, and there is currently wide variability in rATG regimens with respect to both dose and duration. METHODS: In a retrospective, single-center analysis, 523 patients undergoing heart transplantation during 1996 to 2009 were stratified by cumulative rATG dose: less than 4.5 mg/kg (group A), 4.5 to 7.5 mg/kg (group B) or greater than 7.5 mg/kg (group C). RESULTS: Survival at 1 year after transplantation was 80% in group A, 90% in group B, and 88% in group C (P = 0.062). Incidence of acute rejection per 1000 patient-years was significantly higher in group A (hazards ratio [HR], 54.8; 95% confidence interval [95% CI], 33.9-83.8) compared to groups B (19.6; 95% CI, 11.4-31.4) and C (23.6; 95% CI, 17.5-31.3). Incidence of severe infection 10 years after transplantation was higher in group C (45%) than groups A (37%) or B (23%) (P < 0.001); cytomegalovirus infection rates were 35%, 20% and 23%, respectively (P = 0.009). Multivariable Cox regression showed an HR of 0.51 (95% CI, 0.25-1.02) for acute rejection with group B versus group A, and 0.54 (95% CI, 0.33-0.88; P = 0.013) for severe infection. The rate of malignancy per 1000 patient-years was higher in groups B (13.85) and C (14.95) than group A (7.83). CONCLUSIONS: These retrospective data suggest that a cumulative rATG dose of 4.5 to 7.5 mg/kg may offer a better risk-benefit ratio than lower or higher doses, with acceptable rates of infection and posttransplant malignancy. Prospective trials are needed.

Detection of alloantibody-mediated complement activation: A diagnostic advance in monitoring kidney transplant rejection?

OBJECTIVE: Antibody-mediated rejection (ABMR) is an important cause of kidney allograft injury. In the last two decades, detection of complement split product C4d along transplant capillaries, a footprint of antibody-mediated classical complement activation, has evolved as a useful diagnostic marker of ABMR. While it was recognized that ABMR may occur also in the absence of C4d, numerous studies have shown that C4d deposition may indicate a more severe rejection phenotype associated with poor graft survival. Such studies suggest a possible diagnostic benefit of ex vivo monitoring the complement-activating capability of circulating alloantibodies. DESIGN AND METHODS: We reviewed the literature between 1993 and 2015, focusing on in vivo (biopsy work-up) and in vitro detection (modified bead array technology) of HLA antibody-triggered classical complement activation in kidney transplantation. RESULTS: Precise HLA antibody detection methods, in particular Luminex-based single antigen bead (SAB) assays, have provided a valuable basis for the design of techniques for in vitro detection of HLA antibody-triggered complement activation reflected by C1q, C4 or C3 split product deposition to the bead surface. Establishing such assays it was recognized that deposition of complement products to SAB, which critically depends on antibody binding strength, may be a cardinal trigger of the prozone effect, a troublesome in vitro artifact caused by a steric interference with IgG detection reagents. False-low IgG results, especially on SAB with extensive antibody binding, have to be considered when interpreting studies analyzing the diagnostic value of complement in relation to standard IgG detection. Levels of complement-fixing donor-specific antibodies (DSA) were shown to correlate with the results of standard crossmatch tests, suggesting potential application for crossmatch prediction. Moreover, while the utility of pre-transplant complement detection, at least in crossmatch-negative transplant recipients, is controversially discussed, a series of studies have shown that the appearance of post-transplant complement-fixing DSA may be associated with C4d deposition in transplant capillaries and a particular risk of graft failure. CONCLUSIONS: The independent value of modified single antigen bead assays, as compared to a careful analysis of standard IgG detection, which may be affected considerably by complement dependent artifacts, needs to be clarified. Whether they have the potential to improve the predictive accuracy of our current diagnostic repertoire warrants further study.

Calcineurin-inhibitor minimization protocols in heart transplantation.

Cardiac transplantation has become an established method for end-stage heart disease. A calcineurin-inhibitor (CNI)-based regimen is the cornerstone of immunosuppressive therapy after cardiac transplantation. CNIs have reduced acute rejection and infection and markedly increased survival of cardiac transplantation patients. However, the dose- and time-dependent nephrotoxic effects of CNIs can limit long-term survival, and chronic renal failure is a major cause of morbidity and mortality in long-term cardiac transplant patients. Early experience on withdrawal of CNIs (and maintenance of patients on azathioprine and steroids) in patients, who developed chronic renal dysfunction, resulted in rejection episodes with, sometimes, fatal outcome. The introduction of newer immunosuppressive drugs, like thymoglobulin, anti CD-25 monoclonal antibodies, mycophenolate mofetil, everolimus or sirolimus into clinical practice, has given transplant physicians new tools to adapt immunosuppression to patients' needs. Changes of immunosuppressive protocols by using new drugs early and late after transplantation and simultaneous reduction or weaning of CNIs have become attractive options. The aim of this article is to review strategies to delay, reduce or prevent CNIs after cardiac transplantation as means to improve short- and long-term outcome mainly by protecting renal function.

Safety and efficacy of statin therapy in patients switched from cyclosporine a to sirolimus after cardiac transplantation.

INTRODUCTION: Statins are an established therapy after cardiac transplantation. Sirolimus (Srl) has been used successfully in cardiac transplant patients. However, potential side effects are hyperlipidemia and interactions with statins. The aim of the study was to evaluate the safety and efficacy of statin therapy after switch to a Srl-based immunosuppression. PATIENTS AND METHODS: Ninety-eight long-term patients were switched from Cyclosporine A to Srl. Also all patients received mycophenolate mofetil alone or mycophenolate mofetil plus steroid therapy. Reasons for switch were renal dysfunction, graftvasculopathy, or skin cancer. Patients were switched 7.8+/-4.7 years after transplant. Total observation period was 12 months before and after switch, respectively. Safety evaluation consisted of regular measurements of CPK and liver enzymes to evaluate the incidence myopathy and hepatoxicity. Efficacy analysis was performed by serial blood lipid assessments (low-density lipoprotein, high-density lipoprotein, total cholesterol, and triglycerides). RESULTS: Forty-three percentage of patients received atorvastatin, 38% pravastatin, and 18% other drugs or therapy changes. Most lipid blood levels increased significantly after switch (cholesterol: 192.9+/-38.6 mg/dL vs. 221.8+/-49.2 mg/dL, P<0.0001; low-density lipoprotein: 108.0+/-35.6 mg/dL vs. 123.8+/-37.9 mg/dL, P<0.0001; and triglycerides: 178.3+/-88.2 mg/dL vs. 225.5+/-139.1 mg/dL, P<0.0001). Blood lipid levels after switch were not associated with statin type. Overall safety was acceptable, although incidence of myopathy doubled after switch (n=20 vs. 40; P<0.01). However, most cases were asymptomatic CPK elevations in the pravastatin group. Hepatotoxicity rate was 4% and only temporary. CONCLUSION: Statin therapy after switch from cyclosporine A to Srl in long-term cardiac transplant patients is safe. However, regular testing of blood lipids and CPK should be mandatory.