Personalized Risk Prediction and Mechanistic Profiling of Cardiac Allograft Vasculopathy

University Of Pennsylvania, Philadelphia PA

Investigators

Abstract

1 PROJECT SUMMARY: 2 Cardiac Allograft Vasculopathy (CAV) is the leading cause of transplanted heart failure and loss. Despite being 3 a major cause of morbidity for heart transplant (HT) recipients, little progress has been made in tailoring CAV 4 surveillance and mitigation strategies to individual patient risk. All guideline-endorsed methods for CAV detection 5 rely on repurposing techniques initially developed for evaluating native coronary artery disease. As such, these 6 techniques focus on estimating the flow of blood through the large, epicardial coronary arteries. Despite the 7 large-vessel focus of established CAV screening, the pathobiology of CAV also involves significant microvascular 8 inflammation, with distinct histologic changes which precede overt, large-vessel manifestations. Currently, there 9 are no objective means for measuring these microscopic changes, forcing clinicians to rely on tests which only 10 detect CAV after it has progressed to macroscopic stages. We hypothesize that the application of digital tissue 11 analysis methods in CAV will lead to both a clinically viable risk-assessment tool, and to the discovery of novel 12 mechanistic biology. Human tissue samples contain a wealth of information which is underutilized in conventional 13 clinical and scientific workflows. However, advanced digital technologies capable of extracting and quantifying 14 the spatial information contained within residual tissues are poised to change this. Computer-aided image anal- 15 ysis of digital pathology (DP) slides can extract novel ‘morphologic biomarkers’, quantifying the sizes, patterns, 16 and spatial relationships which comprise the cardiac microarchitecture. Combining DP morphologic analysis with 17 new, spatial-molecular profiling techniques adds additional insights, enabling deep, mechanistic interrogations. 18 Our team is a leader in cardiac digital pathology (DP) analysis, having developed numerous first-in-heart pipe- 19 lines for generating clinically important diagnoses and risk-assessments. We are also leaders in spatial-molec- 20 ular profiling of human cardiac tissues, with important contributions in both HT and non-HT diseases. As de- 21 scribed in a recent publication in Circulation, we have developed the ‘integrated-CAV Prediction’ (iCAV-Pr) sys- 22 tem. iCAV-Pr uses clinical data and residual tissue from routine post-HT biopsies to predict which patients will 23 go on to develop CAV years before overt disease onset. To fill unmet needs in heart transplant care, we propose 24 a multicenter study aimed at validating iCAV-Pr predictive performance, at establishing it’s potential role in clin- 25 ical practice, and at defining the biology that underlies it’s predictions. In Aim 1, we will enhance the iCAV-Pr via 26 extraction of new morphologic features, followed by performance validation in a diverse, multinational, retrospec- 27 tive cohort. In Aim 2, we will assess the efficacy of iCAV-Pr vs. intravascular ultrasound, the most sensitive 28 existing method for detecting early changes in CAV. And, in Aim 3, we will perform high-plex digital spatial 29 proteomic profiling on serially collected EMB pathology slides to deeply explore the mechanisms underlying CAV 30 development. Successful completion of these aims could open a new frontier in personalized post-transplant 31 care, enabling providers to tailor both CAV screening and treatment strategies to individual patient risk.