Abstract 4363330: 4D Cardiac Magnetic Resonance Identifies Differences in Regional Strain Patterns Among Pediatric Heart Transplant Patients with Acute Rejection or Cardiac Allograft Vasculopathy

Abstract

Background: Acute rejection (AR) and cardiac allograft vasculopathy (CAV) are leading causes of morbidity and mortality after pediatric heart transplant (PHT). Cardiac catheterization with endomyocardial biopsy (EMB) is the gold standard for diagnosing AR and CAV but it is invasive and costly. Cardiac magnetic resonance (CMR) has potential as a non-invasive screening modality, but standard software requires selection of one slice for base, mid, and apex, resulting in a significant loss of data. Hypothesis: We hypothesized that a recently developed regional 4D (3D+time) CMR strain analysis method would identify regional strain differences between healthy PHT patients and those with AR or CAV. Methods: PHT patients undergoing cardiac catheterization with EMB and concomitant CMR were prospectively enrolled. For each patient, we used a custom graphical user interface in MATLAB to compile short-axis CMR image stacks into four-dimensional sequences. Using feature tracking, we generated a deformable 3D mesh throughout the cardiac cycle, which allowed us to calculate regional surface area and transmural peak strain, systolic strain rate and diastolic strain rate. Continuous variables between groups (Healthy PHT, AR, or CAV) were compared with a Kruskal-Wallis test followed by a Wilcoxon rank-sum test for pairwise comparisons. Results: Thirty-eight patients were included (median age 17 years [IQR 15-19]), nine (24%) with AR and eight (21%) with CAV. AR patients had worse surface area peak strain and systolic strain rate compared to the healthy PHT group at multiple basal and mid-LV segments (Table 1); no apical segments reached significance. Surface area peak strain and late diastolic strain rate were worse in AR compared to CAV, though the CAV group was overall similar to healthy PHT. AR patients also had worse transmural peak strain and systolic strain rate in several basal and mid-LV segments compared to healthy PHT and CAV groups (Table 2). Conclusion: Patients with AR have worse surface area peak strain and systolic strain rate compared to healthy PHT patients, highlighting the utility of CMR as a non-invasive screening modality for AR. Further evaluation of strain patterns that distinguish CAV, AR, and healthy PHT patients is necessary.