Despite improvements in early post-transplant survival, the rate of progression to allograft loss after pediatric HT has not changed since the 1980s.1 Most late deaths are attributed to “graft failure” and “chronic rejection,”1 and on histologic analysis, many of these failed allografts are characterized by coronary vasculopathy and fibrosis.2 CMR-derived ECV fraction is a validated noninvasive measure of DMF in the absence of edema or infiltrative disease.3 ECV quantifies the full spectrum of myocardial fibrosis and is highly correlated with histologic measures of the CVF (ie, diffuse fibrosis) of the LV in adults.4-8 ECV is also highly reproducible6,9-13 and may predict heart failure outcomes and mortality in adults.14,15 While ECV is a robust metric of DMF, very little has been reported about whether ECV can detect DMF and DMF-associated allograft dysfunction in HT recipients. In this exploratory study, our aims were to validate the association of ECV with histologic myocardial fibrosis in a cohort of prospectively recruited pediatric HT recipients and to explore associations of ECV with clinical indicators of allograft function. We hypothesized that ECV is correlated with histologic myocardial fibrosis, hemodynamics, and serum BNP level.

Fibrosis is commonly described in heart allografts lost late after transplantation. CMRderived ECV is a validated measure of DMF in native adult hearts that may predict heart failure and mortality. We explored associations of ECV with histologic myocardial fibrosis and clinical features after pediatric heart transplantation. Twenty-five recipients (7.0±6.3 years at transplant and 10.7±6.5 years post-transplant) were prospectively recruited for CMR and BNP measurement at the time of surveillance biopsy. All had normal ejection fractions and lacked heart failure symptoms. Fibrosis was quantified on biopsy after picrosirius red staining as CVF. ECV was quantified using contemporaneous hematocrit on basal and mid-short-axis slices. ECV was moderately correlated with CVF (r=.47; P=.019). We found no associations of ECV with hemodynamics, ischemic time, time since transplantation, or number of prior biopsies or acute rejections. Compared to healthy non-transplant controls, there was no significant difference in ECV (25.1±3.0 vs 23.7±2.0%, P=.09). Log-transformed BNP was correlated with ECV (recipients: r=.46, P=.02; recipients and controls: r=.45, P=.006). These findings suggest ECV quantifies DMF and relates to biological indicators of cardiac function after pediatric heart transplantation.

After approval from the University of Pittsburgh Institutional Review Board (PRO13020024), we prospectively enrolled consecutive pediatric HT recipients who were ≥13 years of age for ECV quantification at the time of clinically indicated, surveillance EMB. Because factors besides the accumulation of excess collagen can expand the ECM,3 recipients with clinically significant acute rejection (≥grade 2R/ pAMR1)16,17 on concurrent EMB were excluded from analysis. We also did not recruit recipients who were <9 months post-HT to reduce the possibility of confounding from ischemia-reperfusion injury that occurs with the HT procedure. Patients who were not able to undergo a complete, contrast-enhanced CMR scan without sedation, including those with glomerular filtration rate ≤30 mL/min/1.73 m2 or for other reasons (eg, retained pacemaker leads), were not approached for consent. Because normal ECV ranges have not been established across various CMR platforms, we also recruited 12 healthy, non-HT individuals between the ages of 18 and 30 years to serve a comparison group. Individuals with known medical issues, regular medication use, or history of smoking were not enrolled. IRB stipulations did not allow us to enroll individuals aged <18 years to serve as controls for this protocol.