1356898 - Trends on serial CMR in pediatric heart transplantation based on risk of chronic graft failure

Chronic graft failure (CGF) in pediatric heart transplant patients (PHT) is multifactorial and includes history of significant rejection and cardiac allograft vasculopathy (CAV). Multiparametric CMR may be a useful tool to identify early myocardial changes related to CGF. We sought to evaluate differential changes on serial CMR in cohorts of PHT with and without clinical risk factors for CGF.

Methods:

Retrospective evaluation of PHT undergoing clinically indicated comprehensive structure-function CMRs with stress perfusion (using regadenoson) was performed. Patients with at least 2 CMRs more than 0.9   years apart were included. PHT were defined as high-risk if there was a concerning transplant history including CAV and history of moderate-severe (M-S) rejection prior to 1st CMR or new significant clinical event between CMR2 and CMR1 (i.e., new M-S rejection, new CAV or CAV progression). In the absence of the above, PHT were classified as low risk. Changes in CMR variables (including biventricular volumetry, EF, myocardial T1/T2/ECV, presence of LGE and myocardial perfusion reserve index) were assessed between the two groups. We compared parameter changes from CMR1 to CMR2 between the two groups. Separately, in the low-risk group, we also assessed whether significant changes in parameters occurred between the two exams. 

Results:

Sixty-six CMRs (33 PHT) were evaluated. Mean age at CMR1 was 13.1 (SD 4.3) years with a mean time difference of 2.8 (SD 1.6) years between CMR1 and CMR2. Table 1 demonstrates the demographic and clinical characteristics of the study cohort and the age at transplant/gender matched low-risk (n=18) and high-risk (n=15) groups. Across repeat exams in the high- risk group, global ECV (low risk group: 28% CMR1 to 27.2% CMR2; high risk group 29.9% on CMR1 to 31.4% CMR2; p=0.03) increased significantly and there were signs of LV/RV remodeling (increased LVESVi p=0.05, decreased RVEDVi p=0.07 and decreased RVESVi p=0.06) with no significant decline in EF in the high-risk group (Table 2). At a mean interval of 3.1 years (SD 1.5), the low-risk group showed a decline in T1 (1050 to 1017 ms; p =0.02) with no significant worsening of CMR parameters (Table 3); 39% of CMR2 were performed in this group to follow up abnormalities on CMR1. The most common abnormality in the low-risk group was focal LGE which remained stable during this period.

Conclusion:

ECV increases over time in PHT with risk factors for CGF; long term clinical outcomes related to this finding need to be further assessed. In addition, PHT with low-risk for CGF show stable CMR findings over a 3.1-year interval; LGE is a common stable finding in this group. Further study of longitudinal data may help determine frequency of CMR surveillance in low- risk PHT.