Vascular endothelial growth factor A is associated with the subsequent development of moderate or severe cardiac allograft vasculopathy in pediatric heart transplant recipients

Abstract

Background: Cardiac allograft vasculopathy (CAV) is the leading cause of chronic allograft loss after pediatric heart transplantation. We hypothesized that biomarkers of endothelial injury and repair would predict CAV development in pediatric heart transplant recipients.

Methods: Blood was collected from pediatric heart transplant recipients at the time of routine annual coronary angiography, and the concentrations of 13 angiogenesis-related molecules were determined. The primary end point was the presence of moderate or severe CAV by angiography during a 5-year follow-up period.

Results: The study enrolled 48 recipients (57% male) with a median age of 15.5 years (range, 2-22 years) and median time post-transplant of 5.8 years (range, 2-15 years). Eight recipients developed moderate/severe CAV at a median follow-up of 4.7 years, of whom 3 died, 3 underwent retransplantation, 1 had a myocardial infarction, and 1 was listed for retransplantation. Clinical characteristics associated with the development of moderate/severe CAV included prednisone use at enrollment (p = 0.03) and positive recipient cytomegalovirus immunoglobulin G at the time of transplant (p = < 0.01). Multivariable Cox proportional hazards regression identified plasma vascular endothelial growth factor (VEGF)-A concentration greater than 90 pg/ml at the time of blood draw as a significant predictor of time to moderate or severe CAV (hazard ratio, 14.3; 95% confidence interval, 1.3-163). Receiver operating characteristic curve analysis demonstrated that VEGF-A shows moderate performance for association with the subsequent development of CAV (area under the curve, 0.77; 95% confidence interval, 0.61-0.92).

Conclusions: VEGF-A levels in pediatric heart transplant recipients are associated with clinically important CAV progression within the subsequent 5 years.

Keywords: VEGF; allograft vasculopathy; biomarkers; coronary artery disease; heart transplantation.

Figure 1. Cross sectional analysis among proteins involved in vascular injury and repair and mild CAV at the time of blood sampling

A total of 13 proteins were assayed in plasma by multianalyte profiling and/or by ELISA. Data for all proteins are illustrated in Table 2, and representative box-and-whisker plots of selected proteins are displayed below. By univariate analysis, angiogenic proteins were not significantly elevated (P=NS) in association with mild CAV (CAV₁) at the time of blood sampling.

Figure 2. Plasma VEGF-A concentrations are associated with the development of moderate/severe CAV

A. Box-and-whisker plot of VEGF-A levels at the time of enrollment and the subsequent development of no/mild CAV (N=36) or moderate/severe CAV (N=8; P=0.019, Mann-Whitney U-test). B. Kaplan Meier plot of freedom from moderate or severe CAV (ISHLT CAV₂ or CAV₃) stratified by the median VEGF-A concentration of 90 pg/mL. There were 22 patients in each strata (total 44) and patients with VEGF-A concentrations above 90 pg/mL had a shorter time to development of moderate or severe CAV (P=0.02). In Panels A and B, the median follow up time was 4.7 years. C. Receiver-operating characteristic (ROC) curve analysis for VEGF-A levels and the subsequent development of moderate or severe CAV.