Rapid Fire Abstracts
Richard M. Friesen, MD
Assistant Professor
Children's Hospital Colorado
Richard M. Friesen, MD
Assistant Professor
Children's Hospital Colorado
Lorna P. Browne, MD
Professor
University of Colorado and Children's Hospital Colorado
Erin Englund, PhD
Assistant Research Professor
University of Colorado Anschutz Medical Campus
Emily M. Bucholz, MD, PhD, MPH
Assistant Professor of Pediatrics
Children's Hospital Colorado
Alex J. Barker, PhD
Associate Professor
Children's Hospital Colorado, University of Colorado Anschutz Medical Campus
Cardiac pathophysiology plays an important role in the management of patients with CDH. However, while postnatal echocardiographic measurements of ventricular size and function are associated with outcomes in CDH, prenatal predictors of outcomes and severity are limited. 1 Fetal cardiac MRI may provide cardiopulmonary imaging biomarkers to better predict postnatal adverse outcomes and severity.
Methods:
With IRB approval, 25 prospectively recruited healthy volunteers (33.9 ± 2.1 weeks) and 19 pregnant patients (34.0 ± 1.8 weeks gestation) with prenatal diagnosis of CDH (5 right, 14 left) underwent clinically indicated fetal MRI with IRB approved Doppler ultrasound (DUS Smartsync, Northh Medical) gated cardiac sequences at 1.5T and 3T (Philips Ingenia). Mitral and tricuspid annuli, right (RV) and left (LV) ventricular dimensions were measured from axial and short axis bSSFP sequences. RV and LV longitudinal strain (GLS) were measured by cvi42 (Circle Cardiovascular Imaging, Alberta CA) (Figure 1). Measurements were compared between patients with CDH and healthy volunteers using student t-tests. Measurements were then compared to pre- and post-natal metrics of CDH severity and composite adverse events which included need for ECMO or death (fetal or postnatal) or use of milrinone postnatally.
Results:
19 CDH patients had successful DUS-gated CMR acquisitions. Comparative results are shown in Table 1. Average LV GLS in fetuses with CDH was significantly lower than controls. RV to LV short axis diameter ratio was not significantly different in CDH vs controls, and biventricular shortening fractions were not significantly different between CDH and controls. Eight patients required milrinone postnatally, and the LV GLS was significantly lower in this cohort compared to controls and CDH patients without milrinone (-12.4 ± 2.4% vs -17.5 ± 3.6%, p< 0.0001). Eleven patients had an adverse event (Fetal demise = 1, ECMO = 7, Postnatal death = 3) and were evaluated against controls and CDH patients without adverse events for predictive analysis. RV:LV ratio (AUC = 0.61), LV shortening fraction (AUC = 0.72) and LV GLS (AUC = 0.86) demonstrated the highest AUC coefficients predictive of adverse outcomes (Figure 2).
Conclusion:
This study provides evidence that fetal CMR-derived markers of ventricular function are altered prenatally in patients with CDH despite preserved traditional measures of ventricular function. In particular, reduced left ventricular longitudinal strain correlates with the need for milrinone postnatally as well as adverse outcomes of ECMO or death. Comprehensive prenatal cardiopulmonary evaluation with fetal CMR may provide prognostic biomarkers to improve prenatal counseling and postnatal management.
Figure 1. Fetal CMR measurements in a healthy control (left) and fetus with CDH (right). a) LV diastolic measurements for global longitudinal strain, atrioventricular valve annular length, and biventricular length b) Short axis ventricular dimensions for determination of biventricular dimensions, shortening fraction, and RV:LV ratio. 
Figure 2. Comparison of CDH patients with adverse events (ECMO = black triangles; Death = red triangles) vs healthy controls (black circles) and CDH patients without adverse events (green triangles) by RV:LV ratio, LV shortening fraction (LV SF) and LV longitudinal strain (LV GLS). Receiver operating characteristic (ROC) curves show the area under the curve (AUC) for each comparison.
