Noninvasive quantification of left-to-right shunt in pediatric patients: phase-contrast cine magnetic resonance imaging compared with invasive oximetry

Circulation. 2001 May 22;103(20):2476-82. doi: 10.1161/01.cir.103.20.2476.

Abstract

Background: Blood flow can be quantified noninvasively by phase-contrast cine MRI (PC-MRI) in adults. Little is known about the feasibility of the method in children with congenital heart disease.

Methods and results: In 50 children (mean age 6.2 years, range 1.1 to 17.7 years) with an atrial- or ventricular-level shunt, blood flow rate in the great vessels was determined by PC-MRI, and the ratio of pulmonary to aortic flow (Qp/Qs) was compared with Qp/Qs by oximetry. We found a difference of 2% and a range of -20% to +26% (limits of agreement, mean +/-2 SD). In another 7 children with congenital heart disease but no cardiac shunting (mean age 7.9 years, range 1.3 to 13.5 years), Qp/Qs by PC-MRI was 1.02 (SD +/-0.06). No difference between systemic venous and aortic flow volumes was found (range -17% to +20%, n=37). Blood flow through a secundum atrial septal defect as assessed by PC-MRI (n=24) overestimated the shunt compared with the difference between pulmonary and aortic flows. The mean difference between 3 repeated PC-MRI measurements in each location was 5.3% (SD +/-4.0%, n=522), demonstrating good precision. The interobserver variability was low. The accuracy of PC-MRI was confirmed by in vitro experiments.

Conclusions: Determination of Qp/Qs by PC-MRI in children is quick, safe, and reliable compared with oximetry. Systemic venous flow can be quantified by PC-MRI, whereas through-plane shunt measurement within an atrial septal defect is inaccurate.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Child
  • Child, Preschool
  • Coronary Circulation*
  • Female
  • Heart Defects, Congenital / pathology
  • Heart Defects, Congenital / physiopathology*
  • Humans
  • Magnetic Resonance Imaging, Cine
  • Male
  • Oximetry
  • Reproducibility of Results
  • Sensitivity and Specificity