Feasibility of a 7-min rapid CMR protocol for cine, perfusion, and scar using3D whole-ventricle dynamic ferumoxytol-enhanced MR Multitasking

Presenting Author(s)

• 

  Anthony G. Christodoulou

  Associate Professor
  University of California, Los Angeles (UCLA)
  Los Angeles, California, United States

Primary Author(s)

• XC

  Xi Chen, PhD

  Post doc
  David Geffen School of Medicine at UCLA
  Los Angeles, California, United States

Co-Author(s)

• MJ

  Mana Jameie

  Post doc
  UCLA, United States
• 

  Kim-Lien Nguyen, MD

  Associate Professor of Cardiovascular Medicine and Radiology
  David Geffen School of Medicine at UCLA and VA Greater Los Angeles Healthcare System
  Los Angeles, California, United States

Background: Long exam times are a major issue that limits clinical practice of cardiovascular magnetic resonance (CMR). Rapid protocols are of interest, and a recent white paper suggests perfusion, cine and LGE imaging can answer most common questions in clinical cardiovascular practice [1]. Ferumoxytol is a novel off-label iron-based intravascular contrast agent with long blood half-life (10–14h). Steady-state ferumoxytol-enhanced imaging has been used to acquire cine and measure fractional myocardial blood volume [2,3]; its role in scar imaging has not been thoroughly investigated. This work performs a feasibility study of an abbreviated 7-min protocol for cardiac phase-resolved dynamic ferumoxytol-enhanced imaging using MR multitasking [4], which has the potential to provide information related to cine, perfusion and scar, with 3D whole-ventricle coverage during free breathing.

Methods:
We performed in vivo experiments at 3T at rest on n=15 subjects (3 healthy volunteers, 12 patients). Participants received either a 3-min scan as an add-on to other protocols, during which 2mg/kg ferumoxytol was slowly infused, or a dedicated protocol with a 7-min scan and 4mg/kg ferumoxytol. No breath-holding or ECG was required. 3D stack-of-stars radial sampling with inversion recovery (IR) preparations was used. Spatial resolution=1.7×1.7×8 mm³, number of slices=12–24. We compared left ventricle ejection fraction (LVEF) from standard bSSFP cine and multitasking derived cine in 11 subjects (2 subjects excluded for the lack of standard cine, 2 for poor reconstruction). We performed semi-quantitative upslope analysis of the normalized signal intensity-time dynamics in the LAD/LCX/RCA coronary territories of 4 subjects (of the 3 patients with ischemia and without scar, one was excluded for poor reconstruction; 2 healthy subjects), where PET was used as the reference for ischemia. We evaluated scar tissue correspondence on multitasking derived T1w images and T1 maps using magnitude LGE images as reference.





Results:
Cine: Fig. 1 shows the Bland-Altman plot of the LVEF. Mean bias was –2.5%. Perfusion: There was a significant difference in the upslope of signal intensity dynamics between ischemic and non-ischemic territories (P < 0.001, Fig. 2a). Fig. 2b shows examples of an ischemic patient. Scar: Fig. 3 shows the reference LGE images, postcontrast multitasking T1w images and T1 maps of a patient with scar. The hypo-intensity in T1w images and abnormally high T1 correlate with scars in LGE (red arrows).





Conclusion:
This study demonstrates the feasibility of a 7-min protocol for 3D whole-ventricle dynamic ferumoxytol-enhanced imaging. It can potentially provide information related to cine, perfusion and scar. Stronger evidence has been shown for cine/perfusion; myocardial scar is still investigational and requires further analysis. The multitasking T1map does not have good correlation with the LGE images and shows a larger area of scar tissue, which needs to be determined in follow-up studies.




Fig.1. Bland-Altman plot of the LVEF calculated on standard bSSFP cine and multitasking cine in 11 subjects.
Fig.2. a) Upslope values in 4 subjects. b) The normalized signal intensity-time dynamics and images at two example time points of subject 3. Both RCA and LCX coronary territories are ischemic based on PET reference.
Fig.3. LGE, postcontrast multitasking T1w images and T1 maps of a patient with myocardial scar.