Focus Session
SCMR 22nd Annual Scientific Sessions
Juliano Fernandes, MD, PhD
Physician Researcher
Jose Michel Kalaf Research Institute / Radiologia Clinica de Campinas
.png "Hui Xue, PhD photo")
Hui Xue, PhD
Staff Scientist
National Heart Lung and Blood Institute (NHLBI), National Institutes of Health
Peter Kellman, PhD
Senior Scientist
NHLBI, NIH
Background:
Successful myocardial blood flow (MBF) increase with vasodilation stress is key to ischemia evaluation in cardiovascular magnetic resonance (CMR). Splenic switch-off has been described as a practical method to assess these changes but it does not allow for direct evaluation of the myocardium and only provides information after contrast injection. T2* reflects different levels of myocardial blood oxygenation and may be useful in providing pre-contrast evaluation of vasodilation in the heart. We sought to study the correlation of myocardial T2*, splenic perfusion and MBF changes at different levels of vasodilation using quantitative measurements for all variables.
Methods:
Patients scheduled for a non-stress CMR study were selected and underwent vasodilation stimulus with different doses of dipyridamole (0.18mg/kg to 0.56mg/kg) or no vasodilation (repeat perfusion) in a Siemens Skyra 3T scanner. Myocardial T2*/R2* were acquired before and after vasodilation, pre-contrast, using a free-breathing sequence with automated maps using the Gadgetron framework. MBF was obtained from a quantitative automated map also generated with Gadgetron, after contrast infusion at rest and stress, with myocardial perfusion (MP) ratios calculated from stress/rest maps. Splenic switch off was quantified using the peak signal intensity at stress and rest, corrected for pre-contrast signal, and calculating the spleen signal intensity ratio (SSIR). Correlation coefficients between MBF, MP ratios, changes in T2*/R2* and SSIR were evaluated and ROC curves were used to assess the accuracy of both T2*/R2* maps and splenic switch-off to determine effective vasodilation.
Results:
Forty patients (62.9±12.2 years, 55% males) participated in the study. Changes in T2* and SSIR with vasodilation stimulus and at rest were significantly correlated to MP ratio (r=0.56 [0.31 to 0.75] and r=-0.49 [-0.7 to -0.22]; p=0.0001 and p=0.001 respectively). Both variables also significantly correlated to the maximum obtained MBF (r=0.62 [0.39 to 0.78] and r=-0.55 [-0.74 to -0.29]; p 1.4 as the cutoff for a positive vasodilation stimulus, a decrease in R2* ≤ -1.54Hz provided better discrimination for identifying successful stimulus compared to changes in SSIR (AUC = 0.978 vs 0.827, P=0.04) – Figure 1). The use of SSIR misclassified 8/40 patients in comparison to 2/40 cases with T2* maps (P=0.04). An example of a normal case (no stimulus), matching T2* and SSIR changes and mismatches between T2* and SSIR findings are shown in Figure 2.
Conclusion: Myocardial T2* values before and after vasodilation stimulus significantly correlate to changes in MBF and may provide better discrimination of successful stimulus in comparison to quantitative splenic switch-off analysis.