1355678 - Technical optimization of dual-bolus scanning protocol for quantitative perfusion cardiovascular magnetic resonance – initial results

Quantitative perfusion cardiovascular magnetic resonance (QP CMR) enables absolute assessment of myocardial blood flow (MBF), by acquisition of signal-intensity curves of first pass gadolinium-based contrast agent (GBCA) and tracer kinetic analysis. However precise MBF quantification is influenced by a nonlinear relationship between the obtained signal intensity and GBCA concentration. Dual-bolus first-pass perfusion CMR uses the combination of both low and high GBCA concentration to achieve the required linearity of signal intensity allowing for accurate MBF measurement. A general concern is the signal-to-noise ratio since this effects the QP analysis. Hereby we propose an optimized dual-bolus scanning protocol using an adjusted contrast-administration scheme to enhance the signal-to-noise ratio without associated  contrast-induced artefacts.

Methods:

Adenosine stress perfusion CMR imaging at 3T was performed in subjects with 3 different medical conditions. The analysis included QP CMR was performed in 3 patients with obstructive coronary artery disease (CAD) as confirmed by invasive coronary angiogram, 3 patients with invasively diagnosed microvascular dysfunction (MVD) and 3 healthy controls. Myocardial perfusion images were acquired using a gradient-recalled echo acquisition at basal, mid and apical level.

Stress images were acquired using a saturation recovery turbo spoiled gradient echo sequence after at least 3 minutes of constant intravenous infusion of adenosine (140 μg/kg/min). Rest images were acquired at least 10 minutes after adenosine infusion. The arterial input function (AIF) images were acquired during intravenous administration of a pre-bolus of GBCA (DOTAREM®, Guerbet, Villepinte, France 0.5 mmol/ml) at a dose of 0.0075 mmol/kg 3ml/sec. Perfusion images were acquired using a main bolus of GBCA at a dose of 0.075 mmol/kg 3ml/sec. (Figure 1). The images were analysed offline using cvi42 software (Circle Cardiovascular Imaging Inc, Calgary, Canada) equipped with a fully automated pixel-wise QP CMR module.

Results: The proposed scanning protocol provided high quality images, with a high signal-to-noise ratio, which could be post-processed by the fully automated QP-framework. Assessment of pixel-wise stress MBF maps enabled diagnosis of ischemia in patients with obstructive CAD and MVD. No perfusion defect was visible on pixel-wise stress MBF maps in a healthy control patient (Figure 2).

Conclusion: The optimized dual-bolus scanning protocol seems to achieve a good linearity of signal intensity, resulting in high quality images suitable for QP post-processing without hampering of the signal-to-noise-ratio. This might improve the applicability of QP in diagnosing ischemia in patients with obstructive CAD and MVD.