Rapid T1rho dispersion imaging for improved characterization of myocardial tissue using synthetic dispersion reconstruction

Introduction

Over the past decade, CMRI has become the method of choice for characterizing fibrotic scars. Native T1ρ mapping offers an alternative to conventional T1 and T2 quantification techniques due to its high sensitivity to low-frequency processes. In addition, there is the possibility of T1ρ dispersion imaging, which could be used as a sensitive biomarker for assessing myocardial fibrosis [1]. However, due to a very long measurement time, T1ρ dispersion quantification in myocardium can hardly be done in the limited time of a small animal study. In this work we present a concept for rapid T1ρ dispersion quantification based on the new approach of synthetic dispersion reconstruction (SynDR).

Theory

A T1ρ map is calculated by measuring Nt T1ρ weighted images using different spin lock (SL) times. T1ρ dispersion quantification requires Nf T1ρ maps with different SL amplitudes. Hence the measurement time is very time consuming, because it requires the acquisition of Nt*Nf images (full mapping). With our new approach (SynDR), only a single T1ρ reference map and a series of dispersion weighted images need to be acquired. The T1ρ dispersion can be reconstructed by synthetically generated maps, whereby each map is calculated from the reference map and the dispersion weighted images, only requiring Nt+Nf images.

Methods

All measurements were performed on a 7T small animal scanner. The method was based on an optional cartesian/radial gradient echo sequence using large flip angles (45°) and an optimized readout sorting. The quantification accuracy of SynDR was compared with full mapping measurements in a phantom experiment and validated in vivo on mice. The synthetic T1ρ maps were used to perform a dispersion analysis in myocardium.

Results

The comparison between SynDR and the full mapping reference in phantoms showed a very high quantification accuracy with a mean/maximum deviation of 1.1% and 1.7%. Fig. 1 shows synthetic T1ρ maps (a) in healthy mice and the obtained dispersion map (b) using SynDR. In the dispersion analysis (c) a T1ρ slope of 5.6±1.5ms/kHz was obtained for myocardium. Here an acceleration factor of 4 could be realized in comparison to full mapping. In further measurements, an acceleration of 7.4 could be reached using a radial readout with KWIC filter view sharing.

Discussion

In this work, a novel T1ρ dispersion imaging method was presented that far exceeds the speed of conventional full mapping methods. The acceleration is based on avoiding unnecessary measurements of T1ρ weighted images through more efficient mathematical modeling. Further acceleration could be achieved using an optimized radial data acquisition. The method shows good image quality and high quantification accuracy both in phantom and in vivo. Based on the promising results, further studies in mice are planned to investigate the dispersion character of healthy and diseased tissues.

Reference

[1] Yin Q et al. Magn Reson Imaging. 2017 Oct; 42:69–73.

SynDR method and T1ρ dispersion analysis

Funding Acknowledgement

Type of funding source: Public grant(s) – National budget only. Main funding source(s): BRD, Bundesministerium für Bildung und Forschung

4092Magnetic resonance imaging of Fabry disease cardiomyopathy in patients receiving oral chaperone therapy

Background

Fabry disease is a lysosomal storage disorder with multiple organ involvement. Renal and cardiac symptoms can lead to dialysis and myocardial hypertrophy with fibrosis, responsible for heart failure with preserved ejection fraction (HFpEF). Enzyme replacement therapy (ERT) is available for all patients with Fabry disease since 2001, requiring infusions every other week. Since May 2016, the chaperone migalastat represents a novel form of specific therapy as the first oral therapy available for certain Fabry patients. Through this molecule the function of the mutated enzyme α-galactosidase A can be restored. Recent trials have shown positive cardiac effects of chaperone therapy using echocardiography; however, MRI investigations further evaluating these findings are not available yet.

Objective

To evaluate cardiac effects of migalastat therapy in patients with amenable α-galactosidase A mutations in the prospective monocentric HEAL-FABRY registry (NCT03362164).

Methods and results

Comprehensive clinical investigations including serial MRI were conducted at baseline before initiation of migalastat therapy and at least one year thereafter in all patients without contraindications such as pacemakers or ICDs. Out of 29 patients included in the study (mean age at start of therapy 52.8±14 years, total range 20–74 years), until then 12 patients with MRI data completed the 1-year follow-up. At 1 year, enzyme activity in leucocytes increased from 0.06 to 0.21 nmol/min/mg protein (p=0.001). Distinctive changes over time were observed not only in diastolic but also systolic parameters. The systolic myocardial mass index was reduced by 2.39% (p=0.10). In the AHA segment number 5, most important for classification of severe myocardial damage in Fabry patients, late gadolinium enhancement was reduced by 8.58% in all 5 patients with verified progressive fibrosis (p=0.14). One patient stopped migalastat therapy due to personal reasons. No significant side effects were observed.

Analysis of LGE (systolic phase)

Conclusion

These preliminary MRI data show positive effects of migalastat therapy in patients with Fabry disease and cardiac involvement. Compared to echocardiography, MRI has the potential to allow for comprehensive additional analyses regarding both cardiac morphology and function.