Synopsis

Cardiac PCr/ATP ratios measured by ³¹P MRS are a measure of the biochemical “energy reserve” in the heart. They decrease early in the progression of a wide range of diseases, making them a valuable biomarker. Logistically, it is often important for a long-running study to be able to compare data acquired on different MRI scanners. In this study, we show that comparable results may be acquired with 3T Prisma and 3T Trio scanners by scanning 10 healthy volunteers with ³¹P-CSI and ³¹P-DRESS on both systems. It is legitimate to pool 3T Prisma and 3T Trio data for clinical studies.

Purpose

In most major heart diseases the ratio of phosphocreatine (PCr) to adenosine triphosphate (ATP) concentrations, i.e. PCr/ATP, is impaired. So, PCr/ATP is a useful indicator of the energetic state of the heart.^1,2 PCr/ATP can be assessed using 3T ³¹P-MRS. Two widely-used localization sequences are 3D chemical shift imaging (3D-CSI) and 1D depth-resolved spectroscopy (DRESS).³ Many studies run longitudinally, or can gain value from comparing to substantial historic datasets. However, the pace of MR technology continues and we, as well as many other centres who apply ³¹P-MRS, have recently installed a new 3T Prisma (Siemens) MRI scanner to complement, or replace, an ageing 3T Trio (Siemens) scanner. These new systems are not always equipped with the same RF-coils making pooling of existing and newly acquired data even more questionable.

Therefore, the aim of this study was to assess whether PCr/ATP measurements obtained using two different models of 3T MRI scanner, together with different surface coils from different manufacturers can be pooled together. Data from this work will be important for the design of future clinical studies.

Methods

Data acquisition 10 healthy volunteers (4F, age 30 ± 4yrs, BMI 24.4 ± 4.0 kg/m²) were recruited according to local regulations. The scanning protocol is given in Figure 1. Six of the 10 subjects completed the full protocol; four subjects were only scanned with the 3D UTE-CSI sequences. The UTE-CSI sequence followed a published protocol,⁴ with the following parameters: matrix size 8x16x8; FOV 240x240x200mm³; acquisition weighting with 10 averages at k=0, and a shaped excitation pulse. Three B₁-insensitive train to obliterate signal (BISTRO), saturation bands were placed covering anterior chest wall (to suppress signal from skeletal muscle) and liver.⁵ The TR was ~720ms (varied to give maximum SAR), giving a total acquisition time of approximately 10 minutes. The DRESS sequence used an 800µs duration sinc excitation pulse; slice thickness 20mm; TE = 1ms; TR = 1000ms and 180 averages, leading to an acquisition time of 3 minutes.

We used three ³¹P RF coils in our comparison:

1. A “heart-liver” coil (a 1.5T Siemens,retuned for 3T), comprising a large rectangular (28x26cm²) element which acts as a local transmit/receive ¹H coil and the ³¹P transmit coil; a second “loop-butterfly” quadrature pair acts as the ³¹P receive element. This coil was used to acquire the “Trio-CSI-HL” data set.
2. A 10cm loop coil (PulseTeq, Chobham, UK), comprising a ³¹P only transmit/receive coil with an inductively driven single circular 10 cm diameter loop element. This coil is used to acquire the “Trio-DRESS-10cm” data set.
3. A “flex coil” (Rapid Biomedical, Rimpar, Germany), comprising a ¹H butterfly transmit/receive element and a ³¹P 11cm diameter transmit/receive circular loop. This was used to acquire the "Prisma-" data sets.

Data analysis: Data were fitted using the Oxford Spectroscopy Analysis (OXSA) toolbox⁶ for all subjects. Prior knowledge specified 11 Lorentzian peaks, linewidths constrained relative to the PCr linewidth, and literature values for the scalar couplings for the multiplets. Blood contamination and partial saturation was then corrected using T₁ values and the blood DPG/ATP ratio from the literature.⁷

Assessment of reproducibility: The coefficient of reproducibility (CR) was calculated according to:

$$CR = SD\tiny intrasubject\normalsize \times1.96$$

Results and discussion

There were no significant differences in PCr/ATP measured using either scanner, coil, or sequence (Figure 3). We found good reproducibility (Fig. 4) of the 3D-CSI sequence on the Prisma (CR=0.35), and good agreement in measurements between both scanners (CR=0.37) meaning that measurements from these two systems can be compared.

There is greater scatter in PCr/ATP measurements using DRESS compared to the 3D-CSI sequence, this is to be expected owing to the shorter acquisition time of these data-sets. We believe that some of our DRESS measurements, e.g. those with PCr/ATP >3 are likely to have skeletal muscle contamination in the supposedly “cardiac” spectra. We did not use any saturation bands here; in future DRESS studies, we recommend applying a saturation band to suppress signal from the anterior chest wall to reduce the chance of skeletal muscle contamination.

In this study, we were able to test only Siemens scanners operating at 3T. Further validation would be required to pool data across field strengths, or to other makes/models of scanners.

Conclusion

Acknowledgements

References

1. Neubauer S. The failing heart--an engine out of fuel. N Engl J Med. 2007;356(11):1140–1151. doi:10.1056/NEJMra063052.

2. Bottomley PA, Bottomley, A. P. MRS Studies of Creatine Kinase Metabolism in Human Heart. In: eMagRes. Chichester, UK: John Wiley & Sons, Ltd; 2016:1183–1202. doi:10.1002/9780470034590.emrstm1488.

3. Bottomley PA, Foster TB, Darrow RD. Depth-resolved surface-coil spectroscopy (DRESS) for in Vivo1H, 31P, and 13C NMR. J Magn Reson. 1984;59(2):338–342. doi:10.1016/0022-2364(84)90179-3. 4. Tyler DJ, Emmanuel Y, Cochlin LE, et al. Reproducibility of 31 P cardiac magnetic resonance spectroscopy at 3 T. NMR Biomed. 2009;22(4):405–413. doi:10.1002/nbm.1350.

5. Luo Y, de Graaf RA, DelaBarre L, Tannús A, Garwood M. BISTRO: An outer-volume suppression method that tolerates RF field inhomogeneity. Magn Reson Med. 2001;45(6):1095–1102. doi:10.1002/mrm.1144.

6. Purvis LAB, Clarke WT, Biasiolli L, Valkovič L, Robson MD, Rodgers CT. OXSA: An open-source magnetic resonance spectroscopy analysis toolbox in MATLAB. Motta A, red. PLoS One. 2017;12(9):e0185356. doi:10.1371/journal.pone.0185356.

7. Rodgers CT, Clarke WT, Snyder C, Vaughan JT, Neubauer S, Robson MD. Human cardiac 31 P magnetic resonance spectroscopy at 7 tesla. Magn Reson Med. 2014;72(2):304–315. doi:10.1002/mrm.24922.