1H-MRS ultra-long echo time PRESS detects omega-3 fatty acids in vivo at 7T
Martin Gajdošík1,2, Lukas Hingerl1, Petr Šedivý3, Miloslav Drobný3, Monika Dezortová3, Michael Krebs4, Siegfried Trattnig1,2, Wolfgang Bogner1, and Martin Krššák1,2,4

1High-field MR Centre, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Vienna, Austria, 2Christian Doppler Laboratory for Clinical Molecular MR Imaging, Vienna, Austria, 3MR Unit, Institute for Clinical and Experimental Medicine, Prague, Czech Republic, 4Division of Endocrinology and Metabolism, Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria

Synopsis

Omega-3 (n-3) fatty acids are essential to human health. However, the knowledge about their tissue concentrations is not well established. The spectral dispersion at 7T yields a sufficient spectral separation of n-3 (n-3 CH3) and non n-3 methyl (CH3) groups. Large differences in the T2 relaxation and J-modulation of outer triplet lines allowed for excellent separation of the n-3 CH3 signal and for estimation of its relative concentration in adipose tissue for the detection by 1H MR PRESS with TE of 1000ms.

Introduction

Omega-3 (n-3) fatty acids (FA) are essential to human health. However the knowledge about their tissue in vivo concentrations is not well established to this day. Proton (1H) MR Spectroscopy (MRS) allows non-invasive detecting of methyl groups of n-3 and non n-3 FA chains at two different frequencies, 0.98 and 0.90ppm respectively. The methyl signals are triplets with characteristic J-coupling modulations of their two outer signals. The detection of the left outer n-3 methyl signal (1.08ppm) with PRESS and echo time (TE) of 540ms at 1.5T was already shown1. Škoch et al. proposed the utilization of J-difference editing via oMEGA-PRESS2 that allowed for the detection of n-3 methyl left outer triplet signal at 1.05ppm with TE of 199.5ms at 3T. However at 7T, the left outer triplet signal of the n-3 methyl group resonates at 1.01 ppm and due to the limitations of B0 shimming performance it can hardly be resolved from its central line in vivo. On the other hand, the spectral dispersion at 7T yields better resolution of both central signals of n-3 (n-3 CH3) and non n-3 methyl groups (CH3) (Figure 1). The concentrations of n-3 FA are very low (approximately 1:75 in adipose tissue2). Besides the J-modulation, the n-3 CH3 has a longer T2 relaxation time than CH31, which could be an advantage for its detection with long echo times. In this work we assessed the differences in T2 relaxation times of CH2, CH3 and n-3 CH3 signals in oils and in vivo and aimed for detecting the n-3 FA with PRESS and ultra-long TE of 1000ms in thigh adipose tissue at 7T.

Materials and Methods

MRS measurements were performed on 7T Magnetom scanner (Siemens Healthcare, Erlangen, Germany) with 28 channel knee coil (Quality Electrodynamics, Mayfield Village, OH, USA). In order to minimize the RF power requirements and chemical shift displacement error of the PRESS sequence the RF pulses were modified for 7T application: Hermite 2.6ms excitation pulse and Hermite 3.0ms refocusing pulse. The bandwidth was set to 3000Hz. The oil measurements were performed on three samples with different n-3 FA concentrations (corn, soybean and fish oil, Sigma-Aldrich, St. Louis, MO, USA). The T2 times were calculated from sixty TE’s (30…1200ms) signal decays (TR=5s, NA=6). The n-3 FA contents of the oils were compared with the literature data (USDA, National Nutrient Database). In vivo measurements were performed on 4 healthy volunteers (1 female, age 29.8±2.9y, BMI 22.8±3.1kg.m-2). The VOI of PRESS (TE=1000ms, TR=5s, NA=32) was set to 2x2x2cm3 and placed in subcutaneous adipose tissue in the lower part of the thigh. T2 times were calculated on three volunteers (nine TEs : 30…1200ms; TR=5s; NA=16). The T2 of n-3 CH3 was calculated from TE range of 350ms to 1200ms. The MRS signals were fitted in jMRUI with AMARES algorithm3 using Gaussian lineshape. The T2s were calculated in MATLAB (MathWorks, Natick, MA) with monoexponential function. The content of n-3 FA was calculated as ratio of n-3 CH3 to the sum of both CH3 signals after T2 relaxation correction.

Results

The J-modulated n-3 CH3 signal from the soybean oil measured with PRESS and TE of 1000ms is depicted in Figure 2A and B. The T2 times measured in the oils and in vivo are in Table 1. The comparison of the n-3 FA content in different oils and literature is depicted in Figure 3. The results of in vivo measurements are provided in Figure 4A and B. n-3 CH3 resonance can clearly be resolved with another advantage of the suppression of the strong CH2 signal, which would otherwise bias the quantification of the CH3 signals. The mean content of n-3 FA in vivo was 0.021±0.002 a.u..

Discussion

In oils, T2 times of n-3 CH3 were found to be twice as long as for the CH3 signal. Similarly we measured 1.8-times longer n-3 CH3 T2 time in vivo. Applying this spin-spin relaxation correction yields expected n-3 FA fraction of approximately 2% [4]. Further in vivo measurements and biochemical validation from biopsy samples of broader population is necessary.

Conclusion

The antiphase J-modulation of CH3 outer triplet signals observed with TE of 1000ms and the differences in T2 relaxation times between n-3 CH3 and CH3 signal allowed for the identification of well resolved n-3 CH3 resonance making ultra-long TE PRESS a method suitable for detection of n-3 FA in vivo at 7T.

Acknowledgements

The authors wish to thank Dr. Rolf Pohmann for programming of the Hermite pulses, Dr. Vladimír Mlynárik for help with the oil experiments and Anniversary Fund of Austrian National Bank (P15363), as well as BMWFW WTZ Mobility (CZ11-2015) for financial support.

References

[1] Lundbom et al. PRESS echo time behavior of triglyceride resonances at 1.5T: detecting omega-3 fatty acids in adipose tissue in vivo. J Magn Reson. 2009; 201(1):39-47

[2] Skoch et al. The in vivo J-difference editing MEGA-PRESS technique for the detection of n-3 fatty acids. NMR Biomed. 2014; 27(11):1293-9

[3] Vanhamme et al. Improved method for accurate and efficient quantification of MRS data with use of prior knowledge. J Magn Reson. 1997; 129(1):35-43

[4] Lundbom et al. Characterizing human adipose tissue lipids by long echo time 1H-MRS in vivo at 1.5 Tesla: validation by gas chromatography. NMR Biomed. 2010; 23(5):466-72

Figures

Figure 1: Simulation of different spectral dispersions for n-3 and non n-3 methyl triplet signals for three different field strengths (scaling of absolute frequency adapted to respective field strength) and linewidths expected in in vivo experiments. Note better separtion of n-3 CH3 left outer tripplet at 1.5T and better separation of central signals at 7T.

Figure 2: A – Soybean oil spectrum measured with TE of 1000ms (1 average) at 7T. B – The same soybean spectrum with 8Hz Lorentzian filter applied to simulate in vivo conditions.

Figure 3: Correlation between n-3 FA content in three oils measured with PRESS (TE=1000ms) and the literature data (linewidths adjusted to in vivo conditions).

Figure 4: A – VOI (yellow) placement in the subcutaneous adipose tissue of thigh. B – In vivo signal measured with TE of 1000ms.

Table 1: T2 relaxation times in ms of oil samples and in vivo (mean values) measured with PRESS at 7T.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4014