In vivo proton and phosphorus MR spectroscopy can both be applied in monitoring breast cancer treatment response. Proton MR spectroscopy can be used to measure total choline, while ³¹P MR spectroscopy can be used to measure the individual ³¹P containing metabolites of total choline. Here we applied both methods in a group of eight healthy volunteers and also investigated the reproducibility of the two methods to quantify choline composition.

A group of eight healthy female volunteers (ranged 21-30yr) underwent two MRI scan sessions of the right breast with a dual tuned ¹H-³¹P quadrature coil¹ on a 7 tesla scanner (Philips, Cleveland, USA). Each volunteer was scanned twice on the same day, with the same protocol consisting of basic imaging, ³¹P AMESING² MRSI and metabolite cycled^{3 1}H semi-laser spectroscopy. Phosphorus MRSI was done by adiabatic multi-echo spectroscopic imaging, AMESING², with T­R=6s; ΔTE=45ms; matrix 8x8x8; 4x2x4cm³ voxels; acquiring 1 FID and 5 full echoes; with a bandwidth of 8200 Hz and 256 datapoints for the echoes.

Proton spectroscopy was focused on measuring total-choline signal and was done with a water and fat suppressed semi-laser proton sequence with metabolite cycling and TE averaging (118,119,120,121ms) for the overlapping choline resonances at 3.2ppm and a voxel of 15x15x15mm3 carefully placed in fibroglandular tissue.

The group averaged ³¹P FID and echo spectra based on the spectra of the 8 volunteers are shown in Figure 1. Note that the peak labeled GPtC (glycerophosphatidylcholine) is gone from Echo 2 onwards, comparable to the quick disappearance of γ-ATP. Also note that the peak labeled GPC+GPtE loses a substantial intensity from FID to the first echo (GPtE glycerophosphatidylethanolamine has a short T₂), after which the decrease in intensity goes much slower, leaving only the slowly decaying GPC signal. In the FID spectrum there appears to be some signal between the peaks labeled PE and PC that seems to disappear in the echoes, this signal could be AMP⁴.

Figure 2 shows the T₂-fits for the different ³¹P metabolites. Note that the FID signal amplitudes of PE+AMP, PC+AMP, and GPC+GPtC are all well above the fitted mono-exponential curve. The additional bi-exponential fit for these resonances shows a better agree­ment with the data.

Relative metabolite abundance expressed as a percentage of the total ³¹P signal, for this group of volunteers measured twice, as obtained from the data in Figures 1 and 2 are: PE=8%; AMP=5%; PC=4%; Pi=29%; GPE~ 0%; GPC=4%; GPtE=10%; GPtC=17%; γ-ATP=13%; α-ATP=8%; NADP=2%. A measure for the reproducibility of the ³¹P measurements and data analyses of the individual volunteers is given by the mean relative error $$$\langle\overline{RE}_{A_{m}}\rangle$$$ in amplitude A of a metabolite m defined as: $$\langle\overline{RE}_{A_{m}}\rangle=100\cdot\frac{\langle\overline{sd}_{A_{m}}\rangle}{\langle{\overline{A}_{m}}\rangle}, [1] $$

which is shown in Figure 3, for ³¹P and ¹H MRS. Here$$${\langle{\overline{A}_{m}}\rangle}$$$ is the group average of the mean amplitude of metabolite m of the individual duplicate measurements for the eight volunteers, and $$${\langle\overline{sd}_{A_{m}}\rangle}$$$ is the square root of the group average variance from the duplicate measurements of the eight volunteers.

Errors on the individual monoester signals are largest because these signals have on average (see Figure 1) the lowest SNR, overlap partially, and are also influenced by the presence of a disturbing signal from AMP. The sum of monoesters, however, shows only small variability particularly when compared to ¹H tChol. Note the reduction in relative errors and standard deviations of relative errors when making use of T₂-weighted spectra that combines FID and echo spectra instead of FID spectra alone. Figure 4 shows an example of ³¹P MR spectra and ¹H MR tChol edited spectra for volunteer 2 measured in duplicate. The measured tChol concentrations are shown in Table I.

The average tChol concentration(based on 9 protons per molecule) for the group of volunteers is 0.5mM with a mean relative error of 35% that is calculated analogous to that of the ³¹P metabolites. Combined with the ³¹P data and assuming only phosphorylated choline compounds (no free choline) and no mobile phospholipid (MPL) signal in tChol (T₂(MPL)<<120 ms), this would amount on average to the following concentrations of the phosphomonoesters and phosphodiesters in healthy fibroglandular tissue: [PE]=0.4 mM; [PC]=0.2 mM; [GPC]=0.2 mM.

In vivo ³¹P MR spectroscopy offers a better window on cell membrane metabolism and energy metabolism than ¹H MR spectroscopy: ³¹P MRS reproducibility was similar to ¹H tChol MRS and gave insight into the ³¹P metabolite tChol composition. Estimates of phosphomonoester and phosphodiester concentrations in healthy fibroglandular tissue based on a combination of ³¹P and ¹H MR spectroscopy are PE=0.4mM (2 protons); PC=0.2mM (9 protons); GPC=0.2mM (9 protons).