Ratiometric NOE(−1.6) contrast in brain tumors

Abstract

Recently, a new nuclear Overhauser enhancement (NOE)‐mediated saturation transfer effect at around −1.6 ppm from water, termed NOE(−1.6), was reported to show hypointense signals in brain tumors. Similar to chemical exchange saturation transfer or magnetization transfer (MT) effects, which depend on the solute pool concentration, the exchange/coupling rate, the solute transverse relaxation rate, etc., the NOE(−1.6) effect should also depend on these factors. Since the exchange rate is relevant to tissue pH, and the coupling rate and the solute transverse relaxation rate are relevant to the motional property of the coupled molecules, further studies to quantify the contribution from only the exchange/coupling rate and the solute transverse relaxation rate are always interesting. The purpose of this paper is to apply a ratiometric approach to the NOE(−1.6) effect to obtain a metric that is more specific to the NOE coupling rate and the solute transverse relaxation rate than the NOE(−1.6) signal amplitude. Simulations indicate that the ratiometric approach allows us to rule out nearly all of the non‐specific factors including the solute pool concentration, solute and water longitudinal relaxation rates, direct water saturation, and semi‐solid MT effects, and provides a more specific NOE coupling rate‐ and solute transverse relaxation rate‐weighted signal. Animal studies show that the ratiometric NOE(−1.6) decreases dramatically in brain tumors, which suggests that the change in the NOE(−1.6) coupling rate and/or the solute transverse relaxation rate are major contributors to the previously observed hypointense NOE(−1.6) signal in tumors.

Chemical Exchange Saturation Transfer (CEST) Signal at −1.6 ppm and Its Application for Imaging a C6 Glioma Model

“…The amplitudes of rNOE(−1.6) in the z-spectra and AREX resid were reduced on the tumor side ( Figure 1 , Figure 2 and Figure 3 ); this finding is consistent with those reported by other studies [ 9 , 19 , 25 ]. Figure 4 presents the AREX resid (−1.6) images and the hematoxylin and filipin staining results together; they indicate that the tumor region was cholesterol-deficit and that the AREX resid (−1.6) in the tumor region was hypointense.…”

“…A minor absorption appeared on the shoulder of the direct saturation at −1.6 ppm, and it was referred to as the choline group from lipids. The normal side had a higher saturation than the tumor side [ 9 , 19 , 25 ], and this difference increased during the experimental period.…”

“…A total of 25 voxels from normal regions and another 25 voxels from tumor regions were selected and averaged to obtain the representative z-spectra of each animal. The interferences from water, MT, and T 1 were minimized by employing the residuals of apparent exchange-dependent relaxation rate ( AREX resid ) [ 17 , 18 , 19 , 20 , 21 , 22 ]. In brief, AREX resid was obtained using the inversed residue between the measured z-spectrum ( S meas ) and fitted z-spectrum ( S fit ), which was corrected by T 1 and the pool size of MT ( f m ); the equation for obtaining AREX resid is as follows: where S 0 is the normalizer, and ∆ ω is the saturation offset.…”

“…However, a z-spectrum is highly subject to interference from direct water saturation and magnetization transfer (MT), which are usually more substantial than CEST dips. Researchers have proposed several methods for removing the background signals in CEST through pulse sequences (e.g., chemical exchange rotational transfer (CERT) [ 9 , 10 , 11 ] and variable delay multi-pulse (VDMP) [ 12 , 13 ]) or numerical methods (asymmetric MT [ 1 , 6 , 7 , 14 , 15 ], extrapolated semisolid magnetization transfer reference (EMR) [ 15 , 16 ], and deconvolutions with one or multiple Lorentzian functions [ 2 , 17 , 18 , 19 , 20 , 21 , 22 ]). These procedures are crucial for small CEST signals [ 17 ], such as the relayed nuclear Overhauser effect at −1.6 ppm, that is, rNOE(−1.6) [ 9 , 17 , 19 , 20 , 21 , 23 , 24 , 25 , 26 ].…”

Contribution of blood to nuclear Overhauser effect at −1.6 ppm

“…[32][33][34] Previously, we have tentatively assigned the NOE(−1.6) to be from dipolar interactions between phospholipid choline head group protons and water protons. 14,15,19 Recently, we have reported that the NOE(−1.6) signal from rat brain changes significantly with intake of different gases (ie, air, O 2 , and N 2 O), 16 suggesting that the NOE(−1.6) signal may be related to cerebrovascular reactivity. Two other publications also observed the NOE(−1.6) signal in blood.…”