11 applications of proton mrs to study human brain metabolism
DESCRIPTION
neuro chemTRANSCRIPT
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 1/34
Applications of Proton MRS to Study
Human Brain Metabolism
Christopher C. Hanstock and Peter S. Allen
1.
Introduction
Magnetic resonance spectroscopy (MRS) provides information
that is rarely obtainable by other noninvasive means, or even by
invasive methods using radioactive labels. For example, it pro-
vides the means to monitor in time and in space changes in vari-
ous metabolic pools and allows one to think in terms of the
biochemistry of these pools. In this sense, MRS is quite unique,
and, although it cannot be said to be highly specific in diagnosing
individual diseases, it nevertheless enables changes in many criti-
cal and characteristic parameters to be observed noninvasively
for a broad range of metabolic abnormalities. By its nature MRS
lends itself more toward the evaluation of diffuse brain diseases
rather than that of focal lesions. For example, typical applications
of MRS have been (1) to assess the regional distribution of neu-
ronal dysfunction or death, (2) to evaluate distributions in the
oxidative state of the brain, or (3) to detect regions of membrane
abnormality. This list is growing as new MRS technology emerges.
The adoption of MRS as a routine diagnostic and patient manage-
ment tool in clinical medicine has, however, been quite slow when
compared to the rapid acceptance of magnetic resonance imaging
(MRI) several years ago. To understand this one must acknowl-
edge that not only is MRS more demanding technically than MRI,
but the clinical significance of spectroscopic findings has not al-
ways been recognized prior to the in vivo application of the nuclear
magnetic resonance (NMR) spectroscopic techniques. The signifi-
cance of N-acetylaspartate (NAA) in brain is a case in point
(Blakely, 1988). Although its exact role in neurons is still not fully
From Neuromethods vol. 33 Ce ll Neurobiology Techm ques
Ed A A Boulton G B Baker and A N Bateson 0 Humana Press Inc
347
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 2/34
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 3/34
Applications of Proton MRS 349
N-Trimethyl
Chqphosphoryl-Cho.
glycerophos~horyl-ChoI
PR
CHZ
AH,
I+
CH,-I;I-CH,
Cr / PCr
N-Acetylaspartate
H
CH,COOH
HN NHR
NC/
'c'
I
HOOC
3.2 2.8 2.4
2.0
Chemical Shij? ppm)
Fig. 1. Proton MR spectrum from normal human brain (PRESS, TE 80
ms) showing the three prominent singlet peaks arising from the N-
trimethyl compounds, from Cr/PCr and from NAA (including other N-
Acetyl compounds). The resonances from coupled spin multiplets
between 2.0 and 2.7 ppm are not observable at this value of TE.
metabolites are of particular interest because of their roles in neu-
rotransmission and for Glu and Asp in their excitotoxicity. Con-
sequently, the ability of ‘H MRS to detect changes in their
concentrations may provide insights into the etiology of neuro-
degenerative disease that could not have otherwise been obtained
VanderKnaap et al., 1992). It must be borne in mind, however,
that the limitations in spatial resolution and sensitivity inherent
in MRS will prevent the discrimination between the amino acids
residing in different cellular subpools, e.g., intracellular vs extra-
cellular synaptic compartments. However, the observation of
changes in GABA concentrations resulting from antiepileptic treat-
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 4/34
350 Hanstock and Alien
ments (Preece et al., 1991; Confort-Gouny et al., 1993) have begun
to be reported. The observation of Gln, on the other hand, could
shed light on its activity as a Glu precursor, as well as its role in
the metabolism of ammonia in diseases where abnormalities of
ammonia are present (Preece et al., 1991; Ross, 1991; Kreis et al.,
1992b; Confort-Gouny et al., 1993). Other coupled spin metabo-
lites that are being studied include myo-inositol, a sugar involved
in several mechanisms including a secondary messenger system
(Berridge et al., 1989); glucose, the basic substrate of brain
metabolism, and the observation of which has been reported under
both normal (Gruetter et al., 1992) and hyperglycemic conditions
(Kreis et al., 1992a); and lactate, whose concentration has been
shown to increase as a result of anaerobic metabolism. The devel-
opment of methods for observing coupled spins is covered in Sub-
heading 3.
2. localized ‘H Spectra of Methyl Singlets in Human Brain
2.1. Methods
Localized single voxel ‘H spectroscopy studies of the human
brain have increasingly relied on the use of two localization pulse
sequences, namely, the STEAM (stimulated echo-acquisition
mode) (Frahm et al., 1989a) and the PRESS (point resolved spec-
troscopy) (Gordon et al., 1984; Bottomley, 1987) schemes. Both of
these techniques allow for localization to a volume whose dimen-
sions, and orientation in space are defined by the three orthogo-
nal slice-selective pulses present in each of the sequences. The
location of that volume is at the intersection of these slices. In
addition to the localization pulses, additional pulses are usually
included in both of the sequences to bring about water suppres-
sion. This can be done either by frequency selective saturation
(Haase et al., 1985; Frahm et al., 1989a) or by inversion nulling
(Patt et al., 1972). Water suppression is necessary because of the
large difference in concentration between the water (50 M) and
the metabolites (up to -10 mM) to be measured.
As well as the methods for acquiring single voxel spectra, there
are several methods that allow a spatial mapping of a metabolite
over a predefined brain slice. These fall into the category of spec-
troscopic or chemical shift imaging methods (Brown et al., 1982;
Maudsley et al., 1983; Pykett et al., 1983; Dixon, 1984). Through
postprocessing, such techniques can provide either a complete ‘H
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 5/34
Applications of Proton MRS 351
spectrum from each adjacent voxel within a slice or, alternately,
an image representing the concentration map of a single metabo-
lite resonance throughout the slice. Generally, the much longer
acquisition period required for these methods, as well as the sig-
nal processing, extracts a significant time penalty.
For the majority of studies reported in the literature, the spec-
troscopic pulse sequences are applied using a circumscribing
radiofrequency (RF) head coil, often a birdcage coil (Hays et al.,
1985; Tropp, 1989; Vu110 et al., 1992), which facilitates the acquisi-
tion of preparatory NMR images used for volume selection or reg-
istration. A much smaller number of studies have used a surface
coil for both transmission and signal reception (Ackerman et al.,
1980), primarily to take advantage of the higher receiver sensitiv-
ity of the surface coil for selected volumes that lie close to the
surface of the head. In addition, the lower RF power requirements
of surface coil transmission are advantageous in studies performed
at higher magnetic field strengths (3-4 T), where typical RF power
levels of a circumscribing coil would have exceeded safety guide-
lines (Athey, 1992).
One consequence of the time required to execute all the
RF
and
gradient pulses necessary for localization of’H spectroscopy in vivo
is that a significant part of the available signal can be lost through
various relaxation mechanisms. The reported resonance peak ratios
of the various metabolites are therefore weighted by the respective
transverse relaxation rates (T,s> (Hanstock et al., 1988; Frahm et al.,
1989b) of the metabolites in question. The variation in T2s between
the metabolite resonances thus makes the concentration ratio mea-
surement strongly dependent on the spin-echo time or TE that was
used in the experiment. Because there is also a magnet field depen-
dence of Tz, care must be exercised when comparing data from one
field strength/laboratory to another. The TE chosen for the pulse
sequence also affects any additional signal loss resulting from
molecular diffusion in any field inhomogeneity, a loss which is gov-
erned by the nature of the pulse sequence used.
Methods for absolute concentration quantification have been
reported (Kreis et al., 1993a), and have made use of both internal
(e.g., Cr, water) and external (water) concentration references. Such
methods make use of corrections for differences in the
T2
relax-
ation rates of metabolites and for partial volume effects caused
by regions of CSF falling within the selected volume.
For metabolites that are freely mobile, the fundamental factor
affecting the sensitivity or signal-to-noise (S/N) of the methyl sin-
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 6/34
352 Hanstock and Allen
glet peaks is the total metabolite concentration present in the
selected volume. Bound metabolites cannot be observed by a typi-
cal in vivo MRS spectrometer. The MRS visible concentration
determines both the minimum volume size accessible in a given
time and the minimum amount of signal averaging that will be
required to give an adequate S/N in a time tolerable for the sub-
ject. A second factor affecting S/N is the magnet field homogene-
ity within the selected volume. Manual and automatic shim
routines allow for optimization of field homogeneity, however,
one cannot loose site of the fact that placement of the voxel adja-
cent to certain structures, e.g., near to bone or air interfaces, can
substantially limit one’s ability to shim. This becomes more seri-
ous at higher magnetic field strengths where the dlstortlons m
the field homogeneity at tissue interfaces due to susceptibility
effects become more significant. The increase in signal strength
and hence the improvement in S/N obtained by using a higher
magnetic field is partially offset by the reported shortening of
metabolite T2s, particularly where longer TE experiments are
described. Typical acquisition times for single volume spectra are
in the 2-10 min range, whereas for a spectroscopic image the
acquisition time may be over 1 h.
2.2. Distribution of the Resonances NA, Cr, and Cho
in Human Brain
The distribution of the three most easily observable metabolite
resonances in the proton spectrum of the human brain, shown in
Fig. 2, has been studied on both the macroscopic and the cellular
levels. For example, several studies have demonstrated the differ-
ences in metabolite levels between gray and white matter, whereas
others have reported the metabolite complement in different cell
types grown in culture, as a guide for in vivo observations.
The NA resonance has been proposed and has gained consid-
erable acceptance as an index of the neuronal pool size, since a
growing body of evidence suggests that the amino acid NAA is
confined to neurons (Birken et al., 1989; Urenjak et al., 1992;
Urenjak et al., 1993). The only reported exception has been that
detected in oligodendrocyte type-2A progenitor cells (O-2A) cul-
tured in vitro (Urenjak et al., 1992; Brand et al., 1993; Urenjak et
al., 1993). Because this latter cell type would not be expected to
contribute significantly to MR spectra acquired from healthy adult
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 7/34
Applications of Proton MRS 353
NAA + N-Ace@?
Cho
.i L
I
3.5
I I I
3 2.5 2
Chemical Shift (ppm)
Fig 2 Proton MR spectrum from human brain acquired at 3 T using
the PRESS pulse sequence with an intermediate TE = 30 ms, from the
temporal lobe of a normal volunteer. Because of the intermediate nature
of TE, the overlapping multiplet resonances of the coupled spin metabo-
htes can be clearly seen in the 2-O-2.7 ppm region of the spectrum.
brain, it may be an issue in studies of developing brain and in
injured brain. Specifically, injured brain has been shown to have
increased activity of platelet derived growth factor and fibroblast
growth factor, both of which induce O-2A adult cells to exhibit
characteristics of O-2A perinatal cells (Wolswijk et al., 19921, which
in turn may modulate the NA peak measured by MRS.
In normal brain the distribution of NAA is reported to be 5-
25 higher in gray matter than in white matter (Kreis et al., 1993a;
Michaelis et al., 1993; Hetherington et al., 1994a; Kreis, 1994), pos-
sibly indicating an increased concentration in the nerve cell bod-
ies compared to the axons. It has also been suggested that this
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 8/34
3.54 Hanstock and Allen
apparent concentration gradient may be due to the higher axonal
activity of the enzymes NAA-aminohydrolase and L-aspartate-N-
transferase (Burri et al., 1991). The action of these enzymes would
result in a faster turnover of NAA, an important element in its
proposed role as an acetyl reservoir in lipid synthesis (Matalon et
al., 1989; Burri et al., 1991; Kunnecke et al., 1993; Petroff et al.,
1994). The NA resonance peak at 2.02 ppm has signal contribu-
tions from N-acetylaspartylglutamate (NAAG) as well as NAA.
Immunohistochemical studies have shown that while NAA and
NAAG both stain positively to carbodiimide in the brain, they
appear to be exclusive in location, with NAAG in mterneurons and
NAA in pyramidal neurons (Moffett et al., 1993). Because both NAA
and NAAG are exclusive to neuronal elements, then their sum, as
measured by in vivo MRS, continues to remain a potential marker
for the overall neuronal pool within the voxel of interest.
Because of its ubiquity and suggested uniform distribution in
normal brain (Frahm et al., 1989a), many MRS studies have used
total Cr (Cr + PCr) as an internal concentration reference when
exploiting metabolite peak ratios as a means to quantify apparent
concentration changes. In contrast, several quantitative MRS mea-
surements of direct concentration in vivo (Kreis et al., 1993a;
Michaelis et al., 1993; Hetherington et al., 1994a; Kreis, 1994) and
in tissue extract (Petroff, 1989) indicate that there is a variation in
the total Cr concentration levels, where gray matter has a 25-30
higher concentration than white matter. Moreover, it has been
reported that creatine concentration increases in the proportion
1:2:4 between neurons:astrocytes:oligodendrocytes grown in cul-
ture (Urenjak et al., 1992; Urenjak et al., 1993) the latter pair of
these being located predominantly in gray matter. While these
three cell types occupy a significant proportion of the cytoplas-
mic space in normal brain, studies of injured brain have revealed
that elevated levels of macrophages may be observed Petroff et
al., 1992; Lopez-Villegas et al., 1995). This is relevant because mac-
rophages in culture have been shown to possess elevated PCr con-
centrations when activated, and would contribute to the total Cr
pool measured by ‘H MRS (Seguin et al., 1990; Seguin et al., 1991).
The MRS peak designated as Cho is the sum of contributions
from several choline derivatives including free choline, phos-
phorylcholine, and glycerophosphorylcholine, as well as those of
noncholine origin from betaine and carnitine. Several quantita-
tive MRS studies comparing the Cho concentration in gray and
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 9/34
Applrcations of Proton MRS 355
white matter have found there to be no significant difference
between them (Kreis et al., 1993a; Michaelis et al., 1993; Hetherington
et al., 1994a; Hetherington et al., 1996). Owing to the involvement
of cholines with membrane lipids, particularly myelin in brain
tissue, it has been suggested that choline containing compounds
would be expected to rise in conditions of membrane disruption,
as may be experienced following brain injury (Brenner et al., 1993;
Szigety et al., 1993).
2.3. Observation of Spectral Changes of NA, Cr, and Cho
from ‘H MRS in Human Subjects
2.3.1, Brain Development and Aging
Several studies have been performed that explored the changes
in the ‘II MR spectrum resulting from early development and from
aging of the brain (Kreis et al., 199313;Chang et al., 1996; Ashwal
et al., 1997). The most marked changes occur in the first 6 mo after
birth, when a rapid increase in the NA/Cr ratio is observed,
accompanied by a decrease in the Cho/Cr ratio at a similar rate.
The factor of 2 increase in the NA intensity from birth to adult
brain was interpreted as neuronal development (Kreis et al.,
1993b), whereas the elevated Cho was thought to reflect acceler-
ated myelination in the first few months of life. For a group of
adults in the age range 19-78, a quantitative study estimating
metabolite concentrations in frontal white matter found that while
the NA was relatively stable, there were increases with aging in
the Cr and Cho resonances in gray matter, whereas in white mat-
ter, no significant changes in metabolite concentrations were
observed (Chang et al., 1996).
2.3.2. Neurodegenerative Diseases
Neurodegenerative diseases such as Alzheimer’s disease (AD),
amyotrophic lateral sclerosis (ALS), and multiple sclerosis (MS),
which involve the regional loss of neuronal tissue, are potentially
fertile areas for study by ‘H NMR. The modulation of the NA peak
has been the focus of attention owing to its postulated role as a
neuronal marker. Observations of a decline in the NA peak, typi-
cally reported as changes in the NA/Cr or NA/Cho ratios, have
been ascribed to a depletion in neurons. It is important to bear in
mind, however, that in order to observe a significant decrease in
these ratios, a rather substantial decrease in neurons per unit vol-
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 10/34
356 Hanstock and Allen
ume is required. The reason for this stems from the large cyto-
plasmic volume occupied by neurons (-80 ) compared to glia,
coupled to the fact that the neuronal metabolite pool has all three
metabolites present, whereas the glia contribute only to the Cr
and Cho peaks.
MRS studies of AD have examined tissue extracts from brain
regions that encompassed a range of senile plaques. In one study,
significant decreases in the NA intensity for AD brain samples
were observed compared to controls, with the largest decreases
correlating with the largest number of senile plaques (Klunk et
al., 1992). In a second study, whereas decreases in the NA inten-
sity (20-30 ) were observed u-t samples from cortical gray matter
regions, no changes were observed in the cortical white matter
samples (Kwo-On-Yuen et al., 1994). In contrast, an in vivo study
using spectroscopic imaging methodology reported significant
decreases in the NA/Cr and NA/Cho ratios for selected volumes
in white matter, but no differences in the Cho/Cr ratio between
the AD brain and controls In the posterior centrum semiovale,
however, NA/Cho and Cho/Cr ratios were both increased and
no change in the NA/Cr ratio was observed (Meyerhoff et al.,
1994a). The authors concluded that their data suggest diffuse
axonal loss accompanied with membrane alterations in both gray
and white matter.
The application of MRS techniques to the study of the rapid
neurodegeneration resulting from ALS has also received atten-
tion recently. Initial reports focused on the neuronal loss m the
primary motor cortex and showed significant decreases in the NA/
Cr ratio (Pioro et al., 1994; Jones et al., 1995; Gredal et al., 1997)
Similar conclusions have been made from studies of the bramstem,
with a strong correlation between upper motor neuron and bul-
bar function loss based on neurological testing and the degree of
NA/Cr depletion (Cwik et al., 1997). A strong correlation between
the extent of motor cortex depletion and brainstem depletion
resulting from ALS, in addition to a progressive reduction of the
NA/Cr ratio has been reported in abstract form (Hanstock et al.,
1997) when longitudinal measurements were made at 2-3-mo
intervals over a 1-yr period.
In vivo MRS studies from brain regions that were hyperintense
on MRI and associated with MS lesions revealed that the NA/Cr
ratio could be decreased by up to 30 , with the largest reductions
observed for those patients who were most severely affected
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 11/34
Applications of Proton MRS 357
(Arnold et al., 1990a; Matthews et al., 1991; Miller et al., 1991;
VanHecke et al., 1991; Arnold et al., 1992; Bruhn et al., 1992; Arnold
et al., 1994; Davie et al., 1994; Husted et al., 1994; Pan et al., 1996a).
Changes in the NA/Cho ratio showed a similar pattern, with a
reduction of similar magnitude to that reported for the NA/Cr
ratio. These observations did not depend on whether the lesions
were acute or chronic. Longitudinal studies for chronic lesions
found that while MR images showed little change in the extent of
the lesions studied, MRS measurements of the NA/Cr ratio
showed further decreases 12-18 months after the initial examina-
tion (Arnold et al., 1994). In contrast, studies of acute MS lesions
reflect a significant decrease in the NA/Cr ratio at the onset of
lesion development, which decreases further over a l-4 mo period,
followed by a recovery toward control values in the subsequent
4-8 month period (Davie et al., 1994). Reductions in the NA/Cr
ratio have also been observed in normal appearing white matter
in patients with either acute or chronic MS lesions, with the mag-
nitude of the reductions being intermediate between control and
lesion values (Davie et al., 1994; Husted et al., 1994). False assign-
ments of MS lesions based on NA/Cr ratio measurements has been
overcome by the use of MR image segmentation to estimate the
gray:white mix in selected MRS voxels (Hetherington et al , 1996;
Pan et al., 1996a).
2.3.3. lschaemia
The modification of metabolite concentrations caused by the
interruption of blood flow to the brain has been evaluated by in
vivo MRS in the case of stroke and cardiac arrest. All studies of
chronic lesions resultmg from stroke have demonstrated that there
is a significant decrease in the NA peak relative to the peaks of
other metabolites, thereby suggesting neuronal loss (Graham et
al., 1992; Petroff et al., 1992; Sappey-Marinier et al., 1992; Graham
et al., 1993; Gideon et al., 1994; Hetherington et al., 1994b). In the
acute stages of lesion development, however, NA levels were
shown to remain in the normal range, and remain so for the first
l-2 wk following the event (Graham et al., 1993). Using spectro-
scopic imaging and taking advantage of a magnetic field strength
of 4.1 T, the spatial distribution of the NA resonance intensity was
determined following a stroke (Hetherington et al., 1994b).
Extracting subspectra from a region within the 6-wk-old infarct
showed a total absence of NA, whereas a volume adjacent to the
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 12/34
358 Hanstock and Allen
infarct showed only a decrease in the NA compared to that in an
equivalently located volume in the contralateral hemisphere.
In a study of coma, resulting from a variety of insults in new-
borns, infants, and children, occipital NA/Cr was lower in the
infants and children, whereas Cho/Cr was elevated in all groups
when compared to age-matched controls (Ashwal et al., 1997). The
extent of NA/Cr and Cho/Cr ratio changes were further increased
in those patients who had elevated lactate. Patient outcome and
recovery was shown to correlate strongly with the extent of these
abnormal metabolite concentration ratios. A similar study for adult
subjects in a coma, and currently under review, has shown that
serial NA/Cr ratio measurements declined following ischemia
with the rate and extent of reduction being predictive of outcome
(Penn et al., 1997). Moreover, postmortem studies on the nonsurvivor
group showed that the largest decreases in the NA/Cr ratios corre-
lated with the largest loss of nerve cell bodies and axons following
histological examination and cell volume estimation.
2.3.4. Cancer
MRS studies of cancer in brain has taken place at both in vivo
and in vitro levels of investigations. Whereas in vitro studies of
cultured human tumor cell lines and excised tumor tissue has
facilitated the identification of tumor-borne metabolites, in vivo
applications have enabled comparisons of metabolite pools to be
measured between regions located within the tumor, adjacent to
the tumor, and in contralateral brain. By using the MRS data from
the contralateral region as an intraindividual control, the metabo-
lite pool changes occurring adjacent to the tumor, which result
from the effects of, for example, compression and peritumor
edema, have been evaluated. Moreover, the metabolic milieu of
the tumor tissue, and the effects on metabolism as a response to
forms of therapy have also been investigated.
The presence of all three singlet metabolite peaks, which are
routinely observed in normal brain (NA, Cr, Cho), has been dem-
onstrated in extract studies of excised tumor tissue in vitro (Peel-
ing et al., 1992; Kotitschke et al., 1994; Florian et al., 1995; Carpmelh
et al., 1996). However, the relative proportions of these peaks
reflect the differentiation of tumor types. The presence of a NA
peak was considered to stem from residual brain that had been
infiltrated by the tumor and was therefore contaminating the
excised tissue sample (Peeling et al., 1992; Carpinelli et al., 1996).
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 13/34
Applicatrons of Proton MRS 359
Glioblastoma multiforme exhibited an elevated Cr/Cho ratio when
compared to either differentiated or anaplastic astrocytomas
(Carpinelli et al., 1996). Conversely, in tissue derived from astro-
cytoma, gliobastoma on malignant melanoma, the choline level
was reported to vary not only between different tumor types, but
also between tumors of the same type and level of malignancy,
and also between samples from within the same tumor (Kotitschke
et al., 1994).
In vivo studies of tumor metabolism as a means to assess tumor
grade and type have been the focus of several reports and reviews
(Arnold et al., 1990b; Kugel et al., 1992; Barker et al., 1993; Ott et
al., 1993; Usenius et al., 1994a,b; Negendank et al., 1996; Preul et
al., 1996). All report significant changes in the metabolite ratios,
where, as a general rule, the NA/Cr ratio decreased and the Cho/
Cr ratio increased. These data were interpreted in early studies as
a decrease in NA intensity (loss of neurons in the sampled vol-
ume), and an increase in Cho (increase in lipid metabolism or
mobilization) (Arnold et al., 1990b; Kugel et al., 1992; Barker et
al., 1993; Ott et al., 1993). However, as a result of quantitative con-
centration measurements for astrocytomas in vivo (grades I-IV),
it was realized that the Cho concentration remained relatively
constant and that the Cr concentration decreased from grade I-II
tumors through to grade IV (Usenius et al., 1994a,b). Confirma-
tion of these in vivo data was provided by the examination of
tissue extracts derived from excised tumor.
In a large 15 site study, an attempt was made to evaluate and
classify several tumor types based on their Cho/Cr, NA/Cr and
Cho/NA ratios (Negendank et al., 1996). Whereas all tumor types
had ratios distinguishable from normal brain, classification was
imprecise due to the large data scatter, astrocytomas being a par-
ticular problem. Using a method described as metabolic profil-
ing, Preul et al. report the accurate classification of 90/91 tumors,
compared to 71/91 correct diagnoses obtained using the primary
preoperative clinical tests of CT, conventional MRI, and conven-
tional angiography (Preul et al., 1996). This spectroscopic metabo-
lite profiling requires the concentrations of seven metabolites to
be estimated, six from the tumor (Cho, Cr, NAA, alanine, lactate,
and lipid) and Cr from normal contralateral brain (used as a ref-
erence). The ratios of the six tumor to contralateral Cr are then
expressed as a ratio profile which was shown to be characteristic
for each of the tumor types examined grade II (low grade) astro-
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 14/34
360 Hanstock and Allen
cytoma, grade III (anaplastic) astrocytoma, grade IV astrocytoma
(glioblastoma multiforme), meningioma, and metastases from lung
and breast cancer).
Metabolite ratios in regions adjacent to tumors are reported to
be dependent on the presence or absence of peritumor edema
(Kamada et al., 1994a,b). If peritumor edema is present, then a
significant decrease in the NA/Cr ratio was observed, moreover,
this decrease returns to normal values as the edema dissipates.
The large decrease in the T, relaxation rate for the metabolites
that was observed to accompany edema is suggested as respon-
sible for some of the changes in metabolite ratios.
Several studies have examined the effects of radiation therapy
on the metabolite composition of tumor tissue, and also on that of
the normal brain (Szigety et al., 1993; Sijens et al., 1995, Usenius et
al., 1995). Decreases in the tumor Cho were observed following
radiation therapy, which were suggested to result from a decrease
in cell density (increased interstitial space) (Sijens et al., 1995).
The decrease in NA intensity observed m normal brain, which
had received a substantial dose of radiation (Szigety et al., 19931,
was recently confirmed by more careful quantitative measure-
ments (Usenius et al., 1995).
2.3.5.
HIV
Changes in the metabolite ratios observed in spectra obtained
from patients with HIV, where dementia has been diagnosed,
showed significant reductions in the NA/Cr ratio, and increases in
the Cho/Cr ratio for both gray and white matter regions (Menon et
al., 1990; Chong et al., 1993; Laubenberger et al., 1996; Tracey et al.,
1996). For HIV-positive asymptomatic patients, one study showed
that the NA/Cr was only slightly decreased, while the Cho/Cr ratio
was unchanged (Laubenberger et al., 1996). Conversely, another
study using patients in the early stages of HIV infection showed a
significant elevation of the Cho/Cr ratio, but with no change in the
NA/Cr ratio, suggesting that the elevation of Cho may be a marker
prior to the onset of dementia (Tracey et al., 1996). A study show-
ing changes in the NA/Cr ratio for children with AIDS, and grouped
according to the clinical parameters of encephalopathy (AE) or
nonencephalopathy CANE), showed that in the basal ganglia both
the AE and the ANE groups had reduced NA/Cr ratios compared
to controls (Pavlakis et al., 1995>, whereas in white matter only the
AE group had an NA/Cr ratio lower than the controls.
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 15/34
Applications of Proton M RS
3. The Development of Techniques to Measure
Metabolites with Coupled Spins
3.1. introduction
367
The ubiquity of the proton gives rise at one and the same time
to enormous analytical potential (because each and every metabo-
lite has a proton spectrum), as well as to serious problems of ana-
lytical discrimination (because the small chemical shift range of
the proton often leads to unmanageable overlap of metabolite
spectra, particularly at 1.5 T). The richness of the proton spec-
trum from brain is illustrated by the 2-3 ppm section of a 300-
MHz spectrum from a cat brain extract shown in Fig. 3, where the
narrowness of the lines in aqueous solution, coupled to the
enhanced chemical shift dispersion at - 7 T, give rise to a partial
resolution of all of the metabolite multiplets. However, during in
vivo application of MR spectroscopy, differences in magnetic sus-
ceptibility on a microscopic spatial scale within the tissue milieu
give rise to resonance linewidths (- 0.1 ppm) which tend to obscure
the finer chemical shift separations and particularly the multiplet
splittings. Although brute force enhancement of the chemical shift
dispersion by increasing the magnetic field strength is a viable
option in analytical applications of NMR spectroscopy in vitro, it
is not an option in vivo because of the concomitant increase in
RF
heating and because of technological difficulties in manufactur-
ing large-bore whole-body magnets capable of generating very
high magnetic fields
When trying to extract concentration information for the
severely overlapping resonances of brain metabolites that are often
less concentrated than the NA, Cr, and Cho covered in Subhead-
ing 2., four broad options are available in vivo. The first option is
to carry out a detailed numerical modeling of the whole in vivo pro-
ton spectrum (Provencher, 1993; Provencher et al., 1995; Stanley
et al., 1995), using as fitting parameters the relative metabolite
concentrations and as the basis functions, predetermined indi-
vidual metabolite spectra. The second option
1s
to move to as
high
afield strength as possible (Mason et al., 1994; Gruetter et al., 1996;
Pan et al., 1996b), in the hope that the concomitant increases in
signal-to-noise ratio (SNR) and chemical shift dispersion will
clarify the spectrum sufficiently for metabolite quantification. In
the many cases of metabolites with coupled proton spins, a third
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 16/34
362 Hanstock and Allen
N
N
G B
J
G B
I
I
I I
I
3.2
2.8 2.4 2.0 1.6
Chemical Shift ppm)
Fig. 3. A limited region 2-3 ppm) of the 300 MHz
proton spectrum
from an acid extract of cat brain.
option is to acquire more information by way of a spectvuEly two-
dimensional 2-D) spectrum Berkowitz et al., 1988; Hurd et al., 1991;
Brereton et al., 1994; Dreher et al., 1995; Ryner et al., 1995; Ziegler
et al., 1995; Kreis et al., 1996) that separates into the second NMR
dimension the unique coupling information of all the metabolites
present. The fourth option, also viable only for the metabolites
with coupled spins, is to reduce the information content of the spec-
tral acquisition by editing the one-dimensional 1-D) spectrum
Rothman et al., 1984; Dumoulin, 1985; Hetherington et al., 1985;
Williams et al., 1986; Hanstock et al., 1987; Hanstock et al., 1988;
McKinnon et al., 1988; Sotak et al., 1988; Brereton et al., 1990;
Knuettel et al., 1990; Trimble et al., 1990; Thomas et al., 1991;
Rothman et al., 1992; deGraaf et al., 1993; Rothman et al., 1993), so
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 17/34
Applications of Proton MRS
363
as to observe only a single multiplet from a single metabolite while
suppressing the signal from all but that predefined metabolite.
The technical problem of spectral discrimination is exac-
erbated by the issue of spatial encoding, which is basic to all in
vivo MRS studies. The spectral discrimination problem has in
general been addressed in the two limiting cases of either single
voxel localization or multiple voxel spatial maps spectroscopic
imaging or chemical shift imaging, CSI). In the former limit, either
a complete 1-D spectrum or an edited spectrum is acquired from
the single voxel of interest, In the latter limit, multiple 1-D spec-
tra are obtained, each localized to an individual voxel, Brown et
al., 1982; Adalsteinsson et al., 1993; Meyerhoff, 1994b; Hwang et
al., 1996) and often rendered into an image for a single peak. For
example, the phosphocreatine PCr) peak in the 31Pspectrum or
the Cho, Cr, and NA methyl singlets in the proton spectrum readily
provide metabolite images because they are strong and have a
sufficiently long T,. However, for the weaker, broad, and over-
lapping multiplets of coupled spin systems, e.g., the amino acid
neurotransmitters in the proton spectrum of brain, it is question-
able if at 1.5 T this methodology can give rise to quantitative maps
of concentration that are free from overlap artifacts. The edited
multiplet is, nevertheless, still a viable option at 1.5 T for produc-
ing either a single voxel measurement or a two-spatial-dimension
concentration map of a single coupled-spin metabolite resonance.
3.2. Metabolic Specificity
The MRS signal will be most metabolite specific if its identifi-
cation is made to depend not on the usual single chemical shift
value of one of the resonances of the metabolite of interest, but
instead on a combination of all the chemical shift differences,
together with the indirect scalar couplings, associated with as
many coupled spin multiplets of the target metabolite as possible.
3.2.1. Scalar Coupling
The indirect scalar interaction Abragam, 1961; Gunther, 1995)
is that which couples together the spins of the protons of neigh-
boring molecular groups in the target metabolite molecule, e.g.,
the proton spins of the CH, the CH,, or the CH, groups etc., each
of which has a different chemical shift value. The strength and
sign) of the scalar coupling interaction,f, which in turn determines
the multiplet splitting, is governed by the electronic structure of
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 18/34
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 19/34
Applications of Proton MRS 365
era1 metabolites whose coupling cannot be described as weak, being
strong at 1.5T, but relaxing more towards the weak limit at 4 T.
Into this category fall the important metabolites of Glu and Gln,
collectively designated Glx, as well as the A,B, system of the two
methylene groups in taurine (Tau) and the N2Q coupling in the
AM,N,Q spin system of Ins. However, because of the high cur-
rent interest in Glx, it is worth discussing this case in more detail.
The challenge with Glu and Gln arises not only because the simi-
larity of their molecular structures produces multiplet chemical
shifts that are very similar for both molecules, but also because
the steric effects in the two Glx molecular structures produce
inequivalencies within both pairs of methylene protons in those
molecules and give rise to strong negative J couplings between
the protons of each of the methylene groups. These strong cou-
plings cause the multiplet structure (and hence the overall spec-
tral lineshape) to be quite sensitive to variations in pulse sequence
(both timings and pulse shapes) and magnetic field strength.
Under such circumstances a quantitative interpretation of the spec-
tral intensity at any single frequency is not at all straightforward
and requires a detailed understanding of its origin.
3.2.2. Numerical Modeling
The identification of a metabolite, and the measurement of its
concentration, from a numerical modeling of the complete 1-D
proton spectrum relies on being able determine its individual con-
tribution to the spectrum, usually at the single chemical shift value
of one of its multiplets. It has been practiced with some degree of
success by several workers (Behar et al., 1991; Provencher, 1993;
Provencher et al., 1995; Stanley et al., 1995). However, when spec-
tral overlap occurs the spectral intensity at a single chemical shift
value is no longer a unique measure of a single metabolite and
one may have to assume that any changes observed in the inten-
sity are because of only one of the contributing metabolites chang-
ing with pathology. Greater metabolite specificity may be obtained
by seeking consistency between the changes of more than one
multiplet of the metabolite in question at each of their character-
istic chemical shift values. This makes the numerical fitting rou-
tines very dependent on a detailed understanding of the pulse
sequence dependence and the magnetic field dependence of all
the multiplet lineshapes. Even for a single metabolite, one cannot
assume that the shape and relative intensities of the multiplets
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 20/34
366 Hanstock and Allen
remain the same at all echo times, even if all those multiplets were
to have the same relaxation times. This is illustrated quite strik-
ingly in NAA (Wilman et al., 19961, by a comparison of the NA
singlet (2.02 ppm) and the strongly coupled aspartate ABX mul-
tiplet (2.6 ppm). Because of the strong coupling, the echo time
dependence of the aspartate multiplet is itself field dependent and,
moreover, at any field strength it is markedly different from that
of the uncoupled singlet, which depends only on transverse
relaxation. A lack of appreciation of this point could suggest that
the two NAA resonances were reflecting different concentrations
of NAA. The case of Glx is significantly more involved than NAA
in this regard.
3.2.3. Spectroscopy at High Field In Vivo
The luxury of a high field magnet (4 T, for example) clearly
mitigates several of the more severe difficulties associated with
strongly coupled spin systems at 1.5 T. Glx is a case in point. At
4.1 T, the team at the University of Alabama (Mason et al., 1994;
Pan et al., 1996b) have approximated the Glx response by means
of a weak couplmg approach in which the spin system is regarded
as a AM,X, system. In this approximation, the Alabama group
neglected the inequivalencies of the methylene protons on the C3
and C4 carbons CM2and X,, respectively) and assumed that 6,,/
J
- 38 and 6,x/J MX - 6.6 correspond to the weak coupling limit.
Ir?Fomparison, a full calculation at two different field strengths,
namely, 1.5 T and 4 T (Allen et al., 1997), using all the J couplings
listed in that reference, illustrates some of the consequences of
making the weak coupling approximation even at 4 T. Neverthe-
less, the Alabama group have been able to provide quantitative
estimates of Glu m a number of human subjects.
The high field strength of 4 T has also enabled Gruetter et al.
(1996) to separate from the water peak, and subsequently observe
the 5.23 ppm peak of glucose. A comparison of the normally sought
3.44 ppm glucose peak, however, showed that even at 4 T the 3.44
ppm peak is still partially overlapping a 3.49 resonance assigned
to myo-Ins.
3.2.4. 2-D Spectroscopy In Vivo
The exploitation of 2-D spectroscopic methods to unravel com-
plicated 1-D spectra is standard practice in chemical applications
of high-spectral resolution NMR (Ernst et al., 1987). For example,
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 21/34
Applicatrons of Proton MRS
367
J-resolved spectroscopy can provide a spectral map that separates
the multiplet structure from the chemical shift structure along
orthogonal axes of presentation, thereby eliminating the cause of
much overlap. COSY (Gunther, 1995>, on the other hand, gives a
2-D spectral map in which the peaks along the diagonal reflect
the 1-D spectrum and the off- diagonal peaks represent all the
connectivities. The exploitation of the same techniques in vivo
would be an ideal proposition. However, some of the constraints
of working in vivo have proved to be substantial handicaps. For
example, the extensive data acquisition period (previously 30 min
or more, but now as short as 15 min) which accommodates the
incrementation of the so-called t, interval, renders the technique
quite susceptible to subject motion, a serious issue when those
subjects are from neurodegenerative brain patient populations.
Moreover, the longer values of the f, increment that are needed to
provide the appropriate sweep width at in vivo field strengths,
push out the total t, periods to values that can be comparable to or
greater than T, for metabolites in vivo. Signal loss due to trans-
verse relaxation is therefore also a significant problem in vivo.
Nevertheless in certain casesJ-resolved spectroscopy has had some
success at 1.5 T, as recently demonstrated both in human brain
(Ryner et al., 1995) and human muscle (Kreis et al., 1996).
3.2.5. Spectral Editing In Vivo
The question of whether or not to edit for a particular metabolite
resonance is one whose answer is dependent both on the metabolite
and on the magnetic field strength available. Overlap is the key
issue and even in the midst of some very crowded spectral regions,
e.g., 2.0 to 3.0 ppm and 3.3 to 4.3 ppm, the decision of whether or
not to edit is a subjective one. Once the decision to edit has been
made, the criteria by which a viable editing sequence must be judged
are as follows. First, the sequence must provide excellent back-
ground discrimination against overlapping resonances. Second, It
must be sufficiently fast to have a low vulnerability to motion arti-
facts, and third, the sequence length must be sufficiently short to
ensure that all editing and localization procedures can be accom-
plished well within T, in order to preserve signal strength.
A simple form of editing and one which has been used very
successfully in the weak coupling limit by the Yale group
(Rothman et al., 1984; Hetherington et al., 1985; Hanstock et al.,
1988; Rothman et al., 1992; Rothman et al., 19931, is that of difeer-
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 22/34
368
Hanstock and Allen
ence spectroscopy. Their principal applications have been to the lac-
tate AX, system and the A,M,X, spin system of GABA. Although
this method takes advantage of the existence of scalar coupling
between groups within the target metabolite molecule, its metabo-
lite specificity stems primarily from the uniqueness of the two
chemical shift values associated with the two coupled multiplets.
The Yale group has combined difference spectroscopy with sur-
face coil (Bendall et al., 1985) and with ISIS (Ordidge et al., 1986)
localization techniques, as well as with several water-suppression
strategies (Rothman et al., 1993). The method is highly metabolite
specific and the intrinsic signal loss from the method is small when
the multiplicity is low. Its strength is in its simplicity, and it has
been applied most notably to groups of epilepsy patients to moni-
tor the efficacy of GABA-enhancing drugs (Petroff et al., 1995;
Petroff et al., 1996). It is, however, quite vulnerable to background
subtraction artifacts arising from patient motion between sub-
tracted scans, hardware instabilities, and minor differences in spin
dynamics owing to differences in the two pulse sequences. Cer-
tain arbitrary adjustments in one spectral intensity have been used
(Rothman et al., 1993) to optimize background cancellation. As a
result the efficacy of the singlet background elimination is much
more modest than that of the multiple quantum filters treated
below. Other groups have also used this technique to monitor
GABA (Preece et al., 1995; Keltner et al., 19961, though the latter
reference incorporates difference spectroscopy into a PRESS
sequence. The PRESS variant of difference spectroscopy has also
been proposed for lactate editing (Bunse et al., 1995). A variant of
the difference method (reported for a Glx phantom-only experi-
ment [Lee et al., 19951) uses differential transverse relaxation at 4
T to provide the difference between two spectra. It is based on the
estimation of a particularly short T, for Glx (-50 ms), in contrast
to the T,s (-few hundred milliseconds) of other metabolites that
are present in the 2.00-3.00 ppm range of the proton spectrum.
An alternative to the dfirence spectroscopy approach is multiple
qtlantum coherencefiltering (Ernst et al., 1987; Gunther, 1995; Lee et
al., 1995). By using magnetic field gradients to filter out all but a
single order of multiple quantum coherence (MQC), the goal is to
produce a “single shot” editing method, which in addition to being
highly metabolite specific, is also far less vulnerable to patient
motion than editing methodologies that subtract successive scans.
The term “single shot” does not exclude signal averaging. It is
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 23/34
Applrcations of Proton MRS 369
simply meant to convey the notion that all information is obtained
from a single sequence. MQC filters also provide far greater back-
ground discrimination against uncoupled spin magnetization,
such as the intense singlets of water, NA at 2.02 ppm, Cr at 3.02
ppm, and Cho at 3.2 ppm, which can easily be made as much as
three orders of magnitude in phantoms (McKinnon et al., 1988;
Wilman et al., 1993). In vivo, however (Keltner et al., 19971, sin-
glet suppression has fallen far short of this.
Although MQC filters surpass difference editing in several
respects, they clearly suffer from weaknesses of their own. For
example, one weakness is the potential signal loss associated with
an MQC filter. This signal loss can arise for two main reasons.
The first is because of the limited inherent yield of the filter. The
more coherences there are to share the spin information, the
smaller the magnetization that can be derived from any one. The
second loss mechanism is transverse relaxation, the seriousness
of which stems from the short T,s of metabolites in vivo relative
to the length of the filter sequence. Another weakness, and one
shared with difference editing, is the difficulty in suppressing
unwanted coupled-spin background when specific multiplets of
the target metabolite spectrum cannot be excited selectively with-
out exciting background multiplets at the same time. This is a prob-
lem that is worse at low fields (1.5 T), and for backgrounds arising
from larger groups of coupled spins. Finally, because of the reli-
ance on the careful manipulation of coherences by the RF pulses,
MQC editing in vivo is probably more demanding of
RF
pulse
integrity than is difference editing. The issue of B, inhomogeneity
caused by surface coil transmission (Shen et al., 1991) has been
thoroughly dealt with by Garwood and coworkers (Garwood et
al., 1991; deGraaf et al., 1995a,b) through the development of adia-
batic pulses. The issue of self-refocussing in a spatially uniform
B, in order to maintain the relative phases of all coherences was
described by Geen and Freeman (Geen et al., 1991).
Procedures to mitigate the weaknesses mentioned above can
be illustrated by reference to two problems that have dominated
the MQC filter literature over the last decade. The first is the mea-
surement of lactate in the presence of a lipid background and the
second is the measurement of the amino acid neurotransmitters.
Lactate measurement benefits from a simple weakly coupled spin
system, a long T, (Blamire et al., 1994; vanderToorn et al., 19951,
and a chemical shift difference between the coupled methine and
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 24/34
370 Hanstock and Allen
methyl groups (2.78 ppm or -178 Hz at 1.5 T), which facilitates
selective excitation of the methine spins without perturbation of
the lipid spins. Because of lactate’s simple coupled-spin system,
it has been possible to derive a procedure to deal with the inher-
ent yield problem (Trimble et al., 1990), which combines different
orders of coherence and recoups the full magnetization, thereby
giving an inherent yield of unity. To improve background lipid
suppression and maintain inherent yield, this procedure has been
incorporated into a series of sequences elegantly exploiting even
more spectral selectivity by the Johns Hopkins’ group (He et al.,
199513; 1996) and applied at 4.7 T to the measurement of lactate
and Iproplatin in tumors in rats and mice (He et al., 1995a). How-
ever, with the amino acid neurotransmitters, the coupled-spin
systems are more complex, having a greater number of coupled
spins as well as strong coupling in several cases. Nevertheless,
strategies for mitigating signal loss due to both inherent yield and
to transverse relaxation have been proposed for the weakly
coupled GABA (Wilman et al., 1995b) and the strongly coupled
Glx (Thompson et al., 1997) and demonstrated in vivo on the nor-
mal human brain (Keltner et al., 1997; Thompson et al., 19971,
though the level of the success falls short of that achieved in the
simple lactate case.
The performance of MQC filters on phantoms is in little doubt,
largely because of the narrow linewidths and long T,s. The trans-
lation of this performance to an in vivo capability has not yet been
so well demonstrated. The crux of the matter seems to be the
incorporation of spatial encoding into the filter sequence without
undermining the filter specificity and sensitivity.
When the spin system is amenable to its use, e.g., lactate and
GABA, difference spectroscopy is simple and easy to use. Bearing
in mind its vulnerability to small patient movements, it provides
a broad measure of metabolite changes due to pathology or drug
therapy. When the spin system does not provide well-separated
multiplets, which are also weakly coupled, MQC filtration looks
much more promising. However, it is ironic that as one pushes
the MQC filter to the most demanding of tasks, e.g., Glx with its
coupled spin background, one finds oneself taking refuge in tech-
niques such as spectral modeling, which one originally developed
the filter to avoid. Nevertheless, it should be realized that after
MQC filtration, the residual spectrum contains many fewer com-
ponents than the unfiltered spectrum and, moreover, it is the larg-
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 25/34
Applications of Proton MRS
est of the background peaks (i.e., the singlets) that are suppressed
most efficiently.
Acknowledgments
The authors are grateful to the Medical Research Council of
Canada for ongoing support of their spectral editing program.
References
Abragam, A (1961) Electron-nucleus interactions, in The principles @ nuclear
mugnetrsm. Oxford Umversrty Press, Oxford, pp. 159-215.
Ackerman, J H , Grove, T H , Wong, G G., Gadlan, D G , and Radda, G K
(1980) Mapping of metabolites m whole animals by 31PNMR using surface
coils Nature (London) 283, 167-170
Adalstemsson, E , Spielman, D M , Wright, G A, Pauly, J M , Meyer, C H ,
and Macovskr, A. (1993) Incorporatmg lactate/lipid drscrrmination mto a
spectroscoprc image sequence Map Reson Med 30,124-130
Allen, P S , Thompson, R. B , and Wilman, A H (1997) Metabolic-specific NMR
spectroscopy m vlvo NMR in Btomedicine 10,435-444
Arnold, D. L., Matthews, P M , Francis, G., and Antel, J (1990a) Proton magnetic
resonance spectroscopy of human brain m vrvo m the evaluation of multiple
sclerosis. Assessment of the load of drsease Map Reson Med 14,154-159
Arnold, D. L , Shoubrrdge, E A, Villemure, J. G., and Femdel, W (1990b) Proton
and phosphorus magnetic resonance spectroscopy of human astrocytomas m
viva Prehminary observatrons of tumor grading. NMR m Blamed 3,184-189
Arnold, D L , Matthews, P M, Gordon, F S, O’Connor, J , and Antel, J I?.
(1992) Proton
magnetic
resonance spectroscoprc rmagutg for metabobc
characterisation of demyelmatmg plaques Ann Neurol 31,235-241
Arnold, D L , Riess,G T , Matthews, P M , Francis, G S , Collins, D L , Wolfson,
C , and Antel, J P. 1994)Useof proton magnetic resonancespectroscopy for
monitoring disease rogressron m multrple sclerosis Ann Neural 36,76-82
Ashwal, S , Holshouser, B A , Tomasr,L. G , Shu, S , Perkm, R M , Nystrom G
A, and Hmshaw, D B 1997) ‘H-Magnetrc resonancespectroscopy-deter-
mined cerebral lactate and poor neurological outcome in children with cen-
tral nervous system disease.Ann Neural. 41,470-481.
Athey, T W 1992) Current FDA guidance for MR patient exposure and con-
siderations for the future PYOC. Nat1 Acad. SCI USA 649,242-257
Barker, I’ B , Ghckson, J D , and Bryan, R N. 1993) In vivo magnetic reso-
nance spectroscopy of human brain tumors. Toptcs zn Magnettc Resonance
Imaging 5,32-45
Behar, K L., Hollander, J A. d., Stromski, M. E , Ogmo, T, Shulman, R G ,
Petroff, 0 A C., and Prrchard, J W. 1983)High-resolution ‘H nuclear mag-
netic resonancestudy of cerebral hypoxia m vivo
PYOC
Nat1 Acad SCI USA
80,4945-4948
Behar, K L and Ogino, T 1991)Assignment of resonancesm the ‘H spectrum
of rat bram by two-dimensional shift correlated and J-resolved NMR spec-
troscopy. Map ResonMed 17,285-303
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 26/34
372
Hanstock and Allen
Bendall, M R and Pegg, D T (1985) Theoretical description of depth pulse
sequences, on and off resonance, mcludmg improvements and extensions
thereof. Magn Reson Med 2,91
Berkowitz, B A , Wolff, S D., and Balaban, R S. (1988) Detectron of metabohtes
m vlvo using 2D proton homonuclear correlated spectroscopy J Magn Reson
79,547-553
Berridge, M J and Irvine, R F (1989) Inositol phosphates and cell signaling
Nature 341,197-205
Birken, D. L and Oldendorf, W H (1989) N-acetyl-L-aspartic acid A literature
review of a compound prominent in ‘H-NMR spectroscopic studies of brain
Neuroscr Blobehav Rev 13,23-31
Blakely, R D. (1988) The neurobiology of N-Acetyl-aspartic Acid* A literature
review of a compound prominent in ‘H-NMR spectroscopic studies of the
brain. Neurobrol 30,39-100
Blamue, A M , Graham, G D., Rothman, D. L , and Prichard, J W. (1994) Pro-
ton spectroscopy of human stroke assessment of transverse relaxation times
and partial volume effects m single volume Steam MRS. Map Reson Imag-
ing 12,1227-1235.
Bottomley, P A (1987) Spatial locahzatlon m NMR spectroscopy in vwo Ann
N Y Acad. Scl 508,333-348
Brand, A., Richter-Landsberg, C , and Liebfritz, D (1993) Multmuclear NMR
studies on the energy metabolism of gl ial and neuronal cells Dev Neurosct
15,289-298.
Brenner, R. E , Munro, P M , Wllhams, S C , Bell, J D , Barker, G. J , Hawkins,
C I’, Landon, D N., and McDonald, W I (1993) The proton NMR spectrum
m acute EAE The significance of the change in the Cho Cr ratio Magn Reson
Med 29,737-745
Brereton, I M., Rose, S E , Galloway, G J , Moxon, L N , and Doddrell, D M
(1990) In viva volume selective metabohte editing via correlated Z-order
Magn Reson Med 16,460-469.
Brereton, I. M , Galloway, G. J , Rose, S E., and Doddrell, D M (1994) Local-
ized two-dimensional shift correlated spectroscopy m humans at 2 Tesla
Magn Reson Med 32,2X-257
Brown, T. R , Kmcald, B. M , and Ugurbrl, K (1982) NMR chemical shift imag-
ing in three dimensions.
Proc
Nat1 Acad Scz USA 79,3523-3526
Bruhn, H., Frahm, J , Merboldt, K. D , Hanrcke, W., Hanefleld, F., Christen, H
J , Kruse, B , and Bauer, H. J (1992) Multiple sclerosis m children Cerebral
metabolic alterations monitored by locahsed proton magnetic resonance spec-
troscopy m vivo. Ann Neurol 32,140-150.
Bunse, M , Jung, W.-I., Schick,F., Dietze, G J., and Lutz, 0 (1995) HOPE, a new
lactate editing method J Magn Reson BlOY, 270-274
Burri, R , Steffan, C., and Herschkowrtz, N (1991) N-Acetyl-L-Aspartate IS a
malor source of acetyl groups for lipid synthesis during rat brain develop-
ment Dev Neuroscl 13,403-411
Carpinelh, G , Carapella, C M., Palombi, L , Raus, L , Caroli, F , and Podo, F
(1996) Differentiation of glioblastoma multiforme from astrocytomas by m
vitro ‘H MRS analysis of human brain tumors Antrcancer Res 16,1559-1563
Chang, L , Ernst, T , Poland, R. E , and Jenden, D. J. (1996) In viva proton mag-
netic resonance spectroscopy of the normal aging human brain Life Scrences
58,2049-2056
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 27/34
Applications of Proton M RS
373
Chong, W K , Sweeney, B., Wilkinson, I D , Paley, M , Hall-Craggs, M A,
Kendall, B. E., Shepard, J, K, Beecham, M., Miller, R. F., and Weller, I. V.
(1993) Proton spectroscopy of the brain in HIV infection correlation with
clinical, immunologic, and MR imaging findings Radzology 188, 119-124
Confort-Gouny, S., Vion-Dury, J , Nicoli, F., Dano, I’., Donnet, A, Grazziam, N.,
Gastaut, J. L , G&oh, F , and Cozzone, P J (1993) A multiparametric data analysis
showing the potential of locahsed proton MR spectroscopy of the brain m the
metabolic defuution of neurological diseases. J AJeurol Ser.118,123-133
Cwik, V A., Hanstock, C C., Allen, P. S., and Martin, W. R. W. (1998) Estima-
tion of brainstem neuronal loss m amyotrophic lateral sclerosis with in viva
proton magnetic resonance spectroscopy. Neural 50,72-77.
Davie, C A., Hawkins, C I’., Barker, G. J., Brennan, A., Tofts, P S., Miller, D. H
and McDonald, W I (1994) Serial proton magnetic resonance spectroscopy
in acute multiple sclerosis lesions Brain 117,49-58
deGraaf, A A., Luyten, P. R., Hollander, J. A. d., Heindel, W., and Bovee, W M.
M J (1993) Lactate imaging of the human brain at 1.5T usmg a double quan-
tum filter Magn Reson. Med. 30,231-235.
deGraaf, R A, Luo, Y., Terpstra, M , and Garwood, M (1995a) Spectral edrtmg
with adiabatic pulses J Magn Resort. B109, 184-193.
deGraaf, R A., Luo, Y, Terpstra, M, Merkle, H, and Garwood, M (1995b) A new
localization method usmg an adiabatic pulse, BIR-4 J Magn Reson B106,245-252
Dixon, W T (1984) Simple proton spectroscopic imaging. Radrology 153,189-194
Dreher, W. and Leibfritz, D (1995) On the use of two-dimensional-J NMR mea-
surements for m viva proton MRS measurement of homonuclear decoupled
spectra without the need for short echo times. Magn. Reson Med 34,331-337
Dumouhn, C L (1985) The application of multiple quantum techniques for the
suppression of water signals in ‘H NMR spectra. 1. Magn Reson 64,38-46
Ernst, R R , Bodenhausen, G., and Wokaun, A. (1987) Prmcrples of nuclear mag-
netic resonance in one and two dimensions Clarendon Press, Oxford, U.K.
Florian, C. L, Preece, N E , Bhakoo, K. K , Williams, S. R, and Noble, M. D
(1995) Cell type-specific fingerprmtmg of menmgroma and menmgeal ceils
by proton nuclear magnetic resonance spectroscopy Cancer Res. 55,420-427.
Frahm, J., Bruhn, H, Gyngell, M L , Merboldt, K. D, Hanicke, W , and Sauter, R
(1989a) Localised high-resolution proton NMR spectroscopy using stimulated
echoes Initia l apphcations to human brain in vivo Magn Reson Med 9,79-93
Frahm, J., Bruhn, H , Gyngell, M. L , Merboldt, K. D., Hanicke, W , and Sauter,
R. (1989b) Locahsed proton NMR spectroscopy in different regions of the
human brain in vivo Relaxation times and concentrations of cerebral me-
tabohtes. Magn Reson Med 11,47-63.
Gadian, D G., Wilhams, S R, Bates, T E, and Kaupine, R. A (1993) Brain
damage studied by NMR and other methods. NMR spectroscopy, current
status and future possibilities. Acta Neurochzr. 57 1-S.
Garwood, M. and Ke, Y. (1991) Symmetric pulses to induce arbitrary flip angles
with compensation for RF mhomogeneity and resonance offsets. 1 Magn.
Reson 94,511-525.
Geen, H and Freeman, R (1991) Band selective radio frequency pulses J Magn
Reson 93,93-141
Gideon, I’., Sperling, B., Arlien-Soborg, P , Olsen, T. S., and Henriksen, 0 (1994)
Long-term follow-up of cerebral infarction patients with proton magnetic
resonance spectroscopy Stroke 25 967-973
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 28/34
374 Hanstock and Allen
Gordon, R E and Ordldge, R J (1984) Volume selection for high resolution
NMR studies Proc Sot Map ResonMed 272
Graham, G D , Blamire, A M, Howseman, A. M., Rothman, D. L., Fayad, I’ B ,
Brass, L M , Petroff, 0 A , Shulman, R. G , and Prichard, J W (1992) Proton
magnetic resonance spectroscopy of cerebral lactate and other metabohtes
in stroke patients Stroke23,333-340.
Graham, G D , Blamve, A M , Rothman, D L , Brass, L M , Fayad, P B , Petroff,
0 A, and Prichard, J W (1993) Early temporal variation of cerebral
metabolites after human stroke. A proton magnetic resonance spectroscopic
study
Stroke
24,1891-1896
Gredal, 0, Rosenbaum, S , Topp, S , Karlsborg, M., Strange, I’, and Werdelm,
L. (1997) Quantification of brain metabohtes m amyotrophic lateral sclerosis
by localised magnetic resonance spectroscopy Neural 48,878-881
Gruetter, R , Rothman, D L , Novotny, E. J , Shulman, G I, Prlcard, J W , and
Shulman, R G (1992) Detection and assignment of the glucose signal m Hl
NMR difference spectra o f the human brain. Magn ResonMed 27,183-188
Gruetter, R , Garwood, M , Ugurbil, K , and Seaqulst, E R (1996) Observation
of resolved glucose signals m ‘H NMR spectra of the human bram at 4 tesla
Magn ResonMed 36, -6
Gunther, H (1995) Two-dimensional nuclear magnetic resonance spectroscopy,
m
NMR SpectroscopyBasicprinciples, conceptsand applzcationsn chemistry
Wiley, Chichester, U K , pp 273-334
Haase, A, Frahm, J , Hanicke, W ,and Matthew, D (1985) ‘H NMR chemical
shift selective (CHESS) imaging.
Phys Med Bzol 30,341-344
Hanstock, C C , Bendall, M R , Hethermgton, H P , Boisvert, D P , and Allen,
P S (1987) Localized in vivo proton spectroscopy using depth pulse spec-
tral editing 1 Magn Reson71,349-354
Hanstock, C C , Rothman, D L, Prichard, J W , Jue, T, and Shulman, R G
(1988) Spatially localized ‘H NMR spectra of metabohtes m the human brain
Proc Nat1 Acad Scl USA 85,1821-1825
Hanstock, C C , Cwik, V A, Martin, W R W , Boyd, C , Brooke, M H , and Allen,
I’ S (1997) Brain stem and motor cortex neuronal loss m amyotrophic lateral
sclerosis (ALS) as measured by IH MRS
Proc Int Sot Magn ResonMed 1187
Hays, C. E , Edelstein, W. A., Schenck, J. F., Mueller, 0 M , and Eash, M (1985)
An efficient, highly homogeneous radiofrequency coil for whole-body NMR
imaging at 1.5T J Magn Reson63,622-628.
He, Q , Bhulwalla, Z M , Maxwell, R. J , Griffiths, J R , and Ghckson, J D (1995a)
Proton NMR observation of the antmeoplastic agent Iproplatm m viva by
selective multiple quantum coherence trasnfer (Sel-MQC)
Magn ResonMed
33,414-416
He, Q, Shungu, D C., 21~1, P C M. v., Bhulwalla, Z M, and Ghckson, J D
(199513) Single scan m viva lactate editing with complete lipid and water
suppression by selective multiple quantum coherence transfer (Sel-MQC)
with application to tumors. y
Magn Reson
B106,203-211
He, Q , Bhujwalla, Z M , and Glickson, J D (1996) Proton detection of cholme
and lactate m EMT6 tumors by spm echo-enhanced selective multiple quan-
tum coherence transfer J
Magn Reson
B112,18-25
Hethermgton, H I’, Avison, M J , and Shulman, R G (1985) ‘H homonuclear
edltmg of rat brain using semi selective pulses Proc Nat1 Acad Scz USA 82,
3115-3118
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 29/34
Appilcations of Proton MRS
375
Hethermgton, H I’., Mason, G F., Pan, J. W, Ponder, S L., Vaughan, J. T , Tweig,
D B , and Pohost, G. M (1994a) Evaluation of cerebral gray and white matter
metabolite differences by spectroscopic imaging. Magn Reson Med 32,565-571
Hethermgton, H. P , Pan, J W , Mason, G F, Ponder, S. L ,Tweig, D B , Deutsch,
G , Mountz, J., and Pohost, G M (1994b) 2D ‘H spectroscopic imaging of the
human brain at 4 1T Magn Reson. Med. 32‘530-534
Hethermgton, H P , Pan, J, W , Mason, G F , Adams, D., Vaughn, M. J , Tweig,
D B , and Pohost, G. M (1996) Quantitative ‘H spectroscopic imaging of
human brain at 4.1T usmg image segmentation. Magn Reson Med 36,21-29
Hugg, J. W , Dmjn, J H , Matson, G B., Maudsley, A. A, Tsuruda, J S , Gelinas,
D F , and Weiner, M. W (1992) Laterahzation of human focal eprlepsy by P-
31 magnetic resonance spectroscopic imaging Neurology 42,2011-2018.
Hugg, J W , Laxer, K. D, Matson, G B, Maudsley, A A, and Weiner, M W
(1993) Neurons loss localizes focal epilepsy by proton MR spectroscopic
imaging Ann. Neural 34,788-794
Hurd, R. E. and Freeman, D (1991) Proton editing and imaging of lactate NMR
in Blamed 4,73-80
Husted, C A, Goodm, D. S., Hugg, J. W., Maudsley, A A, Tsuruda, J S , DeBie,
S H , Fem, G , Matson, G D , and Werner, M W. (1994) Biochemical alter-
ations m multiple sclerosis lesions and normal appearing white matter
detected by m viva 31Pand ‘H spectroscopic imaging Ann. Neural 36,157-165.
Hwang, J.-H., Graham, G D , L.Behar, K , Alger, J R , Prichard, J W , and
Rothman, D L (1996) Short echo time proton magnetic resonance spectro-
scopic imaging of macromolecule and metabohte signal intensities in human
brain Magn Reson Med 35,633-639
Ikeda, Y and Lond, D M (1990) Molecular basis of brain injury and brain edema
the role of oxygen and free radicals Neurosurgery 27, l-11
Jones, A. I’., Gunawardena, W J, Coutinho, C M. A, Gatt, J A., Shaw, I C ,
and Mitchell, J D (1995) Prehmmary results of proton magnetic resonance
spectroscopy m motor neuron disease (amyotrophic lateral sclerosis) Neuvol
Scz 129 (Suppl), 85-89
Kamada, K , Houkin, K., Hida, K , Matsuzawa, H , Iwasaki, Y , Abe, H., and
Nakada, T (1994a) Localised proton spectroscopy of focal bram pathology
in humans, significant effects of edema on spin-spin relaxation time Magn.
Reson Med 31‘537-540.
Kamada, K , Houkin, K , Iwasaki, Y, Abe, H , and Kashiwaba, T (1994b) In
vivo proton magnetic resonance spectroscopy for metabohte changes of
human brain edema Neurolgla Medzco-Chirurgxa 34,676-681
Kay, L and McClung, R E (1988) Product operator description of AB and ABX
spin systems ] Magn Reson 77,258-273
KeItner, J. R , Wald, L W., Christensen, J. D., Maas, L. C., Moore, C. M., Cohen,
B M, and Renshaw, I’. R. (1996) A technique for detecting GABA m the
human brain with PRESS localization and optimized refocussmg spectral
editing radiofrequency pulses Magn Reson Med 36,458-461.
Keltner, J R., Wald, L L , Frederick, B d B., and Renshaw, P F (1997) In viva
detection of GABA m human brain using a localized double quantum filter
technique Magn Reson Med 37,366-371.
Klunk, W., Panchalingen, K , Moossy, J., McClure, R , and Pettegrew, J (1992)
N-Acetyl-L-aspartate and other amino acid metabohtes in Alzheimer’s dis-
eased brain. A prelimmary proton nuclear magnetic resonance study Neuvol
42‘1578-1585
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 30/34
376 Hanstock and Allen
Knuettel, A and Klmmlch, R (1990) A phase sensitive single scan method for
volume selective editing of NMR signals using cyclic polarization transfer
m vlvo determination of lactate I Magn Reson86, 253-263.
Kotitschke, K., Jung, H., Nekolla, S , Haase, A , Bauer, A., and Bogdahn, U (1994)
High-resolution one- and two-dimensional IH MRS of human brain tumor
and normal gllal cells. NMR in Biomed 7, 111-120
Kreis, R and Ross, B D (1992a) Cerebral metabolic disturbances in patients
with subacute and chronic diabetes melitus detection with proton MR
spectroscopy
Radzology
84,123-130.
Kreis, R., Ross, B. D., Farrow, N A , and Ackerman, Z (1992b) Metabolic disor-
ders of the brain u-t chronic hepatic encephalopathy detected with H-l MR
spectroscopy
Radzology
82,19-27
Krels, R., Ernst, T ,and Ross, B D (1993a) Absolute quantltatlon of water and
metabohtes m human brain II Metabohte concentrations 7 Magn Reson
B102,9-19.
Kreis, R , Ernst, T , and Ross, B D (1993b) Development of the human brain In
vlvo quantlflcatlon of metabolite and water content with proton magnetic
resonance spectroscopy. Magn ResonMed 30,424-437
Krels, R (1994) Quantltatlon m locahzed proton MR spectroscopy Proc Sot
Magn Reson I,3
Kreis, R and Boesch, C (1996) Spatially localized, One and Two dlmenslonal NMR spec-
troscopy and In Vlvo application to human muscle 1
Magn Reson
113,103-118
Kugel, H , Hemdel, W , Ernestus, R, Bunke, J., duMesm1, R , and Friedmann,
G (1992) Human bram tumors. spectral patterns detected with locahsed H-
1 MR spectroscopy
RadIology
183,701-709
Kunnecke, B , Cerdan, S , and Seelig, J (1993) Cerebral metabohsm of [1,2-
‘3C21glucose and [U-‘3C,13-hydroxybutyrate m rat brain as detected by 13C
nmr spectroscopy NMR in Blamed ,403-411.
Kwo-On-Yuen, P , Newmark, R., Budinger, T., Kaye, J , Ball, M , and Jagust, W
(1994) Brain N-acetyl-L-aspartic acid in Alzhelmer’s disease A proton mag-
netic resonance study. Brain Res 667,167-174.
Laubenberger, L , Haussmger, D , Bayer, S , Thielemann, S , Schneider, B ,
Mundmger, A, Hennig, J , and Langer, M (1996) HIV-related metabolic
abnormahties in the bram. depiction with proton MR spectroscopy with short
echo times.
Radiology
199,805~810
Lee, H K , Yaman, A , and Nalcloglu, 0 (1995) Homonuclear J-refocussed spec-
tral editing technique for quantiflcatlon of glutamme and glutamate by ‘H
NMR spectroscopy. Magn. Reson.Med 34‘253-259
Lopez-Vlllegas, D , Lenkinskl, R. E , Wehrh, S L., Ho, W Z , and Douglas, S D
(1995) Lactate production by human monocytes/macrophages determined
by proton MR spectroscopy Magn. ResonMed 34,32-38
Mason, G. F, Pan, J W., Ponder, S L, Twieg, D B , Pohost, G M , and
Hethermgton, H P (1994) Detectlon of brain glutamate and glutamme m
spectroscopic images at
4
1T
Magn ResonMed 32,142-145
Matalon, R, Kaul, R , Casanova, J , Mlchals, K , Johnson, A, Rapm, I, Gashkoff,
P , and Deanchmg, M (1989) Aspartoacylase deficiency The enzyme defect
m Canavans disease J Inher Metab DLS12,329-331
Matthews, P , Francis, G , Antel, J., and Arnold, D (1991) Proton magnetic reso-
nance spectroscopy for metabolic characterisation of plaques m multiple scle-
rosls
Neural 41,1251-1256
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 31/34
Applications of Proton M RS
Maudsley, A. A., Hrlal, S K., Perman, W. H , and Simon, H. E (1983) Spatial ly
resolved high resolutron spectroscopy by four-dimensional NMR I Mugn
Reson
51,147-152
McClung, R. E. and Nakashima, T 1986)Simulation of 2D NMR spectra using
product operators in a sperical basis 1.
Magn Reson 70,187-203
McKmnon, G. C. and Boesiger,P. 1988)A one shot lactate editing sequence or
locahzed whole body spectroscopy Magn ResonMed 8,355-361
Menon, D K., Baudoum, C. J., Tomlinson, D., and Hoyle, C 1990) Proton MR
spectroscopy and imaging of the brain in AIDS Evrdence of neuronal loss m
regions that appear normal with imaging J CompufAss Tomog 14,882-885.
Meyerhoff, D. J , Mackay, S., Constans, J.-M., Norman, D., VanDyke, C., Fein,
G., and Werner, M. W. 1994a)Axonal mlury and membrane alterations in
Alzheimer’s diseasesuggestedby m viva proton magnetic resonancespec-
troscopic imaging.
Ann. Neural. 36,40-47
Meyerhoff, D J. 1994b)Magnetic resonance pectroscopicmagmg,mNMR in Physz-
ology
and Blomedlcine
R. J. Gillies) Academic Press,SanDiego, pp. 169-184.
Michaelis, T , Merboldt, K D , Bruhn, H., Hanicke, W., and Frahm, J 1993)
Absolute concentrations of metabohtes m the adult human brain m viva
quantification proton MR spectra.
Radzology 187,219-227
Miller, D, Austin, S., Connelly, A., Youl, B., Gadian, D., and McDonald, W.
1991)Proton magnetic resonancespectrocopy of an acute and chronic lesson
m multiple sclerosesLancef 337, 58-59.
Moffett, J R, Namboodiri, M A. A., and Neale, J H 1993) Enhanced
carbodirmide fixation for immunohrstochemistry: Apphcatron to the com-
parative distributions of N-acetylaspartylglutamate and N-acetylaspartate
rmmunoreactivitres m rat brain JHtsfochem Cyfochem 41,559-570
Negendank, W. G , Sauter, R , Brown, T R., Evalhoch, J L., Falmr, A, Gotsis, E.
D , Heerschap, A., et al 1996) Proton magnetic resonancespectroscopy m
pateints with glial tumors a multicentre study J Neurosurg 84,449-458
Ordrdge, R J , Connelly, A., and Lohman ,J. A. 8. 1986) mage selected n Vrvo
spectroscopy ISIS. A new technique for spatially selective NMR spectros-
copy J Magn Reson 66,283-294
Ott, D , Henrug, J., and Ernst, T 1993)Human brain tumors. assessmentwith
in vivo proton MR spectroscopy. Radiology 186,745-752
Pan, J W., Hethermgton, H I’., Vaughn, J. T, Mitchell, G , Pohost, G M , and
Whitaker, J. N. 1996a)Evaluation of multiple sclerosisby ‘H spectroscopic
Imaging at 4 1T
Magn Reson Med 36, 72-77
Pan, J W , Mason, G F , Pohost, G M., and Hetherington, H. P 1996b) Spec-
troscopic imaging of human brain glutamate by water supressed -refocussed
coherence transfer at 4 1T.Magn Reson Med 36,7-12
Patt, S L and Sykes, B D 1972) Solvent suppressionusing the WEFT water
eliminated Fourier transform) method ]
Chem Phys 56,3182
Pavlakis, S G , Lu, D , Bakshr, S , Pahwa, S , Barnett, T A., Porricolo, M E ,
Gould, R J , Nozyce, M L , and Hyman, R. A. 1995) Magnetic resonance
spectroscopy m childhood AIDS encephalopathy
Pedzaf Neural
12,277-282
Peeling, J and Sutherland, G. 1992)High-resolution IH NMR spectroscopy stud-
ies of extracts of human cerebral neoplasms.Magn Reson Med. 24,123-136
Penn, A M W , Hanstock, C C , Lr, Y., and Zhu, G. 1997) Correlation of ‘H
magnetic resonancespectroscopy with quantrtatrve hrstology and outcome
in coma Neuvol , in review
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 32/34
378
Hanstock and Allen
Petroff, 0 A C. (1989) High-field proton magnetic resonance spectroscopy of human
cerebrum obtained during surgery for epilepsy Neurology39,1197-1202
Petroff, 0. A C , Graham, G. D , Blamrre, A M, al-Rayess, M , Rothman, D L ,
Fayad, P B , Brass, L M., Shulman, R G , and Prichard, J. W (1992) Spectro-
scopic imaging of stroke m humans histopathology correlates of spectral
changes. Neurology 12,1349-1354
Petroff, 0 A C , Rothman, D L , and Behar, K L (19941 Metabohte and macro-
molecule changes withm the MS plaque measured m vi va with serial ‘H NMR
spectroscopy
Proc Sot Magn Reson 2,586
Petroff, 0 A C , Rothman, D L , Behar, K L , and Mattson, R H (1995) Irutial
observations on ef fect of vigabatrin on in viva ‘H spectroscopic measure-
ments of y-Aminobutyric Acid, Glutamate and Glutamine m human brain
Epzlepsza 6, 457-464
Petroff, 0 A C , Rothman, D L , Behar, K L , Lamoureux, D , and Mattson, R
H (1996) The ef fect of Gabapentm on brain GABA m patients with epilepsy
Ann Neurol 39,95-99
Pioro, E , Antel, J , Cashman, N , and Arnold, D (1994) Detection of cortical
neuron loss m motor neuron disease by proton magnetic resonance spectro-
scopic imagmg in vivo
Neurol 44,1933-1938
Preece, N E , Willlams, S R, Jackson, G , Duncan, J S , Houseman, J , and
Gadian, D G (1991) 1H NMR studies of vlgabatrin induced increase m cere-
bral GABA
Proc SMR, 20th Annual Meeting, San Franczsco,USA 1000
Preece,
N. E , Jackson, G. D., Houseman, J , Duncan, J S, and Williams, S R
(1995) Nuclear magnetic resonance detection of increased cortical GABA m
the vigabatrm treated rat u-r vivo
Epzlepsia 5,431-436
Preul, M , Caramanos, Z., Collins, D., Villemure, J , Leblanc, R, Olivier, A,
Pokrupa, R , and Arnold, D (1996) Accurate, nonmvasive diagnosis of human
bram tumors by using proton magnetrc resonance spectroscopy
Nature Medz-
czne
2,323-325
Provencher, S W (1993) Estimation of metabohte concentrations from local-
ized m viva proton NMR spectra Magn ResonMed 30,672-679
Provencher, S W , Harucke, W , and Micheahs ,T (1995) Automated quantitatlon
of localized ‘H MR spectra in vrvo capabilities and hmitatrons Proc Sot
Map ResonMed 3,1952
Pyket,t I L and Rosen, B. R (1983) Nuclear magnetic resonance m vlvo proton
chemical shift imaging Radiology149,197-201
Ross, B D (1991) Biochemical considerations m ‘H spectroscopy Glutamate
and glutamine, myo-mositol and related metabolites NMR Bzomed4,59-63
Rothman, D L , Behar, K L., Hethermgton, H P , and Shulman, R G (1984)
Homonuclear ‘H-double resonance difference spectroscopy of the rat brain
in vzvo Proc Nat1 Acad Scz USA81, 6330-6334
Rothman, D L, Hanstock, C C , Petroff, 0. A C , Novotny, E J, Prichard, J
W., and Shulman, R G (1992) Localized ‘H spectra of glutamate in human
brain
Magn ResonMed
25,94-106
Rothman, D. L , Petroff, 0 A C , Behar, K L , and Mattson, R
H
(1993) Local-
ized ‘H NMR measurements of y-ammobutyric acid m human brain m vlvo
Proc Nat1 Acad Scz USA 90,5662-5666
Ryner, L N , Sorenson, J. A, and Thomas, M A (1995) 3D localized 2D NMR
spectroscopy on an MRI scanner
] Magn Reson
B107,126-137
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 33/34
Applmtions of Proton MRS 379
Sappey-Marmier, D , Calabrese, G , Hetherington, H. P , Fisher, S. N., Deicken,
R , VanDyke, C , Fern, G , and Werner, M W. (1992) Proton magnetic reso-
nance spectroscopy of human brain* Applrcations to normal white matter,
chronic infraction, and MRI white matter signal hypermtensmes. Magn Reson
Med 26,313-327
Segum, F , Jubault, C., Grrvet, J I’., and LePape, A (1990) 3’P NMR study of mtracel-
lular pH during the respiratory burst of macrophages Exp. Ceil
Res
186,188-191
Seguin, F , Grrvet, J. I’. , Akoka, S., Jubault, C , and LePape, A (1991) Blochemr-
cal events occurmg during the respiratory burst o f macrophages A 31P and
13C study
Exp Cell Res
196,141-145
Shen, J F. and Allen, P S (1991) The effect of an inhomogeneous RF field on
double quantum filtering. j Magn Reson 92,550-559
Glens, P E., Vecht, C J , Levendag, P C , vanDi)k, P , and Oudkerk, M (1995)
Hydrogen magnetic resonance spectroscopy follow-up after radiation therapy
of human brain cancer lnvestzgatwe Radiology 30, 738-744.
Sotak, C H and Freeman, D. (1988) A method for volume localized lactate
editing using zero quantum coherence created m a stimulated echo pulse
sequence J Map Reson 77,382-388
Stanley, J A, Drost, D J , Wrllramson, P C , and Thompson, R T (1995) The
use of a prrorr knowledge to quantify short echo In Vrvo ‘H MR spectra.
Magn Reson Med 34,17-24
Szrgety, S K , Allen, P S , Huyser-Wrerenga, D , and Urtasun, R C (1993) The
effect of radiation on normal human CNS as detected by NMR spectroscopy
lnt J Radlat Oncol Bzol Phys 25,695-701
Thomas, M A, Hethermgton, H P , Moyerhoff, D J , and Twerg, D B (1991) Lo-
calized double quantum filtered ‘H NMR spectra J Mugn. Reson 93,485-496
Thompson, R B. and Allen, P S. (1998) A new quantum filter design procedure
for use on strongly coupled spin systems found m-vrvo its application to
glutamate Magn Reson Med 39,762-771
Tracey, I, Carr, C. A, Gurmaraes, A. R , Worth, J L , Navra, B A, and Gonzales,
R. G. (1996) Bram cholme-contammg compounds are elevated rn HIV-posr-
tive patients before the onset of AIDS dementia complex A proton magnetic
resonance spectroscoprc study Neurology 46,783-788
Tremble, L A, Shen, J F , Wrlman, A H , and Allen, P S (1990) Lactate editing
by means of selectrve pulse filtering of both zero and double quantum
coherences J Magn Reson 86,191-198
Tropp, J (1989) The theory of the brrdcage resonator J Magn
Reson82,51-62
Urenlak, J , Wrllrams, S R , Gadian, D G., and Noble, M. (1992) Specif ic expres-
sion of N-acetylaspartate in neurons, oiigdendrocyte-type-2 astrocyte pro-
genitors, and immature olrgodendrocytes in vitro. J Neurochem 59,55-61
Urenjak, J., Wdhams, S. R , Gadran, D G , and Noble, M. (1993) Proton nuclear
magnetic resonance spectroscopy unambrguously identifies different neural
cell types J Neuroscl 13,981-989
Usemus, J I’, Kauppmen, R A , Vamro, P A, Hernesmemr, J, A , Vapalahtr, M
P , Pallarvr, L A, and Sormakallro, S (1994a) Quantitative metabohte pat-
terns of human brain tumors detection by IH NMR spectrscopy m vrvo and
in vit ro J Comput Ass Tomog 18,705-713
Usenms, J P , Vamio, P , Hernesmemr, J , and Kauppinen, R A. (1994b) Cho-
lme-contammg compounds m human astrocytomas studred by ‘H NMR spec-
troscopy in vrvo and in vitro J Neurochem
63,1538-1543
7/21/2019 11 Applications of Proton MRS to Study Human Brain Metabolism
http://slidepdf.com/reader/full/11-applications-of-proton-mrs-to-study-human-brain-metabolism 34/34
380 Hanstock and Allen
Usenms, T., Usenms, J I’., Tenhunen, M , Vamio, P , Johansson, R , Sormakalho,
S., and Kauppinen, R. (1995) Radiation-induced changes m human brain
metabohtes as studied by ‘H nuclear magnetic resonance spectrscopy m viva
Int J Xadzat. One Bzol Phys. 33,719-724
VanderKnaap, M S , Grond, J. V. d , Luyten, I’. R , Hollander, J. A d , Nauta, J
J, P., and Valk, J. (1992) ‘H and 31Pmagnetic resonance spectroscopy of bram
degenerative cerebral disorders Neural 31,202-211
vanderToorn, A., Dqkhmzen, R. M., Tulleken, C A., and Nrcolay, K. (1995) T,
and T, relaxation times of the major ‘H-containing metabohtes in rat brain
after focal rschemia. NMR zn Bzomedzczne8, 245-252
VanHecke ,P , Marchal, G., Johanmk, K., Demaerel, P , Wilms, G , Carton, H ,
and Baert, A L. (1991) Human brain proton localised NMR spectroscopy m
multiple sclerosis Magn Reson Med 18, 199-206
Vullo, T., Zipagan, R. T., Pascone, R., Whalen, J, P, and Cahill , P T (1992)
Experimental design and fabrication of birdcage resonators for magnetic reso-
nance imaging. Magn Reson Med 24,243-252
Williams, S R., Gadran, D G., and Proctor, E. (1986) A Method for lactate
detection in VIVO by spectral edrtmg without the need for double irradiation
J Magn Reson 66,562-567.
Wilman, A H and Allen, P S (1993) In vivo NMR detectron strategies for y-
ammobutyrrc acid, uti lizing proton spectroscopy and coherence pathway
filtermg with gradients J Magn. Reson BlOl, 165-171
Wllman, A H. and Allen, I’. ‘5. (1995a) The response of the strongly coupled AB
system of citrate to typical IH MRS localization sequences I Magn Reson
B107,25-33
Wilman, A. H and Allen, P. S (1995b) Yield enhancement of the double quan-
tum filter sequence designed for the detection of GABA m proton spectros-
copy of brain J Magn Reson B109,169-174
Wllman, A. H and Allen, P S (1996) Observmg N-Acetyl Aspartate via both its
N-Acetyl and its strongly coupled aspartate groups m In Viva proton mag-
netic resonance spectroscopy J, Magrz Reson B113,203-213
Wolswilk, G and Noble, M (1992) Cooperation betwen PDGF and FGF converts
slowly dividing O-2A adult progemtor cells to rapidly dividing cells with char-
acteristics of O-2A permatal progenitor cells. J Cell Bzol
118, 889-900.
Ziegler, A, Izquierdo, M , Remy, C., and Decorps, M (1995) Optimrzation of
homonuclear Two dimensional correlation methods for In Viva and Ex Vrvo
NMR J Magn Reson B107,10-18.