Hypoxia imaging in brain tumors

Hypoxia imaging in brain tumors

F. Zerrin Yetkin, MD*, Dianne Mendelsohn, MD

Division of Neuroradiology, Department of Radiology, University of Texas Southwestern Medical Center at Dallas,

Dallas, TX 75390-8896, USA

Assessment of the oxygenation status of brain

tumors has been of increasing interest under the light

of recent advances in oncology. Extensive experi-

mental and in vivo studies have documented that

tissue oxygen tension is a critical factor influencing

the tumor response to radiation and some chemo-

therapeutic agents. Hypoxia compromises the effect

of irradiation on tumor cells. Oxygen increases the

efficiency of x or gamma radiation [1]. The oxygena-

tion status of a tumor is also a factor in the regulation

of gene expression for malignant progression [2].

Furthermore, in head and neck tumors, pretreatment

oxygenation status of the tumor predicts the radiation

response and regional tumor control [3,4]. As for

brain tumors, high-grade gliomas have large propor-

tions of hypoxic tissue that contribute to resistance to

radiation [5,6].

A number of strategies to increase tumor oxygena-

tion have been developed to improve outcome. Ad-

ministration of nicotinamide [7,8], inhalation of 100%

oxygen [9], hyperbaric oxygen [10–12], or high

oxygen mixtures, such as carbogen (95% oxygen

and 5% carbon dioxide) [7,8,13–17] are suggested

to increase the tumor’s sensitivity to radiation therapy.

Different types of tumors display variable hypoxic cell

proportions and response to reoxygenation strategies

[18,19]. Radiation therapy, however, changes the

oxygenation status of the tumor tissue along with

the vascularization and perfusion of tumors that are

tied closely to oxygen distribution. Oxygen tension in

a tumor may decrease or increase during radiation

therapy [20–22]. Additionally, successful results of

experimental studies on tumor oxygenation strategies

do not apply to clinical trials [23–25] necessarily.

Existing methods to measure the efficacy of a tumor

oxygenating treatment in clinical studies are confound-

ed by the heterogeneity of the patients and tumor

response [22,24]. The lack of information regarding

the behavior of individual tumors limits evaluation of

treatment strategies. Noninvasive imaging techniques

capable of detecting and measuring oxygenation sta-

tus are needed to assess hypoxic fraction of tumors.

The advent of new imaging techniques can facilitate

studying the efficacy of adjuvant therapies increasing

tumor oxygenation, identifying patients who might

benefit from radiosensitizers, and planning treatment

such as fractionation of radiation and hypoxia selec-

tive drugs [22,26].

Diagnostic imaging studies for brain tumors

mainly has investigated morphological characteristics

of the tumors and evaluated the response to therapy

based on the changes in tumor size. As the complex

nature of tumor metabolism is investigated, the need

for additional information on the metabolic features of

brain tumors is recognized increasingly. In recent

years, the emphasis of imaging studies is shifted to

evaluation of metabolic activity of brain tumors for

diagnostic purposes and assessment of the response to

therapeutic regimens. Diagnostic and prognostic val-

ues of several tumor characteristics such as vascular-

ization, perfusion, glucose consumption, and oxygen

use have been shown for various tumors. A variety of

invasive and noninvasive methods have been used to

evaluate the oxygenation status of tumors such as

polarographic oxygen microelectrodes [26–29], sin-

gle photon emission tomography [30–33], positron

emission tomography [34–48], MR spectroscopy

[49–53], and MR imaging [24,54–57]. Polarographic

needle allows direct measurement of tumor oxygena-

1052-5149/02/$ – see front matter D 2002, Elsevier Science (USA). All rights reserved.

PII: S1052 -5149 (02 )00029 -1

* Corresponding author.

E-mail address: [email protected]

(F.Z. Yetkin).

Neuroimag Clin N Am 12 (2002) 537–552

tion; however, this technique is invasive, and only a

portion of the tumor can be sampled [26–29]. Sen-

sitivity and accuracy of the polarographic needle

measurements have been reported at varying levels

[27]. Existing neuroimaging modalities provide an

indirect evaluation of the brain tumor oxygenation.

Following sections include relevant information on

tumor oxygenation obtained using neuroimaging

modalities and present emerging techniques designed

to image cerebral oxygen saturation in normal and

diseased states.

Radionuclide imaging modalities

Positron emission tomography (PET) is used com-

monly to evaluate metabolic components of brain

tumors. Various radioisotopes are used to evaluate

different aspects of tumors. In an effort to evaluate

tumor oxygenation, nuclear medicine studies have

focused on the development of tissue hypoxia mark-

ers. Azomycin arabinoside, an indicative of hy-

poxia, has been studied in patients with decreased

perfusion as determined with technetium-hexamethyl-

propyleneamine oxime (99mTc-HMPAO) and

patients exhibited an increased avidity for azomycin

[58,59]. Several technical issues are yet to be

resolved, however, such as distinguishing between

marker avidity arising from hypoxic mechanisms

versus tumor perfusion [30].

Positron emitters such as carbon11 [51,60–62],

oxygen15 [41,46,63,], and [18F]Fluoromisonidazole

([18F]FMISO) have been used to assess tumor metab-

olism in PET studies and [18F] flurodeoxygluxcose

has been the most widely used radiotracer to evaluate

glucose metabolism of brain tumors [64]. Tumors

frequently exhibit increased glucose consumption

compared with normal tissue. The suggested mecha-

nisms of increased glucose metabolism in tumors

include increased anaerobic metabolism resulting

from oxygen starvation [35,50,51]. Technical prob-

lems associated with the use of flurodeoxygluxcose

include the difficulties in differentiating normal brain

accumulation from tumor uptake and the detection of

graded uptake of flurodeoxygluxcose in various

lesions such as necrosis, infarct, and inflammatory

lesions [35]. Regional cerebral blood flow, cerebral

blood volume, oxygen extraction fraction, and oxygen

and glucose use in brain tumors are measured quanti-

tatively using the oxygen-15 steady state inhalation

technique [46,65–68]. Several studies have reported

reduced oxygen extraction fraction in tumor tissue.

Lammertsma et al [41], however, showed that tissue

heterogeneity might affect the results of oxygen

extraction fraction in tumors and suggested that the

oxygen-15 steady state has limited value in the assess-

ment of pathophysiology of tumors.

[18F]FMISO is a hypoxia binding radiopharma-

ceutical used for imaging hypoxic regions in tumors

[45]. In vitro studies using [18F]FMISO have shown

the need for very low levels of oxygen for significant

uptake, however, high proportion of hypoxia in a

variety of tumors was reported. The heterogeneity of

oxygenation in individual tumors also was docu-

mented using PET of [18F]FMISO [45].

MR spectroscopy

Proton (1H) and phosphorous (31P) MR spectros-

copy are used to detect in vivo metabolite differences

between the normal brain tissue and tumor. Associa-

tion between the distribution of various metabolite

concentrations and the tumor histology and tumor

grade has been investigated widely [12,50 –53,

69–76]. A limited number of clinical studies have

investigated the correlation between the brain tumor

oxygenation and MR spectroscopy findings.

Proton MR spectroscopy studies include investi-

gations of the association between the MR spectros-

copy findings and the indirect measures of tumor

oxygenation such as regional metabolites [51] and

cerebral blood volume [77]. Proton MR spectroscopy

shows the distributions of various metabolite concen-

trations. N-acetyl-L-aspartate, N-phosphocholine

(choline), and lactate are evaluated frequently for brain

tumors [69,71,76,78–80]. Increased choline and/or

creatinine and decreased N-acetyl-L-aspartate and/or

creatinine reflect the neuronal tissue loss and increased

membrane synthesis that are known to occur in glio-

blastomas and astrocytomas [51]. Lactate is an end

product of glycolysis, a frequent feature of gliomas,

and is reported to increase in brain tumors. The

increased lactate/creatinine ratio in the core of malig-

nant gliomas may reflect the anaerobic glycolysis

possibly in conjunction with hypoxia, whereas, in

the edematous tissue, higher lactate/creatinine ratios

may be the result of hypoxia caused by compromised

regional perfusion [51]. Lactate, however, is not spe-

cific to hypoxia, since it can reflect the dominance of

glycolysis, a frequent feature of glioblastomas. The

lack of correlation between increased tumor metabo-

lism and lactate accumulation as detected with 1H-MR

spectroscopy also was reported, and this finding was

questioned as a possible indication of oxygen depriva-

tion in tumor tissue [50]. In 19 patients with glioma,

comparison of the 1H-MR spectroscopy findings with

the regional cerebral blood volume, an indirect param-

F.Z. Yetkin, D. Mendelsohn / Neuroimag Clin N Am 12 (2002) 537–552538

eter of tumor oxygenation, revealed that regional

cerebral blood volumes are proportional to the choline

values in high-grade tumors [77].

Animal models employing different lines of

tumors studied 31P-MR spectroscopy as a tool to

detect tumor metabolism. 31P-MR spectroscopy was

used in evaluating tumor growth [81,82], oxygen

tension (pO2) [83], perfusion [84], and intravascular

oxyhemoglobin concentration [85]. Metabolism of

cell membrane phospholipids is important in monitor-

ing tumors. Phosphomonoester is a precursor, and

phosphodiester is a product of membrane phospho-

lipid metabolism. The ratio of phosphomonoester to

phosphodiester is related to cell membrane synthesis

and metabolic turnover, which are different in tumors

then normal tissue [76]. A negative correlation

between the tumor oxygen tension and ratio of phos-

phomonoester to nucleoside triphosphate was

reported, suggesting that 31P-MR spectroscopy can

detect changes in tumor bioenergetic status. Further

investigation of the rate of glycolysis in tumors under

hypoxic and well-oxygenated conditions in tumor

models, using tumor specific 13C-CP nuclear mag-

netic resonance spectroscopy showed that some line

of tumors can maintain the same level of bioenergetic

status as detected with b nucleoside triphosphate to Piratio under hypoxic conditions, however. Decrease in

the ratio of b nucleoside triphosphate to Pi and

increased lactate levels were not observed until tumor

blood flow was decreased by 90% [86]. The effect of

oxygen and nutrient supply on the global bioenergetic

status of a tumor as measured with the Pi to nucleoside

triphosphate ratio was studied using a glioma model.

The change of tumor metabolism from aerobic to

anaerobic was observed at intravascular distances

larger than approximately 200mm. The Pi to nucleo-

side triphosphate ratio was increased as the mean

intravascular distance increased. Moreover, a linear

relationship between the increased ratio of Pi to

nucleoside triphosphate, and the fraction of the non-

perfused tumor areas was detected [53]. These find-

ings possibly indicate that a major determinant of

tissue oxygenation in the glioma is the limited supply

of oxygen. The results of 31P-MR spectroscopy

experiments obtained with high strength magnets are

not yet translated to applications in humans, however.

MR imaging

Contrast-enhanced MR imaging

There are numerous parameters to characterize

tumor vasculature, including vascular volume, blood

flow, and vascular permeability [87–89]. Neovascu-

larization is essential for tumor growth and is induced

through release of various angiogenic factors such as

vascular endothelial growth factor. Vascular perme-

ability also is promoted by vascular endothelial

growth factor [89,90]. Incidentally, a hypoxic environ-

ment provokes production of vascular endothelial

growth factor. The close linkage between tumor me-

tabolism, hypoxia, angiogenesis, and vascular per-

meability places a special emphasis on enhancement

patterns of brain tumors [52,87,91–93]. T1-weighted

MR images obtained with intravenous gadolinium

diethylenetriamine pentaacetic acid compound con-

trast material administration show the fraction of

tumor capillaries with decreased permeability. Gado-

linium diethylenetriamine pentaacetic acid used for

contrast enhancement is a low molecular weight

agent that easily can leak into interstitial space in

the presence of brain-blood barrier disruption. The

leakage of contrast agent from tumor microvascula-

ture may be limited by decreased flow and perme-

ability of the capillaries [87]. Capillary permeability in

a tumor can be quantified using dynamic T1 measure-

ments [88,89,93,94].

There is a positive correlation between the degree

of decreased permeability of tumor vasculature and

the metastatic potential of a tumor [89,90,95,96]. The

rate of gadolinium diethylenetriamine pentaacetic acid

uptake is proportional to the number of functional

microvessels with a disrupted brain-blood barrier in

gliomas. In a recent study using dynamic T1 mapping,

it was shown that the number of capillaries with

increased permeability correlated with the hypoxic

fraction. Rapidly growing tumors are reported to have

high vascular permeability compared with tumors

with slow growth rates [89].

Gadolinium diethylenetriamine pentaacetic acid

enhancement of tumors is not specific to neovas-

cularization. Presence of enhancement of necrotic

brain tissue decreases the use of contrast enhance-

ment to evaluate the degree of tumor vascular

permeability. Another limitation of T1-weighted con-

trast enhanced MR images is that tumor microvessel

density cannot be determined accurately in the

presence of capillaries without brain-blood barrier

disruption [52,91,97].

Dynamic susceptibility contrast MR imaging

Basic information regarding the pathophysiology

and metabolism of brain tumors is essential to use the

dynamic contrast imaging modalities. The complex

relationship between tumor vascularization, metabo-

lism, hypoxia, and behavior is affected with tumor

F.Z. Yetkin, D. Mendelsohn / Neuroimag Clin N Am 12 (2002) 537–552 539

type, grade, and status of treatment [5,20,29,54,

96,98–101].

Tumor hypoxia is related closely to insufficient

blood flow caused by chaotic and partially nonfunc-

tional tumor vasculature [100]. Two forms of hypoxia

are described in tumor tissue: (1) diffusion limited

chronic hyperoxia where the oxygenation of tumor

cells are decreased as their distance from the vessel is

increased and (2) acute hypoxia originating from the

irregularities of the blood flow leading to limited

perfusion around blood vessels [98,100]. In addition

to the chaotic organization and nonfunctional charac-

ters of tumor vessels, a transient perfusion also is

shown in animal tumor models. Tumor vessels that

appear nonfunctional or closed may be open tran-

siently and functional at different times. Existence of

viable tumor cells near nonfunctional tumor vessels

might be explained with this phenomenon. Fluctua-

tions in tumor vessel patency indicate that hypoxia

may occur without the total vascular stasis [98]. The

degree of vascularization and perfusion in a tumor is

not identical. The proportion of vascular structures that

show perfusion varies among animal tumor lines,

ranging from 20% to 80%. The perfusion fraction

per unit tumor area is not a determining factor for

tumor tissue perfusion, since it is not known what

proportion of nonperfused vessels is permanently or

temporarily nonfunctional [96]. The fact that most of

the vascular structures are not perfused at a given time

suggests that extensive areas of tumors are hypoxic

and not accessible by therapeutic agents causing

decreased response to treatment. The relationship

between the morphologic characters of the tumor

vasculature, perfusion, and the level of oxygenation

is a complex issue.

Rijkin et al [100] have provided a sophisticated

digital imaging system that documents the architec-

ture and the functional status of the tumor vascular

network and oxygenation level. The spatial relation-

ship of each physiologic parameter was displayed by

combining the histological images obtained with

markers for vessel, hypoxia, and perfusion. Fig. 1A

is an example of composite digital images obtained

from a whole tumor cross-section, simultaneously

displaying the relationship of vessels, perfusion, and

hypoxic regions. A detailed photograph of a histologic

section after exposure to three different markers is

shown in Fig. 1B. Hypoxic regions are localized at a

distance of 70 mm to 80 mm from the perfused vessels,

and numerous investigators have shown that viable

tumor cells are found in these hypoxic regions [102].

The distance at which oxygen tension falls to zero is

Fig. 1. (A) Pseudocolored image of a human glioma xenograft (E106). Composite digital image is obtained after sequential

scanning for Hoechst perfusion (blue), hypoxia with pimonidazole as a marker (green), and endothelium (red) in tumor sections.

(B) Detailed photograph of a histologic section after triple exposure to same markers. (From Rijken P, Bernsen H, Peters J, et al.

Spatial relationship between hypoxia and the (perfused) vascular network in a human glioma xenograft: a quantitative multi-

parameter analysis. Int J Radiat Oncol Biol Phys 2000;48:571–82; with permission.)


between 80 mm and 180 mm [5,53,96,102]. In addition

to variations in the fraction of perfused vessels, the

architecture of the vessels (ie, the distance and the

order between the capillaries) is known to affect

oxygen gradient. Intercapillary distance is one of the

parameters used in evaluation of tumor vascularity.

The introduction of the dynamic susceptibility

magnetic perfusion technique provides information

additional to that obtained by contrast MR images

[95,100,103–110]. Dynamic susceptibility contrast

MR imaging is based on tracking the passage of an

intravenously injected bolus of contrast media through

the tissue. Relative cerebral blood volume maps

obtained using dynamic susceptibility contrast MR

imaging has been useful in evaluating tumor vascula-

ture shown to correlate with tumor grade [95]. Cere-

bral blood volume maps obtained using dynamic

susceptibility contrast MR imaging display the func-

tional vessels [105]. It is conceivable that cerebral

blood volume maps reveal the status of oxygenation,

especially acute hypoxia in tumors, to some degree,

since many factors affect tumor oxygenation. Gliomas

are resistant to radiation therapy and characterized by

an exceptionally high degree of vascularization [96].

High proportion of hypoxia also is documented [45].

Tumor vessels are disorganized. Tortuous and mostly

larger than normal vessels contain blind ends and

arteriovenous shunts [90]. Microvascular character-

istics of glioma models are studied using high and low

molecular weight MR contrast agents. High molecular

weight contrast agents remain in the intravascular

compartment and have minimal leakage [89,105].

Good correlation between relative cerebral blood

volume estimates, histology, and quantitative auto-

radiography are shown in experiments using T2-

weighted steady state susceptibility contrast MR

imaging with high molecular weight contrast agents

[105]. Low molecular weight contrast material (gado-

linium diethylenetriamine pentaacetic acid), however,

leaks to the interstitial compartment in the presence of

brain-blood barrier disruption, affecting cerebral

blood volume maps.

The MR imaging acquisition technique selected

for dynamic susceptibility contrast MR imaging (ie,

gradient echo or spin echo) affects the size of the

vessels that can be detected [104,105,109]. Spin echo

technique is sensitive to smaller vessels whereas

gradient echo images can image both small and large

vessels. The accuracy of relative cerebral blood vol-

ume maps, therefore, may be underestimated because

of a number of factors such as contrast extravasation

to interstitium, early arteriovenous shunting, or using

spin echo acquisition for cerebral blood volume map-

ping. Presaturation techniques are reported to over-

come the relaxivity effect of T1 shortening and

increase the accuracy of cerebral blood volume maps

of tumors [97,104,106]. Fig. 2 shows the time course

of signal intensity plot obtained with dynamic sus-

ceptibility contrast MR imaging in a patient with

glioblastoma multiforme. In Fig. 2A, bolus passage

of contrast material through the vascular compartment

cannot be tracked, probably because of early arterio-

venous shunt, increased capillary permeability, and T1

shortening effect. In Fig. 2B, dynamic susceptibility

contrast MR imaging after a presaturation dose allows

detection of the intravascular contrast passage, con-

tributing to accuracy of regional cerebral blood vol-

ume measurement. The dose and timing of the

contrast material given for pre-enhancement may lead

to an increase in the signal intensity and overestima-

tion of cerebral blood volume. T2* measurements

using fast gradient echo images are widely used in

clinical practice and cerebral blood volume is esti-

mated based on relative perfusion parameters.

Blood oxygen level–dependent MR imaging

Since Ogawa demonstrated that T2* signal inten-

sity changes as a result of varying degrees of blood

oxygen saturation, blood oxygen level–dependent

contrast imaging has been used to detect hemodynamic

changes in the brain during activation [111,112].

Blood oxygen level–dependent imaging also is used

to evaluate the physiologic parameters of brain me-

tabolism such as cerebral blood flow and oxygen

consumption. The results of blood oxygen level–

dependent imaging for cerebral blood flow and oxygen

consumption are comparable to those obtained with

animal models and PET [113–115]. Blood oxygen

level–dependent contrast imaging technique exploits

the paramagnetic properties of deoxyhemoglobin,

which acts as a natural intravascular paramagnetic

contrast agent [111,116]. As the concentration of

deoxyhemoglobin increases, apparent transverse

relaxation time (T2*) and transverse relaxation time

(T2) decreases, leading to attenuated signal intensity in

T2* and T2-weighted images [112]. The MR imaging

sequences sensitive to changes in blood oxygen levels

are gradient echo and spin echo technique. Both

sequences are acquired commonly using echoplanar

imaging. Gradient echo sequence is intrinsically sen-

sitive to the susceptibility of deoxyhemoglobin and

hence is used widely for blood oxygen level–depen-

dent contrast imaging [116–118].

Hypoxia and hyperoxia are subject to many studies

as potential methods for producing signal contrast in

MR images of the brain. Using blood oxygen level–

dependent contrast imaging technique, manipulation


of oxygen content in the inhaled air can provide

qualitative and quantitative information on regional

oxygenation of brain. During hypoxia, concentration

of deoxyhemoglobin is increased and causes decreased

signal intensity on T2*-weighted images. Decreased

signal intensity on T2* images produces a contrast

effect similar to gadolinium diethylenetriamine penta-

acetic acid in the intravascular compartment [119–

122]. Oxygenation, on the other hand, increases local

T2*, causing signal intensity increases [116 –

118,123–125]. The sensitivity of blood oxygen

level–dependent effect has been essential to evaluate

cerebral blood oxygen saturation for a variety of

pathophysiologic conditions in addition to functional

MR imaging.

MR imaging during hyperoxia has been success-

fully used to detect oxygenation of brain tumors

[55,56,126–128]. Inhalation of oxygen or carbogen

decreases blood deoxyhemoglobin concentration

resulting in signal intensity changes on T2* weighted

images. Hemoglobin is saturated almost completely

under normal conditions, and inhalation of 95% to

100% oxygen results in only a small increase in

oxygenation of hemoglobin [116–118,123,124,128].

Signal intensity changes observed during oxygen

inhalation also originate from dissolved oxygen in

Fig. 2. T2* signal intensity changes in a glioblastoma multiforme during injection of gadolinium diethylenetriamine penta-acetic

acid are obtained using a spin echo/echoplanar imaging sequence. The time course of signal intensity changes in the selected

pixels from intact cortex and tumor are displayed in red and black respectively. The tumor pixel was selected from a region with

contrast enhancement. Signal intensity values are expressed in arbitrary units. (A) Early T1 shortening obscuring the transit of

contrast material through vascular compartment. With the advantage of presaturation, the subsequent gradient echo/echoplanar

imaging acquisition from the same tumor pixel exhibited the decrease in signal intensity during passage of gadolinium

diethylenetriamine penta-acetic acid through vascular compartment (B).


the plasma, since the increase in oxyhemoglobin

content is limited. The high magnetic moment of

oxygen is suggested as a cause of image contrast,

since molecular oxygen is a biradical with direct ef-

fect on the relaxation of water proton and has bulk

magnetic susceptibility effect [116,118,123,126]. The

contrast mechanism of blood oxygen level–depend-

ent imaging is affected by changes in blood flow and

blood volume in addition to intravascular magnetic

susceptibility [124]. Signal intensity changes caused

by hyperoxia, however, are dominated by blood

oxygenation, and the effect of blood volume changes

on signal intensity are minimal [116,118, 121,122].

The sensitivity of blood oxygen level–dependent

contrast MR imaging makes it a powerful tool to study

tumor oxygenation in animal models. Hyperoxia

models have been successful in employing blood

oxygen level–dependent contrast imaging. Various

tumor types have been studied to evaluate the correla-

tion between MR signal changes and tissue oxygena-

tion. Compared with normal vessels, tumor vessels

have a significantly different response to changes in

systemic blood pressure, blood flow, and hyperoxia

[56,126–128]. Oxygen causes vasoconstriction in

normal vessels, whereas most tumor vessels lack

smooth muscle and do not show vasoconstriction

resulting in relatively increased vascular volume in

the tumor region [126]. Tumors with high vascular

density are expected to show increased oxygenation

and large signal increase changes. Significant and

reproducible increases in T2* have been reported in

tumors during hyperoxia [24,55–57,126–128]. His-

tologic assessments of the regions with maximum

signal increase during hypoxia reveal low vascular

density without necrosis. It is suggested that increased

oxygen availability during hypoxia can cause

increased blood flow and tissue oxygenation com-

pared with oxygenated regions. Also, oxygen con-

sumption in hypoxic regions of tumors increases

when more oxygen is available, leading to increased

tumor oxygen tension and T2* signal intensity

[56,57,127]. These features of tumor vasculature have

been exploited to evaluate tumor oxygenation. Differ-

ent patterns of tumor behavior in response to hyper-

oxia also are reported. Regions of decreased T2*

during carbogen breathing have been observed within

hypoxic areas as identified by histology [128]. Animal

models using T2* MR measurements under normoxic

and hypoxic conditions provide clinically relevant

information on tumor blood flow, oxygen tension,

and metabolism.

In people, normobaric hyperoxia is used as a

respiratory challenge to study blood oxygen level–

dependent signal changes. Hyperoxia increases para-

sympathetic influence in the regulation of the heart,

causing bradycardia; however, none of the studies

inducing normobaric hyperoxia in people reported

any discomfort caused by oxygen inhalation [129–

131]. The confounding effect of gaseous oxygen on

T2*-weighted images should be controlled during MR

imaging of hypoxia. There is an artifactual increase of

T2* signal obtained from the object of interest when

gaseous oxygen spreads in the MR scanner [132].

During hyperoxia experiments, oxygen delivery by

way of a closed system would prevent erroneous T2*

signal intensity changes caused by an oxygen leak.

T2* signal intensity increases at the gray matter, and

large veins on gradient echo images acquired during

100% oxygen breathing were observed [117,118,129].

The effect of 100% oxygen inhalation on the T1

shortening in the various tissues outside the brain also

was reported [131].

Recently, blood oxygen level–dependent contrast

imaging has been used to map brain oxygenation

[133,134]. Images of the brain oxygenation have been

acquired using gradient echo/echoplanar imaging

acquisition during breathing room air and a brief

period of (30 seconds) of 100% oxygen (7 L per

minute) inhalation in normal adults. T2* signal

increase consistent with the timing of oxygen breath-

ing is observed in the cortex and basal ganglia.

Mapping of regions with increased T2* signal inten-

sity during pure oxygen breathing displays cortex and

Fig. 3. Example of mapping the brain regions with T2*

signal intensity increase during 100% oxygen breathing in a

healthy young adult. The images are acquired gradient echo/

echoplanar imaging sequence. Regions with significant T2*

signal intensity change are displayed in blue.



basal ganglia similar to cerebral blood volume maps

obtained using dynamic susceptibility contrast per-

fusion images [135]. The increase in the T2* signal

intensity starting with oxygen inhalation and returning

to baseline of breathing room air after cessation of

oxygen breathing was observed in normal brain tissue.

Fig. 3 shows the mapping of brain regions with

increased oxygenation as detected by an increased

T2* signal.

Under normal conditions, interstitial space and

arterial oxygen levels have a linear relationship.

Hyperoxia does not change cerebral blood flow or

capillary permeability. Capillary permeability to oxy-

gen is reported to decrease slightly in response to acute

or prolonged hyperoxia, however [124]. Arterial oxy-

gen tension, cerebral blood flow, capillary permeabil-

ity to oxygen, and oxygen consumption determine

oxygen tension in interstitial space [114,124]. The

increased signal intensity during hyperoxia can be ex-

plained by decreased deoxyhemoglobin and increased

oxygen saturation in plasma [116,118,123]. MR map-

ping of cerebral oxygenation under hyperoxia also is

used in patients with brain tumor [133,135]. All

patients displayed signal increases in cortex and basal

ganglia as seen in normal subjects. On the other hand,

changes in the signal intensity of tumors varied among

patients, ranging between no change from baseline to a

heterogeneous distribution of increased T2* signal

over the tumor region (Fig. 4A). The temporal course

of signal intensity change obtained from the tumor

regions displayed different features compared with

that of intact cortex (Fig. 4B, C). The response of the

tumor tissue to hyperoxia may be a function of the

tumor blood flow, oxygenation, and metabolism that

was present under normoxic conditions and provides

more clues on variations of tumor metabolism [126,

127]. In a few cases, tumor regions of increased signal

intensity during hyperoxia were within the regions of

increased cerebral blood volume as detected by sus-

ceptibility perfusion imaging. In some cases, however,

different regions of tumor tissue were displayed for ce-

rebral blood volume maps and oxygen maps, probably

indicating the mismatch between perfusion and oxy-

genation (Fig. 5). This observation of signal increase

caused by oxygen inhalation in regions of tumor with-

out perfusion or no oxygenation in perfused regions

suggests that inhaled oxygen is subject to a different

pathophysiological mechanism than intravascular con-

trast agent. The significance of the additional informa-

tion obtained with oxygen mapping under hyperoxic

conditions merits further investigation.

Animal models and studies on humans successfully

demonstrated that changes in cerebral blood oxygen

saturation can be detected using blood oxygen level–

dependent contrast imaging. Technical difficulties

include variations in the regional cerebral blood flow

and volume [136–138]. The relationship between

various physiologic parameters and cerebral blood

oxygen saturation is studied using blood oxygen

level–dependent effect under a variety of conditions.

Experiments using blood oxygen level–dependent

imaging detected that the relative changes in cerebral

blood flow and cerebral oxygen consumption are

coupled in an approximate ratio of 2:1 and showed

the linear relationship between arterial oxygen tension

and interstitial oxygen tension [114,124]. As different

aspects of the biophysical basis of blood oxygen

level–dependent contrast mechanism are understood

better, the potential for clinical applications will be

increased. MR imaging studies have evaluated cere-

bral blood oxygen saturation in the context of relative

measurements. Attempts to obtain absolute measure-

ments showed the effect of regional cerebral blood

volume variations on the interpretation and quantifica-

tion of T2 and T2 * signal intensity and the importance

of hematocrit level in order to measure blood oxygen

[137,139]. Recently, cerebral oxygen saturation esti-

mated with MR imaging was shown to correlate with

direct measurements of oxygen saturation in the aorta

[140]. Quantification of changes in brain oxygen-

ation would be extremely useful in clinical assess-

ment of ischemic brain disease and brain tumors alike.

Recently, in vivo quantitative mapping of cerebral

oxygenation was accomplished using MR imaging

[141]. In this report, using multiecho gradient and spin

echo sequences, cerebral blood oxygen saturation was

determined as 58.4% F 1.8%, which is in line with

known cerebral blood oxygen saturation under normal

conditions. Their results showed that spatial variations

in cerebral blood oxygen saturation maps were caused

by noise. An overall uniform mapping of cerebral

blood oxygen saturation was obtained, however, and

Fig. 4. (A) Example of blood oxygen level–dependent contrast imaging during 100% oxygen breathing in a patient with

glioblastoma multiforme. (B, C) Regions with significant T2* signal intensity change are displayed in blue. Mapping brain

regions with increased T2* signal intensity shows the increased oxygenation in the cortex and some regions of tumor. The time

course of signal intensity changes in the selected pixels from intact cortex and tumor are displayed. Selected pixel from intact

cortex shows T2* signal intensity change starting with breathing oxygen and returning to baseline immediately after cessation of

oxygen administration (B). (C) Pixel from the tumor region also shows increased signal intensity starting with oxygen

administration. T2* signal intensity, however, displays slow decrease and delayed return to baseline values.


a lower value for cerebrospinal fluid compared with

that of brain parenchyma was observed. Mapping of

brain oxygenation as detected and quantified using

blood oxygen level–dependent contrast imaging is

shown in Fig. 6. The distribution of cerebral blood

oxygen saturation is displayed in color, red corres-

ponding to regions in which the estimation of cerebral

blood oxygen saturation could not be determined.

Quantification of cerebral blood oxygen saturation

using MR imaging has a promising clinical prospect

as the technical difficulties and confounding factors

are resolved.

Summary

Assessment of the oxygenation status of brain

tumors has been studied increasingly with imaging

techniques in light of recent advances in oncology.

Tumor oxygen tension is a critical factor influencing

the effectiveness of radiation and chemotherapy and

malignant progression. Hypoxic tumors are resistant

to treatment, and prognostic value of tumor oxygen

status is shown in head and neck tumors.

Strategies increasing the tumor oxygenation are

being investigated to overcome the compromising

Fig. 5. Examples of cerebral blood volume mapping obtained with dynamic susceptibility contrast MR imaging and oxygenation

mapping acquired with blood oxygen level–dependent contrast imaging during 100% oxygen breathing in a patient with

recurrent glioblastoma multiforme. (A) Heterogenous contrast enhancement on T1-weighted image. Note the different regions of

brain displayed with cerebral blood volume (B) and oxygenation mapping (C).


effect of hypoxia on tumor treatment. Administration

of nicotinamide and inhalation of various high oxygen

concentrations have been implemented. Existing

methods for assessment of tissue oxygen level are

either invasive or insufficient. Accurate and noninva-

sive means to measure tumor oxygenation are needed

for treatment planning, identification of patients who

might benefit from oxygenation strategies, and assess-

ing the efficacy of interventions aimed to increase the

radiosensitivity of tumors.

Of the various imaging techniques used to assess

tissue oxygenation, MR spectroscopy and MR

imaging are widely available, noninvasive, and clin-

ically applicable techniques.

Tumor hypoxia is related closely to insufficient

blood flow through chaotic and partially nonfunctional

tumor vasculature and the distance between the capil-

laries and the tumor cells. Information on character-

istics of tumor vasculature such as blood volume,

perfusion, and increased capillary permeability can

be provided with MR imaging. MR imaging tech-

niques can provide a measure of capillary permeability

based on contrast enhancement and relative cerebral

blood volume estimates using dynamic susceptibility

MR imaging. Blood oxygen level dependent contrast

MR imaging using gradient echo sequence is intrin-

sically sensitive to changes in blood oxygen level.

Animal models using blood oxygen level–dependent

contrast imaging reveal the different responses of

normal and tumor vasculature under hyperoxia. Nor-

mobaric hyperoxia is used in MR studies as a method

to produce MR contrast in tissues. Increased T2 * sig-

nal intensity of brain tissue has been observed using

blood oxygen level–dependent contrast MR imaging.

Dynamic blood oxygen level–dependent contrast MR

imaging during hyperoxia is suggested to image tumor

oxygenation. Quantification of cerebral oxygen sat-

uration using blood oxygen level–dependent MR

imaging also has been reported. Quantification of

cerebral blood oxygen saturation using MR imaging

has promising clinical applications; however, tech-

nical difficulties have to be resolved.

Blood oxygen level dependent MR imaging is an

emerging technique to evaluate the cerebral blood

oxygen saturation, and it has the potential and versa-

tility to assess oxygenation status of brain tumors.

Upon improvement and validation of current MR

techniques, better diagnostic, prognostic, and treat-

ment monitoring capabilities can be provided for

patients with brain tumors.

References

[1] Cline JM, Thrall DE, Rosner GL, et al. Distribution of

the hypoxia marker CCI-103F in canine tumors. Int J

Radiat Oncol Biol Phys 1944;28:921–33.

[2] Hoeckel M, Schlenger K, Aral B, et al. Association

between tumor hypoxia and malignant progression in

advanced cancer of the uterine cervix. Cancer Res

1996;56:4509–15.

[3] Bussink J, Kaanders JHAM, Rijken PFJW, et al. Vas-

cular architecture and microenvironmental param-

eters in human squamous cell carcinoma xenografts:

effects of carbogen and nicotinamide. Radiother On-

col 1999;50:173–84.

[4] Nordsmark M, Overgaard M, Overgaard J. Pretreat-

ment oxygenation predicts radiation response in ad-

vanced squamous cell carcinoma of the head and

neck. Radiother Oncol 1996;41:31–9.

[5] Bernsen HJJA, Rijken PFJ, Peters H, et al. Hypoxia

in a human intracerebral glioma model. J Neurosurg

2000;93:449–54.

[6] Rockwell S. Oxygen delivery: implications for the bi-

ology and therapy of solid tumors. Oncol Res 1997;

9:383–90.

[7] Fatigante L, Ducci F, Cartei F, et al. Carbogen and

nicotinamide combined with unconventional radio-

therapy in glioblastoma multiforme: a new modality

treatment. Int J Radiat Oncol Biol Phys 1997;37:

499–504.

[8] van der Maazen RWM, Thijssen HOM, Kaanders

Fig. 6. Mapping cerebral blood oxygen saturation using MR

imaging in a healthy volunteer. Color scale indicates the

quantitative measurement of cerebral blood oxygen satu-

ration ranging from 0% to 100%. (From An H, Lin W.

Quantitative measurements of cerebral blood oxygen

saturation using magnetic resonance imaging. J Cereb Blood

Flow Metab 2000;20:1225–36; with permission.)


JHAM, et al. Conventional radiotherapy combined

with carbogen breathing and nicotinamide for malig-

nant gliomas. Radiother Oncol 1995;35:118–22.

[9] Grau C, Horsman MR, Overgaard J. Improving the

radiation response in a C3H mouse mammary carci-

noma by normobaric oxygen or carbogen breathing.

Int J Radiat Oncol Biol Phys 1992;22:415–9.

[10] Dische S. Hyperbaric oxygen. The medical research

council trials and their clinical significance. Br J Ra-

diol 1978;51:888–94.

[11] Saunders M, Dische S. Clinical results of hypoxic cell

radiosensitisation from hyperbaric oxygen accelerated

radiotherapy, carbogen, nicotinamide. Br J Cancer

1996;74:S271–8.

[12] Shimuzu H, Kumabe T, Tominaga T, et al. Noninva-

sive evaluation of malignancy of brain tumors with

proton MR spectroscopy. AJNR Am J Neuroradiol

1996;17:737–47.

[13] Falk SJ, Ward R, Bleehen NM. The influence of

carbogen breathing on tumour tissue oxygenation in

man evaluated by computerized pO2 histography.

Br J Cancer 1992;66:919–24.

[14] Kruuv JA, Inch WR, McCredie JA. Blood flow and

oxygenation of tumours in mice: effects of breathing

gases containing carbon dioxide at atmospheric pres-

sure. Cancer 1967;20:51–9.

[15] Martin L, Lartigau E, Weeger P, et al. Changes in the

oxygenation of head and neck tumors during carbo-

gen breathing. Radiother Oncol 1993;27:123–30.

[16] Powell MEB, Collingridge DR, Saunders MI, et al.

Improvement in human tumour oxygenation with car-

bogen breathing. Radiother Oncol 1999;50:167–71.

[17] Rojas A. ARCON: accelerated radiotherapy with

carbogen and nicotinamide. Br J Radiol 1992;24:

174–8.

[18] Fenton BM, Raubertas RF, Boyce DJ. Quantification

of micro-regional heterogeneities in tumor oxygena-

tion using intravascular HbO2 saturations. Radiat Res

1995;141:49–56.

[19] Oikawa H, Al-Hallaq HA, Lewis MZ, et al. Spectro-

scopic imaging of water resonance with short repeti-

tion time to study tumor response to hyperoxia. Magn

Reson Med 1997;38:27–32.

[20] Bussink J, Kaanders J, Rijken P, et al. Changes in

blood perfusion and hypoxia after irradiation of a

human squamous cell carcinoma xenograft tumor

line. Radiat Res 2000;153:398–404.

[21] Raleigh JA, Zeman EM, Calkins DP, et al. Distribu-

tion of hypoxia and proliferation associated markers

in spontaneous canine tumors. Acta Oncol 1995;34:

345–9.

[22] Stone HB, Brown JM, Phillips TL, et al. Oxygen in

human tumors: correlations between methods of

measurement and response to therapy. Radiat Res

1993;136:422–34.

[23] Dische S, Rojas A, Rugg T, et al. Carbogen breath-

ing: a system for use in man. Br J Radiol 1992;65:

87–90.

[24] Hallaq-Al HA, River J, Zamora M, et al. Correlation

of magnetic resonance and oxygen microelectrode

measurements of carbogen-induced changes in tumor

oxygenation. Int J Radiat Oncol Biol Phys 1998;41:

151–9.

[25] Rubin P, Hanley J, Keys HM, et al. Carbogen breath-

ing during radiation therapy. Int J Radiat Oncol Biol

Phys 1979;5:1963–70.

[26] Rampling R, Cruickshank G, Lewis AD, et al. Direct

measurement of pO2 distribution and bioreductive

enzymes in human malignant brain tumors. Int J


[27] Kavanagh MC, Sun A, Hu Q, et al. Comparing

techniques of measuring tumor hypoxia in different

murine tumors: Eppendorf pO2 histograph, [3H]

Misonidazole binding and paired survival assay.

Radiat Res 1996;145:491–500.

[28] Nozue M, Lee I, Yuan F, et al. Interlaboratory varia-

tion in oxygen tension measurement by Eppendorf

‘‘Histograph’’ and comparison with hypoxic marker.

J Surg Oncol 1997;66:30–8.

[29] Pogue BW, Paulsen KD, O’Hara JA, et al. Estimation

of oxygen distribution in RIF-1 tumors by diffusion

model-based interpretation of pimonidazole hypoxia

and eppendorf measurements. Radiat Res 2001;155:

15–25.

[30] Chapman JD, Engelhardt EL, Stobbe CC, et al. Mea-

suring hypoxia and predicting tumor radioresistance

with nuclear medicine assays. Radiother Oncol 1998;

46:229–37.

[31] Groshar D, McEwan A, Parliament MB, et al. Imag-

ing tumor hypoxia and tumor perfusion. Nucl Med

1993;34:885–8.

[32] Hulshof M, Rehmann CJ, Booij J, et al. Lack of

perfusion enhancement after administration of nicoti-

namide and carbogen in patients with glioblastoma:

a 99mTc-HMPAO SPECT study. Radiother Oncol

1998;48:135–42.

[33] Kahn D, Follett KA, Bushnell DL, et al. Diagnosis

of recurrent brain tumor: value of 201Tl SPECT vs.18F-fluorodeoxyglucose PET. AJR Am J Roentgenol

1994;163:1459–65.

[34] Blasberg RG. Prediction of brain tumor therapy

response by PET. J Neurooncol 1994;22:281–6.

[35] Conti PS. Introduction to imaging brain tumor metab-

olism with positron emission tomography (PET).

Cancer Invest 1995;13:244–59.

[36] Di Chiro G, DeLaPaz RL, Briiks RA. Glucose uti-

lization of cerebral gliomas measured by [18] fluoro-

deoxyglucose and positron emission tomography.

Neurology 1982;32:1323–9.

[37] Di Chiro G, Brooks RA, Patronas NJ. Issues in the in

vivo measurement of glucose metabolism of human

central nervous system tumors. Ann Neurol 1984;15:

S138–46.

[38] Eskey CJ, Korestky AP, Domach MM, et al. Role

of oxygen vs. glucose energy metabolism in a

mammary carcinoma perfused ex vivo: direct mea-

surement by 31P NMR. Proc Natl Acad Sci U S A

1993;90:2646–50.


[39] Glantz M, Hoffman JM, Coleman RE, et al. Iden-

tification of early recurrence of primary central

nervous system tumors by [18F]fluorodeoxyglucose

positron emission tomography. Ann Neurol 1991;29:

347–55.

[40] Gross MW, Weber WA, Feldmann HJ, et al. The val-

ue of F-18-fluorodeoxyglucose PET for the 3-D radi-

ation treatment planning of malignant gliomas. Int J


[41] Lammertsma AA, Jones T. Low oxygen extraction

fraction in tumours measured with the oxygen-15

steady state technique: effect of tissue heterogeneity.

Br J Radiol 1992;65:697–700.

[42] Mineura K, Sasajima T, Kowada M, et al. Perfusion

and metabolism in predicting the survival of patients

with cerebral gliomas. Cancer 1994;73:2386–94.

[43] Mineura K, Shioya H, Kowada M, et al. Blood

flow and metabolism of oligodendrogliomas: a

positron emission tomography study with kinetic

analysis of 18-F-fluorodeoxyglucose. J Neurooncol

1999;43:49–57.

[44] Perlmutter JS, Powers WJ, Herscovitch P, et al. Re-

gional asymmetries of cerebral blood flow, blood vol-

ume, and oxygen utilization and extraction in normal

subjects. J Cereb Blood Flow Metab 1987;7:64–7.

[45] Rasey JS, Koh WJ, Evans ML, et al. Quantifying

regional hypoxia in human tumors with positron

emission tomography of [18F]Fluotromisonidazole:

a pretherapy study of 37 patients. Int J Radiat Oncol

Biol Phys 1996;36:417–28.

[46] Tyler JL, Diksic M, Villemure J, et al. Metabolic and

hemodynamic evaluation of gliomas using positron

emission tomography. J Nucl Med 1987;28: 1123–33.

[47] Witte O, Levivier M, Violon P, et al. Prognostic value

of positron emission tomography with [18F]Fluoro-2-

deoxy-D-glucose in the low-grade glioma. Neurosur-

gery 1996;39:47–77.

[48] Witte O, Lefrane F, Levivier M, et al. FDG-PET as a

prognostic factor in high-grade astrocytoma. J Neuro-

oncol 2000;49:157–63.

[49] Alavi A, Alavi JB, Lenkinski RE. Complementary

roles of PET and MR spectroscopy in the manage-

ment of brain tumors. Radiology 1990;177:617–8.

[50] DeSouza BX, Duhaney MO, Inscoe SW, et al. Me-

tabolism of human gliomas: assessment with H-1 MR

Spectroscopy and F-18 Fluorodeoxyglucose PET. Ra-

diology 1990;177:633–41.

[51] Go KG, Kamman RL, Mooyaart EL, et al. Localised

proton spectroscopy and spectroscopic imaging in

cerebral gliomas with comparison to positron emis-

sion tomography. Neuroradiology 1995;37:198–206.

[52] Nelson SJ, Vigneron DB, Dillon WP. Serial evalua-

tion of patients with brain tumors using volume MRI

and 3D 1H MRSI. NMR Biomed 1999;12:123–38.

[53] Van der Sanden B, Rijken P, Heerschap A, et al. In

vivo 31P magnetic resonance spectroscopy and mor-

phometric analysis of the perfused vascular architec-

ture of human glioma xenografts in nude mice. Br J

Cancer 1997;75:1432–8.

[54] Bussink J, Kaanders J, Rijken P, et al. Multiparam-

eter analysis of vasculature, perfusion and prolifera-

tion in human tumour xenografts. Br J Cancer 1998;

77:57–64.

[55] Hallaq-Al HA, Zamora M, Fish BL, et al. MRI mea-

surements correctly predict the relative effects of tu-

mor oxygenating agents on hypoxic fraction in rodent

BA1112 tumors. Int J Radiat Oncol Biol Phys 2000;

47:481–8.

[56] Karczmar GS, River JN, Vijayakumar S, et al. Effects

of hyperoxia on T *2 and Resonance frequency

weighted magnetic resonance images of rodent tu-

mours. NMR Biomed 1994;7:3–11.

[57] Robinson SP, Howe FA, Griffiths JR. Noninvasive

monitoring of carbogen-induced changes in tumor

blood flow and oxygenation by functional magnetic

resonance imaging. Int J Radiat Oncol Biol Phys

1995;33:855–9.

[58] Parliament MB, Chapman JD, Urtasun RC, et al.

Noninvasive assessment of human tumor hypoxia

with 123I-iodoazomycin arabinoside: preliminary re-

port of a clinical study. Br J Cancer 1992;65:90–5.

[59] Urtasun RC, McEwan AJ, Parliament MB, et al.

Measurement of hypoxia in human tumors by SPECT

imaging iodoazomycin arabinoside. Br J Cancer

1996;74:209–12.

[60] Derlon JM, Bourdet C, Bustany P, et al. [11C]l-me-

thionine uptake in gliomas. Neurosurgery 1989;25:

720–8.

[61] Lilja A, Bergstrom K, Hartvig P, et al. Dynamic study

of supratentorial gliomas with L-methyl-11C-methi-

onine and positron emission tomography. AJNR Am

J Neuroradiol 1985;6:505–14.

[62] Mosskin M, Von Holst H, Bergstrom M, et al. Posi-

tron emission tomography with 11C-methionine and

computed tomography of intracranial tumours com-

pared with histopathologic examination of multiple

biopsies. Acta Radiol 1987;28:673–81.

[63] Mineura K, Sasjima T, Itoch Y, et al. Blood flow and

metabolism of central neurocytoma. A positron emis-

sion tomography study. Cancer 1995;76:1224–32.

[64] Leskinen S, Lapela M, Lindholm P, et al. Metabolic

imaging by positron emission tomography in oncol-

ogy. Ann Med 1997;29:271–4.

[65] Ito M, Lammertsma AA, Wise RJS, et al. Measure-

ment of regional cerebral blood flow and oxygen uti-

lisation in patients with cerebral tumours using 15O

and positron emission tomography: analytical tech-

niques and preliminary results. Neuroradiology 1982;

23:63–74.

[66] Lammertsma AA, Wise RJS, Heather JD, et al. Cor-

rection for the presence of intravascular oxygen15-in

steady-state technique for measuring regional oxygen

extraction ratio in the brain: Results in normal sub-

jects and brain tumour and stroke patients. J Cereb

Blood Flow Metab 1983;3:425–31.

[67] Lammertsma AA, Wise RJS, Cox TCS, et al. Mea-

surement of blood flow, oxygen utilisation, oxygen

extraction ratio, and fractional blood volume in hu-


man brain tumours and surrounding oedematous tis-

sue. Br J Radiol 1985;58:725–34.

[68] Rhodes CG, Wise RJS, Gibbs JM, et al. In vivo

disturbance of the oxidative metabolism of glucose

in human cerebral gliomas. Ann Neurol 1983;14:

614–26.

[69] Arnold DL, Shoubridge EA, Villemure JG, et al.

Proton and phosphorus magnetic resonance spectro-

scopy of human astrocytomas in vivo. Preliminary

observations on tumor grading. NMR Biomed

1990;3:184–9.

[70] Arnold DL, Emrich JF, Shoubridge EA, et al. Char-

acterization of astrocytomas, meningiomas, and pitui-

tary adenomas by phosphorus magnetic resonance

spectroscopy. J Neurosurg 1991;74:447–53.

[71] Chang L, McBride D, Miller BL, et al. Localized in

vivo 1H magnetic resonance spectroscopy and in vitro

analyses of heterogenous brain tumors. J Neuroimag-

ing 1995;5:157–63.

[72] Edzes HT, Teerlink T, van der Knaap MS, et al.

Analysis of phospholipids in brain tissue by 31P

NMR at different compositions of the solvent system

chloroform-methanol-water. Magn Reson Med 1992;

26:46–59.

[73] Hugg JW, Matson GB, Twieg DB, et al. Phospho-

rus-31 MR spectroscopic imaging (MRSI) of normal

and pathological human brains. Magn Reson Imaging

1992;10:227–43.

[74] Lehnhardt FG, Rohn G, Ernestus RI, et al. 1H and

31P-MR spectroscopy of primary and recurrent hu-

man brain tumors in vitro: malignancy-characteristic

profiles of water soluble and lipophilic spectral com-

ponents. NMR Biomed 2001;14:307–17.

[75] Merchant TE, van der Ven LT, Minsky BD, et al. 31P

NMR phospholipid characterization of intracranial tu-

mors. Brain Res 1994;649:1–6.

[76] Peeling J, Sutherland G. High resolution 1H NMR

spectroscopy studies of extracts of human cerebral

neoplasms. Magn Reson Med 1992;24:123–36.

[77] Henry RG, Vigneron DB, Fischbein N, et al. Com-

parison of MRSI and cerebral blood volume imaging

of gliomas. In: Programs and Abstracts of the 5th

Annual Meeting International Society of Magnetic

Resonance in Medicine. Vancouver, British Colum-

bia, Canada; 1997. p. 1123.

[78] Alger JR, Frank JA, Bizzi A, et al. Metabolism of

human gliomas: assessment with H-1 MR spectros-

copy and F-18 fluorodeoxyglucose PET. Radiology

1990;177:633–41.

[79] Damaerel P, Johannik K, Van Hecke P, et al. Local-

ized 1H NMR spectroscopy in fifty new cases of

newly diagnosed intracranial tumors. J Comput Assist

Tomogr 1991;15:67–76.

[80] Kugel H, Heindel W, Ernestus RI, et al. Human brain

tumors: spectral patterns detected with localized H-1

MR spectroscopy. Radiology 1992;183:701–9.

[81] Okunieff PG, Koutcher JA, Gerweck L, et al. Tumor

size dependent changes in murine fibrosarcoma: use

of in-vivo 31P NMR for non-invasive evaluation of

tumor metabolic status. Int J Radiat Oncol Biol Phys

1986;12:793–9.

[82] Wendland MF, Sujata BI, Karen KF, et al. Correla-

tions between in-vivo 31pMRS measurements, tumor

size, cell survival, and hypoxic fraction in the murine

EMT6 tumor. Magn Reson Med 1992;25:217–32.

[83] Vaupel P, Okunieff P, Kallinowski F, et al. Correla-

tions between 31p-NMR spectroscopy and tissue O2

tension measurements in a murine fibrosarcoma. Ra-

diat Res 1989;120:477–93.

[84] Evelhoch JL, Sapareto SA, Nussbaum GH, et al. Cor-

relations between 31P NMR spectroscopy and 15O

perfusion measurements in the RIF-1 murine tumor

in-vivo. Radiat Res 1986;106:122–31.

[85] Rofstad EK, Demuth P, Fenton BM, et al. 31P nuclear

magnetic resonance spectroscopy studies of tumor

energy metabolism and its relationship to intravessel

oxyhemoglobin saturation status and tumor hypoxia.

Cancer Res 1988;48:5440–6.

[86] Nielson FU, Daugaard P, Bentzen L, et al. Effect of

changing tumor oxygenation on glycolytic metabo-

lism in a murine C3H mammary carcinoma assessed

by in vivo nuclear magnetic resonance spectroscopy.

Cancer Res 2001;61:5318–25.

[87] Taylor JS, Tofts PS, Port R, et al. MRI of tumor

microcirculation: promise for the new millennium.

J Magn Reson Imaging 1999;10:903–7.

[88] Tofts PS, Brix G, Buckley DL, et al. Estimating ki-

netic parameters from dynamic contrast-enhanced T1-

weighetd MRI of a diffusible tracer: standardized

quantities and symbols. J Magn Reson Imaging

1999;10:223–32.

[89] Verhoye M, van der Sanden BPJ, Rijken PFJW, et al.

Assessment of the neovascular permeability in glioma

xenografts by dynamic T1 MRI with gadomer-17.

Magn Reson Med 2002;47:305–13.

[90] Gillies RJ, Bhujwalla ZM, Evelvoch J, et al. Applica-

tions of magnetic resonance in model systems: tumor

biology and physiology. Neoplasia 2000;2:139–51.

[91] Barbier EL, Boer JA, Peters AR, et al. A model of the

dual effect of gadopentate dimeglumine on dynamic

brain magnetic resonance images. J Magn Reson

Imaging 1999;10:242–53.

[92] Goda F, Bacic G, O’hara JA, et al. The relationship

between partial pressure of oxygen and perfusion in

two murine tumors after X-ray irradiation: a com-

bined gadopentate dimeglumine dynamic magnetic

resonance imaging and in vivo electron paramagnetic

resonance oximetry study. Cancer Res 1996;56:

3344–9.

[93] Van der Sanden BPJ, Rozijin TH, Rijken PFWJ, et al.

Noninvasive assessment of the functional neovascu-

lature in 9L-Glioma growing in rat brain by dynamic

1H magnetic resonance imaging of gadolinium up-

take. J Cereb Blood Flow Metab 2000;20:861–70.

[94] Tofts PS, Kermode AG. Measurement of the blood-

brain barrier permeability and leak space using dy-

namic MRI. Fundamental concepts. Magn Reson

Med 1991;17:357–67.


[95] Aronen HJ, Gazit IE, Louis DN, et al. Cerebral blood

volume maps of gliomas: comparison with tumor

grade and histologic findings. Radiology 1994;191:

41–51.

[96] Bernsen H, Rijken P, Oostendorp T, et al. Vascularity

and perfusion of human gliomas xenografted in athy-

mic nude mouse. Br J Cancer 1995;71:721–6.

[97] Donahue K, Krouwer HGJ, Rand SD, et al. Utility of

simultaneously acquired gradient-echo and spin-echo

cerebral blood volume and morphology maps in brain

tumor patients. Magn Reson Med 2000;43:845–53.

[98] Dewhirst MW, Kimura H, Rehmus SW, et al. Micro-

vascular studies on the origins of perfusion-limited

hypoxia. Br J Cancer 1996;27:S247–51.

[99] Fenton BM, Paoni SF, Lee J, et al. Quantification of

tumour vasculature and hypoxia by immunohisto-

chemical staining and HbO2 saturation measure-

ments. Br J Cancer 1999;79:464–71.

[100] Rijken P, Bernsen H, Peters J, et al. Spatial relation-

ship between hypoxia and the (perfused) vascular net-

work in a human glioma xenograft: a quantitative

multi-parameter analysis. Int J Radiat Oncol Biol

Phys 2000;48:571–82.

[101] Vaupel P, Kelleher DK, Hockel M. Oxygenation sta-

tus of malignant tumors: pathogenesis of hypoxia and

significance for tumor therapy. Semin Oncol 2001;

28:29–35.

[102] Van der Sanden B, Heerschap A, Simonetti AW, et al.

Characterization and validation of noninvasive oxy-

gen tension measurements in human glioma xeno-

grafts by 19F-MR relaxometry. Int J Radiat Oncol

Biol Phys 1999;44:649–58.

[103] Boxerman JL, Rosen BR, Weiskoff RM. Signal-to-

noise analysis of cerebral blood volume maps from

dynamic NMRI studies. J Magn Reson Imaging

1997;7:528–37.

[104] Kassner A, Annesley DJ, Zhu XP, et al. Abnormalities

of the contrast re-circulation phase in cerebral tumors

demonstrated using dynamic susceptibility contrast-

enhanced imaging: a possible marker of vascular tor-

tuosity. J Magn Reson Imag 2000;11:103–13.

[105] Le Duc G, Peoch M, Remy C, et al. Use of T2-

weighted susceptibility contrast MRI for mapping

the blood volume in the glioma-bearing rat brain.


[106] Levin JM, Kaufman MJ, Ross MH, et al. Sequential

dynamic susceptibility contrast MR experiments in

human brain: residual contrast agent effect, steady

state, and hemodynamic perturbation. Magn Reson

Med 1995;34:655–63.

[107] Neeman M, Provenzale JM, Dewhirst MW. Magnetic

resonance imaging applications in the evaluation of

tumor angiogenesis. Semin Radiat Oncol 2001;11:

70–82.

[108] Petrella JR, Provenzale JM. MR perfusion imaging of

the brain: techniques and application. AJR Am J

Roentgenol 2000;175:207–19.

[109] Sugahara T, Korogi Y, Kochi M, et al. Perfusion sen-

sitive MRI of gliomas: comparison between gradient-

echo and spin-echo echo-planar imaging techniques.

AJNR Am J Neuroradiol 2001;22:1306–15.

[110] Wong JC, Provenzale JM, Petrella JR. Perfusion MRI

of brain neoplasms. AJR Am J Roentgenol 2000;174:

1147–57.

[111] Kim S, Ugurbil K. Functional magnetic resonance

imaging of the human brain. J Neurosci Methods

1997;74:229–43.

[112] Ogawa S, Lee TM, Kay AR, et al. Brain magnetic

resonance imaging with contrast dependent on blood

oxygenation. Proc Natl Acad Sci U S A 1990;87:

9868–72.

[113] Hoge RD, Atkinson J, Gill B, et al. Investigation of

BOLD signal dependence on cerebral blood flow and

oxygen consumption: the deoxyhemoglobin dilution

model. Magn Reson Med 1999;42:849–63.

[114] Hyder F, Kennan RP, Kida I, et al. Dependence of

oxygen delivery on blood flow in rat brain: A 7 tesla

nuclear magnetic resonance study. J Cereb Blood

Flow Metab 2000;20:485–98.

[115] Zhu X, Kim S, Anderson P, et al. Simultaneous oxy-

genation and perfusion imaging study of functional

activity in primary visual cortex at different visual

stimulation frequency: quantitative correlation be-

tween BOLD and CBF changes. Magn Reson Med

1998;40:703–11.

[116] Kennan RP, Scanley BE, Gore JC. Physiologic ba-

sis for BOLD MR signal changes due to hypoxia/

hyperoxia: separation of blood volume and mag-

netic susceptibility effects. Magn Reson Med

1997;37:953–6.

[117] Kwong KK, Wanke I, Donahue K, et al. EPI imaging

of global increase of brain MR signal with breath-

hold preceded by breathing O2. Magn Reson Med

1995;33:448–52.

[118] Rostrup E, Larsson HB, Toft PB, et al. Signal

changes in gradient echo images of human brain in-

duced by hypo- and hyperoxia. NMR Biomed 1995;

8:41–7.

[119] deCrespigny A, Wendland MF, Derugin NN, et al.

Real-time observation of transient focal ischemia

and hyperemia in cat brain. Magn Reson Med 1992;

27:391–7.

[120] Houston GC, Papadakis NG, Carpenter TA, et al.

Mapping of the cerebral response to hypoxia mea-

sured using graded asymmetric spin echo EPI. Magn

Reson Imaging 2000;18:1043–54.

[121] Jezzard P, Heineman F, Taylor J, et al. Comparison of

EPI gradient-echo contrast changes in cat brain

caused by respiratory challenges with direct simula-

tion evaluation of cerebral oxygenation via a cranial

window. NMR Biomed 1994;7:35–44.

[122] Turner R, Le Bihan D, Moonen C, et al. Echoplanar

time course MRI of cat brain oxygenation changes.


[123] Berkowitz BA. Role of dissolved plasma oxygen hy-

peroxia-induced contrast. J Magn Reson Imaging

1997;15:123–6.

[124] Duong T, Iadecola C, Kim S. Effect of hyperoxia,


hypercapnia, and hypoxia on cerebral interstitial oxy-

gen tension and cerebral blood flow. Magn Reson

Med 2001;45:61–70.

[125] Zhong J, Kennan RF, Fulbright RK, et al. Quantifi-

cation of intravascular and extravascular contribu-

tions to BOLD effects induced by alteration in

oxygenation or intravascular contrast agents. Magn

Reson Med 1996;40:526–36.

[126] Karczmar GS, Kuperman VY, River JN, et al. Mag-

netic resonance measurement of response to hyper-

oxia differentiates tumors from normal tissue and

may be sensitive to oxygen consumption. Invest Ra-

diol 1994;29:S161–3.

[127] Kuperman VYu, River JN, Lewis MZ, et al. Changes

in T2*-weighted images during hyperoxia differenti-

ate tumors from normal tissue. Magn Reson Med

1995;33:318–25.

[128] Robinson SP, Collingridge DR, Howe FA, et al. Tu-

mour response to hypercapnia and hyperoxia moni-

tored by FLOOD magnetic resonance imaging. NMR

Biomed 1999;12:98–106.

[129] Kolbitsch C, Schocke M, Hormann C, et al. Effects of

hyperoxia and hypocapnia on regional venous oxygen

saturation in the primary visual cortex in conscious

humans. Br J Anaesth 1999;83:835–8.

[130] Lund VE, Kentala E, Scheinin H, et al. Heart rate

variability in healthy volunteers during normobaric

and hyperbic hyperoxia. Acta Physiol Scand 1999;

167:29–35.

[131] Tadamura E, Hatabu H, Li W, et al. Effect of oxygen

inhalation on relaxation times in various tissues.

J Magn Reson Imaging 1997;7:220–5.

[132] Bates S, Yetkin Z, Jesmanowicz A, et al. Artifacts in

functional magnetic resonance imaging from gaseous

oxygen. J Magn Reson Imaging 1995;4:443–5.

[133] Yetkin FZ, Mendelsohn D, Purdy P, et al. Magnetic

resonance imaging of oxygenation in human brain

and tumors. Radiology 2001;218:614.

[134] Yetkin Z, Mendelsohn D, Purdy P. Oxygen as a con-

trast agent in brain imaging: technique and MRI char-

acteristics. In: Proceedings of the 39th Annual

Meeting of American Society of Neuroradiology,

Boston: 2001. p. 114.

[135] Yetkin Z, Mendelsohn D, Purdy P, et al. Evaluation of

brain tumor oxygenation using MRI. In: Proceedings

of the 39th Annual Meeting of American Society of

Neuroradiology, Boston (MA): 2001. p. 43.

[136] Lin W, Paczynski R, Celik A, et al. Experimental

hypoxemic hypoxia: Changes in R2* of brain paren-

chyma accurately reflect the combined effects of

changes in arterial and cerebral venous oxygen satu-

ration. Magn Reson Med 1998;39:474–81.

[137] LinW, Celik A, Paczynski RP, et al. Quantitative mag-

netic resonance imaging in experimental hypercapnia:

improvement in the relation between changes in brain

R2* and the oxygen saturation of venous blood after

correction for changes in cerebral blood volume.

J Cereb Blood Flow Metab 1999;19:853–62.

[138] Wright GA, Hu Bob S, Macovski A. Estimating oxy-

gen saturation of blood in vivo with MRI at 1.5 T.

Magn Reson Imaging 1991;1:275–83.

[139] Zijl PCM, Eleff SM, Ulatowski JA, et al. Quantitative

assessment of blood flow, blood volume and blood

oxygenation effects in functional magnetic resonance

imaging. Nat Med 1998;4:159–67.

[140] Foltz WD, Merchant N, Downar E, et al. Coronary

venous oximetry using MRI. Magn Reson Med

1999;42:837–48.

[141] An H, Lin W. Quantitative measurements of cerebral

blood oxygen saturation using magnetic resonance

imaging. J Cereb Blood Flow Metab 2000;20:

1225–36.


Date post:	30-Dec-2016
Category:	Documents
Upload:	dianne
View:	212 times
Download:	0 times

Hypoxia imaging in brain tumors

Documents