Journal of Cellular Biochemistry Supplement 39:45–53 (2002)
Quantitative Assessment of Tumor Oxygen Dynamics:
Molecular Imaging for Prognostic Radiology
Ralph P. Mason,1* Sophia Ran,2and Philip E. Thorpe2
1Department of Radiology, U.T. Southwestern Medical Center, Dallas, Texas
2Department of Pharmacology, U.T. Southwestern Medical Center, Dallas, Texas
now available to characterize tumor oxygenation. This Prospect will consider a new method, Fluorocarbon Relaxometry
using Echo planar imaging for Dynamic Oxygen Mapping (FREDOM), which we have recently developed for oximetry,
Biochem. Suppl. 39: 45–53, 2002.
? 2002 Wiley-Liss, Inc.
One of the fundamental molecules governing the survival of mammalian cells is oxygen. Oxygen has
Key words: magnetic resonance imaging; hexafluorobenzene; oxygen; tumor; tissue factor; antibody targeting
The human genome has been sequenced and
it is now realized that cells are ultimately char-
acterized by gene expression and proteomics,
which govern phenotype. There is also an in-
creasing realization that environmental factors
(epigenetic) can strongly influence cell develop-
ment and response to therapy. This is partic-
ularly relevant to oncology. It has long been
appreciated that hypoxic tumor cells are more
resistant to radiotherapy [Gray et al., 1953].
Indeed, a threefold increase in radio resistance
may occur when cells are irradiated under hy-
poxic conditions compared with pO2>15 Torr
for a given single radiation dose. However,
recent modeling has indicated that the propor-
tion of cells in the range 0–20 Torr may be most
significant in terms of surviving a course of
1997]. Increasingly, there is evidence that
hypoxia also influences such critical character-
istics as angiogenesis, tumor invasion, and me-
Harris, 2001]. Moreover, repeated bouts of
intermittent hypoxic stress may be important
in stimulating tumor progression [Cairns et al.,
2001]. Thus, the ability to measure pO2non-
or chronic interventions becomes increasingly
Early work examined cells in vitro, where
ambient oxygen concentrations are readily con-
trolled. In vivo, hypoxia may be achieved by
clamping the blood supply to a tumor, but other
levels of oxygenation reflect the interplay of
supply and consumption. Robust fine needle
polarographic electrodes opened the possibility
of measuring pO2in tumors in situ in vivo to
define local pO2under baseline conditions or
with respect to interventions. In early work,
Cater and Silver [Cater and Silver, 1960]
showed the ability to monitor pO2at individual
locations in patient’s tumors with respect to
breathing oxygen. Later, Gatenby et al. 
showedthatpO2in atumor wascorrelatedwith
clinical outcome. Tumor oximetry received its
greatest boost with the development of the
Eppendorf Histograph polarographic needle
electrode system. This computer-controlled
distributions of tumor oxygenation and has
? 2002 Wiley-Liss, Inc.
Grant sponsor: The National Cancer Institute; Grant
numbers: RO1 CA74951, CA54168, CA79515; Grant spon-
sor: Cancer Imaging Program; Grant number: P20 CA
86354; Grant sponsor: NIH BRTP Facility; Grant number:
*Correspondence to: Ralph P. Mason, PhD, C. Chem,
Department of Radiology, U.T. Southwestern Medical
Center, 5323 Harry Hines Blvd., Dallas, TX 75390-9058.
Received 4 October 2002; Accepted 8 October 2002
Published online in Wiley InterScience
been applied extensively to clinical trials. Many
reports have now shown that tumors are highly
heterogeneous and have extensive hypoxia:
strong correlations have been shown in cervix
and headandnecktumorsbetween median pO2
or hypoxic fraction and survival or disease free
survival [Brizel et al., 1996; Ho ¨ckel et al., 1996;
Nordsmark et al., 1996; Fyles et al., 1998].
strong prognostic indicator and this device has
laid a convincing foundation for the value of
measuring pO2 in patients. However, the
Histograph is highly invasive and it is not pos-
sible to make repeated measurements at indi-
vidual locations, precluding dynamic studies
to assess the influence of interventions on
Thus, many other techniques have been
developed to assess tumor oxygenation more
or less directly (Table I) [Stone et al., 1993].
Magnetic resonance methods are attractive,
since MRisinherently non-invasive.The pione-
pO2could be imaged in various organs based on
the19F NMR spin-lattice relaxation rate (R1) of
perfluorocarbon reporter molecules following
i.v. infusion. Prompted by these studies, we
surveyed a number of PFCs and identified that
hexafluorobenzene (HFB, Fig. 1) has many
virtues as a pO2reporter [Mason et al., 1996].
Symmetry provides a single narrow19F NMR
signal and the spin lattice relaxation rate is
highly sensitive to changes in pO2, yet mini-
mally responsive to temperature. From a
TABLE I. Comparison of Tumor Oximetry Techniques
32 G needle
locations each 4
Global or map
[Hunjan et al., 2001;
Zhao et al., 2001a,
[Mason et al., 1994]
[Stone et al., 1993]
23 G needle i.t.
1 s to 6.5 min
10 min PEDRI
Current Needle electrode
Oxygen 26 G needle Single location?1 s
[Cater and Silver,
et al., 1988;
Zhao et al., 2002b;
Mason et al.,
Oxygen Current 26 G needleMultiple tracks 1 s per location [Brizel et al., 1996;
Ho ¨ckel et al.,
1996; Fyles et al.,
[Zhao et al., 2001b]
26 G needleTypically 2–4
[Liu et al., 2000]
[Howe et al., 1999]
[Stone et al., 1993]
[Stone et al., 1993]
[Stone et al., 1993]
31P NMR PCr, ATP, Pia
Non-invasive 2–10 cc
[Stone et al., 1993]
COMET DNA strand
[Stone et al., 1993]
46 Mason et al.
practical perspective, HFB is readily available
commercially in high purity, cheap and well
characterized in terms of lack of toxicity [Zhao
et al., 2001a]. Indeed, HFB exhibits no muta-
genicity, no teratogenicity or fetotoxicity and
the manufacturer’s material data safety sheet
indicates LD50>25 g/kg (oral-rat) and LC50
95 g/m3per 2 h (inhalation-mouse).
Recognizing that tumors are heterogeneous
and that pO2may fluctuate, we developed a
procedure, which allows repeated quantita-
tive maps of regional pO2to be achieved with
multiple individual locations simultaneously in
have applied FREDOM to diverse tumor types
and interventions [Hunjan et al., 2001; Zhao
et al., 2001a,b, 2002a,b; Song et al., 2002], as
discussed in detail below, but here we describe
the assessment of a human tumor xenograft
with respect to respiratory challenge and ad-
ministration of a vascular targeting agent.
Human L540 Hodgkin’s tumors were im-
planted subcutaneously in the flanks of male
mice and allowed to grow to ?1.5 cm diameter
(3–6 mm thick). Mice were anesthetized using
ketamine/xylazine. HFB (?40 ml, Lancaster)
was injected directly in both central and peri-
pheral regions of the disc-like tumors using a
Hamilton syringe (Reno, NV) with a custom
made fine sharp needle (32 G) using our stan-
was placed within a 2 cm single turn solenoid
coil in anOmega CSI 4.7 T horizontal bore mag-
net system with actively shielded gradients.
confirm the distribution of HFB. Tumor oxyge-
nation was estimated on a regional basis using
19F echo planar imaging (EPI) relaxometry of
a pulse burst saturation recovery (PBSR) pre-
(t), a single spin echo planar image was acquir-
ed. Applying the ARDVARC protocol (Alter-
Acquisitions to Reduce
[Hunjan et al., 2001] enhanced the quality of
the relaxation data, and typically a precision
of 2–5 Torr at each of 20–50 voxels was achiev-
ed within a tumor in 6.5 min. The spin-lattice
relaxation rate (R1¼1/T1) was calculated
for each voxel using a three-parameter fit of
by the Levenberg–Marquardt least-squares
algorithm. Selection criteria were applied
to the data: data were accepted provided
that the relative standard
(sT1/T1)<25% and 0<sT1<2.0 s. Oxygen
tension was estimated using the relation-
(0.018?0.00017T), where T is rectal tempera-
ture in 8C and R1 is the spin-lattice relaxation
rate in s?1[Le et al., 1997].
The inhaled gas could be manipulated via a
1 dm3/min). In other cases, the air-breathing
mice were injected i.v. with anti-VCAM1-tTF
(truncated tissue factor) coaguligand (200 ml¼
40 mg protein) to produce tumor-specific infarct
[Huang et al., 1997]. The coaguligand was
19F MR images were obtained to
and a typical pO2map from a Hodgkin’s lymphoma xenograft in a SCID mouse.
Hexafluorobenzene,the linearoxygendependence ofthe19F NMRspin-latticerelaxation rate(R1)
Dynamic Assessment of Tumor Oxygenation47
constructed by conjugating the MK2.7 mono-
clonal antibody directed against mouse VCAM-
1 to the extracellular domain of human tissue
factor (TF), as described previously [Ran et al.,
1998].Thrombosisoftumor vesselswas verified
histologically in experiments parallel to NMR
[Ran et al., 1998]. Briefly, mice were selected at
specific time points before and after adminis-
tration of coaguligand, anesthetized and per-
fusedwithheparinized saline. The tumorswere
Sections were cut, stained with hematoxylin-
eosin and the number of thrombosed vessels
Statistical significance of changes in oxyge-
nation was assessed using analysis of variance
(ANOVA) on the basis of Fisher PLSD and data
are quoted as mean?SE of the mean.
Typical baseline oxygenation ranged from
hypoxia to 50 Torr with mean 9.7?1.4 (SE)
Torr and median 7.2 Torr (Fig.2): no significant
changes occurred during three repeat measure-
ments while breathing air over a period of
24 min. Within 8 min of altering the inspired
gas to carbogen, therewas achangein oxygena-
tion with elevation in mean (P<0.01) and
median pO2, and decreased hypoxic fraction
(Fig. 2). Oxygen tension continued to rise some-
times reaching a plateau after 16 min with
mean pO2 significantly above baseline (P<
0.00001). Upon returning the inhaled gas to
air, tumor pO2gradually declined, but after
with baseline (P<0.01).
Administration of the anti-VCAM-1.TF co-
aguligand caused a rapid reduction in pO2in
Torr) with mean declining from 15.5?2.7 to
4.2?2.4 Torr (P<0.01) after 16 min and to
?1.5?2.4 Torr (P<0.001) after 70 min. Little
change was observed in regions, which were
initially poorly oxygenated (initial pO2<10
Torr), although some locations indicated in-
creased oxygenation (Fig. 3). Histology showed
open blood vessels in the untreated tumors, but
substantial thromboses at 4 h.
In the brief experiments described here, we
have demonstrated the ability to monitor both
baseline pO2distributions in a tumor in vivo
and dynamic changes with respect to interven-
tions. This approach allows multiple specific
locations tobeinterrogated simultaneouslyand
sequentially. The data indicate the importance
oftemporaland spatialresolution ininvestigat-
ing tumor physiology. During the period of
oxygenation in response to respiratory chal-
since it reveals distinct heterogeneity in tumor
each case data have been pooled from three repeat determina-
tions. Lower: Histograms of baseline with mouse breathing air.
Mean (x) pO2¼9.7?1.4 (SE) Torr, median pO2(m)¼7 Torr.
Center: Mouse breathing carbogen (95% O2/5% CO2): mean
pO2¼41?3.9 Torr, median 39 Torr. Top: Oxygenation of
individual voxels from respective pO2maps. The general trend
slower, upon return to breathing air. This emphasizes the need
for rapid time resolution and danger of pooling data. Variation
in mean (*) and median (D) pO2with respect to respiratory
challenge is overlaid. Each measurement required 6.5 min.
48Mason et al.
oxygenation and differential response to inter-
vention. In particular, the initially well-oxyge-
nated regions of the tumor showed a significant
decline in pO2(hypoxiation) following adminis-
tration of the coaguligand, whereas the less
well-oxygenated regions showed minimal re-
sponse (Fig. 3). This could be attributed to dif-
ferential vascular efficiency. One would expect
reduction in pO2 in well-oxygenated tumor regions (initial
Torr (P<0.01) after 16 min and to ?1.5?2.4 Torr (P<0.001)
after 70 min. Considering two representative individual tumor
regions (voxels) * and D with initially high pO2, both showed a
were initially poorly oxygenated (initial pO2<10 Torr), for
a: Administration of the coaguligand caused a rapid
example, *, although in some regions pO2was found to rise
significantly ~. b: H & E stained sections show open blood
vessels in control tumor, but distinct thromboses 4 h following
administration of coaguligand. c: Maps of pO2obtained using
FREDOM 5 and 50 min after infusion of coaguligand. Arrows
indicate regions showing particularly dramatic decrease or
increase in pO2.
Dynamic Assessment of Tumor Oxygenation49
pO2to be highest in well-perfused regions and
delivery of the coaguligand should also be most
efficient there. Meanwhile, less well-oxyge-
nated regions, and hence, by inference less
well-perfused regions would tend to be less
likely to beinfarcted. Indeed, one region (Fig.3)
to this region resulting from infarct elsewhere.
In future, it will be important to correlate pO2
measurements with blood flow as provided by
such non-invasive techniques as1H Dynamic
Contrast Enhanced (DCE) or Blood Oxygen
histological investigations. Specifically, many,
but not all, tumor capillaries were found to be
occluded. The great potential advantage of
tumor serves as its own control, and indeed,
individual voxels reveal both baseline hetero-
geneity and differential response to interven-
tions. Moreover, non-invasive
minimize the number of animals and quantity
of agent required. Since antibodies may be
savings could greatly enhance the efficiency of
We have previously applied the FREDOM
approach to various tumor types including
et al., 2001a,b, 2002a,b]. In diverse sublines of
the Dunning prostate R3327 tumor, we have
shown distinctly different oxygenation pat-
terns, which were related to tumor differentia-
tion, vascularity, and growth rate [Zhao et al.,
2000, 2002b]. Thus, the undifferentiated AT1
subline with volume doubling time of 5 days
showed distinct heterogeneity with pO2values
ranging from>50 Torr to hypoxia when anes-
thetized rats breathed air. In response to
breathing elevated oxygen (either oxygen or
carbogen) those regions initially well oxyge-
nated showed a rapid and significant response,
while initially poorly oxygenated regions show-
ed little, if any, response [Hunjan et al., 2001].
had greater baseline hypoxia and similar re-
differentiated HI subline (VDT 9 days) showed
very different behavior. Baseline oxygenation
was significantly higher, yet a substantial frac-
tion of large tumors was hypoxic. Remarkably,
all regions irrespective of initial pO2showed
rapid and significant increase in pO2 with
respect to breathing elevated oxygen [Zhao
(VDT 20 days) had significantly higher baseline
oxygenation, but the response to increased
oxygen breathing was sluggish compared to HI
application of the technique to mammary
13762NF adenocarcinomas [Song et al., 2002].
Baseline pO2was similar to the Dunning pro-
state AT1, but all regions responded to oxygen
intervention, though often slowly with con-
tinual increases over 45 min. Such data have
suggested that a given tumor type has char-
acteristic baseline pO2 distributions and re-
sponse to an intervention. However, baseline
pO2alone has not been a good indicator of the
potential response to an intervention between
different tumor types. If such data are con-
firmed in the clinic, they indicate the potential
importance of measuring pO2in the tumors of
individual patients, so that therapy may be
individualized and optimized with inclusion of
appropriate adjuvant interventions.
Histograph data have shown value in evalu-
patients may be stratified into well-oxygenated
and poorly oxygenated groups for differential
therapy. A next step should be the evaluation of
the tumors, to see whether their pO2can be
modulated by a relatively innocuous interven-
tion, such as breathing oxygen. Elevation of
tumor oxygenation might bring these hypoxic
tumors into the well-oxygenated range and
should show animprovedresponsetoradiation.
Non-responders (persistently hypoxic tumors)
could be considered for hypoxia selective cyto-
toxin therapy, such as tirapazamine [Brown,
1999] in an attempt to render them more
sensitive to treatment.
Of greater relevance to the biochemical com-
munity is the ability to screen interventions in
non-invasive techniques become increasingly
important. As combinatorial approaches gain
application, many more agents are available,
but typically in small quantities. In the past,
assessment of the efficacyof novel coaguligands
has relied on histological endpoints (destruc-
tive) or gross anatomical evaluation (slow). The
FREDOM procedure, described here, allows in-
dividual locations within a tumor to be followed
50 Mason et al.
for a period of hours. These procedures can be
used to rapidly screen new agents and suggest
evaluation. This will both spare animals and,
perhaps more significantly, reduce the amount
of drug required. The non-destructive imaging
technique will accelerate the discovery process
for targets, efficacious conjugates, and success-
ful development of drugs.
As shown in Table I, there are many alter-
native techniques available to investigate
tumor oxygenation. Electrodes have been con-
sidered by some to be a ‘‘gold standard’’ and we
have shown that pO2 distributions assessed
with the Eppendorf Histograph are commensu-
rate with FREDOM [Mason et al., 1999a]. We
have also shown similar dynamic response to
interventions such as hyperoxic gas inhalation
using fiber optic probes or electrodes or FRE-
DOM [Zhao et al., 2001b, 2002b]. Electrodes,
fiber optic probes, mass spectrometry probes,
of a sizable needle into tumors and report limit-
ed regions raising issues of sampling. While
our current implementation of FREDOM does
require aneedle forintratumoral(i.t.)injection,
the mobile fluid (HFB) allows a very fine sharp
at multiple locations simultaneously. Sampling
remains an issue, since we typically interrogate
5–10% of a given tumor volume. However, the
highly consistent intertumor behavior between
multiple tumors of a given type (and size)
HFB is highly volatile and clears from tumors
within 24 h, repeated measurements on sub-
highly consistent data achieved in tumors with
such successive measurements indicates the
effective representation of the true distribution
of oxygen tensions within the tumors [Zhao
reporter molecules can also be delivered i.v.
(e.g., oxygen sensitive phosphorescent dyes,
ESR sensitive agents, NMR sensitive fluorocar-
bon emulsions), but such an approach may bias
measurements towards vascular oxygenation
and particularly towards well-perfused regions
of a tumor [Mason et al., 1994; McIntyre et al.,
Specific classes of reporter molecule can
reveal hypoxia as a surrogate for pO2 (e.g.,
pimonidazole, EF5, CCI-103F) [Stone et al.,
1993]. Following i.v. infusion these agents
become reduced in tissues and are trapped.
However, in the presence of oxygen, they are
reoxidized and ultimately clear from the body.
Histological assessment of the distribution of
these agents provides microscopic indications
currently being tested in clinical trials. In
addition, various labeled derivatives have been
developed which are NMR or PET active. Other
nuclear imaging agents have been developed
specifically to image hypoxia, for example,
Cu-ATSM and the galactopyranoside IAZA
While many successful reporter molecules
have been developed to investigate tumor oxy-
genation, an alternative approach is to exploit
endogenous indicators of oxygenation. Near
of variation in tumor hematocrit (viz. blood
saturation [Liu et al., 2000]. Cheap instrumen-
tation can reveal rapid changes, but current
technology provides very coarse spatial resolu-
MRI approaches provide high spatial and
temporal resolution related to hemoglobin oxy-
gen saturation, but signal also responds to
variations in vascular volume, and flow [Howe
et al., 1999]. Moreover, changes in vascular
oxygenation may not coincide with pO2, which
is a balance between oxygen delivery and con-
sumption and clearance. Ultimately, we must
recognize that quantitative estimates of tumor
tissue pO2 have been shown to be directly
related to clinical outcome in several tumor
types, while other parameters such as vascular
of their prognostic value.
Since many biochemical pathways are under
oxygen regulation, they can provide an elegant
window on hypoxia, for example, induction
of Hif-1 and Glut-1 together with secondary re-
sponses such as increased production of VEGF,
NIP3, and tumor associated macrophage activ-
ity [Knowles and Harris, 2001]. While such
molecules could themselves indicate hypoxia, a
more versatile approach is to adopt the hypoxic
response elements as promoter sequences
coupled to reporter genes such as GFP (green
fluorescent protein) [Cao et al., 2001].
From a clinical perspective,19F MR remains
added to a clinical MR scanner for about $150k.
Dynamic Assessment of Tumor Oxygenation51
For preclinical biochemical studies, NMR sys-
tems have routine19F capabilities at 4.7, 7, or
9.4 T. Indeed, diverse reporter molecules have
been created to access such diverse parameters
as transmembrane pH, transmembrane chlo-
ride potential, [Ca2þ], temperature, and of
course pharmacokinetics of drugs such as 5-
FU [Mason, 1999].19F NMR can also be applied
to assess gene therapy, as demonstrated by
others, for monitoring the conversions of 5-FC
(fluorocytosine) to 5-FU (5-fluorouracil) and by
us with the novel class of agent PFONPG as a
substrate for b-galactosidase activity [Mason
et al., 2002].
Ultimately, the value of a technique depends
on its robustness, ease of use, and widespread
implementation. To date, few labs had adopted
the FREDOM approach because efficient inves-
sequence. With the recent upgrade of our own
instrumentation to the Varian Unity INOVA,
the software is now available on this popular
platform facilitating ready implementation
elsewhere. While the FREDOM technique is
currently limited to preclinical investigations,
approval for HFB to facilitate future clinical
applications. As with electrode approaches, we
foresee initial applications to readily accessible
tumors, for example, head and neck, cervical,
breast, and prostate. The ability to map pO2
in <8 min makes this technique a practical
proposition for application to patients, and the
value of monitoring dynamic changes in tumor
oxygenation has the potential to influence
In terms of research applications, it is known
to many acute interventions ranging from
irradiation to photodynamic therapy, various
chemotherapies, and experimental new appro-
aches such as vascular targeting agents shown
assessing the dynamic time course of such
interventions to provide clear insight into the
mode of action of therapeutic approaches and
aid in the high throughput screening of new
This work was supported in part by The
NationalCancer Institute(RO1 CA74951
(PET), CA54168 (PET), and CA79515 (RPM)
and in conjunction with Cancer Imaging Pro-
gram P20 CA 86354 and NIH BRTP Facility
Ballinger JR. 2001. Imaging hypoxia in tumors. [Review].
Semin Nucl Med 31:321–329.
Brizel DM, Scully SP, Harrelson JM, Layfield LJ, Bean JM,
Prosnitz LR, Dewhirst MW. 1996. Tumor oxygenation
predictsforthelikelihoodofdistant metastases inhuman
soft tissue sarcoma. Cancer Res 56:941–943.
Brown JM. 1999. The hypoxic cell: A target for selective
cancer therapy. Cancer Res 59:5863–5870.
Cairns RA, Kalliomaki T, Hill RP. 2001. Acute (cyclic)
hypoxia enhances spontaneous metastasis of KHT
murine tumors. Cancer Res 61:8903–8908.
Cao Y, Li C, Zhang X, Dewhirst M. 2001. A study of
hypoxia-induced gene expression in human tumors: 48th
Annual meeting of Radiation Research. San Juan, PR,
Cater D, Silver I. 1960. Quantitative measurements of
oxygen tension in normal tissues and in the tumors of
patients before and after radiotherapy. Acta Radiol 53:
Fyles AW, Milosevic M, Wong R, Kavanagh M-C, Pintile M,
Sun A, Chapman W, Levin W, Manchul L, Keane TJ, Hill
RP. 1998. Oxygenation predicts radiation response and
survival in patients with cervix cancer. Radiother Oncol
Gatenby RA, Kessler HB, Rosenblum JS, Coia LR,
Moldofsky PJ, Hartz WH, Broder G. 1988. Oxygen dis-
tribution in squamous cell carcinoma metastases and
its relationship to the outcome of radiation therapy. Int
J Radiat Oncol Biol Phys 14:831–838.
Gray L, Conger A, Ebert M, Hornsey S, Scott O. 1953. The
concentration of oxygen dissolved in tissues at time of
irradiation as a factor in radio-therapy. Br J Radiol 26:
Howe FA, Robinson SP, Rodrigues LM, Griffiths JR. 1999.
Flow and oxygenation dependent (FLOOD) contrast MR
imaging to monitor the response of rat tumors to
Huang X, Molema G, King S, Watkins L, Edington TS,
Thorpe PE. 1997. Tumor infarction in mice by anti-body
directed targeting of tissue factor to tumor vasculature.
HunjanS,Zhao D,Constantinescu A,HahnEW,AntichPP,
Mason RP. 2001. Tumor oximetry: Demonstration of an
enhanced dynamic mapping procedure using fluorine-19
echo planar magnetic resonance imaging in the Dunning
prostate R3327-AT1 rat tumor. Int J Radiat Oncol Biol
Ho ¨ckel M, Vaupel P. 2001. Tumor hypoxia: Definitions and
current clinical, biologic, and molecular aspects. J Natl
Cancer Inst 93:266–276.
Ho ¨ckel M, Schlenger K, Aral B, Mitze M, Scha ¨ffer U,
Vaupel P. 1996. Association between tumor hypoxia and
malignant progression in advanced cancer of the uterine
cervix. Cancer Res 56:4509–4515.
Knowles HJ, Harris AL. 2001. Hypoxia and oxidative stress
in breast cancer. Hypoxia and tumourigenesis. [Review].
Breast Cancer Res 3:318–322.
52 Mason et al.
Le D, Mason RP, Hunjan S, Constantinescu A, Barker BR, Download full-text
Antich PP. 1997. Regional tumor oxygen dynamics: 19F
PBSR EPI of hexafluorobenzene. Magn Reson Imaging
Liu H, Song Y, Worden KL, Jiang X, Constantinescu A,
Mason RP. 2000. Noninvasive investigation of blood
oxygenation dynamics of tumors by near-infrared spec-
troscopy. Appl Optic 39:5231–5243.
Mason RP. 1999. Transmembrane pH gradients in vivo:
Measurements using fluorinated vitamin B6 derivatives.
Curr Med Chem 6:481–499.
Mason RP, Antich PP, Babcock EE, Constantinescu A,
Peschke P, Hahn EW. 1994. Non-invasive determination
of tumor oxygen tension and local variation with growth.
Int J Radiat Oncol Biol Phys 29:95–103.
Mason RP, Rodbumrung W, Antich PP. 1996. Hexafluor-
obenzene: A sensitive 19F NMR indicator of tumor
oxygenation. NMR Biomed 9:125–134.
Mason RP, Constantinescu A, Hunjan S, Le D, Hahn EW,
Antich PP, Blum C, Peschke P. 1999a. Regional tumor
oxygenation and measurement of dynamic changes.
Radiat Res 152:239–249.
Mason RP, Constantinescu A, Ran S, Thorpe PE. 1999b.
Oxygenation in a human tumor xenograft: Manipulation
through respiratory challenge and anti-body directed
infarction. Proc 27th ISOTT. Dartmouth, NH, p 43.
MasonRP, OttenP, Li Y, Koeneman K. 2002.Genereporter
molecules: A novel approach revealing b-galactosidase
activity: ISMRM 10th Scientific Meeting. Honolulu,
Hawaii, USA, p 384.
McIntyre DJO, McCoy CL, Griffiths JR. 1999. Tumour
oxygenation measurements by 19F MRI of perfluorocar-
bons. Curr Sci 76:753–762.
Nordsmark M, Overgaard M, Overgaard J. 1996. Pretreat-
ment oxygenation predicts radiation response in advanc-
ed squamous cell carcinoma of head and neck. Radiother
Ran S,Gao B,Duffy S, Watkins L, RoteN, ThorpePE. 1998.
Infarction of solid Hodgkin’s tumors in mice by antibody-
directed targeting of tissue factor to tumor vasculature.
Cancer Res 58:4646–4653.
Song Y, Constantinescu A, Mason RP. 2002. Dynamic
breast tumor oximetry: The development of prognostic
radiology. Technol Cancer Res Treat (in press).
Stone HB, Brown JM, Phillips T, Sutherland RM. 1993.
Oxygen in humantumors: Correlations between methods
ofmeasurement andresponseto therapy. RadiatRes136:
Thomas SR. 1988. The biomedical applications of Fluorine-
19 NMR. In: Partain CL, Price RR, Patton JA, Kulkarni
MV, James AEJ, editors. Magnetic resonance imaging.
London: W.B. Saunders Co. p 1536–1552.
WoutersBG, BrownJM.1997.Cells at intermediateoxygen
levels can be more important than the ‘‘hypoxic fraction’’
in determining tumor response to fractionated radio-
therapy. Radiat Res 147:514–520.
Zhao D, Hahn EW, Constantinescu A, Mason RP. 2000. A
comparison of oxygen dynamics during respiratory chal-
lenge in two Dunning prostate tumor sublines having
diverse Tpots: Radiation Research Soc. Meeting. Albu-
Zhao D, Constantinescu A, Jiang L, Hahn EW, Mason RP.
2001a. Prognostic radiology: Quantitative assessment of
tumor oxygen dynamics by MRI. Am J Clin Oncol 24:
Zhao Z, Constantinescu A, Hahn EW, Mason RP. 2001b.
Tumor oxygen dynamics with respect to growth and
respiratory challenge: Investigation of the Dunning
prostate R3327-HI tumor. Radiat Res 156:510–520.
Zhao D, Constantinescu A, Hahn EW, Mason RP. 2002a.
Measurement of tumor oxygen dynamics predicts bene-
ficial adjuvant intervention for radiotherapy in Dunning
prostate R3327-HI tumors. Radiat Res (in press).
Zhao D, Constantinescu C, Hahn EW, Mason RP. 2002b.
Differential oxygen dynamics in two diverse Dunning
prostateR3327rattumorsublines (MAT-Luand HI)with
respect to growth and respiratory challenge. Int J Radiat
Oncol Biol Phys 53:744–756.
Dynamic Assessment of Tumor Oxygenation53