Original Research
Comparative Study of Methods for Determining
Vascular Permeability and Blood Volume in Human
Gliomas
Judith U. Harrer, MD,
1
Geoff J.M. Parker, PhD,
2
Hamied A. Haroon, MSc,
2
David L. Buckley, PhD,
2
Karl Embelton, MSc,
2
Caleb Roberts, BSc,
2
Danielle Bale ´ riaux, MD,
3
and Alan Jackson, MBChB, PhD
2
*
Purpose: To characterize human gliomas using T
1
-
weighted dynamic contrast-enhanced MRI (DCE-MRI), and
directly compare three pharmacokinetic analysis tech-
niques: a conventional established technique and two
novel techniques that aim to reduce erroneous overestima-
tion of the volume transfer constant between plasma and
the extravascular extracellular space (EES) (K
trans
) in areas
of high blood volume.
Materials and Methods: Eighteen patients with high-grade
gliomas underwent DCE-MRI. Three kinetic models were
applied to estimate K
trans
and fractional blood plasma vol-
ume (v
p
). We applied the Tofts and Kermode (TK) model
without arterial input function (AIF) estimation, the TK
model modified to include v
p
and AIF estimation (mTK), and
a “first pass” variant of the TK model (FP).
Results: K
TK
values were considerably higher than K
mTK
and K
FP
values (P 0.001). K
mTK
and K
FP
were more com-
parable and closely correlated (= 0.744), with K
mTK
gen-
erally higher than K
FP
(P 0.001). Estimates of v
p(mTK)
and
v
p(FP)
also showed a significant difference (P 0.001); how-
ever, these values were very closely correlated (= 0.901).
K
TK
parameter maps showed “pseudopermeability” effects
displaying numerous vessels. These were not visualized on
K
mTK
and K
FP
maps but appeared on the corresponding v
p
maps, indicating a failure of the TK model in commonly
occurring vascular regions.
Conclusion: Both of the methods that incorporate a mea-
sured AIF and an estimate of v
p
provide similar pathophys-
iological information and avoid erroneous overestimation of
K
trans
in areas of significant vessel density, and thus allow
a more accurate estimation of endothelial permeability.
Key Words: dynamic contrast-enhanced MRI; brain tu-
mor; permeability measurements; blood– brain barrier;
perfusion imaging; cerebral blood volume
J. Magn. Reson. Imaging 2004;20:748 –757.
© 2004 Wiley-Liss, Inc.
TUMOR MICROVASCULATURE is characterized by a
disproportionate fraction of blood vessels in compari-
son to the tissue fraction, abnormal vessel morphology
and routing, and altered blood flow. Of particular inter-
est is the endothelial permeability of the newly devel-
oped vessels, since these characteristically show large
intercellular gaps that allow the passage of medium-
and large-sized molecules from the intravascular to the
extravascular extracellular space (EES) (1,2).
Several pharmacokinetic parameters, such as the
volume transfer constant (K
trans
) between plasma and
the EES, the fractional volumes of the EES (v
e
), and the
plasma (v
p
) can be derived from contrast agent (CA)
concentration curves obtained from T
1
-weighted dy-
namic contrast-enhanced MRI (DCE-MRI) after fitting a
pharmacokinetic model of CA distribution (3–7). Of the
various different approaches to analyze DCE-MRI data,
the most commonly applied is the Tofts and Kermode
(TK) model described in 1991, which is based on an
assumed arterial input function (AIF) that is derived
from a sample of the normal population (3). Its wide use
is most likely explained by its excellent stability and
simplicity of application (8,9). However, this technique
has been shown to be subject to a number of potential
errors. One problem is the assumption that the ob-
served CA concentration change in each voxel solely
reflects CA leakage into the EES. This leads to errone-
ously high K
trans
values caused by intravascular CA,
which itself contributes to the signal increase that often
affects many voxels (10 –13). This artifact has therefore
been referred to as “pseudopermeability.” Another
cause of error in the conventional TK model is the ap-
plication of a standardized vascular input function
1
Department of Neurology, Aachen University Hospital, Aachen, Ger-
many.
2
Imaging Science and Biomedical Engineering, University of Manches-
ter, Manchester, UK.
3
Service de Radiologie, Ho ˆpital Erasme, Cliniques Universitaires de
Bruxelles, Bruxelles, Belgium.
Contract grant sponsor: German Society of Clinical Neurophysiology
and Functional Imaging.
*Address reprint requests to: A.J., Imaging Science and Biomedical
Engineering, Faculty of Medicine, University of Manchester, Stopford
Building, Oxford Road, Manchester, M13 9PT, United Kingdom.E-mail:
Alan.Jackson@man.ac.uk
Received February 2, 2004; Accepted June 23, 2004.
DOI 10.1002/jmri.20182
Published online in Wiley InterScience (www.interscience.wiley.com).
JOURNAL OF MAGNETIC RESONANCE IMAGING 20:748 –757 (2004)
© 2004 Wiley-Liss, Inc. 748