Neuroradiology (2004) 46: 118–121
DOI 10.1007/s00234-003-1121-y

D. Sarma
R. I. Farb
D. J. Mikulis
K. G. terBrugge

Received: 6 March 2003
Accepted: 29 July 2003
Published online: 16 January 2004
Ó Springer-Verlag 2004
D. Sarma Æ R. I. Farb Æ D. J. Mikulis
K. G. terBrugge (&)
Fell Pavilion 3-210, Toronto Western
Hospital, 399 Bathurst Street,
Toronto, ON M5T 2S8, Canada
E-mail: karel.terbrugge@uhn.on.ca
Tel.: +1-416-6035800 Extn. 5163
Fax: +1-416-6034257

DIAGNOSTIC NEURORADIOLOGY

Reversal of restricted diffusion in cerebral
venous thrombosis: case report

Abstract We report a patient with
extensive cerebral venous thrombosis who showed complete reversal
of restricted diffusion on MRI,
accompanied by excellent clinical
recovery. The implications of these
findings in relation to interpretation
of diffusion changes in cerebral
venous thrombosis are discussed,
and differences with arterial stroke
are highlighted.

Introduction
Diffusion sensitive MR techniques have revolutionized
imaging of stroke. In contrast to arterial stroke, however,
the evolution of diffusion abnormalities in venous thrombosis and infarction is less well understood. We present
a patient with extensive cerebral venous thrombosis
leading to loss of consciousness and seizures. Following
treatment there was complete reversal of restricted diffusion initially seen on MRI, accompanied by excellent
clinical recovery. We discuss the implications of these
abnormalities, and highlight the need for better understanding of the relationship between diffusion changes
and pathophysiology of cerebral venous thrombosis.

Case report
A 23-year-old woman presented with a 5-day history of increasing
bifrontal headaches, followed by nausea and vomiting for 2 days.
She had started taking oral contraceptives 1 month prior to presentation. There was no known prior illness or history of recurrent
headaches. CT showed high density along the superior and inferior
sagittal and straight sinuses and vein of Galen, indicating extensive
dural sinus and deep vein thrombosis (Fig 1a, b). Intravenous
heparin was initiated.

Keywords Intracranial sinus thrombosis Æ Diffusion-weighted imaging Æ Cerebral infarct

On arrival at our institution the patient had an episode of tonicclonic seizures, treated with phenytoin. She was in a postictal state
at the initial examination. Heparin was continued and she underwent MRI the next day which included contrast-enhanced autotriggered elliptic centric-ordered venography [1], which revealed
more extensive venous thrombosis (Fig 1c, d). In addition to the
sites identified on CT, there was involvement of the internal cerebral veins, torcular and transverse sinuses. The petrosal sinuses and
right superior ophthalmic vein were dilated, indicating alternative
channels of venous drainage.
On diffusion weighted imaging (DWI) there were areas of high
signal in the lentiform and caudate nuclei, thalami, left frontal and
parietal and right parietotemporal regions. There were corresponding dark areas on the apparent diffusion coefficient (ADC)
maps, indicating restricted diffusion (Fig 1e, f). No hemorrhage
was seen on CT in these sites. The patient steadily recovered on
heparin, with a gradually increasing level of consciousness. Neurological examination was normal on the fifth day of therapy. She
was discharged on warfarin.
On follow-up visits she had no headache or neurological deficits and returned to all former activities. Her only complaint was
mild difficulty in word-finding, which was not noticeable to her
family members, who described her speech as normal. Investigations revealed no factors predisposing to cerebral venous
thrombosis.
MR at 10 weeks showed patency of the dural sinuses except for
a residual filling defect in the superior sagittal sinus (Fig 2a, b).
DWI and ADC maps showed complete resolution of restricted
diffusion (Fig 2c, d). The areas of involvement showed no residual
signal change on MRI.

119

Fig. 1a–f Initial studies. a, b CT shows high-density thrombus in
the internal cerebral veins, vein of Galen, straight and superior
sagittal sinuses (arrows). c, d Contrast-enhanced MR venogram
shows absent flow in superior and inferior sagittal sinuses, torcular
and transverse sinuses, in addition to the sites of thrombus on CT.
A dilated superior ophthalmic vein (arrowheads) provides collateral
venous outflow. e Diffusion-weighted imaging (DWI) shows high
signal in basal ganglia and thalamus bilaterally. f Map of apparent
diffusion coefficients (ADC) shows these areas as dark, indicating
restricted diffusion

Discussion
Cerebral venous thrombosis (CVT) has multiple causes,
often related to hereditary or acquired predispositions
[2], including hypercoagulable states such as oral contraceptive use. Presentation is variable and initial
symptoms and signs are nonspecific. However, slow,
stepwise progression, starting with headaches and
vomiting and leading to neurological deficits and
seizures is often seen [2].
MRI is now used for detection of CVT. Contrastenhanced MR venography, such as we used [1], has
obviated conventional angiography for diagnosing the

condition or defining its extent. DWI is routinely added
to MRI investigation of CVT to identify cytotoxic
edema and venous infarcts, as in arterial stroke.
The concepts underlying tissue water diffusion and its
role in stroke have been of interest since the introduction
of DWI. The latter allows assessment of molecular
motion of water in biological tissues. This led to a major
advance in ischemic stroke imaging when it was observed that restricted water movement occurred. DWI
shows water proton diffusion, that occurs freely in the
extracellular compartment, but is relatively restricted
within cells. Shifting of water molecules from extra- to
intracellular compartments and their subsequent intracellular restriction is thought to be responsible for
ischemia-related abnormalities on DWI, believed to
represent cytotoxic edema [3, 4, 5, 6, 7, 8].
In arterial stroke cytotoxic edema is caused by failure
of ATP-dependent ion transport of sodium and calcium
across the cell membrane. With rapid accumulation of
sodium in ischemic cells, water follows to maintain
osmotic equilibrium. Acute cytotoxic edema may be
reversible if reperfusion occurs, as in animal models [9,
10] and also reported in humans after thrombolysis [11].
However, in most clinical situations, cytotoxic edema

120

Fig. 2a–d Imaging after treatment. a, b MR venogram shows
patency of dural sinuses and deep cerebral veins (cf Fig 1c, d).
c, d DWI and ADC map show complete reversal of restricted
diffusion seen earlier (Fig 1e, f)

quickly progresses to cell death and necrosis. Restriction
of diffusion, as seen on ADC maps, is claimed to be
irreversible, and is taken to represent the infarcted zone
in arterial stroke.
Data on the time-course of DWI changes and
infarction in CVT are limited to case reports and small
series. Vasogenic edema is commonly seen, but so to are
areas of reduced ADC, suggesting cytotoxic edema [7,
12]. Vasogenic edema, with normal or increased ADC,
was the most common finding in 14 patients with confirmed CVT. However, low ADC suggesting cytotoxic
edema was also often seen [13].
In another series of nine patients it was concluded
that the pattern of DWI/ADC brain lesions in CVT is
more heterogeneous than previously thought [14]. ADC
were increased in five patients and decreased in three;
one patient showed reduction in ADC to a degree usually reported with arterial infarcts, but the changes later
resolved almost completely.
Reversal of ‘‘cytotoxic edema’’ in a case of CVT has
been reported following endovascular thrombolysis [15].
Another report mentions marked resolution of diffusion
and clinical abnormalities following intravenous
heparin, though the findings refer to trace images and
not ADC maps [16].
Interpretation and prognostic value of reduced ADC
suggesting ‘‘cytotoxic edema’’ in CVT do not necessarily parallel those of arterial stroke, as demonstrated

by our patient and other reported cases. Concepts
applying to arterial infarcts may not be applicable to
CVT. Venous thrombosis is a slower process compared
to the sudden drop in perfusion seen with arterial
stroke. Moreover, alternative outflow channels may
allow some perfusion to occur, albeit at lower rates. It
has been suggested that the presence and efficiency of
venous collaterals, as well as recanalization, may have
an impact on brain tissue damage and hence on the
prognosis of acute CVT [17].
In arterial stroke, the change in diffusion is thought
to be due to failure of cell water homeostasis caused by
disruption of the energy-dependent membrane pump
that prevents intracellular accumulation of water against
an osmotic gradient. However, there is no direct relationship between the onset of irreversal biomolecular
events signaling cell death and restriction of intracellular
water diffusion. Cell swelling begins as the effectiveness
of the membrane pump diminishes with decreasing
delivery of energy substrates. If blood flow is restored
before irreversible events occur, it is conceivable that
diffusion abnormalities could also reverse.
It has been speculated that diffusion changes in venous infarcts may be initially due to an increase in
transcapillary and interstitial pressure, causing restriction of diffusion, rather than ischemia leading to failure
of the ion pump [15, 16]. There may be little or no energy
failure, at least initially, contributing to a wider time
window in which diffusion changes are reversible.
Neuronal loss depends not only on the degree of
blood flow reduction, but also on its duration. In
penumbral tissues surrounding the core of infarction,
irreversible events have not yet occurred because flow
reduction is less profound. There is thus a ‘‘period of
grace’’ wherein timely restoration of blood flow permits
cell recovery—a concept that forms the basis of modern
stroke management.
In venous thrombosis, the perfusion deficit is less
profound. However, if it persists for long enough, cellular homeostasis can fail, leading to cytotoxic edema
and cell death. Since the perfusion deficits are less catastrophic, it is conceivable that a large population of
cells exists in which ‘‘intracellular edema’’ with restricted
diffusion is present, but the biomolecular events leading
to irreversible cell death have not yet occurred. We believe that the brain affected by venous occlusion behaves
more like the penumbral tissue in arterial ischemia.
Reopening of venous occlusion can then lead to restoration of cell homeostasis, reversal of ADC reduction,
and prevention of cell death in much the same way as in
the penumbral region in arterial stroke.
We speculate that the hemodynamics in CVT initially
lead to ‘‘intracellular edema’’ which is reversible for an
undefined, variable time. This would explain the complete disappearance of restricted diffusion seen in this
case, despite extensive venous thrombosis and gross

121

neurological deterioration for several days before recovery. The dilated petrosal sinuses and superior ophthalmic
vein probably provided an alternative venous outlet,
allowing enough perfusion to maintain cell viability,

despite the presence of restricted diffusion. We suggest
‘‘intracellular’’ rather than cytotoxic edema be used when
referring to restricted diffusion in CVT, as the latter is
generally claimed to imply irreversibility and cell death.

References
1. Farb RI, Scott JN, Willinsky RA,
Montanera WJ, Wright GA, terBrugge
KG (2003) Intracranial venous system:
gadolinium-enhanced three-dimensional
MR venography with auto-triggered
elliptic centric-ordered sequence: initial
experience. Radiology 226: 203–209
2. Lee SK, Kim BS, terBrugge KG (2002)
Clinical presentation, imaging and
treatment of cerebral venous thrombosis (CVT). Intervent Neuroradiol
8: 5–14
3. Le Bihan D, Breton E, Lallemand D,
Grenier P, Cabanis E, Laval-Jeantet M
(1988) MR imaging of intravoxel incoherent motion: application to diffusion
and perfusion in neurologic disorders.
Radiology 161: 401-407
4. Benveniste H, Hedlung LW, Johnson
GA (1992) Mechanism of detection
of acute cerebral ischemia in rats by
diffusion-weighted magnetic resonance
microscopy. Stroke 23: 746—754
5. Berry I, Gigaud M, Manelife C (1992)
Experimental focal cerebral ischemia
assessed with IVIM-MRI in the acute
phase at 0.5 tesla. Neuroradiology 34:
135—140

6. Warach S, Chien D, Li W, Ronthal M,
Edelman RR (1992) Fast magnetic resonance diffusion-weighted imaging of
acute human stroke. Neurology 42:
1717—1723
7. Yoshikawa T, Abe O, Tsuchiya K, et al
(2002) Diffusion-weighted magnetic
resonance imaging of dural sinus
thrombosis. Neuroradiology 44:
481–488
8. J. Fiebach, O. Jansen, P. Schellinger, S.
Heiland, W. Hacke, K. Sartor (2002)
Serial analysis of the apparent diffusion
coefficient time course in human stroke.
Neuroradiology 44: 294–298
9. Li F, Silva MD, Sotak CH, Fisher M
(2000) Temporal evolution of ischemic
injury evaluated with diffusion-,
perfusion-, and T2-weighted MRI.
Neurology; 54: 689–696
10. Singhal AB, Dijkhuizen RM, Rosen
BR, Lo EH (2002) Normobaric hyperoxia reduces MRI diffusion abnormalities and infarct size in experimental
stroke. Neurology 58: 945–952
11. Krueger K, Krugel H, Grond M, Thiel
A, Maintz D, Lackner K (2000) Late
resolution of diffusion-weighted MRI
changes in a patient with prolonged
reversible ischemic deficit after thrombolytic therapy. Stroke 31: 2715–2718

12. Forbes KPN, Pipe JG, Heiserman JE
(2001) Evidence for cytotoxic edema in
pathogenesis of cerebral venous infarction. AJNR 22: 450–455
13. Ducreux D, Oppenheim C, Vandamme
X, et al (2001) Diffusion-weighted
imaging patterns of brain damage
associated with cerebral venous thrombosis. AJNR 22: 261–268
14. Chu K, Kang DW, Yoon BW, Roh JK
(2001) Diffusion-weighted magnetic
resonance in cerebral venous thrombosis. Arch Neurol 58: 1569–1576
15. Manzione J, Newman GC, Shapiro A,
Santo-Ocampo R (2000) Diffusion- and
perfusion-weighted imaging of dural
sinus thrombosis. AJNR 21: 68–73
16. Jenkins J, Hussain N, Lee D, Jog MS
(2001) Reversible parkinsonism and
MRI diffusion abnormalities in cortical
venous thrombosis. Neurology
57: 364–366
17. Stolz E, Gerriets T, Bodeker RH,
Hugens-Penzel M, Kaps M (2002)
Intracranial venous hemodynamics is a
factor related to a favorable outcome in
cerebral venous thrombosis. Stroke 33:
1645–1650