Journal of the Neurological Sciences 193 Ž2001. 43–47
www.elsevier.comrlocaterjns

Short communication

A CADASIL case with normal skin biopsy and without mutations in
exons 3 and 4 of the Notch3 gene q
G.R. de Freitas a , J. Miklossy b, S. Christen-Zach
¨ b, M. Reichhart a, J. Bogousslavsky a,)
a

Department of Neurology, Centre Hospitalier UniÕersitaire Vaudois, Lausanne CH1011, Switzerland
b
Department of Neuropathology, Centre Hospitalier UniÕersitaire Vaudois, Lausanne, Switzerland
Received 12 March 2001; received in revised form 3 July 2001; accepted 5 July 2001

Abstract
The diagnosis of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy ŽCADASIL. is usually
confirmed by genetic testing or skin biopsy. We here report the case of a 69-year-old woman with recurrent transient ischemic attacks
ŽTIAs. and strokes, seizures, and dementia without any mutations in exons 3 and 4 of the Notch3 gene and with a normal skin biopsy, but
who showed characteristic CADASIL abnormalities on brain pathological examination. Our findings suggest that negative results in these
two tests do not exclude the disease and a leptomeningeal biopsy or a second skin biopsy should be considered in such cases. q 2001
Elsevier Science B.V. All rights reserved.
Keywords: Cerebrovascular disorders; CADASIL; Diagnosis; Skin biopsy; Notch3 gene; Electron microscopy

1. Introduction
Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy ŽCADASIL. is an
adult-onset hereditary disorder characterized by recurrent
transient ischemic attacks ŽTIAs. and strokes, migraine,
seizures, and dementia w1x. The pathological hallmark of
this disease is the presence of characteristic granular osmiophilic material ŽGOM. within the basal membrane of
brain vascular smooth muscle cells; these vascular changes
were also later reported in nerve, striated muscle, and skin
w2–4x. In 1997, mutations within the Notch3 gene were
identified in these patients and the diagnosis of CADASIL
can now be confirmed by genetic testing w5x. In cases in
which no mutations are found, ultrastructural examination
of a skin biopsy should be considered w6x.
We here report the case of a woman who presented
recurrent TIAs and strokes, seizures, and dementia in
whom the skin biopsy was normal and DNA analysis
failed to detect mutations in the exons 3 and 4 of the
Notch3 gene. However, the diagnosis of CADASIL was
q

This work was supported in part by a European Neurological Society
grant to Dr. Gabriel R. de Freitas.
)
Corresponding author. Tel.: q41-21-314-1230; fax: q41-21-3141231.
E-mail addresses: gfreitas@nitnet.com.br ŽG.R. de Freitas.,
julien.bogousslavsky@chuv.hospvd.ch ŽJ. Bogousslavsky..

established by postmortem cerebral examination. This case
shows that standard tests for CADASIL can give falsenegative results and that further investigations are warranted in some patients.

2. Case presentation
In May 1995, a 69-year-old female patient suddenly
developed right hemiparesis and hypoesthesia, recovering
in about 10 min. Four years previously, she had suffered a
similar episode. The patient had bilateral glaucoma. Thirteen years previously, laminectomy was performed because of disc herniation. Amigdalectomy and surgery for
lower member varices had been performed several years
previously. There was no history of hypertension, diabetes,
hypercholesterolemia, cardiac disease, migraine, cigarette
smoking, or alcohol consumption. Her father had died at
the age of 55 due to pneumonia following a stroke, her
mother at the age of 93 from unknown causes, and her
brother at the age of 66 from colon carcinoma. She had
one daughter with a normal MRI who, from the age of 14
years, had suffered from seizures that were controlled with
phenobarbital. There was no other family history of cerebrovascular diseases, seizures, or migraine. Neurological
examination revealed only mild brachial right side weakness. Routine hematological and biochemical tests, clotting

0022-510Xr01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 5 1 0 X Ž 0 1 . 0 0 5 7 9 - 2

44

G.R. de Freitas et al.r Journal of the Neurological Sciences 193 (2001) 43–47

function tests, lipid metabolism, electrocardiogram,
echocardiogram, chest radiograph, extracranial Doppler ultrasonogram of the carotid and vertebral arteries, MRI
angiography, and transcranial Doppler were normal. MRI
revealed diffuse periventricular white matter abnormalities.
In July 1997, she had a right-sided partial motor seizure
with secondary generalization and was admitted to our
hospital. An electroencephalogram revealed generalized
low amplitude slowing with frontal intermittent rhythmic
delta activity predominantly in the left fronto-parieto-occipital areas. Blood and urine amino acids, serum lactate,
pyruvate, arylsulfatase, betagalactosidase, protein electrophoresis, immunoglobulin, cortisol, ceruloplasmin, carnitine, protein C, protein S, antithrombin III, vitamin B12,
and thyroid hormone values were normal. DNA analysis
for mitochondriopathies and tests for striated muscle antibodies, anticardiolipin antibodies, antinuclear antibodies,

and rheumatoid factor were negative. Low levels of vitamin B6 Ž0.3 mgr100 ml, normal range 3–83. and folic
acid Ž4 nmolrl, normal range 7–39. were found. Homocysteine levels were high Ž29.6 mmolrl, normal range
5–15. and vitamins were administered. Examination of the
cerebrospinal fluid showed 2471 erythrocytesrmm3 and
340 mgrl of protein, and serologic tests for herpes simplex, syphilis, and Lyme disease were negative. SPECT
Tc99 showed hypoperfusion in the frontal and temporal
lobes bilaterally and in the left parietal lobe. Nerve conduction studies and muscle biopsy results were normal.
Genomic DNA was extracted from peripheral blood leukocytes for analysis for mutations in exons 3 and 4 of the
Notch3 gene. Electron microscopy of the skin biopsy was
normal. Cognitive deterioration progressed. In October
1997, neurological examination showed a lethargic, disoriented patient. Gait and constructive apraxia were noted.

Fig. 1. Illustration of the characteristic white matter and vascular changes in our CADASIL case. ŽA. Small lacunar infarcts Žarrows. in the parietal white
matter. Paraffin section from the parietal region, stained for myelin. Loyez stain Ž=20.. ŽB. Thickening and fibrosis of the wall of a small deep-penetrating
artery. HE stain Ž=300.. ŽC. Thickening, reduplication, and fragmentation of the internal elastic lamina of two small arteries. VGE Ž=150.. ŽD.
Perivascular cuff of macrophages around a small cerebral vessel. PAS Ž=300.. ŽE. Concentric thickening of the arterial wall showing marked fibrosis.
VGE Ž=200.. ŽF. Microphotograph showing a perivascular inflammatory infiltrate. The inflammatory cells are positive for leucommun. Some of the
immunoreactive cells are infiltrating the vessel wall Ž=200..

G.R. de Freitas et al.r Journal of the Neurological Sciences 193 (2001) 43–47

Deep-tendon reflexes were brisk. Hypertonia, bilateral
grasp reflexes, and primitive reflexes were observed. Plantar responses were flexor. She became bedridden and died
in March 1998 of pneumonia.

3. Materials and methods
The brain was fixed in 10% formalin and 19 blocks
were taken from different regions of the cerebral cortex,
white matter, basal ganglia, thalamus, brainstem, and cerebellum. After embedding in paraffin, 7-mm-thick sections
were prepared and stained using the haematoxylin–eosin
ŽHE., van Gieson–elastin ŽVGE., Loyez, periodic acid–
Schiff ŽPAS., toluidine blue, Congo Red, thioflavin S, Oil
red O, and Sudan black techniques. The avidin–biotin–
peroxidase technique was used to study the binding of
monoclonal antibodies specific for vimentin ŽDAKO,
M725; diluted 1:50., human leukocyte common antigen
ŽDAKO, M701; diluted 1:200., a-smooth muscle actin
ŽSigma, A-2547; diluted 1:1000., CD 68 ŽDAKO, M814;
diluted 1:100., and T cells ŽDAKO, M742; diluted 1:100..
Polyclonal antibodies specific for ubiquitin ŽDAKO,
Z0458; diluted 1:300. or GFAP ŽDAKO, Z 334; diluted
1:500. were also used. Small samples taken from the deep
white matter and small fragments of the leptomeningeal
arteries were processed for ultrastructural analysis; the
ultrathin sections were contrasted with uranyl acetate and
lead citrate and examined using a Philips CM-10 electron
microscope.
Direct sequencing of DNA extracted from leukocytes
while the patient was alive was performed. Exons 3 and 4

45

were amplified by using polymerase chain reaction with
the following sets of primers: forward primer N1F, 5XTGTGCTGCCCAACCAAGCGA-3X , and reverse primer
N1R, 5X-ACTGACCACACCCCCGACTA-3X , and forward
primer N2AF, 5X-TAGTCGGGGGTGTGGTCAGT-3X , and
reverse primer N3AR, 5X-CCTCTGACTCTCCTGAGTAG-3X , for exons 3 and 4, respectively. Amplicons were
sequenced in both directions by means of dye-labeled
terminators on an ABI377 DNA sequence apparatus ŽPerkin-Elmer, Applied Biosystems, Paris, France. w7x.

4. Results
4.1. Neuropathology
The weight of the brain was 1030 g. Gross examination
revealed moderate atherosclerosis of the large cerebral
arteries. On coronal brain slices, multiple, sometimes confluent, small, granular, grayish, lacunar infarcts were seen
in the deep white matter of both hemispheres. Some
lacunar infarcts were also found in the basal ganglia and
thalamus. The brainstem, cerebellum, and spinal cord were
spared.
Histological analysis showed multiple lacunar infarcts
in the deep white matter ŽFig. 1A. associated with diffuse
pallor of the myelin. Vascular changes consisted of thickening and fibrosis of the wall of small- and medium-sized
leptomeningeal arteries and arterioles penetrating the white
matter and basal ganglia ŽFig. 1B.. Fragmentation of the
elastic lamina of some arteries was also present ŽFig. 1B.,
but the most consistent change was destruction of the

Fig. 2. Illustration of the accumulation of GOM in the leptomeningeal arteries. ŽA. Electron micrograph Ž=12,000. showing smooth muscle cells, in the
tunica media of the affected vessel wall, surrounded by thickened basal membrane-like material Žarrow. and numerous GOM. ŽB. Higher magnification
view of the granular, electron-dense material in the media Žarrowhead., associated with cell debris Ž=24,000.. ŽInsets in C and D. High magnification of
GOM deposit pushing away the basal membrane-like material that accumulated around smooth muscle cells. Magnification of insets: =24,000 and
=32,000, respectively.

46

G.R. de Freitas et al.r Journal of the Neurological Sciences 193 (2001) 43–47

muscular layer, which had a smudged granular appearance
and showed marked loss of smooth muscle cells. The
remaining muscle cells were clear and ballooned. The
eosinophilic granular material, deposited in the muscular
layer, was slightly positive with PAS, VGE, and Sudan
black stains and negative for amyloid ŽCongo Red and
thioflavin S.. Smooth muscle cell immunoreactivity was
severely reduced in the muscular layer due to the severe
muscle cell loss. The granular material of the media was
immunopositive for vimentin and fibronectin. Some
perivascular cuffs of inflammatory cells, mainly T lymphocytes, were present ŽFig. 1F.. No senile plaques were
found in the cerebral cortex and no beta-amyloid deposits
were seen in the leptomeningeal and cortical arteries. A
few neurofibrillary tangles were found in hippocampus and
entorhinal cortex.
Ultrastructural analysis of the brain showed marked
destruction of smooth muscle cells with an accumulation
of GOM in the muscular layer ŽFig. 2.. These extracellular
granular deposits were numerous, especially around smooth
muscle cells, and sometimes embedded in small vacuoles
in the basal lamina-like material. The small arteries of the
white matter showed severe destruction, sometimes with
complete disappearance of the tunica media, which was
replaced by cell debris and macrophages containing electron-dense granular material; in some areas, smooth muscle cells and GOMs were no longer recognizable.
4.2. Molecular genetics
There were no mutations in exons 3 and 4 of the
Notch3 gene.

5. Discussion
The diagnosis of CADASIL was established in our
patient by characteristic changes on brain pathological
examination. Both skin biopsy and standard genetic testing
failed to reveal the characteristic changes of CADASIL.
The clinical presentation of our patient, with recurrent TIA
and subcortical strokes, seizures, and dementia, was suggestive of CADASIL. Although the age of onset of
CADASIL in our patient is later than often reported, it
falls within the reported range of 27–65 years w5x. Moreover, subtle clinical findings may have been present for
years before the patient sought medical attention.
Data concerning the sensitivity and specificity of tests
for CADASIL are still lacking. In 50 unrelated patients
strongly suspected of having CADASIL, no mutation of
the Notch3 gene was seen in 5 w5x. In two studies, GOM
was found in the skin biopsy of all CADASIL patients Ž8
and 14 patients, respectively. w8,9x and in none of the
control group w8x, and skin biopsy has been suggested as
the method of first choice for the diagnosis of CADASIL
w9x. However, a CADASIL case with GOM in the

meningeal vessels, but without changes in the skin and
striated muscle arteries, has been reported w10x. The same
authors reported a further case with a clinical and family
history highly suggestive of CADASIL, but with a negative skin biopsy. Unfortunately, genetic analysis was not
performed on these two patients.
The family history of our patient is inconclusive. Her
father died aged 55 after a stroke. Her daughter has
epilepsy and had a normal MRI at the age of 42; since
seizures are rarely the initial symptom of CADASIL and
MRI penetrance of CADASIL is almost complete between
30 and 40 years of age w11x, the diagnosis of CADASIL is
unlikely. A sporadic case has been reported w7x. However,
the small number of family members and the lack of a
satisfactory family history do not allow us to clearly define
our case as sporadic.
It could be argued that abnormalities in the skin biopsy
could be focal w12x, and, therefore, false-negative biopsies
could result from the use of insufficient or inadequate
material. However, the existence of negative skin biopsies
in CADASIL has been confirmed by a recent genetically
confirmed Canadian CADASIL family, in which skin
biopsy was negative in all eight members affected w13x. It
could also be claimed that the our genetic screening was
incomplete, since, in some CADASIL patients, the mutation is not located on exons 3 or 4 but on one of the 21
remaining exons of Notch3 gene. However, screening for
the additional 21 exons is very time consuming and rarely
performed. Moreover, in some patients, genetic testing
may fail to show changes, even after complete screening.
Although negative genetic testing or negative skin biopsies have already been reported in CADASIL patients, we
were unable to find a case in which both tests were
negative. Our findings therefore suggest that skin biopsy
and DNA examination do not detect all affected patients.
Negative results do not exclude the disease, and biopsy of
leptomeningeal arteries or a second skin biopsy should be
performed in these cases.

Acknowledgements
The authors thank Dr. E. Tournier-Lasserve for Notch3
genetic screening; Dr. L. Chimelli for help with electron
microscopy; S. Gros, S. Trepey, and A. Montoro for
technical assistance; and S. Burki for help with photography.

References
w1x Tournier-Lasserve E, Joutel A, Melki J, et al. Cerebral autosomal
dominant arteriopathy with subcortical infarcts and leukoencephalopathy maps to chromosome 19q12. Nat Genet 1993;3:256–9.
w2x Ruchoux M-M, Guerouaou D, Vandenhaute B, Pruvo J-P, Vermersch P, Leys D. Systemic vascular smooth muscle cell impairment in

G.R. de Freitas et al.r Journal of the Neurological Sciences 193 (2001) 43–47
cerebral autosomal dominant arteriopathy with subcortical infarcts
and leukoencephalopathy. Acta Neuropathol 1995;89:500–12.
w3x Schroder
¨ JM, Sellhaus B, Jorg
¨ J. Identification of the characteristic
vascular changes in a sural nerve biopsy of a case with cerebral
autosomal dominant arteriopathy with subcortical infarcts and
leukoencephalopathy ŽCADASIL.. Acta Neuropathol 1995;89:116–
21.
w4x Goebel HH, Meyermann R, Rosin R, Schlote W. Characteristic
morphologic manifestation of CADASIL, cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy,
in skeletal muscle and skin. Muscle Nerve 1997;20:625–7.
w5x Joutel A, Corpechot C, Ducros A, et al. Notch 3 mutations in
CADASIL, a hereditary adult-onset condition causing stroke and
dementia. Nature 1996;383:707–10.
w6x Dichgans M, Mayer M, Uttner I, et al. The phenotypic spectrum of
CADASIL: clinical findings in 102 cases. Ann Neurol 1998;44:431–
9.
w7x Joutel A, Dodick DD, Parisi JE, Cecillon M, Tournier-Lasserve E,

47

Bousser MG. De novo mutation in the Notch3 gene causing
CADASIL. Ann Neurol 2000;47:388–91.
w8x Ebke M, Dichgans M, Bergmann M, et al. CADASIL: skin biopsy
allows diagnosis in early stages. Acta Neurol Scand 1997;95:351–7.
w9x Mayer M, Straube A, Bruening R, et al. Muscle and skin biopsies
are a sensitive tool in the diagnosis of CADASIL. J Neurol 1999;
246:526–32.
w10x Rubio A, Rifkin D, Powers JM, et al. Phenotypic variability of
CADASIL and novel morphologic findings. Acta Neuropathol
1997;94:247–54.
w11x Chabriat H, Vahedi K, Iba-Zizen MT, et al. Clinical spectrum of
CADASIL: a study of 7 families. Lancet 1995;346:934–9.
w12x Schultz A, Santoianni R, Hewan-Lowe K. Vasculopathic changes of
CADASIL can be focal in skin biopsies. Ultrastruct Pathol 1999;23:
241–7.
w13x Demaerschalk B, MacLeod PM, Lee DH, et al. A Canadian
CADASIL kindred: how sensitive is a skin biopsy? Stroke
2001;32:331 wabstractx.