126

PostScript

Although we do not yet know the mechanism by which TMS works, we are very much
interested in the increase in cerebellar blood
flow after TMS. This finding suggests that
TMS over the cerebellum may activate the
decreased cerebellar function. This effect may
be caused by direct stimulation to the cerebellum as we used the maximum strength of
stimulation at 2.5 times the motor threshold.
Another possibility is that sensory input from
the contracted muscles by TMS might occur.
After the TMS trial we continued TMS in
some patients. Patients receiving TMS once or
twice a week kept their improved condition at
least for 6 months, but patients receiving TMS
every 2 weeks returned to their baseline in 2
weeks. Therefore, we think that the effects of
TMS with our method are maintained for
about 1 week. We conclude that TMS over the
cerebellum is a promising treatment for
spinocerebellar degeneration.

Acknowledgement

(1)

(2)

(3)

(4)

dog
(inu)

wool
(keito)

trash
(haizara)

pencil
(enpitsu)

(5)

Kanji

Meaning
Pronunciation

newspaper
(shinbun)

Kana

This work was financially supported by the Magnetic
Health Science Foundation.

Y Shiga, T Tsuda, Y Itoyama
Department of Neurology, Tohoku University School
of Medicine, 1–1 Seiryo-machi, Aoba-ku, Sendai
980–8574, Japan

H Shimizu
Department of Neurology, Yamagata National
Hospital, Japan

K-I Miyazawa
Department of Neurology, Kohnan Hospital, Japan

K Jin

Figure 1 Examples of dictation. This patient showed intact Kana writing and difficulties in
Kanji writing: 1, no response; 2, partial completion; 3, partial substitution; 4, neographism;
5, correct.

Department of Neurology, Miyagi National
Hospital, Japan

T Yamazaki
Department of Radiology, Tohoku University School
of Medicine, Japan
Correspondence to: Dr Y Shiga;
yshiga@neurol.med.tohoku.ac.jp

References
1 Pascual-Leone A, Rubio B, Pallardo F, et al.
Rapid rate transcranial magnetic stimulation of
the left dorsolateral prefrontal cortex in
drug-resistant depression. Lancet
1996;348:283–7.
2 Mally J, Stone TW. Improvement in
parkinsonian symptoms after repetitive
transcranial magnetic stimulation. J Neurol Sci
1999;162:179–84.
3 Shimizu H, Tsuda T, Shiga Y, et al.
Therapeutic efficacy of transcranial magnetic
stimulation for hereditary spinocerebellar
degeneration. Tohoku J Exp Med
1999;189:203–11.
4 Matsuda H, Yagishita A, Tsuji S, et al. A
quantitative approach to technetium-99m ethyl
cystinate dimer: a comparison with
technetium-99m hexamethyl propylene amine
oxine. Eur J Nucl Med 1995;22:633–7.

Agraphia in Kanji after a
contusional haemorrhage in the
left temporo-occipital lobe
The Japanese language has two different
writing systems, Kana (Japanese syllabograms) and Kanji (Japanese ideograms), and
Japanese sentences usually consist of combinations of both. Recently, there has been
speculation that different pathways are involved in Kanji and Kana reading and
writing.1 It has been suggested that the
processing of Kanji and Kana involves different intrahemispheric mechanisms, as judged
mainly through studies on patients with

www.jnnp.com

alexia and agraphia.2 Wernicke’s area and its
surrounding left middle temporal lobe may
play the most important part in Kanji reading
when visual information is transmitted by
any pathway. We recently had the opportunity
to examine a patient who had mild transient
aphasia and persistent agraphia of Kanji after
a discrete contusional haemorrhage in the left
temporo-occipital lobe.
A 66 year old right handed man (pensioner)
with 12 years of school education was
involved in a traffic accident and sustained a
closed head injury. He was admitted to our
hospital the next day. He had no history of
neurological problems. His parents, brothers,
and sisters are all right handed. On examination, he was alert and cooperative. Neurological examination disclosed no motor or sensory
disturbance.
Formal language assessment was undertaken 3 days after injury using the standard
language test for aphasia (SLTA). His spontaneous speech was fluent. He had no difficulty
initiating speech, articulated normally, and
did not have logorrhoea. Echolalia was not
seen and phonological structure was clear.
Object naming was moderately impaired,
with considerable paraphasia. Word fluency
(animal naming) was 3/min. Repetition was
excellent. Reading aloud was good; however,
writing of both Kanji and Kana was impaired, and considerable paragraphia was
evident.
In addition, he demonstrated constructional apraxia using figure copying of a three
dimensional cube but not oral, ideomotor, or
ideational apraxia. There were no signs of
motor impersistence or unilateral spatial
neglect. The score on Raven’s coloured progressive matrices was 26/36.
Brain CT on admission showed an area of
high density consistent with contusional

haemorrhage in the left temporo-occipital
lobe. Brain MRI 5 days after admission
showed contusional haemorrhage in the left
temporo-occipital lobe involving the posterior
part of the temporal cortex and the adjacent
white matter. There were no abnormalities in
the left angular gyrus.
Aphasia was gradually resolved, his scores
on SLTA 8 weeks after the onset becoming
normal. Despite this, his writing ability
remained clearly disturbed. We asked him to
read and write on dictation 221 Kanji and 76
Kana, all of which are taught in the first 2
years of primary school in Japan. He was
unable to write 83 of the 221 Kanji (37.6%),
which he reported that he knew but could not
recall. He tended to have difficulty with Kanji
characters, which were more complex and
had been learnt later (fig 1). Error types
included partial substitution (two), partial
completion (seven), neographism (29), and
no response (45). On the other hand, he made
no mistakes in reading Kanji or in reading and
writing Kana letters. On the revised Wechsler
adult intelligence scale (WAIS-R), his verbal
intelligence quotient (IQ) was 105 and his
performance IQ was 101, with an overall IQ of
103. Brain MRI 3 months after admission
showed a subcortical lesion in the left
temporo-occipital lobe (fig 2).
Although this patient experienced aphasia
immediately after cerebral contusion, the
aphasia resolved over a subsequent 8 week
period. He was left with an isolated disturbance of Kanji writing. In the absence of any
dementia, aphasia, or disturbance of consciousness, his condition was classified as one
of pure agraphia.
Soma et al2 described “pure agraphia of
Kanji” in three patients, who had a lesion in
the left posterior temporal area extending to
the angular gyrus on CT. Although they

PostScript

127

Figure 2 Brain MRI 3 months after admission showed the subcortical lesions in the left
temporo-occipital lobe.

exhibited amnestic aphasia, alexia, and
agraphia in the acute phase, disturbances
other than agraphia of Kanji disappeared in a
few months. Yokota et al3 also reported a case
with pure agraphia of Kanji, and suggested
that the process of writing Kanji involves a
different pathway from that which mediates
Kanji reading in the left temporal lobe.
Iwata1 proposed a hypothetical neuronal
mechanism in the writing of Kanji and Kana
as follows: the Broca’s and motor association
areas are the final coordination centres, but
the angular gyrus plays an important part in
sending the graphic information to the motor
areas. However, spontaneous writing and
responses to dictation are usually initiated by
Wernicke’s area, which gives rise to two
different pathways to the angular gyrus. One
is the auditory somaesthetic association pathway leading directly from Wernicke’s area to
the angular gyrus; Kana writing mainly
depends on the intactness of this route. The
second pathway is from Wernicke’s area to the
occipital lobe by way of the posterioinferior
temporal area. This is the pathway involved in
selecting the correct Kanji graphemes according to the meaning of the word, and thus
recalled visual engrams of letters are sent to
the angular gyrus.
Our patient demonstrated that initial amnestic aphasia, and agraphia of Kanji were
associated with a lesion in the left temporal
lobe. Because the posterior temporal region is
located in close proximity to the angular gyrus
and Wernicke’s area, the pathological process
in the first area affects the second two regions
in its acute phase. We conclude that the
persistent symptom of pure agraphia for Kanji
in this study was caused by the left posterior
temporal lesion which disconnected the pathway for Kanji writing selectively.
S Maeshima, A Ueyoshi, T Matsumoto
Department of Physical Medicine and
Rehabilitation, Wakayama Medical University,
Wakayama, Japan

R Okita, H Yamaga, F Ozaki, H Moriwaki
Division of Neurological Surgery, Hidaka General
Hospital, Wakayama, Japan

P Roger
School of Communication Sciences and Disorders,
University of Sydney, Australia

Correspondence to: Dr S Maeshima;
maeshima@wakayama-med.ac.jp

References
1 Iwata M. Kanji versus Kana.
Neuropsychological correlates of the
Japanese writing system. Trans Neurosci
1984;7:290–3.
2 Soma Y, Sugishita M, Kitamura K, et al.
Lexical agraphia in the Japanese language.
Pure agraphia for Kanji due to left
posterioinferior temporal lesions. Brain
1989;112:1549–61.
3 Yokota T, Ishiai S, Furukawa T, et al. Pure
agraphia of kanji due to thrombosis of the
Labbe vein. J Neurol Neurosurg Psychiatry
1990;53:335–8.

Successful autologous stem cell
transplantation in a patient with
chronic inflammatory
demyelinating polyneuropathy
A patient with chronic inflammatory demyelinating polyneuropathy is reported on who
never had spontaneous remissions for 10
years, but who is in remission for 2 years after
an autologous stem cell transplantation
(ASCT). Before ASCT he needed at least 20 mg
prednisone/day and in addition intravenous
immunoglobulin (IVIg) treatment at regular
intervals. An ASCT was considered in this
patient because of serious side effects of
immunosuppressive treatment.
Chronic
inflammatory
demyelinating
polyneuropathy (CIDP) is characterised by
weakness with sensory impairments in the
arms and legs. The reflexes in the arms and
legs disappear. The onset is insidious. There
may be progressive deterioration or a course
with remissions and exacerbations. Routine
blood examination is usually normal. Protein
in CSF is increased in most patients. Electrophysiological studies may show slow nerve
conduction velocities, conduction blocks, or
dispersion.1 Inflammatory demyelinating
polyneuropathy is considered to be an autoimmune disease which is supported by the
presence of inflammatory cells in sural nerve
biopsies and the beneficial response to immunosuppressive treatment in most patients.2 We
describe a patient with CIDP who has had this
disease for 10 years without spontaneous
remissions. He responded to immunosuppressive therapy, but needed high doses and had

severe side effects. Therefore, we decided to
try to induce a long lasting remission by
autologous stem cell transplantation (ASCT).
This patient was 38 years old when, in 1988,
he began to have tingling and numbness in
his fingers. The sensory symptoms progressed
to his arms and legs, followed by weakness.
Initially his symptoms were attributed to too
much stress but at the end of 1990 when
weakness had worsened to such an extent
that he was no longer independent in his daily
activities, it was decided that it was time for a
neurological evaluation. At that time he had
weakness of his arms and legs, MRC grade 4.
There was atrophy of his intrinsic hand muscles and fasciculations were seen in the muscles of his arms. Except for diminished knee
reflexes, there was areflexia. He had numbness of his arms and legs in a glove and stocking distribution. Electrophysiological studies
showed slowed nerve conduction velocities of
the median, ulnar, tibial posterior, and peroneal nerves in the range of 15–25 m/s with
prolonged distal latencies and F wave responses. The distal compound muscle action
potentials were small. Few muscles had signs
of denervation. Blood examination was unremarkable and CSF total protein was 0.67 g/l
without cells. Sural nerve biopsy was consistent with CIDP, showing inflammatory cells
and demyelination.
In 1990 he was started on 60 mg/day prednisone. After 2 weeks improvement began.
This improvement continued to normal
strength leaving only slight numbness in his
fingers. During tapering off of the prednisone
dose he had mood disturbances at every
change of dose. Repeatedly he had a relapse at
doses lower than 20 mg/day. Subsequently we
tried to replace prednisone by 150 mg/day
azathioprine but after 2 years of treatment
with azathioprine it was still not possible to
decrease the dose of prednisone to less than
20 mg/day. Azathioprine was replaced by
methotrexate (7.5 mg/week) for 6 months
without a beneficial response. Thereafter he
was treated with IVIg. After two cycles of 30g/
day IVIg for 5 days he needed infusions every
other week to maintain his improved condition. We tried to prolong these intervals by
adding prednisone. With 20 mg/day prednisone he remained free of relapses for about
4 months. However, these intervals became
gradually shorter; he developed arthralgias
after administration of IVIg, and had repeated
gastric pain. After 8 years of immunosuppressive treatment, without spontaneous remissions, we discussed the possibility of autologous stem cell transplantation (ASCT).
In May 1998 peripheral blood stem cells were
harvested by leukopheresis after mobilisation
with cyclophosphamide (4 g/m2) and granulocyte colony stimulating factor (G-CSF; 5
µg/kg). CD 34+ cells were positively selected by
immunomagnetic beads (Clinimax®, purity
98.7%) and cryopreserved. Before this procedure the patient had had IVIg every other week
together with 20 mg/day prednisone. With this
regular IVIg treatment he was in an optimal
condition: he had no weakness and only mild
sensory impairment of his fingers. After the
mobilisation procedure, prednisone could be
tapered off to 8 mg/day. Even without IVIg he
remained in this improved condition for 5
months before he relapsed. Weakness in the
arms and legs returned together with sensory
impairment. Electrophysiological studies were
repeated. We decided firstly to improve his
neurological condition by IVIg treatment before myeloablative treatment was started. This
treatment resulted in considerable improvement; only mild sensory impairment of the
fingers remained.

www.jnnp.com