609

J7ournal of Neurology, Neurosurgery, and Psychiatry 1994;57:609-613

Naming difficulties in alexia with agraphia for
kanji after a left posterior inferior temporal lesion
Yasuhisa Sakurai, Katsuyuki Sakai, Manabu Sakuta, Makoto Iwata

Abstract
The case is described of a patient with
alexia and agraphia for kanji, and severe
anomia after a subcortical haemorrhage
in the left posterior inferior temporal
area. Magnetic resonance imaging at four
months after onset showed a lesion in the
inferior temporal and fusiform gyri,
extending from the temporo-occipital
junction toward the anterior third of the
temporal lobe. Comparison with other
reported cases of alexia with agraphia
and anomia made it clear that when
accompanied by severe anomia, the
lesions extended either forward to the
anterior part of the middle temporal
gyrus or medially to the parahippocampal gyrus. It is suggested that the disconnection of association fibres between the
parahippocampal, fusiform, middle, and
inferior temporal gyri, especially between
the parahippocampal gyrus and the other
temporal gyri, or the cortical damage to
the posterior part of these gyri is essential
for the production of anomia.
(7 Neurol Neurosurg Psychiatry 1994;57:609-613)

Department of
Neurology, Institute
for Brain Research,
University of Tokyo
School of Medicine,
Tokyo, Japan
Y Sakurai
M Iwata
Department of
Neurology, Japanese
Red Cross Medical

Center, Tokyo, Japan
K Sakai
M Sakuta
Correspondence to:
Dr Yasuhisa Sakurai,
Departnent of Neurology,
Institute for Brain Research,
University of Tokyo School
of Medicine, 7-3-1, Hongo,

Bunkyo-ku, Tokyo 113,
Japan.
Received 28 January 1993
and in revised form
14 July 1993.
Accepted 28 September 1993

Since Iwatal first described a case of alexia
with agraphia only for kanji (Japanese morphograms) due to a left posterior inferior temporal lesion, there have been many similar
case reports stressing the importance of this
area in the reading and writing of kanji.'-"
Alexia with agraphia for kanji due to a left
posterior inferior temporal lesion has two
main features. Firstly, it is often accompanied,
to a greater or lesser degree, by naming difficulties (anomia). Little is known as to why
anomia is caused by this lesion. Secondly, in
some patients kana reading is more impaired
than kanji reading, though kanji writing disturbance is always worse and persists longer
than kana writing. For this reason it is called
pure agraphia for kanji by some workers.215
In this paper we describe a Japanese man
with alexia and agraphia for kanji accompanied by severe naming difficulties due to a left
posterior inferior temporal lesion, and discuss
the mechanism underlying the production of
anomia by comparing this patient with
reported cases.
Case report
Clinical details of the patient during the acute
phase have been described elsewhere.'7 The

patient, a 71-year-old right-handed man, was
educated to senior high school level. He had
no history of serious illness. On 22 June 1991
he suddenly felt a slight left-sided headache.
He then noticed that he could not remember
the names of objects or his family and could
not read newspapers or write anything but his
own name. He was admitted to the Japanese
Red Cross Medical Center on 22 June with a
diagnosis of cerebrovascular disease.
He was fully conscious on admission.
Cranial nerve function, motor system, cutaneous and deep sensation, and coordination
were all intact except for a mild hearing loss in
both ears. There was no visual field defect.
Blood chemistry was normal. Audiometry
showed mild hearing loss in both ears for
speech discrimination (right ear, 75%; left
ear, 80%; 80-85 dB).
NEUROPSYCHOLOGICAL EXAMINATION

The patient's verbal IQ with the Wechsler
adult intelligence scale-revised (WAIS-R)
was 80, and his perforrnance IQ was 86.
The Western aphasia battery (WAB;
Japanese edition'8 (table 1)) at two months
showed that his spontaneous speech was fluent with preserved prosody. He showed
marked word finding difficulty and overused
demonstrative pronouns, but there were few
paraphasias. He made three correct responses
out of 20 in object naming and, given phonemic cues, made another four correct
responses. Circumlocution was observed in
most of the incorrect responses. He could
explain how to use objects and could point to
Table 1 Scores for the Western aphasia battery
August 1991 October 1992

Spontaneous speech

Information content

Fluency
Naming total
Object naming
Repetition
Comprehension total
Yes/no questions
Auditory word recognition
Sequential commands
Reading total
Comprehension of sentences
Reading commands
Kanji to object matching
Kana to object matching
Kanji to picture matching
Kana to picture matching
Writing total
Copying
Kanji writing to dictation
Kana writing to dictation
Sentence writing to dictation
Praxis total
Construction total

7/10
8/10
2-4/10
13/60
8-6/10
7/10
51/60
41/60
48/80
3-9/10
4/40
7-5/20
3/3
3/3
3/3
3/3
6-55/10
10/10
0/6
4/6
7/10
9-8/10
8/10

9/10
9/10
5 7/10
31/60
9-6/10
9-45/10
54/60
59/60
76/80
8-05/10
40/40
13/20
3/3
3/3
3/3
3/3
7-5/10
10/10
2/6
5/6
9/10
9-8/10
7 9/10

Sakurai, Sakai, Sakuta, Iwata

610

objects on command fairly well (17/20).
Repetition was almost normal. Auditory comprehension was slightly impaired. Kanji reading was poor, which resulted in a low score for
comprehension of written language, whereas
kana reading was generally good, though with
a few phonemic paraphasias (this difference
was made clear in the reading of sentences
including both kanji and kana, but was not
reflected in the WAB score). In kanji-to-object
matching he could point to the appropriate
objects correctly, though he could only read
one of six kanji word items. This implies that
semantics were preserved even when he could
not read kanji words, characteristic of deep
dyslexia.'9 Also, in writing to dictation, kanji
was inferior to kana. Agraphia of kanji due to
word recall difficulty and phonemic paragraphias of kana were noted. The patient's
Figure 1 CT scan on
admission showing a large
high density area in the left
temporal subcortical
structure extending to the
anterior part of the
temporal lobe. The left is
the patient's left side.

speech profile was rated as one of anomic
aphasia and alexia with agraphia, relatively
selective for kanji. HIe did not show apraxia, a
construction disturbance, nor colour agnosia.
To evaluate the extent of the reading and
writing difficulties for kanji compared with
kana more precisely, at three months the
patient was given reading aloud and writing to
dictation tasks involving 100 kanji and the
corresponding kana (readings of kanj), all of
which are taught in the first three years of primary school. The results showed that in both
reading (correct response in kanji, 20%; kana,
75%; p < 0 001 on the two-tailed Fisher's
exact method) and writing to dictation (correct response in kanji, 4%; kana, 59%; p <
0 001 on the two-tailed Fisher's exact
method), kana function was significantly
superior to kanji function. Most of the errors

Naming difficulties in alexia with agraphia for kanji

comprised no response in kanji tasks and
phonemic paralexias or paragraphias in kana
tasks. A similar tendency was observed on
re-examination 15 months after onset with
significant improvement (reading: correct
response in kanji, 46%, p < 0 005; kana,
91%, p < 0 05; writing to dictation: correct
response in kanji, 12%, p < 0 005; kana,
77%, p < 0 005, on the McNemar test). On
the whole, he could read in kana more than
90% of the items he read correctly in kanji, as
well as in writing. He could also read almost
Figure 2 MRI T2weighted coronal view (0 5
T; TR = 2500 ms; TE =
120 ms) through the
atrium of the lateral
ventricle (A) and
hippocampus (B, C) at
four months. The right is
the patient's left side. A
low intensity area
suggesting haemosiderin
deposition is present in the
cortico-subcortical tissue of
the left fusiform, inferior
temporal gyri from the
occipital to the anterior
third of the temporal lobe.

61

all kanji or kana he could write in the reexamination, though there were some kanji or
kana which he could not read but could write,
or vice versa, in the first test (number of items
with both correct reading and correct writing
were two for kanji and 45 for kana).
Severe naming difficulties and alexia with
agraphia for kanji still persisted at 16 months,
though there was some improvement (table
1).
CT on admission showed a high density
area in the left temporal subcortical structure,
thus the patient was diagnosed as having a
cerebral haemorrhage (fig 1). There was also a
low density area suggesting an old infarction
in the anterolateral portion of the left putamen. MRI at the same time confirmed these
lesions and there was no other change in
intensity affecting the left thalamus. In MRI
at four months the lesion was evident in the
cortico-subcortical tissue of the left inferior
temporal and fusiform gyri, probably affecting
the inferior longitudinal fasciculus and
extending from the temporo-occipital junction toward the anterior third of the temporal
lobe (fig 2). Single photon emission CT with
iodoamphetamine labelled with iodine- 123
(123I-IMP-SPECT) at four months showed
decreased blood flow in the lower and medial
parts of the left temporal lobe, including the
hippocampal region, whereas the regional
cerebral blood flow in the left basal ganglia
and thalamus was preserved.

Discussion
The patient presented with alexia with
agraphia for kanji and severe naming difficulties after a subcortical haemorrhage of the left
posterior inferior temporal area. Anomia can
also be observed in a thalamic lesion.20 The
haematoma and its surrounding oedema in
the acute phase could have some effect on the
left thalamus, but MRI at this time showed
that there was no change in intensity in the
thalamus. In addition, the patient's severe
anomia was still present at 12 months.
Therefore, we cannot attribute anomia to the
acute dysfunction of the thalamus.
The patient had an old infarction in the
anterolateral portion of the left putamen. A
basal ganglia lesion can also produce atypical
aphasia." In such a case, however, the lesion
affects more widespread areas including the
head of the caudate nucleus, the putamen,
and the internal capsule. The patient's putaminal lesion was too small to cause aphasia,
and the regional cerebral blood flow of the left
basal ganglia was preserved. Furthermore, the
patient did not show anomia until admission,
thus, the old infarction in the putamen cannot
cause the anomia.
There have been more than 20 case reports
of alexia with agraphia due to a left posterior
inferior temporal lesion. Severe naming difficulties often occurred in those patients in
whom kanji reading was more impaired,
whereas in patients with more impaired kana
reading naming difficulties were, if present,
only slight. Thus to correlate the lesion with

612

Table 2 Classification of alexia with agraphia for kanji with significant difference in
reading between kanji and kana. Values are mean (%) scores

Sakurai, Sakai, Sakuta, Iwata

reading and severe anomia (less than 80%
correct in confrontation naming).'2-"
Reading
We determined which was more impaired,
kanji or kana, in each patient based on the
Patient group*
Naming
kana
kanJi
Lesiont
scores for the individual kanji and kana reading
Group A
test
and the authors' evaluation. We then
Shiota et al3
sI.
20/36(56)
20/24(83)
PIT#
Kawamura'
chose
patients with significantly more
Patient TN
sI.
263/416(63)
402/416(92) Fu, inf T, T-O#1
impaired kana or kanji reading from each
Mochizuki and Ohtomo6 N
8/48(17)
74/100(74) InfT, 0ii
group (p < 0'05 or less on the Fisher's exact
Group B
Oka et al 9
16/20(80%) 752/800(94)
674/800(84) InfT, Fu#
method; table 2) and traced their lesion on
Kawahata et al 10
18/20(90%) 24/25(96)
2/25(8)
PIT4
CT axial slices. Figure 3 shows the superimGroup C
Imamura et al 12
0/20(0%) 28/30(93)
9/30(30)
HC, OR, inf T-Os41t
lesion images of each group. We can
posed
Notoya et al 1"
11/20(55%) 20/20(100)
12/20(60)
HC, inf T, Ful
see
from
this figure that in group A patients
Soma etal!5
33/50(66%) 219/221(99)
201/221(91) Mid T, infTjt
Patient 2
the
lesion
extends from the left posterior infeNotoya et al 16
7/20(35%) 20/20(100)
14/20(70)
Mid T, inf T, T-O0
rior temporal area backward to the lateral
This patient
3/20(15%) 75/100(75)
20/100(20) InfT, Fut
occipital gyrus. In alexia with agraphia for
*Group A, more impaired kana reading; group B, more impaired kanji reading with slight or no
kanji without profound anomia (group B) the
anomia; and group C, more impaired kanji reading with profound anomia. tBased on t
authors' evaluation. tp < 0 05, t4p < 0-01 on the two-tailed Fisher's exact method between ka nji
lesion is mostly localised in the left posterior
reading and kana reading. Abbreviations: PIT = posterior inferior temporal; inf = inferi,or;
inferior temporal area (almost corresponding
sl. = slightly disturbed; N = normal; T = temporal; 0 = occipital; s = subcortical; Fu = fusifor
HC = hippocampus; OR = optic radiation.
to Brodmann area 37). When severe anomia
accompanies alexia with agraphia for kanji
the symptom we divided the reported cases of (group C), the lesion affects the left posterior
alexia with agraphia due to a posterior inferi4 or inferior temporal area, extending either fortemporal lesion into three groups: group 2 , ward as far as the anterior part of the middle
patients with more impaired kana readinge-7; temporal gyrus or medially to the parahipgroup B, patients with more impaired katzji pocampal gyrus. As this inference is based on a
reading and slight or no anomia (80% or mo: re small number of patients, we need further evicorrect in confrontation naming)8-"; anid dence to confirm it.
This patient showed severe anomia, but he
group C, patients with more impaired ka, Iji
could explain how to use objects and point to
objects almost always correctly on command.
L
R Such a symptom is characteristic of pure
anomic aphasia caused by a localised lesion.22
In contrast, a patient with anomia with a cortical degenerative disease such as Pick's disease
or
Mesulam's slowly progressive aphasia23
A
cannot explain how to use objects, recognise
them, or point to objects on command.24 As
this patient did not improve with tactile cues
and could copy and match figures to sample,
the naming disturbance was not caused by
visual agnosia. This difference is probably
because in anomia caused by a localised
lesion, lexical information is still preserved in
the cortex but its use is impaired, whereas in
anomia observed in a cortical degenerative
disease lexico-semantic information itself is
B
affected.
Comparison of the lesions of group B
patients with those of group C patients
showed that the localised damage to the left
posterior inferior temporal area (posterior part
of the middle and inferior temporal area and
fusiform gyri) did not produce anomia and
that severe anomia occurred only when the
posterior inferior temporal lesion extended
forward to the anterior part of the temporal
C
lobe or medially to the parahippocampal
gyrus. We identified the extent of the lesion in
our patient in more detail on MRI coronal
images. It affected the point at which fibres
from the fusiform, middle, and inferior temporal gyri seemed to converge, extending from
Figure 3 Superimposed schematic CT images of alexia with agraphia due to a left
posterior inferior temporal lesion. The left is the patients' left side. Areas with oblique lines
the temporo-occipital junction to the anterior
denote the left posterior inferior temporal area (almost corresponding to Brodmann's area
third of the temporal lobe. It was continuous
37). (A) Patients with significantly more impaired kana reading.3'6 (B) Patients with
along almost the whole length of the parahipsignificantly more impaired kanji reading and slight or no anomia. 9 10 (C) Patients with
significantly more impaired kanji reading and severe anomia, including our patient. 121315pocampal gyrus as a result, separating the
The lesion in alexia with agraphia for kanji is usually localised in the left posterior inferior
parahippocampal gyrus from the other tempotemporal area (B). When kana reading impairment is more pronounced, the lesion affects
ral gyri (fig 2). A similar lesion was reported in
the lateral occipital area (A). If it accompanies severe anomia, the lesion extends forward
to the anterior part of the temporal lobe or medially to the parahippocampal gyrus (C).
a patient with severe anomia and alexia with

613

Naming difficulties in alexia with agraphia for kanji

agraphia for kanji due to cerebral infarction.'4

These findings suggest that anomia occurs
when the association fibres are disconnected
between the parahippocampal, fusiform, middle, and inferior temporal gyri, or when the
lesion affects the parahippocampal gyrus. The
patient of Mochizuki and Ohtomo6 (group A,
delineated by solid lines in fig 3A) does not
seem to fit our view in that there was no
anomia though the lesion extended forward in
the temporal lobe. In this patient, however,
the damage was, according to the MRI finding, restricted to the occipital lobe and the
inferior temporal gyrus, whereas the lingual,
fusiform, parahippocampal, superior and
middle temporal gyri were spared. This
patient shows that-anomia does not appear if
the posterior part of the parahippocampal
gyrus is intact.
In our patient the parahippocampal gyrus
and hippocampus were apparently spared, but
SPECT showed the decreased blood flow in
the left inferior and medial temporal area,
including the hippocampal region. This
implies that the lesion had some effect on the
hippocampal memory system.
In monkeys the parahippocampal gyrus and
perirhinal cortex have afferent and efferent
connections with several association areas in
the frontal, temporal, and parietal lobes.25
Given these findings, it is assumed that the
disconnection of association fibres between
the parahippocampal, fusiform, middle, and
inferior temporal gyri, especially between the
parahippocampal gyrus and the other temporal gyri, or the cortical damage to the posterior
part of these gyri is essential for the production of pure anomic aphasia.
We found in a review of published work'7
that the more severe the naming difficulties
were, the more impaired kanji writing seemed
to be, and this was dependent on the extent of
the lesion. The combination of anomia and
alexia with agraphia for kanji is observed only
in Japan. The same lesion may result in
anomia and reading and writing difficulties of

irregular words in English-speaking countries.

We are grateful to Professor Ichiro Kanazawa, Department of
School of Medicine, for his
Neurology,
University of Tokyo
We also express our thanks to Dr
support and encouragement.
Shigehiro Imamura, Kumamoto Rosai Hospital for sending us
CT films and to Dr Hiroshi Mochizuki, Tohoku University
School of Medicine for informing us of his case in detail.

1 Iwata M. Kanji versus Kana. Neuropsychological correlates of the Japanese writing system. Trends Neurosci
1984;7:290-3.
2 Yokota T, Ishiai S, Furukawa T, Tsukagoshi H. Pure
agraphia of kanji due to thrombosis of the Labbe vein. J

Neurol Neurosurg Psychiatry 1990;53:335-8.
3 Shiota J, Kawamura M, Isono 0, Hirayama K. Alexia with
agraphia produced by a localized infarction in the inferior posterior region of the left temporal lobe [in
Japanese]. No to Shinkei 1986;38:1051-5.
4 Shimomura T, Tagawa K, Nagata K, Shishido F. A case
of alexia with agraphia due to the left temporo-occipital
lobar hematoma[in Japanese]. Shinkeinaika 1987;26:
57-64.
Kawamura M. Non-classical type alexia without agraphia
[in Japanese]. Shitsugosho Kenkyu 1988;8: 185-93.
6 Mochizuki H, Ohtomo R. Pure alexia in Japanese and
agraphia without alexia in kanji.The ability dissociation
between reading and writing in kanji vs kana. Arch
Neurol 1988;45:1157-9.
7 Shimomura T, Suzuki T, Takahashi S. Two cases of

alexia with agraphia due to the left posterior inferior
temporal lesion [in Japanese]. Shinkeinaika 1989;30:

396-403.
8 Iwata M. Neural mechanisms of reading and writing.
In: Tsukada Y, ed.
Neurogrammatological approach.
Perspectives on neuroscience. From molecule to mind.
Tokyo: University of Tokyo Press, 1985:299-312.
9 Oka H, Asano T, HattoriS, MochioS, Sugishita M. Pure
alexia and alexia with agraphia caused by lesions of the
left inferior temporal and fusiform gyri [in Japanese].
Shinkeinaika 1985;23:73-6.
10 Kawahata N, Tagawa K, Hirata Y, Nagata K, Shishido F. A
case of alexia with agraphia due to the infarction in the
left inferior-posterior temporal lobe [in Japanese].

Rinshoshinkeigaku 1987;27:420-7.

11 Kawamura M, Hirayama K, Hasegawa K, Takahashi N,
Yamamura A. Alexia with agraphia of kanji (Japanese
morphograms). J Neurol Neurosurg Psychiatry 1987;50:
1125-9.
12 ImamuraS, Ohno N, Nakashima A, Yamanaga H, Ideta
T. Alexia with agraphia caused by temporo-occipital
infarcts of the dominant side. Case report [in Japanese].

Shinkeinaika 1985;23:362-8.
13 Notoya M, Suzuki S, Kurachi M, Koyama Y, Hiramatsu
H. A case of alexia with agraphia caused by a left inferior
lesion [in Japanese]. Shinkeishinnigaku 1987;
temporal
3:244-50.
14 Kaneki M, Takeda K, Bandoh M, Iwata M, Mannen T,
Sekiguchi S. A case of amnesic aphasia and alexia with
agraphia caused by a subcortical infarction of the left
inferior temporal area [in Japanese] [abstract].
Shinkeishinnigaku 1988;4:33.
15 Soma Y, Sugishita M, Kitamura K, MaruyamaS, Imanaga
H. Lexical agraphia in the Japanese language. Pure
agraphia for kanji due to left posteroinferior temporal
lesions. Brain 1989;112:1549-61.
16 Notoya M, Tetoriya H, Suzuki S, Fujii H, Furukawa H.
inferior temporal
Alexia with agraphia of kanji of the left Kenkyu
1991;11:
Shitsugosho
gyrus
type
[in
Japanese].
140-6.
17 Sakai K, Sakurai Y, Sakuta M, Iwata M. Naming difficulties
seen in a case of alexia with agraphia caused by a
left postero-inferior temporal lesion [in Japanese].
Rinshoshinkeigaku 1992;32: 1227-31.
18 Kertesz A. The Western aphasia batery. New York: Grune
and Stratton, 1982 (translated into Japanese: Sugishita
M, et al. Tokyo: Igaku-Shoin, 1986).
19 Newcombe F, Marshall JC. On psycholinguistic classifications of the acquired dyslexias. Bull Orton Soc 1981;
31:29-46.
20 Graff-Radford NR, Eslinger PJ, Damasio AR, Yamada T.
infarction of the thalamus: behavioral,
Nonhemorrhagic
anatomic, and physiologic correlates. Neurology 1984;34:
14-23.
21 Damasio AR, Damasio H, Rizzo M, Varney N, Gersh F.
with nonhemorrhagic lesions in the basal
Aphasia and
internal capsule. Arch Neurol 1982;39:
ganglia
15-20.
22 Benson BF. Neurologic correlates of anomia. In: Whitaker
H, Whitaker HA, eds. Studies in neurolinguistics. Vol. 4.
New York: Academic Press, 1979:293-328.
23 Mesulam M-M. Slowly progressive aphasia without generalized dementia. Ann Neurol 1982;11:592-8.
24 Sakurai Y, Momose T, Watanabe T, Bandoh M, Ishikawa
T, Iwata M. Slowly progressive fluent aphasia. Clinical
features and an imaging study including MRI, SPECT
and PET [in Japanese]. Rinshoshinkeigaku 1992;31:
505-11.
25 Squire LR, Zola-Morgan S. Memory: brain systems and
behavior. Trends Neurosci 1988;1 1: 170-5.