Journal of Neurology, Neurosurgery, and Psychiatry 1991;54:841-846

LETTERS TO
THE EDITOR
Alexia without agraphia or hemianopia
in parietal infarction
Alexia without agraphia (pure alexia) may
result from damage to the pathways conveying visual input from both hemispheres to the
dominant angular gyrus, which itself remains
intact but disconnected from the visual
regions.' The most common lesion involves
the left occipital lobe, and may also compromise the splenium of the corpus callosum
(occipital or splenio-occipital alexia)'" Less
often, pure alexia results from a lesion either
in the occipitotemporal paraventricular white
matter,34 or more superiorly and rostrally in
the parieto-occipital or parietal white matter
(subangular or paraventricular alexia). To
date, only three cases of pure alexia without
visual field defect have been described in
patients with proven parieto-occipital
lesions.57 The responsible lesion was surgical
in one case' and intracerebral haematoma in
two cases.67 We now report a patient with
alexia without agraphia or visual field defect,
in whom MRI demonstrated a subcortical
infarction in the left parieto-occipital area.
A right handed normotensive physician
had mitral valve prolapse, paroxysmal atrial
tachycardia, and intermittent atrial fibrillation. He was 66 years old in 1982, when he
suffered a right frontal lobe infarction. Since
the diagnosis of a cystic glioma was also
entertained, he had had open craniotomy,
which confirmed the presence of an infarct.
CT scans, carried out at the time of infarction
and yearly thereafter, showed a large, nonenhancing, low density cystic lesion in the
right frontal region. His neurological examination was normal aside from the cognitive
deficits described below. Extensive psychometric testing in September 1985 demonstrated mild attentional, visuospatial and
constructional deficits, and mild impairment
in his ability to learn new nonverbal material.

Confrontation naming was normal.
On the evening of 21 January 1986, he felt
very anxious, sensed that something was
wrong, and went to bed. The following
morning, on awakening, he was unable to
read the newspaper or the phone book. He
had no difficulty with verbal output. His wife
noticed that he was mildly confused, with
difficulties in verbal comprehension. On
examination four hours later he was oriented
to person, place and time. The quantity and
grammatical content of verbal speech
production were normal, and there were no
paraphasic errors. Repetition of "no ifs, ands,
or buts" and "Methodist Episcopal" was
normal. Comprehension for conversation and
three part commands was normal. He named
visually presented common objects but had
mild difficulty naming small parts of objects.
He could name saturated colours, match
colours, and point to a named colour. Reading
was severely impaired. He had some impairment naming letters and arabic numerals and
understanding some isolated words. He could
not understand the meaning of sentences,
including simple written commands. Compared with his severe alexia, he had only
minimal agraphia when copying written
material and writing to dictation. For
example, when asked to write "Today is a
sunny day in Southern California" he wrote
"Totay is a sussy day in Southern Califortia. "
Although he could not perform written calculations because of difficulty reading Arabic
numbers, he performed mental calculations
well. The remainder of his mental status and
general neurological examination was
unchanged from his baseline. In particular,
confrontation visual fields were normal.
Over the course of the following 48 hours,
the patient's reading ability improved significantly. He could read and understand
simple sentences, although not complex ones.
Writing became entirely normal. Reading
comprehension then gradually improved in
the ensuing weeks, and extensive psychometric testing in April 1986 showed similar
defects to those observed in September 1985.
At this time, reading comprehension was
normal, although the patient stated that his
reading speed was slower than before, and he
had lost interest in reading.
MRI (15 Tesla) brain scan (fig) on the
second day of alexia showed: 1) the pre-

841

existing right frontal lesion; 2) an area of
increased signal intensity in the paraventricular white matter of the left parietooccipital junction, underlying the junction of
the left angular gyrus and left lateral occipital
gyrus (fig A); and 3) an area of increased signal
intensity in the overlying parieto-occipital
cortex (fig B). MRI brain scan carried out four
months later again showed the left subangular
lesion but no longer showed the cortical high
intensity signal. This lesion was seen on T,
weighted but not on T, weighted images. CT
brain scan, carried out on the second day of
alexia, showed the pre-existing right frontal
lesion but failed to demonstrate the left
parieto-occipital lesion that was observed
with MRI. Six weeeks later a CT brain scan
demonstrated the new left subangular infarction. Octopus perimetry done five months
after the stroke was normal.
The patient fulfilled the criteria for the
diagnosis of pure alexia established by
Benson and Geschwind.' He had severe disturbance of reading comprehension,
relatively preserved writing, and absence of
aphasia or dementia. The pre-existing deficits
secondary to the frontal lobe lesion could be
separated from his more recently developed
reading deficit. The impairment of written
calculations, a frequent finding in pure alexia,8
was attributed to difficulties in comprehension of written numbers, since mental calculation abilities were spared. No elements of
the Gerstmann syndrome were present.
MRI demonstrated an area of increased
signal intensity (low T,) in the white matter
immediately underneath the cortex at the left
parieto-occipital junction. The lesion was
observed on T, weighted images on the
second day, when T, weighted images and
CT were negative, a pattern of abnormality
that is characteristically seen in acute cerebral
infarction.9 The CT demonstrated the lesion
six weeks later. The transient increased MRI
signal intensity in the overlying parietooccipital cortex was felt to reflect reactive
hyperaemia associated with the acute infarc-

tion.
To our knowledge, this is the first report of
pure alexia without hemianopia due to a
parieto-occipital ischaemic infarction. The
lesion was confined to the paraventricular
white matter subjacent to the left angular
gyrus and spared the optic radiations. Three
previously reported patients had pure alexia
without hemianopia due to lesions with
similar anatomical localisation.5' Nonetheless, localisation was less precise because the
lesions were acute surgical or intracerebral
haematomas presumably with associated
mass effect.
We would like to express our appreciation to Ms
Barbara Reader for her assistance in the preparation
of this manuscript.

VICENTE J IRAGUI
MARK KRITCHEVSKY
San
Diego VA Medical Center
Neurology Service,
and Department of Neurosciences, University of
California, San Diego, La Jolla, CA, USA

Figure MRI brain scan of pure alexia carried out on day 2. Spin echo, T7-weighted images
(TR 2000 ms, TE = 70 ms). Note subcortical area of increased signal intensity underlying
the left parieto-occipital junction A) and area of increased signal intensity in the overlying
parieto-occipital cortex B).
=

1 Geschwind N. Disconnection_syndromes in
animals and man: Part 1, Brain 1965;88:
237-94.
2 Benson DF, Geschwind N. The alexias. In:
Vinken PJ, Bruyn GW, eds. Handbook of
clinical neurology, Vol 4, Disorders of Speech,
Perception and Symbolic Behavior. Amsterdam: North-Holland, 1969:112-40.
3 Greenblatt SH. Localization of lesions in alexia.

842
In: Kertesz A, ed. Localization in neuropsychology. New York: Academic Press, 1983:
323-56.
4 Damasio AR, Damasio H. The anatomic basis of
pure alexia. Neurology 1983;33:1573-83.
5 Greenblatt SH. Subangular alexia without
agraphia or hemianopia. Brain Lang 1976;3:
229-45.
6 Assal G, Hadj-Djilani M. Une nouvelle observation d'alexie pure sans hemianopsie. Cortex
1976;12:169-84.
7 Pirozzolo FJ, Kerr KL, Obrzut JE, et al.
Neurolinguistic analysis of the language
abilities of a patient with a "double disconnection syndrome": A case of subangular alexia in
the presence of mixed transcortical aphasia. J
Neurol Neurosurg Psychiatry 1981;44:152-5.
8 Friedman RB, Albert ML. Alexia. In: Heilman
KM, Valenstein E, eds. Clinical neuropsychology. New York: Oxford University
Press, 1985:49-73.
9 Kinkel PR, Kinkel WR, Jacobs L. Nuclear
magnetic resonance imaging in patients with
stroke. Semin Neurol 1986;6:43-52.

Minor hemisphere syndrome following
left hemispheric lesion in a right handed
patient

Some degree of unawareness of hemiplegia
occurs in about one third of cases of right
hemisphere patients with left hemiplegia.'
Anosognosia for hemiplegia may be
associated with left hemiasomatognosia. Such
patients behave as if the left half of their body
was no longer part of themselves.3 In contrast
to the relative frequency of such phenomena
following damage to the right hemisphere,
anosognosia and hemiasomatognosia have
rarely been reported following left hemisphere lesions. None of the reported cases has

concerned authentic right-handed patients
and the degree of language impairment has
often been unknown.
We describe a case of a strongly right
handed patient with asomatognosia and
anosognosia for right hemiplegia. She
exhibited other deficits relating to the socalled minor hemisphere syndrome, and had
no language disorders.
On the thirteenth day after aortic valve
replacement for aortic regurgitation, the
patient, a 68 year old right handed woman,
developed an acute hemiplegia. On the fourth
day following onset, neurological examination showed massive right motor deficit affecting the face, the arm and the leg. Right
plantar reflex was extensor, there was severe
hypesthesia and tactile extinction on the right
side of the body, and right homonymous
hemianopia on confrontation. Language and
praxis were normal on bedside evaluation.
The patient had motor impersistence and
strongly tended to keep her head and eyes
turned to the left, even on verbal stimulation
from the right. CT scan at 10 days post onset
revealed a left hemispheric infarct involving
the territory of the middle cerebral artery,
both deep and superficial, and the territory of
the anterior choroidal artery.
The following observations were gathered
during the first three weeks following the
stroke, during which period the neurological
condition of the patient remained essentially
unchanged. Right sided visual neglect was
seen on the dot cancellation test: she failed to
cancel 7/9 dots in the right half of the test
sheet, although she did not miss any of the
eight dots in the left half. The patient was

Letters to the Editor

presented with a list of 40 pairs of items (10
pairs of digit names, 20 pairs of object names,
10 pairs of sentence fragments). She showed
clearcut right auditory neglect: she correctly
reported all of the 40 items presented to her
left ear, but none of those presented to her

right ear.

certainly dominant for language. If the left
hemisphere was dominant for language, or if
language was bilaterally represented, the
extensive left hemispheric softening would
have caused aphasic symptoms, which was
not the case. Second, her left hemisphere was
strongly dominant for manuality, spatial
attention, and -body schema.
Few cases of right asomatognosia or
anosognosia for right hemiplegia have been
reported. Dronkers and Knight8 have described a left handed patient with clear-cut
anosognosia for right hemiplegia, right
spatial neglect and aprosodia, following left
hemispheric infarct. Cambier et al9 have
reported a similar case, but their patient
seems to be ambidextrous. Such is also the
case of the patient reported by Hermann and
Potzl." Among their patients with lefthemispheric lesions, Hecaen and Sauguet"
mention right hemiasomatognosia in two out
of 47 left-handed patients, but in none of 293
that were right-handed. Gross and Kaltenback' have described two patients with such a
denial of right hemiplegia, including one case
with hemiasomatognosia. These patients
were both aphasic, but there is no mention of
their manual dominance. Cutting2 has reported three instances of denial of hemiplegia out
of 22 testable right hemiplegic patients. Their
manual dominance and degree of language
impairment, however, are not reported.

She was largely unaware of her hemiplegia.
When asked if she could move her right hand,
sew, or knit, she answered that she could.
When asked if she could move her right leg,
she answered positively and moved her left
leg as a proof. Sometimes, the patient admitted that she needed some help with walking or
running. Once, when asked if she could cut
her meat and eat all by herself, she affirmed
that she could, if only helped to sit in her bed.
She occasionally showed some abstract
knowledge ofher deficit. She said that doctors
had told her she had hemiplegia, but that she
did not believe it, since she was not paralysed
and could walk.
The patient also showed a variable degree
of right asomatognosia. In several instances,
when shown her right hand, she would
answer that it was the hand of a corpse that
had been introduced into her bed. However,
she sometimes correctly identified the hand as
her own. Touching it with her left hand
apparently facilitated the identification. When
asked to designate her right hand she was
initially unable to find it in the bed. Twelve
days after onset, she could correctly designate
The dominant hemisphere for body
her right arm, leg, eye and cheek.
schema is usually the right hemisphere.
Her spontaneous speech was considerably
However, as we have just seen, considerable
aprosodic and she was initially unable to sing
evidence exists of reversed dominance, as
at all, although she previously used to parrevealed by right asomatognosia and anosoticipate in an amateur choir. Two weeks after
gnosia for right hemiplegia. The left hemithe stroke, her few attempts at singing were
sphere is usually dominant for manuality and
very much out of tune. Moreover, her perceplanguage, although reversed dominance in
tion of melodies was also impaired. Five days
this domain is also well known. But are all the
post-onset, she could not identify common
different patterns of co-dominance concerntunes that were hummed to her, but recoging language, manuality and body-schema
nised them readily as soon as the lyrics were
biologically possible? Or are some of these
added. Three weeks later, she could identify
functions necessarily supported by the same
five out of 10 popular tunes whose melodies
hemisphere (or by opposite hemispheres)?
were sung to her.
We shall consider each pair of functions in
The patient described herselfas completely
turn. First, the left hemisphere is usually
right handed, and denied having been forced
dominant for both language and manuality.
to use her right hand as a child. There was no
Our patient falls into the 4% of those that are
record of any significant cerebral trauma in
right handed with right hemispheric dominbirth or early childhood. She was submitted
ance for language, according to Rasmussen
to a 24 item questionnaire about her preferred
and Milner's figures'2 (for discussion, see
hand or foot in various everyday-life
Geschwind and Galaburda'3). Second, manactivities. She always unambiguously chose
uality and body schema are usually disthe right side. Her parents, grand-parents,
sociated since in most right handed subjects,
and two children were all reported to be right
the right hemisphere is dominant for body
handed.
schema. The patient reported by Dronkers
Language evaluation was normal in all
and Knight8 shows exactly the opposite patrespects. Spontaneous speech, repetition,
tern of dominance. However, Hecaen and
object and picture naming, designation,
Sauguet" mention several left handed
fluency in controlled association, comprehenpatients with left asomatognosia, suggesting
sion, and reading were flawless. She was too
that the right hemisphere can occasionally
awkward with her left hand to allow evaluasupport both functions. The pattern of
tion of writing. There was no left hand
impairment in our patient demonstrates that
apraxia. Visual identification of objects,
that can also occur for the left hemisphere.
colours and faces was normal.
Third, language and body-schema are usually
The patient was a right handed woman,
dissociated, but one hemisphere can
with no family history of left handedness.
occasionally be dominant for both functions,
After a left parieto-temporal infarction, she
as illustrated by the patients of Gross and
showed several symptoms that usually follow
Kaltenback,' who have at the same time
right hemispheric lesions, that is, symptoms
aphasia and anosognosia for hemiplegia folencompassed in the so-called minor hemi- lowing left hemispheric lesion. Our patient's
sphere syndrome. She had right hemispatial
dominance pattern is the exact opposite ofthe
neglect,' right hemiasomatognosia, anoso- pattern which prevails in most of those that
gnosia for right hemiplegia, motor aprosodia6
are right handed, since her left hemisphere is
and a severe impairment in identifying and
dominant for body schema, and her right
producing musical tunes7 but no aphasia.
hemisphere for language.
Two main conclusions can be drawn conThe case favours the idea that there is no
cerning the patient's pattern of cerebral obligatory link between any two of the three
dominance. First, her right hemisphere was cerebral functions we have considered, and