Brain
(1992), 115, 271-291
COGNITIVE PERFORMANCE
IN
MULTIPLE
SYSTEM ATROPHY
by
T. W.
ROBBINS,
1
M.
JAMES,
2
K. W.
LANGE,
3
A. M.
OWEN,
1
N.
P. QUINN
3
and c. D. MARSDEN
3
{From
the
^Department
of
Experimental Psychology, University
of
Cambridge,
the
^Department
of
Psychology,
The
National Hospital
and the
^Department
of
Clinical Neurology, Institute
of
Neurology, London,
UK)
SUMMARY
The cognitive performance
of a
group
of
patients with multiple system atrophy
(MSA) of
striato-nigral
predominance
was
compared with that
of
age
and IQ
matched control subjects, using three tests sensitive
to frontal lobe dysfunction
and a
battery sensitive
to
memory
and
learning deficits
in
Parkinson's disease
and dementia
of
the Alzheimer type.
The MSA
group showed significant deficits
in all
three
of
the tests
previously shown
to be
sensitive
to
frontal lobe dysfunction. Thus,
a
significant proportion
of
patients
from
the
MSA group failed
an
attentional set-shifting test, specifically
at
the stage when an extra-dimensional
shift
was
required. They were also impaired
in a
subject-ordered test
of
spatial working memory.
The
MSA group showed deficits mostly confined
to
measures
of
speed
of
thinking, rather than accuracy,
on
the Tower
of
London task. These deficits were seen
in the
absence
of
consistent impairments
in
language
or visual perception. Moreover,
the
MSA group showed
no
significant deficits
in
tests
of
spatial
and pattern
recognition previously shown
to be
sensitive
to
patients early
in the
course
of
probable Alzheimer's disease
and only
a few
patients exhibited impairment
on the
Warrington Recognition Memory Test. There were
impairments
on
other tests
of
visual memory
and
learning relative
to
matched controls,
but
these could
not easily
be
related
to
fundamental deficits
of
memory
or
learning. Thus,
on a
matching-to-sample task
the patients were impaired
at
simultaneous
but not
delayed matching
to
sample, whereas difficulties
in
a pattern-location learning task were more evident
at its
initial, easier stages.
The MSA
group showed
no consistent evidence
of
intellectual deterioration
as
assessed from their performance
on
subtests
of the
Wechsler Adult Intelligence Scale (WAIS)
and the
National Adult Reading Test (NART). Consideration
of individual cases showed that there
was
some heterogeneity
in the
pattern
of
deficits
in the
MSA group,
with
one
patient showing
no
impairment, even
in the
face
of
considerable physical disability.
The results show
a
distinctive pattern
of
cognitive deficits, unlike those previously seen using
the
same
tests
in
patients with Parkinson's
and
Alzheimer's diseases,
and
suggesting
a
prominent frontal-lobe-like
component.
The
implications
for
concepts
of
'subcortical' dementia
and
'fronto-striatal' cognitive dysfunction
are considered.
INTRODUCTION
In contrast
to the
extensive investigation
of
cognitive deficits
in
Parkinson's disease
(PD),
Huntington's disease (HD)
and
progressive supranuclear palsy
(PSP) (for a
review
see Brown
and
Marsden, 1988), there
has not
been
a
single neuropsychological study
of multiple system atrophy (MSA),
a
condition which accounts
for
5
—10%
of
parkin-
sonians. Multiple system atrophy
is a
term used
to
describe
a
progressive neurological
condition, usually incorporating
a
parkinsonian syndrome (striato-nigral degeneration)
Correspondence to:
DrT. W.
Robbins, Department
of
Experimental Psychology, University
of
Cambridge, Downing
Street, Cambridge
CB2 3EB, UK.
© Oxford University Press
1992
272 T. W. ROBBINS AND OTHERS
and often with additional autonomic failure (Shy-Drager syndrome)
or
cerebellar and/or
pyramidal signs [olivopontocerebellar atrophy (OPCA)] (Berciano, 1982; Oppenheimer,
1983;
Quinn, 1989). Multiple system atrophy appears
to
have distinct neuropathological
signs,
as
shown from
the
recent findings
of
glial cytoplasmic inclusions,
not
only
in
those regions where cell loss
and
gliosis
are
classically found
in MSA,
such
as the
putamen,
but
also
in the
primary
and
premotor cerebral cortex
and
the paracentral gyrus
(Papp
et al.,
1989; Nakazato
et al.,
1990).
As in
PD,
the
brains
of
patients with
MSA
show profound depletion
of
forebrain dopamine, noradrenaline
and
(more variably)
acetylcholine (Spokes
et al.,
1979).
Cognitive deterioration
is not
generally considered
to be an
integral feature
of
MSA
(Quinn, 1989). Although coincidental Alzheimer's disease
can
occur (Trotter,
1973;
Kosaka
et al.,
1981), this
is
seen
at a
rate
no
higher than
in the
general age-matched
population. The clinical impression is that global cognitive impairment
is
not
a
consistent
feature
of
patients with MSA. However,
in our
experience routine neuropsychological
assessment often reveals evidence
of
underfunctioning
of
variable
and as yet
undefined
cognitive deficits.
In this study
we
investigated cognitive function
in a
group
of
patients with
MSA of
the striato-nigral type, concentrating
on
tests that we have found
to be
sensitive to frontal
lobe damage
and PD. In
recent papers (Downes
et al., 1989;
Owen
et al., 1991) we
introduced
an
attentional set-shifting paradigm based
on
learning theory
and
showed
that unmedicated
PD
patients performed
at
least as poorly
as
either medicated patients,
who were later
in
the course
of
PD,
or
patients with neurosurgical lesions of the prefrontal
cortex. However, patients with temporal lobe excisions
or
amygdalohippocampectomy
are
not
impaired
in
this test, showing that
the
frontal lobe deficits have considerable
neural specificity. Another test suitable
for
assessing frontal lobe function
is the
Tower
of London planning task, which Shallice (1982) showed was sensitive to anterior cortical
damage. Owen
et al.
(1990a) confirmed that performance
on a
similar computerized
version of the test
is
impaired,
in
terms of both accuracy
and
speed measures of thinking,
in patients with neurosurgical lesions
of the
frontal
(but not
temporal) cortex.
The
computerized version allowed
the
inclusion
of a
motor control condition, where
the
subject had
to
make moves defined
by
the computer program, thus enabling true thinking
time
on the
test problems
to be
calculated by subtraction. Whereas frontal patients were
found to be less accurate
on
this task,
and
slower to think about the problems subsequent
to
the
first move (Owen
et al.,
1990a), patients medicated
for PD
were no less accurate,
but were significantly slower
in
making their initial move (Morris
et al.,
1988; Owen
et
al.,
19906, unpublished data). Finally, we employed
a
test
of
spatial working memory
that
has
again been found
to be
sensitive
to
frontal lobe damage, partly because
of the
use
of
inefficient strategies
for
reducing
the
load
on
working memory (Owen
et al.,
1990a). Performance
on a
similar test
had
previously been found
to be
normal
in PD
patients
on
medication (Morris
et al.,
1988). Thus,
in
summary,
we
employed one test
(attentional set-shifting) identifying similar deficits
in PD and
frontal lobe patients, another
(Tower
of
London) which produced qualitatively different deficits,
and a
third (spatial
working memory) sensitive
to
frontal lobe damage
but not to
unmedicated
or
mild cases
of medicated
PD.
In order
to
define
the
specificity
of
any deficits found with
the
tests
of
frontal lobe
function,
we
also administered,
to
most
of
the patients, tests from
a
battery which
we
COGNITION IN MULTIPLE SYSTEM ATROPHY 273
had previously shown to be sensitive to memory and learning deficits in PD and
Alzheimer's disease (Sahakian et al., 1988).
SUBJECTS
Multiple system atrophy (MSA) patients
A consecutive series of 16 patients diagnosed as clinically probable MSA of striato-nigral type (MSA-SND)
was studied. All patients were diagnosed at the National Hospital for Neurology and Neurosurgery, Queen
Square, London, according to the criteria set out by Quinn (1989) (see Table 1). Cases
1
-15 had a poor
TABLE 1. CLINICAL FEATURES PRESENT IN MSA GROUP
Case
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Sex
M
M
M
M
M
M
F
M
M
M
M
M
M
M
M
F
Parkinson
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Cerebellar
—
-
+
+
+
+
-
+
+
+
-
-
-
_
+
_
Pyramidal
+
+
+
+
+
+
+
+
+
-
-
+
/-
+
/ —
-
-
_
Autonomic
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Duration of
disease (yrs)
9.5
4.5
7.5
2
7
7
6
9
6
10
6
5.5
2.5
10
2
4
Hoehn and Yahr
stage
III
IV
V
III
III
III
III
V
IV
IV
III
IV
V
III
III
IV
+ = present; +/- = brisk reflexes+equivocal plantar responses; - = absent.
or absent response to levodopa treatment. Case 16 had greater than 50% response with dyskinesia and
fluctuations, but within 2 yrs had developed respiratory stridor, frequent falls, postural faintness and urinary
disturbance with abnormal sphincter EMG. The duration of the disease averaged 6 yrs, taken from the
time of initial symptoms. All but two patients had received a CT scan. One of these two received instead
an MRI scan. Six of the patients also underwent [
l8
F]fluoro-dopa PET scans. Only two of the 16 MSA
patients were not receiving medication at the time of
testing.
All of the remainder were receiving levodopa
preparations or bromocriptine. Two patients were receiving the anti-cholinergic drug benzhexol (Artane).
Normal controls
Control subjects were drawn from a large population of volunteers aged between 35 and 70 yrs from
London, Cambridgeshire and Newcastle upon Tyne (North-East Age Research Panel). These controls
had no history of neurological or psychiatric disorder. From this population, two control groups were
selected separately for the Working Memory and Planning (control group 1; n = 16) and for the Visual
Memory batteries (control group 2; n = 15), matched for age and NART IQ. The performance of the
MSA group on the ID/ED attentional shift task was evaluated by comparison with a control group
(n = 32, two for each MSA patient) chosen on a blind basis, but matched overall for age, NART IQ
and male:female proportion.
Informed consent was obtained from all patients and controls.
274 T. W. ROBBINS AND OTHERS
METHODS
Clinical neuropsychological assessment
All patients and controls were tested on the National Adult Reading Test (NART) (Nelson, 1982) to
obtain an estimate of their optimal intellectual ability. All but one of the patients received routine clinical
assessment in the Department of Psychology, The National Hospital, including a shortened form of the
WAIS or WAIS-R (Wechsler, 1955, 1981). Verbal IQs were prorated from the Vocabulary, Similarities,
Arithmetic and Digit Span subtests. Performance IQs were prorated from the Picture Completion, Block
Design and Picture Arrangement subtests. Similarly, all but one MSA patient received the Recognition
Memory Test (Warrington, 1984). Thirteen patients received tests of visual perception (Unconventional
Views and Incomplete Letters tests) (Warrington and James, 1967; Warrington and Taylor, 1973). For
the McKenna Naming Test (McKenna and Warrington, 1983) scores were available for nine of the 16
patients. Time constraints and physical disability made it impossible for all tests to be administered in
the clinical session.
Experimental computerized tests
The main testing procedures were taken from the Cambridge Neuropsychological Test Automated Battery
(CANTAB), a series of cognitive tests run on an Acom BBC Masterl28 Microcomputer, with a high
resolution Microvitec 30 cm VDU and a Microvitec Touchtec 501 touch-sensitive screen. These were
generally administered in two test sessions, one of approximately 90 min and one of 45 min. Subjects
were seated approximately 0.5 m from the monitor and it was explained that they would have to respond
to stimuli by touching the screen.
They were introduced to the apparatus by way of a 'sensorimotor screening task' in which they were
asked to respond to a series of flashing crosses on the screen by placing the index finger of their preferred
hand on the centre point of each cross. All subjects satisfactorily completed this task and so were allowed
to proceed to the following experimental tests. The first three tests were then taken from the CANTAB
Working Memory and Planning tests (see Owen et al., 1990a).
Spatial short-term span. In this computerized version of the Corsi Block Tapping task, the ability of
subjects to remember a sequence of squares highlighted on the screen was determined. Each trial began
with the same arrangement of nine white 3 cm squares arranged quasi-randomly in the space of the screen.
Subjects were instructed to observe the boxes, as some would change colour one after the other. Their
task was to remember the location and the sequential order of the boxes changing colour. During each
series,
one box would change colour for 3 s and then return to white before the next in the sequence changed
to the same colour. The subject was then prompted by a tone to repeat the sequence by touching the boxes
in the same order. During this response sequence, each selected box changed to the same colour for 1 s
and a feedback tone sounded. Following one demonstration trial by the experimenter, the task began at
the simplest level of difficulty with a two box sequence. After each successful trial, the number of boxes
changing in the next sequence was increased by one to a maximum of nine boxes. After an incorrect attempt
at any particular level, an alternative sequence of the same length was presented. This continued until
the subject had failed three consecutive trials at any one level. During each trial, a number in the bottom
left-hand corner of the screen indicated the length of the current sequence. Also, all boxes changed to
the same colour within each series, although on any two adjacent sequences different colours were used
to minimize interference. The spatial short-term memory span was calculated as the highest level at which
the subject had successfully recalled at least one sequence of
boxes.
The number of errors made in attaining
this level was also recorded.
Spatial
working
memory
task.
In this task the subject was required to 'search through' a number of coloured
3 cm boxes presented on the screen, by touching each one in order to 'open it', thus revealing its contents
(see Fig.
1A).
The colours of the boxes stayed constant within a trial, but varied over trials. The goal
was to collect blue 'tokens' hidden inside the boxes, and once found, to use them to fill an empty column
at the side of the screen. The subjects were instructed that at any one time there would be a single token
hidden inside one of the boxes. Their task was to search until they found it, at which point the next token
would be hidden. The key instruction was that, once a blue token had been found within a particular box,
then that box would not be used again to hide a token for that particular trial. Since every box was used
once,
on each trial the total number of blue tokens to be found corresponded to the number of boxes on
COGNITION IN MULTIPLE SYSTEM ATROPHY
275
•
B
FIG.
1. A, the arrangement of boxes on the sceen for the spatial working memory task. The example shows a six
box trial, in which one of
the
boxes
has
just been chosen to reveal a 'blue token' which can now be added to one already
in the column on the right-hand side of the screen. This leaves the subject with four more tokens to find, in boxes
in which they have not previously been located, B, a sample three move problem for the modified Tower of London
task. The goal position of the balls is shown in the top arrangement, c, sample pairs of compound stimuli and their
arrangement on the screen in the ID/ED attentional set-shifting task.
the screen. In this task, two types of search error are possible. First, a subject may return to open a box
in which a blue counter has already been found (a 'between search' error). Secondly, a subject may return
to a box already opened in the same search sequence (a 'within search' error). Subjects could search the
boxes in any order, but for control purposes, the number of boxes visited (excluding errors), before a