Psychological Medicine, 1975, 5, 239-248

Late recovery from head injury: case report
and review
J. CHRISTIE BROWN1
From the Maudsley Hospital, Denmark Hill, London
SYNOPSIS Some problems of prognosis in patients with head injury are reviewed, and a case of late
and unexpected recovery from post-traumatic dementia presented. The possible mechanism of
recovery is discussed with particular reference to normal pressure hydrocephalus.

Head injuries constitute an increasingly serious
problem today, and are the single most important
cause of death in road accidents. It has been
estimated that there were about 100,000 hospital
admissions of head-injured patients in 1963
(Lewin, 1968), the comparable figure for the
U.S.A. in 1967 being 750,000 with 18,000 deaths
and 13,000 cases of permanent disability
(Critchley, 1969). Among these civilian cases
there is a much lower incidence of penetrating
injury than in wartime ones, with a correspondingly greater emphasis on diffuse rather than
focal damage, both in the acute stage and the
long term.
Since the second world war, considerable
advances have been made in the treatment of all
types of head injury {Lancet, 1968). Rapid
transport to hospital, early surgical treatment of
complications, control of infection, and the
management of secondary respiratory and
metabolic changes in intensive care units have all
played their part. At one centre, for example, the
mortality rate fell from 9% in 1948 to 3.5% in
1966 (Lewin, 1967), but this widespread trend
has left in its train a greater proportion of
disabled survivors (London, 1967). In terms of
prognosis, the ability to predict the occurrence
of permanent sequelae assumes ever-increasing
importance.
Information about prognosis comes from a
number of follow-up series in the literature, but
there are certain general difficulties in interpreting these results, as there is variation
'Present address: Department of Psychological Medicine,
University College Hospital, London W.C.1E 6AU.

between the studies in case selection, in the
range, method, and accuracy of the assessment of
important variables in the acute stage and at
follow-up, in the length of follow-up, and in the
criteria used to denote recovery. There may even
be differences of terminology, 'post-traumatic
syndrome' referring to anything from the
symptomatic triad of headache, giddiness, and
irritability, to a wide constellation of disabilities
including amnesia.
Reports of unselected admissions (Harris,
1970; Lewin, 1970; Steadman and Graham,
1970) give mortality rates from 1.9-3.8%, nearly
all deaths occurring within 48 hours of injury.
The vast majority of patients are back at work
within a few weeks but, in Steadman and
Graham's series at five-year follow-up, 13% had
some residual disabilities and 1% were unable to
work. If those patients with post-traumatic
amnesia lasting 24 hours or more are considered
separately, it is found that 20-30% require less
taxing employment than before injury and 10%
are unemployable.
Some studies deal with patients selected for
more severe injury. In one (London, 1967), of
433 patients admitted to hospital 203 died, and
92 (23%) were disabled five years later. Over half
of these disabled survivors had been unconscious
for more than 24 hours, but about three-quarters
had returned to some kind of employment. In a
second study (Lewin, 1967), a subgroup of
patients 'in coma, or sufficiently unconscious to
be mute and unresponsive to speech and
command, unable to swallow and requiring
artificial feeding, for at least one month' were

239

240

J. Christie Brown

followed up. Such patients represent about 0.4%
of all admissions, and of 102 cases, 39 died. Of
the 63 survivors, 19 returned to their original job
and 29 to a simpler one.
The natural history of developments in the
first few months after severe injury is described
by Carlsson et al. (1968). Four hundred and
ninety-six cases of closed head injury in patients
unconscious for at least 12 hours were studied.
Thirty-four per cent died, most within the first
48 hours, and the numbers followed an exponential curve against time. In the survivors the
course of coma was also exponential; half the
patients regained consciousness within 32 hours,
and the average duration of coma was 57 hours.
Although mortality was related to age, duration
of coma was not. The time taken to return to
work, called the 'restitution time', was also
studied in the survivors. Eighty-two per cent
returned to work, the average restitution time
being 11 weeks, and the curve of this restitution
time was also exponential. Older patients took
longer to return to work, and the data were
summarized by the relationship
Restitution time = Comatimex(15 + 1.5
xage)
It is a characteristic of an exponential curve that,
within any given interval, the probability of an
event, such as the end of coma, remains constant.
So the previous duration of coma cannot be
used, in the acute stage at least, to predict the
likelihood of its ending.
Although this study showed a mean restitution
time of 11 weeks, and it is usually stated that the
disabilities after head injury reach a 'steady state'
in six months (Muller, 1969), very little is known
about the natural history of the sequelae in the
longer term. Miller and Stern (1965) investigated
the 92 survivors of 100 cases with post-traumatic
amnesia of at least 24 hours by follow-up at
three and 11 years. They found improvement in
neurological and social disability between the
first and second examinations, and 74% of
patients judged unfit for work at the first assessment were in employment at the second.
Although return to work is a commonly-used
criterion of recovery, it certainly does not
indicate the absence of sequelae; nor does
failure to work necessarily show the presence of
disabilities directly related to brain tissue

damage. In a study of civilian cases in Luxembourg (Muller, 1969), there was a definite
relationship between failure to return to work
and the availability of social security benefit. In
a series of 281 injured war veterans matched with
controls (Caveness, 1966), 'social and economic
failure' were relatively more common in younger
and less well-educated patients irrespective of
the severity of injury. In cases of penetrating
injury, a more accurate assessment of brain
damage may be made than in closed ones, and
in his investigation of 670 patients Lishman
(1968) used two indices, 'depth of penetration'
and 'total brain tissue destroyed'. Both showed
significant correlations with psychiatric disability found at follow-up, but the author
comments that 'the fact that (they are) statistically significant cannot hide the fact that the
greater part of the co-variance between brain
damage and psychiatric disability is left unaccounted for'.
Despite these difficulties, an overall pattern is
discernible in civilian head injury admissions. In
unselected cases mortality is under 4%, and at
least 95% of the survivors are back in their
accustomed job within three months, only about
1% being severely disabled. In cases with posttraumatic amnesia of more than 24 hours
20-30% require less taxing employment and
10% are unemployable. Even in the cases with
coma lasting one month or more, 40% returned
to their previous employment. Head injury is 'an
eminently recoverable condition' (Lewin, 1968).
In general, the greatest progress is made in the
first six months, but decelerating improvement
may continue over a number of years. The length
of post-traumatic amnesia 'can be accepted as
the single most useful guide to the severity of the
underlying brain injury' (Whitty and Zangwill,
1966).
Unfortunately, these statistically viable
generalizations conceal a disconcerting degree of
individual variation which makes accurate
prognosis in the single case a notoriously difficult
clinical problem. Indeed 'the development
recession and residuum of the clinical sequelae
have a degree of variability as great as the
injuries themselves' (Caveness, 1966). Advanced
age, pre-existing cerebrovascular disease, respiratory and cardiovascular disturbances,
localized intracranial haemorrhages requiring

Late recovery from head injury: case report and review

241

was a laceration over the occiput but no localizing
neurological signs. Radiographs of the skull showed
bilateral linear fractures extending upwards from the
foramen magnum on either side of the external
occipital protruberance.
Two hours after admission he had two generalized
clonic seizures in quick succession followed by
sustained conjugate deviation of the eyes to the right.
He was treated with phenobarbitone and paraldehyde. Later in the evening he became more easily
rousable and showed some purposive movements
but was restless and photophobic.
He remained much the same until the following
afternoon when he climbed over the cot side of his
bed and fell, hitting his head on the floor. There was
no immediate change but by the evening he had
developed neck stiffness. The next day (8 August
1963) head retraction developed and lumbar
puncture showed bloodstained fluid under a pressure
of 210 mm. Over the next 24 hours he became more
deeply unconscious and the plantar responses became extensor.
He was transferred to the Guy's/Maudsley Neurosurgical Unit that evening (9 August). His blood
pressure had risen from 170/80 in the casualty
department to 220/120, and he showed Cheyne
Stokes respiration. Ventriculography and posterior
fossa exploration were performed immediately. The
3rd ventricle was shifted slightly to the right and the
aqueduct kinked forward, suggesting space occupying lesions in the left temporal region and in the
posterior fossa. The intraventricular CSF pressure
was 600 mm. A small extradural haemorrhage was
CASE REPORT
evacuated from over the left hemisphere, likewise a
left intracerebellar haematoma.
The patient was born on 7 April 1928, had a normal
During the next four days he was improved, taking
childhood, and was average at school. He was in
regular employment from the age of 14 years, joining fluids, answering simple questions, and giving his
the police force at 21 and the mounted police at 26. Christian name correctly though giving his age as 19.

surgical intervention, raised intracranial pressure,
papilloedema, focal seizures, and signs of midbrain compression are all associated with poor
prognosis (Goutelle and Mouret, 1970; Harris,
1970). But to none of these factors can be ascribed
a precise and independent value in predicting
outcome.
EEG studies are of only limited prognostic
value. In the acute stage, a 'flat' tracing is
associated with high mortality (Rodin, 1967) and
the appearance of focal slowing may indicate a
developing localized lesion. There is a rough
correlation between EEG abnormality soon after
injury and later psychological disabilities
(Ladurner and Lorenzoni, 1970), but marked
EEG abnormalities may disappear at an early
stage in cases both with and without late sequelae
or may persist in the absence of other sequelae.
Serial records have an important place in
monitoring individual progress; where these are
abnormal, the greatest improvement usually
occurs in the first six months, but the picture
may be complicated by the late appearance of
focal abnormalities.
The following case is described because the
very unusual 'natural history' provides a specific
illustration of the problems of prognosis which
have already been described in general terms.

He married at 23 and a child was born when he was

29. There was no past history of serious illness.
HEAD INJURY

The head injury occurred when he was 35 years old.
At 4.00 p.m. on 6 August 1963 he was riding his
horse in Brixton when it suddenly reared and threw
him. His head struck the side of a van and the
ground. He was probably rendered unconscious
immediately, though he may have recovered for a
brief period during his journey to hospital by
ambulance.
HOSPITAL ADMISSION: 6 A U G U S T - 1 4 NOVEMBER 1 9 6 3

On admission to a teaching hospital casualty department half an hour after the accident, he was unconscious but responding to painful stimuli. There

Some neck retraction persisted, the left plantar

remained extensor, and he showed fine bilateral
nystagmus.
Five days postoperatively (16 August 1963) he
became drowsy and developed weakness of abduction of first the left and then the right eye. A left
frontotemporal burrhole allowed the evacuation of a
subdural haematoma consisting of about 50 ml of
altered blood. Thereafter he improved a little but
remained drowsy and disoriented for time, place, and
person. An EEG on 22 August 1963 showed marked
abnormalities with mixed irregular slow waves and
10-20 second bursts of symmetrical low voltage theta
activity. Further exploration for subdural haematomas by biparietal burrholes was without result.
He remained drowsy, disoriented, and confabulating. He was able to get up and walk a little, but
with gross trunk ataxia and some ataxia of the left

/ . Christie Brown

242

arm. Bilateral anosmia was noted. Ventriculography
was repeated on 9 September 1963 because of his
lack of progress. There was moderate symmetrical
dilatation of the lateral ventricles and still some shift
to the right of the 3rd ventricle. Air did not enter the
4th ventricle. The intraventricular pressure was
30 mm. An EEG on 11 September 1963 remained
grossly abnormal as before.
In September, vomiting became a problem and
continued intermittently for several weeks. He was
noted to have lost 2i stones in weight since admission.
The serum sodium was found to have fallen to
128 mEq per litre (normal 140-152) but was restored
to normal by oral sodium chloride.
During October his appetite and ataxia improved a
little. His speech also improved and he could write
his name. However, he remained severely disoriented in time and place, was unable to recall
day-to-day events, misidentified doctors, and had
difficulty in recognizing his wife. Confabulation

persisted. There was some evidence of nominal
dysphasia and his drawing ability was very poor.
REHABILITATION CENTRE: 14 NOVEMBER 1963-10
JUNE 1964

He was transferred to a rehabilitation centre on
14 November 1963 with a guarded prognosis. It was
anticipated that he might regain orientation in two
or three months' time but the outlook for social
recovery was uncertain. The discharge letter said 'I

have seen patients like this in the past who have
made a complete recovery. However, the longer the
duration of the confusion, the worse, as a rule, is the
ultimate prognosis'.
In January his wife had a second baby. In February
he was noted to be less ataxic, but was facile,
disinhibited, disoriented, and confabulating. By
April he was better at recognizing members of the
family but was said to be emotionally 'explosive'. In
June his physical recovery was regarded as excellent,
with only minimal ataxia but mentally he was still
disoriented and severely confused. He was liable to
outbursts of shouting and obscene language, with
occasional hitting out and episodes of wandering
away. His appetite was voracious and he now
weighed 13 st. 41b.
It was felt that his wife could not be expected to
care for him at home and he was admitted to his
area psychiatric hospital from June to September
1964, pending detailed assessment at the Maudsley
Hospital. There he was doubly incontinent, aggressive, and abusive. He had no memory of any
operation and was uncertain whether he had had an
accident.
INPAT1ENT, MAUDSLEY HOSPITAL: 8 SEPTEMBER-20
NOVEMBER 1964

His mental state fluctuated markedly. His mood was
labile and he was inclined to outbursts of shouting.
He was frequently incontinent of urine at night and

TABLE
PSYCHOLOGICAL TEST RESULTS

Psychological test
Wechsler Bellevue Scale
Full Scale IQ
Verbal IQ
Performance IQ
Wechsler Adult Intelligence Scale
Full Scale IQ
Verbal IQ
Performance IQ
Walton Black New Word Learning
Graham Kendall Memory for Designs
Meyer Paired Associate Learning
AVR
VDRg
Wechsler Logical Memory
Immediate
Delayed
Delayed as % of immediate
Rey Osterrieth Drawing Test
Copy
Delayed Recall
%loss
•Short for. n of WAIS.

Sept. 1964

Jan. 1965

July 1965

89
86
92
8
2

93
93
93
7
2

July 1966

May 1969

June 1971

96'

102
96
109

89
93
84

<4
6

41 +
31

10
1
14
9
7.5
3.0
40%

7.5
5.5
73%

47
26
44.6%

47
30.5
35%

243

Late recovery from head injury: case report and review

sometimes of faeces as well. There were episodes of
coarse behaviour, such as eating with his hands.
His memory also fluctuated. At times he could get
the date within a day or so, but at other times there
was marked disorientation. He was once said to
remember events immediately before the accident
but this was recalled after a conversation with a
police colleague. His general recall of past events was
grossly impaired, and his family history, for example,
was given very inaccurately.
An EEG on 9 September 1964 showed an abnormal
and generally slowed record with symmetrical
bilateral anterior slow waves, indicative of diffuse
cerebral pathology 'of mild degree but sufficient to
account for dementia' (Fig. 1). Repetitions showed
some dubious sharp or slow temporal lobe abnormalities but a sleep record showed no clearly localized
abnormality. In general the records were improved
compared with the year before.
Psychological testing on 10 and 12 November 1964
showed a Wechsler Bellevue Intelligence Quotient of
89, verbal 93, performance 84 (Table). The Wechsler
Memory Scale showed a memory quotient of 83, but
with very poor results in subtests dealing with
information and orientation. Eisenson's tests for
aphasia showed slight difficulty with auditory and
verbal comprehension. Tests of new learning produced results in the category of organic brain
damage (Walton Black New Word Learning Test—
less than 4; Graham Kendall Memory for Designs
Test—raw score 10, predicted 4, difference 6; Meyer
Paired Associate Learning Test—AVR = 41 + ,
VDRg+31). The conclusion was of severe impairment of long term memory and of learning.
An air-encephalogram on 9 October 1964 showed
'cerebral atrophy' with considerable dilatation of all
ventricles but no air over the hemispheres (Fig. 2).
An obstructive element was thought to be possibly
present.
The patient was discharged to his area psychiatric
hospital with a diagnosis of cerebral atrophy and
dementia. The prognosis was regarded as 'hopeless
with cerebral atrophy, apparently permanent disorientation in time and space and features of frontal
lobe syndrome (with extraverted and cold behaviour)
together with confabulation reminiscent of Korsakov
syndrome'.

Kendall tests were now, however, within the normal
range (Table). Repeat testing on 19 July 1965
showed similar results.
A definite clinical improvement was noted by
November 1965, with lessened incontinence and

$Wmri*j*iity^^

IIS?

\WWt*w*t*^^
RP

MyH-^V^W^^

R0

•^v'V^

1971

INPATIENT, AREA PSYCHIATRIC HOSPITAL: 2 0
NOVEMBER 1 9 6 4 - 1 0 NOVEMBER 1 9 6 6

During the first months in hospital the patient's
condition remained much the same with incontinence, outbursts of shouting, disorientation, and
confabulation. Psychological testing on 5 January
1965 showed a WAIS full scale IQ of 89, verbal 86,
performance 92. The Walton Black and Graham

FIG. 1. EEG records of September 1964, May 1969, and
May 1971. R=right, L=left, F=frontal, C=central,
T~ temporal, P=parietal, P-T=parieto-temporal, 0 =
occipital, P-0 =parieto-occipital.

244

J. Christie Brown

(c)
(d)
FIG. 2. Lumbar air-encephalogram of October 1964 showing dilatation of whole ventricular system.

shouting and some improvement in orientation.
Marked memory defects were still clinically apparent,
however, and he was unable to settle in occupational
therapy. Gradual improvement continued so that by
June 1966 he was working well, no longer incontinent or having 'outbursts', and 'more aware of
social and political events'. Psychological testing on

1 January 1966 confirmed sustained improvement on
the Walton Black and Graham Kendall Tests
(Table). The possibility with continued improvement
of some simple work in a protected environment
was entertained.
The emphasis of concern shifted to the problem
of getting him home and to the question of his

Late recovery from head injury: case report and review
relationship with his wife. In July he began talcing his
first nights at home, but by September his wife
proved unwilling for this. He had been sexually
overdemanding, irritable and fussy with the children,
and she no longer felt any love for him as he was 'no
longer the man I married'.
Despite attempts to dissuade him, he discharged
himself from hospital in November 1966. Shortly
before this the General Aptitude Test Battery had
shown scores in the low average range on almost all
aptitude tests, and he was considered as possibly
suitable for work as an electrician's mate. His wife
left him soon after discharge to live with her mother,
but he continued to visit the family nearly every week.
OUTPATIENT EXAMINATION: MAY 1 9 6 9

245

seen on television the day before. There was no
obvious confabulation.
With regard to dysphasia, he had difficulty in
naming a telephone receiver but otherwise there was
no abnormality. Constructions with matches were
satisfactory, as were drawings. Right-left orientation
was normal. All fingers were named except the ring
finger. Simple calculations were satisfactory.
On physical examination he showed bilateral
anosmia, slight nystagmus on upward gaze, some
deafness in the left ear, poor performance on the
heel/shin test (left worse than right), and minimal
unsteadiness when turning sharply while walking.
An EEG (Fig. 1) showed a definite improvement
when compared with that of 1964. A moderate
abnormality persisted over the left central and sylvian
regions (small amplitude 24-26 c/s activity) and there
was some rhythmic 8-9 c/s activity over both central
regions.
Psychological testing showed a verbal IQ of 96 on
the short form of the WA1S and a spatial score of
110 (comparison difficult with the full WA1S). The
Token Test showed no evidence of receptive dysnhasia and all items were correct on the HalsteadWepman battery. The Meyer Paired Associate
Learning Test showed no deficit. The Wechsler
Logical Memory Subtest showed immediate
recall only 0.2 SD below mean normal, but one
hour later delayed recall was considerably impaired at 2.0 SD below normal. The Rey Osterrieth
Drawing Test for non-verbal memory similarly
showed unimpaired immediate recall but below
average recall at 40 minutes. Thus, the only
impairments were in memory after delays of 40
minutes to an hour, and these were not considered
to be far outside the normal range (Table).

His first examination thereafter was in May 1969
when he responded to a follow-up inquiry at the
Maudsley Hospital. He was living alone in his old
house caring for himself and making mortgage
repayments. He had been working since January
1968 as an electrician's mate earning at least £14 per
week. He was on no regular medication and did not
attend a clinic. He had a motor scooter on which he
travelled about London and also visited relatives in
Essex. He said he never got lost and managed
money well. He was waking quite often at 3 a.m. but
did not ruminate and soon fell asleep again. He cried
watching 'sad films about families' on TV which
reminded him of his wife.
On examination, he was well turned out, with an
easy friendly manner but a rather superficial
jocularity. His talk was normal apart from slight
distractibility. The mood was slightly labile. There
were no compulsions, delusions, or hallucinations.
Concentration was subjectively normal and he performed the 100-7 test in 45 seconds with three
OUTPATIENT EXAMINATION: MAY 1 9 7 1
mistakes.
The retrograde amnesia was extensive. There was He was seen again at the Maudsley in May 1971. In
no memory of the accident itself. He could remember general, his condition was unchanged. He had taken
being at Nine Elms Depot where he was 18 months a new job as an animal attendant looking after
before the accident, but he could not remember any horses and had in fact been kicked in the chest by
aspect of his wife's pregnancy (she was four months one of them. There was the same slightly facile
pregnant when the accident occurred). The post- joviality with some lability of mood. There was some
traumatic amnesia appeared to be of about three early morning wakening. Abnormalities on physical
years' duration. He could not remember anything examination were the same as in 1969 except that the
of the Maudsley Hospital, but said he clearly nystagmus had disappeared.
The EEG (Fig. 1) was very similar to that of May
remembered sports day at the area psychiatric
hospital in summer 1966 and from then on could 1969, though possibly with further improvement in
that there were fewer sharp phenomena. Psychologiremember fairly well.
There was no abnormality of current memory on cal testing with the WA1S showed a full scale IQ of
simple tests, and he reported no subjective difficulties. 102, verbal 96, performance 109. The Wechsler
He had arrived at the hospital at the right time and Logical Memory Test showed improved delayed rewas oriented in respect of time, place, and person. call, similarly the Rey Osterrieth Test (Table). Both
He gave a convincing account of his life without verbal and non-verbal memory were now well
inconsistencies and could remember a film he had within normal limits.

/ . Christie Brown

246
SUMMARY OF CASE REPORT

The case history can be conveniently considered
as falling into three stages. The first stage is from
the time of the patient's injury in August 1963
to his discharge to the rehabilitation centre in
November 1963. During this time his grave
physical condition was the main focus of
medical interest, and, although he remained
confused throughout, some guarded optimism
was expressed about the prognosis.
The second stage is from November 1963 until
November 1964 when he was admitted to the
area psychiatric hospital. His general physical
state improved but there was relatively little
psychological improvement, and after the most
careful assessment he was considered to have
reached a 'steady state' of irreversible dementia.
The air encephalogram showed dilatation of the
whole ventricular system but without excess air
over the surface of the brain. It was at the end of
this stage that formal psychological testing was
first carried out, and two tests of verbal learning
and one of short term visual memory all showed
marked impairment. These, and subsequent tests
of memory and learning ability are shown in the
Table.
The third stage is that of recovery. The first
indication of a change came from the psychological testing of January 1965 which showed a
very striking improvement in new word learning
and visual memory which were both within the
normal range. This finding stood out against the
background of the general clinical picture, which
was little changed. From then on improvement
continued and in May 1969 only tests of delayed
visual and verbal recall revealed some persisting
memory disorder. Even this deficit was absent
two years later. In fact, seven years after the
accident only minimal neurological sequelae
could be detected on careful examination, and
his somewhat fatuous manner and lability of
mood were the only obvious psychological
residua of his injury.
The patient recalls sports day at the area
hospital in 1966 as his first clear memory since
the accident, and can remember subsequent
events fairly well. If the 'establishment of
continuous memory' is taken as the mark of the
end of post-traumatic amnesia (Whitty and
Zangwill), its duration in this case was nearly
three years. It is harder to establish the length of

the retrograde amnesia, but the patient's wife
became pregnant four months before the accident
and he always denied any memory of this.
In summary the patient made a spontaneous
and most unexpected recovery which probably
began at some time between November 1964 and
January 1965 and which continued at first fairly
rapidly and then more slowly at least until June
1971.
DISCUSSION

In the early stages after head injury, a number of
changes may take place which are potentially
reversible. Intracranial bleeding may compress
the brain, and respiratory depression lead to
cerebral anoxia and to a secondary rise in intracranial pressure closely related to blood pCO 2
(Crockard and Taylor, 1971). Generalized
cerebral oedema may occur, partly due to
vasodilatation (Langfitt et al., 1966). A number
of other findings show that 'commotio cerebri'
is as much metabolic as mechanical. For example,
changes in water metabolism can produce either
high or low levels of serum sodium—'cerebral
salt retention and cerebral salt wasting'
(McLaurin, 1966). Changes in the CSF include
an increase in concentration of acetylcholine and
potassium (Ward, 1966), a rise in glutamic
oxaloacetic transaminase and possibly lactic
dehydrogenase (Kaltiala et al., 1968), and an
increase of CSF/plasma lactate concentration
ratio (Kurze et al., 1966). In survivors these
disturbances are fairly soon corrected, which
may account for early clinical improvement.
Despite these fragments of information, very
little is known about the processes underlying the
'remarkable powers of recovery of the severely
damaged brain' (Strich, 1969). Particularly
obscure is the slow improvement over months
or years which unfolds after the acute phase is
over. Knowledge of the lesions after injury comes
from post mortem studies which show widely
scattered haemorrhagic and necrotic lesions,
some of which are secondary to post-traumatic
brain displacements and in particular to brainstem compression (Crompton, 1971). Brainstem
and corpus callosum lesions are prominent in
patients dying after protracted coma (Jellinger
and Seitelberger, 1969), but in post-traumatic
dementia cortical lesions are present as well
(Strich, 1969). Demented patients may also show

Late recovery from head injury: case report and review

areas of axon degeneration in the white matter
(Strich, 1961), the nature and distribution of the
lesions suggesting that they are due to shearing
of fibres as a direct result of the cranial impact.
Even if it is likely that similar lesions occur in
survivors, differences in distribution may be as
important as simple quantitative variation. As
cellular regeneration does not occur, the notion
of the 'plasticity' of cerebral function may be
invoked to explain, in part at least, the recovery
that occurs. The work of Illis (1973a, b) has now
provided elegant support for this idea. In a
number of experiments in cats, one posterior
spinal nerve root was cut and spinal neurones
receiving synapses from the root examined at
various intervals. At first there was degeneration
arid loss of boutons termineaux leaving areas of
the cell membrane bare. Then increasing
numbers of small synapses appeared and
eventually, after 20 weeks, the cell surface was
normally populated. In this way new synapses
were apparently developed. It may be that such
events occur in the injured human brain.
One further possible mechanism of recovery
will be considered, as it is particularly relevant
to the case described. In 1965 Hakim and Adams
and their co-workers (Adams et al., 1965;
Hakim and Adams, 1965) described three
patients who presented with an unusual syndrome which they called 'normal pressure
hydrocephalus'. All showed progressive dementia
with early ataxia and urinary incontinence. At
lumbar puncture, CSF pressure was in the high
normal range and withdrawal of fluid produced
some improvement in the patient's condition.
Conversely, lumbar air encephalography led to
a marked but temporary worsening. The air
pictures showed generalized dilatation of the
ventricular system but little or no air in the
subarachnoid space over the hemispheres. In all
cases a ventriculoatrial shunt brought about
marked improvement, two of the three being
patients who had failed to make progress after
severe head injury.
A number of subsequent reports have described similar cases {Lancet, 1970). In some there
is a history of previous head injury or subarachnoid haemorrhage but in others no antecedent
can be found. In general, the more closely the
patients conform to the original syndrome the
more likely they are to benefit from shunting

247

procedures. Salmon (1971) treated nine cases of
'severe post-traumatic encephalopathy' with
ventriculoatrial shunts despite evidence of some
cortical atrophy as well. Two were much
improved and three improved.
The intrathecal injection of radioactive iodine
labelled human serum albumen (RIHSA)
followed by external head scanning shows a
pattern specific to normal pressure hydrocephalus. Instead of passing over the surface of
the brain and being rapidly absorbed, radioactivity is detected in the ventricles and remains
there for up to 72 hours. The RIHSA scan is
probably now the investigation of choice
(Bannister et al., 1967) as it is less disturbing than
AEG.
The pathological mechanism of normal
pressure hydrocephalus is unknown. Hakim and
Adams suggested the 'hydraulic press' hypothesis to explain the ventricular dilatation. They
postulate an initial period of high intraventricular
pressure and some dilatation: then the lower
CSF pressure that follows has a larger surface
area on which to act and can set up a greater
stretching force in the ventricular walls. The
analogy is made with the blowing up of a balloon
which requires the greatest pressure at the
beginning. However, the RIHSA studies have
shown that there is relative CSF stasis in the
ventricles which may lead to the accumulation of
metabolites capable of interfering with cerebral
function.
In the case reported in this paper, the air
encephalogram of November 1964 showed
appearances typical of normal pressure hydrocephalus. Since the publications of Hakim and
Adams in 1965 the condition has become widely
recognized and it is now usual for similar
patients to have ventriculoatrial shunting procedures performed with the expectation of
dramatic improvement.
Although it is possible only to speculate about
the mechanism of recovery in this patient, the
most parsimonious hypothesis is that the normal
pressure hydrocephalus resolved spontaneously:
in some way, ventricular CSF stasis remitted
and more normal CSF circulation was established. However, it might be misleading in the
present state of knowledge to assume that the
abnormalities underlying normal pressure hydrocephalus are ones that are either present or

248

/ . Christie Brown

absent. Disturbances of CSF circulation may
complicate the recovery phase in many cases of
head injury, present in varying degrees and
recovering spontaneously to different extents.
Patients presenting with recognizable normal
pressure hydrocephalus may represent the
extreme end of a continuum, a possibility which
merits further investigation.
Addendum An EMI scan was carried out at Atkinson Morley's
Hospital by Dr. J. Ambrose on 17.5.75. The conclusions were that
there was enlargement of the lateral ventricles, third ventricle, and
fourth ventricle without commensurate enlargement of the basal
cisterns and cerebral sulci. There was also considerable bilateral
frontal lobe atrophy as well as cerebellar atrophy.
I wish to thank all those who cared for the patient and contributed
information on which I have been able to draw. In particular, I am
grateful to Mr. Murray Falconer and Dr. Sabina Strich for much
helpful advice and to Professor Michael Shepherd who suggested the
study in the first place and provided much valuable guidance. I also
thank the patient, who has cooperated with great patience in follow-up.

REFERENCES
Adams, R. D., Fisher, C. M., Hakim, S., Ojemann, R. G.,
and Sweet, W. H. (1965). Symptomatic occult hydrocephalus with 'normal' cerebrospinal-fluid pressure. A
treatable syndrome. New England Journal of Medicine,
273, 117-126.
Bannister, R., Gilford, E., and Kocen, R. (1967). Isotope
encephalography in the diagnosis of dementia due to
communicating hydrocephalus. Lancet, 2, 1014-1017.
Carlsson, C.-A., Essen, C. von, and Lofgren, J. (1968).
Factors affecting the clinical course of patients with severe
head injuries. Journal of Neurosurgery, 29, 242-251.
Caveness, W. F. (1966). Posttraumatic sequelae. In Head
Injury. Conference Proceedings, pp. 209-219. Edited by
W. F. Caveness, A. E. Walker. Lippincott: Philadelphia.
Critchley, M. (1969). Introduction. In The Late Effects of
Head Injury, pp. 3-9. Edited by A. E. Walker, W. F.
Caveness, and M. Critchley. Thomas: Springfield, 111.
Crockard, H. A., and Taylor, A. R. (1971). The importance
of blood-CO2 levels in head-injury management: improvement with hyperventilation. (Abstract.) British Journal of
Surgery, 58, 857.
Crompton, M. R. (1971). Brainstem lesions due to closed
head injury. Lancet, 1, 669-673.
Goutelle, A., and Mouret, P. (1970). Le prognostic des
traumatismes craniens. Acta Psychiatrica Belgica, 70,
445-482.
Hakim, S., and Adams, R. D. (1965). The special clinical
problem of symptomatic hydrocephalus with normal
cerebrospinal fluid pressure. Journal of the Neurological
Sciences, 2, 307-327.
Harris, P. (1970). Current concepts in the management of
brain trauma. Journal of the Royal College of Surgeons of
Edinburgh, 15, 268-282.
Illis, L. S. (1973a). Experimental model of regeneration in the
CNS. 1. Synaptic changes. Brain, 96,47-60.
Illis, L. S. (1973b). Experimental model of regeneration in
the central nervous system. 2. The reaction of glia in the
synaptic zone. Brain, 96, 61-68.
Illis, L. S. (1973c). Regeneration in the central nervous
system. Lancet, 1, 1035-1037.
Jellinger, K., and Seitelberger, F. (1969). Protracted post-

traumatic encephalography. In The Late Effects of Head
Injury, pp. 168-181. Edited by A. E. Walker, W. F.
Caveness, and M. Critchley. Thomas: Springfield, 111.
Kaltiala, E. H., Heikkinnen, E. S., Karki, N. T., and Larmi,
T. K. I. (1968). Cerebrospinal fluid and serum transaminases and lactic dehydrogenase after head injury.
Acta Neurologica Scandinavica, 44, 124-129.
Kurze, T., Tranquada, R. E., and Benedict, K. (1966). Spinal
fluid lactic acid levels in acute cerebral injury. In Head
Injury. Conference Proceedings, pp. 254-259. Edited by
W. F. Caveness and A. E. Walker. Lippincott: Philadelphia.
Ladurner, G., and Lorenzoni, E. (1970). Electroencephalographic findings in the post-traumatic psychiatric syndrome.
Electroencephalography and Clinical Neurophysiology, 28,
326.
Lancet (1968). Leading article: Severe head injuries. 1,
514-516.
Lancet (1970). Leading article: Dementia and normalpressure hydrocephalus. 2, 1074.
Langfitt, T. W., Weinstein, J. D., and Kassell, N. F. (1966).
Vascular factors in head injury. Contribution to brainswelling and intracranial hypertension. In Head Injury.
Conference Proceedings, pp. 172-194. Edited by W. F.
Caveness and A. E. Walker. Lippincott: Philadelphia.
Lewin, W. (1967). Severe head injuries. Proceedings of the
Royal Society of Medicine, 60, 1208-1212.
Lewin, W. (1968). Rehabilitation after head injury. British
Medical Journal, 1, 465-470.
Lewin, W. (1970). Rehabilitation needs of the brain-injured
patient. Proceedings of the Royal Society of Medicine, 63,
28-32.
Lishman, W. A. (1968). Brain damage in relation to psychiatric disability after head injury. British Journal of
Psychiatry, 114, 373-410.
London, P. S. (1967). Some observations on the course of
events after severe injury of the head. Annals of the Royal
College of Surgeons of England, 41, 460-479.
McLaurin, R. L. (1966). Some metabolic aspects of head
injury. In Head Injury. Conference Proceedings, pp. 142157. Edited by W. F. Caveness and A. E. Walker. Lippincott: Philadelphia.
Miller, H., and Stern, G. (1965). The long-term prognosis of
severe head injury. Lancet, 1, 225-229.
Muller, G. E. (1969). Early clinical history, EEG controls,
and social outcome in 1,925 head injury patients. In The
Late Effects of Head Injury, pp. 414-422. Edited by A. E.
Walker, W. F. Caveness, and M. Critchley. Thomas:
Springfield, 111.
Rodin, E. A. (1967). The contribution of the EEG to prognosis after head injury. Diseases of the Nervous System, 28,
595-601.
Salmon, J. H. (1971). Surgical treatment of severe posttraumatic encephalopathy. Surgery, Gynecology and
Obstetrics, 133, 634-636.
Steadman, J. H., and Graham, J. G. (1970). Head injuries:
an analysis and follow-up study. Proceedings of the Royal
Society of Medicine, 63, 23-38.
Strich, S. J. (1961). Shearing of nerve fibres as a cause of
brain damage due to head injury. Lancet, 2, 443-448.
Strich, S. J. (1969). The pathology of brain damage due to
blunt head injuries. In The Late Effects of Head Injury, pp.
501-526. Edited by A. E. Walker, W. F. Caveness, and M.
Critchley. Thomas: Springfield, 111.
Ward, A. A., Jr (1966). The physiology of concussion. In
Head Injury. Conference Proceedings, pp. 203-208. Edited
by W. F. Caveness, A. E. Walker. Lippincott: Philadelphia.
Whitty, C. W. M., and Zangwill, O. L. (1966). Traumatic
amnesia. In Amnesia, pp. 92-108. Edited by C. W. M.
Whitty and O. L. Zangwill. Butterworths: London.