Measurements in Idiopathic
Normal Pressure Hydrocephalus
Computerized neuropsychological test battery
and intracranial pulse waves
Anders Behrens
Department of Pharmacology and Clinical Neuroscience
Department of Radiation Sciences, Umeå Univesity
Umeå 2014
Umeå University Medical Dissertations, New Series No 1689
Responsible publisher under Swedish law: the Dean of the Medical Faculty
This work is protected by the Swedish Copyright Legislation (Act 1960:729)
ISBN: 978-91-7601-155-3
ISSN: 0346-6612
Adaptation of "The Wave" by Alexander Harrison 1885. Source: Wikimedia
Commons
Elektronisk version tillgänglig på http://umu.diva-portal.org/
Printed by: Print & Media
Umeå, Sweden 2014
"Extraordinary claims require extraordinary evidence."
Carl Sagan
"Disorders of intellect [...] happen much more often than superficial observers will
easily believe."
Samuel Johnson: The History of Rasselas, Prince of Abyssinia (1759)
"Nog finns det mål och mening i vår färd. Men det är vägen som är mödan värd."
Karin Boye
To my family Fanny, Astrid and Lovisa
i
TABLE OF CONTENTS
ii TABLE OF CONTENTS
iv ABSTRACT
vi ORIGINAL PAPERS
1 INTRODUCTION
1 Non-invasive intracranial pressure
Transcranial Doppler ultrasonography
ICP pulsations
Idiopathic Normal Pressure Hydrocephalus
Hydrocephalus overview
Idiopathic Normal Pressure Hydrocephalus
2 4 6 6 7 7 Epidemiology
8 Pathophysiology of INPH
9 Diagnosing INPH
9 Treatment of INPH
10 Supplemental tests
11 Cognitive testing in INPH
Neuropsychological testing in INPH
Computerized neuropsychological testing
Test development
12 15 16 16 Research rationale
18 AIMS
19 MATERIALS AND METHODS
19 Ethical approval
19 Patients and controls
19 INPH patients of paper I and II
Cognitively impaired patients of paper III
Hydrocephalus patients of paper III
Healthy elderly of papers III and IV
20 20 20 21 INPH patients of paper IV
Simultaneous infusion test, transcranial Doppler and pressure measurements 22 24 Mathematical model
25 Bench test
26 Neuropsychological test battery
26 Software design
Computerized General Neuropsychological INPH Test (CoGNIT)
Statistics
26 29 31 RESULTS
Transcranial Doppler pulsatility index and intracranial pressure
Intracranial and lumbar pressure waves
Baseline
ii
31 32 33 33 Elevated ICP
34 Time versus frequency analysis
35 CoGNIT
35 Validity
36 Reliability
37 Ability to complete the test
Performance of untreated INPH patients
Neuropsychological improvement after shunt
38 38 40 DISCUSSION
40 Pulsatility index and ICP
41 Pulse pressure amplitudes of the ICP
43 CoGNIT development
44 Test evaluation
44 Completion rates
45 Validity
46 Reliability
Baseline impairment and cognitive change
Limitations
47 48 49 Standardization in INPH
49 Future directions
51 CONCLUSIONS
52 ACKNOWLEDGEMENTS
54 REFERENCES
iii
ABSTRACT
Idiopathic Normal Pressure Hydrocephalus (INPH) is a condition affecting
gait, cognition and continence. Radiological examination reveals enlarged
ventricles of the brain. The pathophysiology is debated, but the disease is
probably related to abnormal pressures and flows in the Cerebrospinal fluid
(CSF) dynamic system. A shunt that drains CSF from the ventricles to the
abdomen often improves the symptoms. Much research on INPH has been
focused on identifying tests that distinguish INPH from other disease
groups, and predict the outcome after shunt surgery. As part of these quests,
there are attempts to find measurement methods of intracranial parameters
that are valid, reliable, tolerable and safe for patients.
Today's technologies for intracranial pressure (ICP) measurement are
invasive, often requiring a burr-hole in the skull. Recently, a method for noninvasive ICP measurements was suggested: the Pulsatile Index (PI)
calculated from transcranial Doppler data assessed from the middle cerebral
artery. In previous studies, the PI is strongly correlated with ICP. The PI
method has been proposed as an initial screening tool for neurointensive
care patients, aiding decisions of who would need additional invasive
pressure monitoring. However, there is great discrepancy between studies in
the suggested calibration of PI and ICP data. In paper I, the relation
between PI and ICP was explored in INPH patients during controlled ICP
regulation by lumbar infusion. The confidence interval for predicted ICP,
based on measured PI was too large for the method to be of clinical utility.
There is a quest for better predictive tests for shunt success in INPH. Recent
studies show promising results with criteria based on ICP wave amplitudes.
The brain ventricular system, and the fluid surrounding the spinal cord are
in contact. In paper II, it was shown that ICP waves could be measured via
lumbar subarachnoid space, with a slight underestimation. Less invasive
than traditional methods of measuring intracranial pressure, this technique
would be associated with fewer complications.
One of the cardinal symptoms of hydrocephalus is cognitive impairment.
Neuropsychological studies have demonstrated specific cognitive domains
that are impaired and improve after shunt surgery in INPH patients. These
include memory, psychomotor speed, attention, manual dexterity and
executive functions. However, there is currently no standardized test battery,
and different studies use different tests and outcome measures. Therefore it
is often difficult to compare results from different research groups. In
response, in this thesis a fully automated computerized neuropsychological
iv
test battery was developed. Included tests were selected based on a literature
search about evidence for neuropsychological testing in INPH. In papers
III and IV the validity, test-retest reliability, responsiveness to
improvement after shunt surgery and feasibility for testing INPH patients
was demonstrated. It was also demonstrated that a group of patients with
INPH were impaired in all subtests, compared to healthy elderly.
v
ORIGINAL PAPERS
This thesis is based on the following papers, which are referred to by their
Roman numerals in the text. Also, correspondence regarding paper I is
included.
I. Behrens A., Lenfeldt N., Ambarki K., Malm J., Eklund A., & Koskinen L.-O.
(2010). Transcranial Doppler pulsatility index: not an accurate method to
assess intracranial pressure. Neurosurgery, 66(6), 1050–1057*
Ia. Behrens A., Lenfeldt N., Ambarki K., Malm J., Eklund A., &
Koskinen L.-O. (2010). In Reply. Neurosurgery, 67(6)*
Ib. Behrens A., Lenfeldt N., Ambarki K., Malm J., Eklund A., &
Koskinen L.-O. (2011). Intracranial pressure and pulsatility index.
Neurosurgery, 69(4)*
II. Behrens A., Lenfeldt N., Qvarlander S., Koskinen L.-O., Malm J., &
Eklund A. (2013). Are intracranial pressure wave amplitudes measurable
through lumbar puncture? Acta Neurologica Scandinavica, 127(4), 233–
241*
III. Behrens A., Eklund A., Elgh E., Smith C., Williams M. A., & Malm J.
(2014). A computerized neuropsychological test battery designed for
idiopathic normal pressure hydrocephalus. Fluids and Barriers of the CNS,
11(1), 22*
IV. Behrens A., Elgh E., Leijon G., Kristensen B., Eklund A., Malm J. The
Computerized General Neuropsychological INPH test (CoGNIT) revealed
improvement in Idiopathic Normal Pressure Hydrocephalus (INPH) after
shunt surgery. In Manuscript.
*Reprinted with permission
vi
vii
INTRODUCTION
This thesis involves measurements of intracranial pressure (ICP), pressure
pulsatility and cognitive functions in patients with idiopathic normal
pressure hydrocephalus (INPH). The introduction will give the relevant
physiological and technical background for understanding the disease and
the methods used in the studies.
Non-invasive intracranial pressure
Measurement of ICP is of vital importance in neurointensive care (1).
Secondary brain injury due to elevated ICP is an important cause of
morbidity in patients with acute neurological and neurosurgical conditions
such as: Traumatic Brain Injury (TBI), intracerebral hemorrhage,
subarachnoid hemorrhage, hydrocephalus, malignant infarction, cerebral
oedema and infections of the central nervous system (2,3). Direct
measurement of ICP requires neurosurgical procedures, implanting a probe
into the skull. External ventricular drainage, where a catheter is placed into
one of the ventricles via a burr hole is regarded as the gold standard (2).
However, the technique is associated with complications such as
hemorrhages and infections (4,5).
Non-invasive measurement of ICP would have several advantages. Obviously
the risk of complications caused by the neurosurgical procedure would
disappear. Also, measurements could be performed in locations without
access to neurosurgeons, such as smaller hospitals, ambulances and even in
space (6). Several methods for non-invasive ICP assessment have been
proposed. Here follows a brief and non-exhausting discussion of methods
representing the principal ways of how non-invasive ICP can be assessed.
The transcranial Doppler sonography method is covered in more depth.
Tympanic Membrane Displacement relies on the communication between
the CSF and the perilymph via the cochlear aqueduct. The method utilizes
tympanometry to measure the effect of the stapedial reflex. As stapes rests
on the oval window, its resulting movement is dependent on the cochlear
fluid pressure, and therefore also the CSF pressure. The resulting tympanic
movement correlates with ICP. However, the predictive limits are
approximately ± 25 mm Hg making this method to imprecise for clinical use
(7).
1
Optic nerve sheath diameter measurements rely on the fact that the
subarachnoid space extends around the optic nerve, meaning that changes in
ICP are reflected in the diameter of the nerve sheath. Thus, ICP can be
assessed via transocular ultrasound. The sensitivity and specificity to detect
ICP>20 mm Hg is within the range of 87–95% and 79–100% respectively
(8). However, the nerve sheath diameter may also be affected by a number of
medical conditions potentially making the method less useful (2).
Magnetic Resonance Imaging (MRI) utilizes motion-sensitive MRI to
measure the blood volumes entering and leaving the cranium during one
cardiac cycle and the resulting CSF-flow. From these values an elastance
index can be derived, that have shown high correlation with ICP in one study
(9).
Transcranial Doppler ultrasonography
Transcranial Doppler ultrasonography (TCD) was first introduced in 1982,
as a way to assess blood flow velocity in the basal cerebral arteries stemming
from the circle of Willis (10). An ultrasonic wave of known frequency is
directed at the blood flow and the reflected signal is measured. The
difference in frequency, the so-called Doppler frequency shift, is assessed
and used to calculate the velocity. The most commonly assessed vessel is the
middle cerebral artery (MCA), because of ease of access and good quality of
the signal. The MCA carries 50-60% of ipsilateral internal carotid blood flow
and is thus the largest artery emanating from the circle of Willis (11). It is
insonated through the temporal bone window, found in an area spanned by a
thought line from tragus to the lateral canthus of the eye, and two cm above
this line.
The flow velocity is pulsatile in synchrony with the heartbeat, and from the
TCD signal, systolic, diastolic and time-averaged velocities can be calculated.
Normal values of mean velocity in healthy adults range between 46-86 cm/s
(10). An increase in flow velocity would indicate decreased cross sectional
area of the MCA, as in vasospasm after subarachnoid hemorrhage (12).
Pulsatility index (PI) is an index calculated from ultrasonic velocity
waveforms. Gosling first introduced the index in 1974, as an indicator of
vascular resistance. He applied PI differences to diagnose peripheral
vascular disease in the popliteal arteries (13). The index is calculated as:
(𝑉!"! − 𝑉!"# ) 2
𝑉!"#$
The terms Vsys, Vdia and Vmean are the maximum, minimum and mean flow
velocities during the cardiac cycle. The main advantage of the PI is that the
nominator and denominator of the ratio are equally affected by the angle of
insonation, and the index is independent of this parameter. This makes for a
stable index, unaffected by small inclination deviances of the ultrasonic
probe.
There have been a number of studies relating PI to ICP, giving hopes for a
good and accessible non-invasive ICP monitoring technique. Results are
reviewed in table 1.
Table 1. Studies on Pulsatility Index (PI) and ICP published before paper I.
Study
Clinical
Number of
Calibration
R2
material
patients
Homburg et al. (14)
TBI
10
ICP = 42* ln PI + 14
0.67
Voulgaris et al. (15)
TBI
37
ICP = 16* PI + 5
0.41
Moreno et al. (16)
TBI
125
ICP = 33* PI - 16
0.69
Bellner et al. (17)
TBI and SH
81
ICP = 11* PI - 1
0.88
TBI = Traumatic Brain Injury, SH = Subarachnoid Hemorrhage, ICP = Intracranial
Pressure, PI = Pulsatility Index.
The most positive results are from the study by Bellner et al (17). They
reported a sensitivity and specificity of 0.89 and 0.92 to detect ICP higher
than 20 mm Hg. The conclusion was that "PI may be of guiding value in the
invasive ICP placement decision in the neurointensive care patient".
There certainly seems to be a strong correlation between PI and ICP.
However, there are large differences between regression lines of the PI and
ICP relationship between studies, giving largely different estimations of ICP.
Before clinical acceptance of the method, the reliability of ICP estimations
needs to be examined. Acceptance of an unreliable method, especially when
used in critically ill patients, may lead to harmful clinical decisions.
It is therefore important to thoroughly evaluate this method. This was the
rationale for paper I. Using lumbar infusion of CSF, ICP could be
controlled, which made it possible to obtain measurement points over a wide
3
range of ICPs in each patient. Thus, making it possible to study the PI/ICP
relation on an individual level.
ICP pulsations
The intracranial cavity mainly consists of three largely incompressible
substances: brain parenchyma, blood and CSF. As the volume of the cavity is
close to constant, a volume change of any of these three substances must be
compensated by a similar negative volume change of the others. This relation
has been termed the Monro-Kellie doctrine (18). Marmarou et al. have
modelled the intracranial response to a volume change, with later
modification by Avezaat and Eijndhoven (19,20):
𝐼𝐶𝑃 = 𝑃! 𝑒 !" + 𝑃!
The volume parameter (V) is the deviation from the equilibrium volume. The
term k describes the elastic properties of the system and is called elastance
and the P0 and P1 are pressure constants. The equation describes the socalled pressure volume curve, displayed in Figure 1. It is evident that there is
an exponential increase in ICP at higher added volumes.
90
∆V
80
ICP [mm Hg]
70
∆P
60
50
40
∆V
30
∆P
20
10
0
2
4
6
8
10
12
14
Units of volume
Figure 1. The intracranial pressure-volume curve. The same amount of volume
entering the scull will produce different pressure amplitudes, depending on the
position along the pressure-volume curve.
In the systole of each heartbeat a bolus of blood is leaving the heart, and
transmitted through the arterial vascular tree. As described by the
4
Marmarou equation, the systolic arterial expansion will, when entering the
scull lead to an increased ICP. The amplitude of the resulting ICP pulse is
dependent on the volume of blood entering the cranium (arterial expansion),
and the subject’s position along the pressure-volume curve. In Figure 1 it is
evident that at higher mean ICP, the same bolus of blood will result in higher
amplitude of the resulting ICP pulse. The transient increase in intracranial
pressure (∆ICP), resulting from the arterial expansion (∆V) related to the
bolus of blood entering the cranium with each heartbeat, can be formulated
in the Marmarou equation as:
𝐼𝐶𝑃 + ∆𝐼𝐶𝑃 = 𝑃! 𝑒 !(!!∆!) + 𝑃!
And after reformulation:
∆𝐼𝐶𝑃 = (𝑒 !∆! − 1)(𝐼𝐶𝑃 − 𝑃! )
Assuming that ∆V is constant, there is a linear increase of ∆ICP with
increasing ICP with the slope (𝑒 !∆! − 1). This relation has been validated in
both humans and animals (20,21). The linear relation holds for baseline and
moderately elevated mean ICP. At lower pressures (typically less than 10 mm
Hg) ∆ICP is constant (21). The relationship of ∆ICP and ICP has been
termed the "pulsatility curve". And accordingly the pulsatility curve has an
ICP independent part at lower ICP and a linear part where ∆ICP increase
linearly with ICP. The slope, and the ICP below which ∆ICP is constant vary
between humans (21).
Direct measurement of ICP and ICP pulsatility requires the neurosurgical
placement of a sensor inside the scull. The two methods most commonly
used in clinical practice are intra-ventricular catheters and intraparenchymal
catheter-tip systems (1). After entering the scull, the intracranial pulsations
are transferred along the CSF pathway to the subarachnoid space and down
through the spinal canal (22,23). This opens up for a potential measurement
route via lumbar space, which would be less invasive, and potentially result
in fewer complications for patients. Also, assessments would be accessible
without a neurosurgical procedure.
Measurements of mean intracranial pressure via lumbar puncture are often
performed by neurologists (24). The access of intracranial pulse pressure
data via the lumbar route would give the neurologist an additional tool to
investigate CSF disorders. This may have special importance in the field of
INPH, where abnormal pulsations have been suggested to predict shunt
outcome (25). However, this needs to be verified in a randomized clinical
trial. Lumbar CSF pressure amplitude measurements may be a useful clinical
5
tool, but it is important to determine the reliability of the method. In
response to this paper II was conducted.
Idiopathic Normal Pressure Hydrocephalus
Hydrocephalus overview
The first scientific description of hydrocephalus has been ascribed to
Hippocrates (466–377 BC) (26). He also named the disease. The word is
constructed from the Greek words hydor = water, and kefalé = head (26).
There are several proposed classifications of hydrocephalus. The condition
can be classified by the pathological condition leading to CSF accumulation
in the head, either by obstruction of the normal CSF flow pattern, or by
perturbations of CSF production/resorption. Another common classification
is that of communicating/non-communicating hydrocephalus. In the
beginning of the 20th century Dandy and Blackfan injected dye into the
ventricles. If dye was subsequently obtained from a lumbar puncture, the
hydrocephalus was termed "communicating". If no dye was recovered it was
termed obstructive or "non-communicating" hydrocephalus (27). Noncommunicating hydrocephalus can be the result of compression of an
anatomical narrowing (e.g. a tumor), of the ventricular system down to the
point where CSF drains into the subarachnoid space. In communicating
hydrocephalus there is no intraventricular obstruction. Communicating
hydrocephalus can be caused by meningitis, subarachnoid hemorrhage,
traumatic brain injury and brain radiation causing collection of debris or
inflammation in the arachnoid granulations (28,29). To mark that there is a
known cause for the disease, the term "secondary hydrocephalus" is used.
Overproduction of CSF by choroid plexus (hyperplasia/papilloma) leading to
hydrocephalus is very rare (28).
To emphasize that most all hydrocephalus, even "communicating" has an
obstructive component, Rekate in 2008 proposed a definition of
hydrocephalus as "Active distension of the ventricular system of the brain
resulting from inadequate passage of CSF from the point of production
within the cerebral ventricles to the point of absorption into the systemic
circulation." (27).
6
Idiopathic Normal Pressure Hydrocephalus
In 1965, Hakim and Adams described a syndrome of cognitive decline,
impaired gait and urinary incontinence in adult patients with
communicating hydrocephalus and CSF mean pressure in the normal range
(30). The syndrome was later called normal pressure hydrocephalus. To
separate this syndrome from hydrocephalus of known cause (secondary
hydrocephalus), the term "idiopathic normal pressure hydrocephalus"
(INPH) is frequently used.
Epidemiology
INPH has a mean age of onset of about 70 years and affects men and women
equally (31,32). Since the symptoms of INPH are common in other diseases
affecting the elderly, the real prevalence is hard to assess and the condition is
likely under diagnosed (32,33). In Norway, Brean et al. conducted a
prospective study that sought INPH patients by advertising and asking
health workers to refer patients. The prevalence was estimated to
22/100000 and the incidence to 5.5/100 000/year (34). This was not a true
population-based study and many INPH patients might not have been
referred. Thus, this study likely gives an underestimation of INPH
prevalence. Jaraj et al. looked at a representative Swedish population older
than 70 years, who had undergone CT-scans and clinical evaluation (32). The
estimated prevalence was 0.2 % in the age 70-79 and 5.9% in persons older
than 80 years. Israelsson et al. looked at patients who had suffered from a
transient ischemic attack and found that 3.9% fulfilled the criteria for INPH
(33). These numbers could be compared to the reported incidence of shunt
surgery for INPH in Sweden (1996-1998) of 1/100000/year (35). Figure 2
shows the number of implanted shunts per 100 000 inhabitants in adults
reported from five out of six neurosurgical centres reporting to the Swedish
hydrocephalus register. Approximately 60 % patients in the register have
INPH (2011), which gives an incidence for shunt surgery of 2.7/100
000/year. It seems that many INPH patients do not receive treatment, even
when conservative estimates of INPH incidence are used.
7
Figure 2. Number of shunt operations in adults per 100 000 inhabitants in five
Swedish neurosurgical centres. The numbers are kindly provided by Nina Sundström
of the Swedish national Hydrocephalus register.
Pathophysiology of INPH
The pathophysiology of INPH remains unknown. It is however widely agreed
that symptoms are the result from a neuronal dysfunction from subcortical
structures, for instance caused by a chronic ischemia (31,36). In addition,
this is also accompanied by disturbances of the CSF system (31). Here
follows a non-exhausting discussion of prevalent pathological findings
encountered in the INPH literature.
The resistance to CSF outflow (Rout) is increased in INPH (37), and have
been proposed as the driving force causing ventricle dilation (38). A Rout > 18
mm Hg/ml/min has been suggested as a limit for shunt surgery (39).
Although a feature of the disease, it is not clear why Rout is increased, neither
how such an increase would cause ventricular dilation (40,41).
Several animal models have highlighted the role of cardiac related
intracranial pulsatility as a possible force for ventricular expansion (40).
Flow-sensitive MRI have shown increased arterial flow pulsations in INPH
8
(42), which has lead to theories of a "water hammer" effect that causes the
ventricular dilation (41,43). It has been proposed that age related arterial
stiffness and less compliance of the brain leads to larger pressure pulsations
in the CSF (44). This is supported by findings that increased blood pressure
and blood pressure pulsations is associated with increasing ventricular size
(45). Elevated blood pressure is also the largest risk factor for INPH (46).
Increased intracranial pulse pressure amplitudes are seen in INPH patients
(47), and probing for these is recently proposed as test for selection of shunt
candidates (48). Also, the patients improving after shunt surgery have a
larger reduction of intracranial amplitudes (49).
Diagnosing INPH
Since 2005, formal guidelines exist for diagnosing INPH (50). To reflect
variability in diagnostic certainty, the guidelines classify INPH as probable,
possible and unlikely. The diagnosis of INPH should be based on history,
neurological examination, physiological findings and a neuroradiological
investigation. The patient’s response to shunt operation is not considered in
the diagnostics, and results of additional “predictive” tests should not be
used for diagnostic purposes (51). The clinical presentation of INPH varies
greatly from patient to patient, and the symptoms mimic diseases including
Parkinson’s, Alzheimer’s, Lewy Body disease, spinal stenosis, and vascular
dementia (52,53). In addition, many INPH patients have comorbidities
(53,54). Major comorbidities affecting cognition are Alzheimer’s and
cerebrovascular disease (54,55). Comorbidities may affect prognosis and
outcome after shunt surgery and should be actively sought (53).
Treatment of INPH
Many INPH patients improve after insertion of a shunt. The shunt drains
CSF, normally from the lateral ventricles, to a low-pressure compartment of
the body, normally the abdomen (31). A CSF shunt will decrease Rout and has
the potential to decrease the ICP as well as the ICP pulse pressure
amplitudes (49,56).
Only one randomized controlled trial including a randomized group
receiving ligated shunts has been published (57). Patients in this study
fulfilled diagnostic criteria of both Binswangers disease and INPH. This
study showed neuropsychological and gait improvement in patients with
9
open shunts (n=7), but not in patients with ligated shunts (n=7). Further, the
European multicentre study on INPH showed that as many as 84 % of
patients (n=142) selected on clinical and radiological criteria alone will
benefit from shunt operation (58). Other studies with similar inclusion but
other outcome measures have reported lower levels of improvement. Malm
et al. found that 72 % of patients improved in gait, and Vanneste et al. found
that only 29% of patients with communicating hydrocephalus improved after
shunt surgery (59,60). Though improved, patients will not regain function in
parity with healthy elderly and the extent of potential recovery diminishes
with disease duration (61,62). Complication rates in the European study was
28%, with 15% needing additional surgery (58). Complications to surgery
include shunt malfunction (20%), subdural hematomas (2-17%),
intracerebral hematomas (3%), shunt infections (6%) and seizures (3-11%)
(63).
Supplemental tests
To predict which INPH patients would benefit from a shunt operation,
several so-called supplemental tests can be used (51). Most common are tests
of CSF drainage, where the effect of a shunt operation is simulated. Hakim
and Adams first reported clinical improvement after a 15 ml spinal tap, and
later Wikkelsö et al. reported that 40-50 ml CSF removal was useful for
predicting shunt responsiveness (30,64). The specificity of the tap test has
been estimated in the range of 33-100% and the sensitivity in the range of
26-100%, largely depending on outcome measures used (51,59,65-68). The
standard interpretation is that the tap test has low sensitivity and that
patients should not be excluded from shunt surgery based on a negative tap
test (51). Sensitivity can be increased by extended lumbar drainage (51), a
procedure where typically ten ml CSF per hour is drained over 72 hours.
Positive prediction values range between 80-100% (51,69,70). Interestingly,
improvement after the test does not seem to be related to change in
ventricular volume (71).
The infusion test is a test that give information about a patients CSF system
(37). In Umeå infusion tests are performed by an in-house development that
has been commercialized (Likvor CELDA®, Likvor AB, Umeå, Sweden). The
equipment works by infusing/draining CSF in the lumbar cistern, and
simultaneously monitoring the CSF pressure via a second needle in lumbar
space. The amount of infused CSF and the pressure response is monitored.
From these numbers physiological parameters are calculated (e.g. Rout). The
infusion pattern used in the scope of this thesis is the constant pressure
10
infusion. The CSF pressure is then regulated and maintained on different
constant levels (see figure 3). Average flow to maintain CSF pressure at each
level is measured. At entry-level, CSF pressure and associated CSF pressure
pulsation amplitudes are measured, without any infusion.
Figure 3. Graph describing a constant pressure infusion (CPI) protocol where the
black lines show the pressure regulation levels. The arrow indicate where the patient
is in seated position for CSF sampling. With kind permission from Sara Qvarlander.
Intracranial pulse pressure amplitudes have been suggested to be used for
selecting INPH patients for shunt surgery. The technique developed by Eide
requires six to 12 hours of continuous ICP measurement. The percentage of
time that the amplitude is above certain thresholds is then calculated.
Elevated pulse pressure is defined as amplitudes either ≥ 4 mm Hg on
average and ≥ 5 mm Hg during more than 10 % of the ICP recording time
(72-74). Cognitive testing in INPH The prevalence of dementia is increasing in the world (75). Proper
diagnostics is important and even more so, identifying treatable or reversible
causes. Cognitive impairment is one of the cardinal features of INPH and its
presence is mandatory for the diagnosis of "probable INPH" according to the
INPH guidelines (50). The cognitive disabilities in INPH vary and patients
are generally moderately cognitively impaired compared to patients with
other dementias (76). The mini mental state examination (MMSE) is a
commonly used screening test for global cognitive function (77). The cut off
score for dementia is < 24 MMSE points (78). Median MMSE-score for
INPH patients at diagnose is 24 out of 30 points and thus 50 % of the
patients score above this limit (31,79), indicating the presence of a ceiling
11
effect when using the MMSE in INPH. This is to some part explained by the
MMSE-tests low sensitivity in dementias of subcortical type and makes the
test less usable in INPH (76,80). More extensive neuropsychological
evaluations show impairments in a majority of INPH patients (81,82).
Neuropsychological testing in INPH
The cognitive profile in INPH is characterized as subcortical impairment,
with no cortical signs such as apraxia, agnosia aphasia or other focal cortical
defects (76). Neuropsychological impairments resemble other subcortical
dementias (83-85). Accordingly, neuropsychological studies in INPH have
shown impairments in memory (81,86), attention (87,88), psychomotor
speed (81,86,89), executive function (81,90,91), and visuospatial skills (92).
Impairments in manual dexterity are also seen (81,93,94). Descriptions of
mental changes include apathy, emotional indifference and bradyphrenia
(95,96). The apathy and bradyphrenia may closely simulate depression (76).
Examples of functional impairments are problems with keeping track of
appointments, driving, managing finances and taking medications properly
(29).
There are a number of neuropsychological studies describing
neuropsychological improvement after shunt surgery. The main larger
studies will be reviewed in the following.
Malm et al. included 35 INPH patients based on conservative clinical criteria
(59). Neuropsychological tests were administered before and three months
after shunt surgery. Tests administered were MMSE, Wechsler Adult
Intelligence Scale-Revisited (intelligence), the Fuld Object Memory (FOM)
Test (consistent retrieval and retrieval failures; memory), the Boston
Naming Test (language) and the Figure copy test (visuoconstructive skills)
from the Alzheimer's Disease Assessment Scale. Significant improvement
after shunt surgery was seen in the FOM retrieval failure and Figure copy
tests.
Thomas et al. administered neuropsychological tests and MMSE to 42 INPH
patients before and at least three months after shunt surgery (86).
Neuropsychological tests were the Wechsler Memory Scale (WMS) (verbal
memory), Rey Osterrieth Complex Figure (RCFT) (visuoconstructive skills
and visual memory), Rey Auditory Verbal Learning Test (RAVLT) (memory
and learning), Line Tracing test, (psychomotor speed), Trail making test
(TMT) B (executive functions), Stroop Color-Word (executive function) and
12
the MMSE. Different forms of the memory tests were administered at
baseline and follow-up to minimize practice effects. Significant improvement
in a subtest was defined as one standard deviation (SD) improvement for the
patient's age, sex and educational level. Overall improvement was defined as
improvement in more than 50 % of administered tests or a four-point
improvement on the MMSE. Overall improvement was seen in 22/42 (52 %)
of the patients. Significant improvement was seen in tests of memory and
psychomotor speed.
The same definition of overall improvement has been applied in the studies
by Duinkerke et al. and Foss et al. (97,98). In the study by Duinkereke 6/10
(60 %) patients improved after shunt surgery. In the study by Foss, the
Dementia Rating Scale was administered. Improvement after shunt surgery
was seen in 12/27 (44 %) patients. Interestingly patients who improved had
significantly higher mean ICP pulse amplitudes in the preoperative
evaluation.
Hellström et al. administered the MMSE and neuropsychological tests to 58
INPH patients and 108 healthy individuals (81). The administered tests were
Target reaction time (wakefulness/attention), Grooved pegboard test
(manual dexterity), tracks (manual dexterity), RAVLT (learning and
memory), Swedish Stroop test (executive functions) and Digit span (working
memory). INPH patients performed significantly worse than healthy in all
administered tasks. Also, patients with vascular risk factors performed worse
than those without. There were also far more correlations between test
scores for patients, than for healthy controls.
In the follow up study by Hellström, the same tests, with the addition of a
simple reaction time test, were administered to 47 patients before and three
months after shunt surgery (62). This time results were compared to 159
healthy individuals. Significant improvement was seen in all tests but in the
simple reaction time and in Digit span forward test. Though improvement,
patients did not reach the performance of the healthy controls in any test.
Adopting the same definition as above, overall improvement was seen in
24/47 (51 %) (99). Instead, defining improvement as improved in ≥ four of
nine administered tests would yield an improvement rate of 82 % (numbers
only presented in dissertation) (99). It was also noted that correlations
between tests weakened or disappeared after surgery.
Solana et al performed the largest study on the topic to date. They included
185 INPH patients and administered MMSE and neuropsychological tests
before and six months after shunting. Included tests were WMS-revised
(information, orientation, visual reproduction and memory), RAVLT
13
(memory), Digit span forward and backward (attention, executive functions),
Phonetic Verbal Fluency Test (executive functions), TMT A (psychomotor
speed), TMT B (executive functions), Purdue Pegboard Test (psychomotor
speed). Significant improvement after surgery was seen in all tests except for
Digits span forward and TMT B. Significant individual improvement was
defined as 1 SD of improvement from the baseline score, adjusted for age,
sex and educational level, or by an increase of at least 20 percentile points
from the baseline score. Number of improved patients in all tests was around
or slightly less than 50 %, except in the TMT B (22 %), Digits span forward
(22 %) and Digits span backward tests (33 %). Improvement by ≥ 4 MMSE
points was seen in 37/144 (26 %) patients. Most impaired cognitive domains
at baseline were seen in tests of memory, executive functions, attention and
psychomotor speed. Most marked improvement was seen in tests of memory
and psychomotor speed.
Hellström et al. followed up their initial neuropsychological study, with the
assessment of a smaller subset of tests (82). The selected tests were the
Grooved pegboard test, the Stroop test and RAVLT including delayed recall.
The tests were administered to 142 INPH patients at baseline but also 3
months and 1 year after shunt surgery. Scores were compared to 108 healthy
controls. Sixty-eight % of patients for the Stroop test and 94 % for the
RAVLT could complete the test as baseline. Discriminative ability between
INPH and healthy at baseline was demonstrated, with AUC values ranging
from 0.86 to 0.95 for included tests. Significant improvement after shunt
was seen in all tests. Number of patients who improved more than 1 SD of
the distribution of the controls ranged between 14 % in the RAVLT learning
test to 62 % in the Grooved pegboard test. The conclusion was that this small
set of tests was expedient, diagnostic discriminative and well suited to
evaluate changes following shunt treatment (82). Accordingly, these tests,
excluding the delayed recall test of RAVLT, were included in a new scale
measuring severity and outcome in INPH (100).
One important aspect of neuropsychological testing is that of practice effects.
Repeated assessments of the same test typically improve results (101). This
would have implications for testing in INPH where repeated testing is
common. The typical retest intervals would be 24 hours (tap test), 72 hours
(extended lumbar drainage) and 3-6 months (follow up after surgery).
Katzen et al. approached the problem of repeated testing by comparing retest
results from a healthy control group. In this small study of 12 INPH patients
and nine controls, improvement was seen in the TMT B and Symbol digit
modalities (102). However, practice effects are affected by population under
study (101). Solana et al. specifically studied performance of repeated testing
in healthy and INPH patients, and found that practice effects seen in healthy
14
were not present in INPH patients (103). Also, interestingly Andrén et al.
found that a cognitive index from the INPH scale referenced above
deteriorated in 33 untreated INPH patients, from 40 points at baseline to 28
points 13.2 (median) months later (61). Implicating that the cognitive decline
in INPH is greater than practice effects, at least over this interval. The matter
how to evaluate change scores in INPH is to current day unresolved.
Adopting the method by Katzen, comparing change scores to those of healthy
controls may underestimate effects of treatment as practice effects are larger
among healthy. Comparing improvement to baseline scores may in turn
overestimate effects, as practice effects may be present. Moreover,
complicating the issue further practice effects may depend on baseline
performance level, retest interval or cognitive domain (101).
Computerized neuropsychological testing
Since the wider adoption of computers there exists computer versions of
classical neuropsychological tests (104). A recent review (2014) identified 17
test batteries for testing in the elderly (105). However, paper and pencil tests
are still the main cognitive assessment tool (106), which requires a
neuropsychologist for selection, administration and interpretation of tests.
Computerized batteries may aid in test selection, with pre-selected tests for
different diseases, as well as pre-recorded instructions may relieve the
neuropsychologist from test administration, making way for more efficient
use of resources. Additional advantages for computerized tests may be
standardization of instructions, scoring with a high level of precision, cost
savings and less influence from examiner bias (105,107). The main criticism
has been lacking reports of reliability, equivalence for the response format particularly for elderly with less computer knowledge, and finally poorly
designed computer-person interfaces. Further, redesign of tests may alter
validity (107). Computer testing may pose a challenge in the elderly.
However, of the different response formats, the touch screen interface has
been proposed as the best choice for testing elderly people (108).
There are certainly pros and cons with computerized cognitive testing
(105,107), but the idea of having a self-administered cognitive battery holds
enough merit to pursue. Especially in INPH where the most often used
cognitive screening instrument lacks sensitivity (76,77). Also, standardized
self-administered test may facilitate research comparability and
collaboration. In the response to this, in addition to the other potential
advantages discussed above, a computerized neuropsychological test battery
customized for INPH was designed and evaluated in papers III and IV.
15
Test development
Before deploying a new test battery several factors must be examined. It
must be assured that tests are valid, reliable and that the intended patient
group can complete the tests.
Validity can be defined as "the degree to which an instrument truly
measures the construct it purports to measure" (109). There are a plethora of
concepts relating to different aspects of validity: Face validity, the degree to
which a construct looks as an adequate reflection of the construct to be
measured; Criterion validity, do the scores reflect a gold standard?
Ecological validity, the degree to which scores in the test says something
about the behaviour in a more natural environment (110). Another aspect of
validity is responsiveness; how well an instrument reflect changes in the
underlying construct (109). It has been argued, because of the different
response format, that discriminant validity, the degree to which different
tests do not co-vary, is important in computerized testing (105).
The reliability of an instrument is defined as "the degree to which a
measurement is free from measurement error" (109). If the measurement
error is too large, statements about impairments or improvements of
cognitive functions become unreliable. The concept is also related to validity
in that covariance of scores from different measurement methods is limited
by the degree of error in the measurements.
Neuropsychological assessment provides non-invasive means of assisting in
diagnostics, tracking disease progression and assessing efficacy of treatment
in INPH. A self-administered neuropsychological test battery with
preselected tests may increase standardization and accessibility for cognitive
testing. A prerequisite is that implemented tests are reliable, valid, and
responsive. Also, to be useful they should show good completion rates.
Research rationale
There are an increasing number of studies about many aspects of INPH, see
figure 4. Each study represents a tremendous amount of effort. However,
what can be concluded from these studies is often limited by small study
16
samples. Pooling over several studies or comparing results is difficult, as
different measurements methods are often used.
Measurements often provide the basis for diagnosis and medical
interventions. There may be limitations due to low quality of measurements
and the use of measurements that are insufficiently characterized. It is
therefore important that measurements are thoroughly evaluated and their
limitations reported and understood.
In response to this, the studies in this thesis contribute by exploring
limitations in ICP measurements. Specifically the PI method of estimating
ICP and the pulse pressure amplitude measurement via the lumbar route are
explored. Also, aiding standardization of cognitive measurements, a selfadministrating computerized neuropsychological test battery is developed.
The hypothesis is that a standardized computerized test battery would be
useful in the cognitive assessments of INPH-patients.
Figure 4. Pubmed search of "Idiopathic Normal Pressure Hydrocephalus". The
graph shows number of publications per year.
17
AIMS
I.
To validate the method of estimating intracranial pressure via
transcranial Doppler sonography pulsatility index of middle cerebral
artery blood flow waves.
II.
To investigate whether cardiac related intracranial pulse pressure
amplitudes in the brain correspond to those measured in lumbar
space.
III.
To develop and evaluate validity and reliability of a computerized
neuropsychological test battery customized for INPH.
IV.
To assess the completion rate for computerized tests in INPH
patients and to test the ability to detect cognitive impairment at
baseline and improvement after a shunt surgery in INPH patients.
18
MATERIALS AND METHODS
Ethical approval
The Regional Ethical Review Board in Umeå approved all the studies
presented in this thesis.
Patients and controls
Table 2. Patient cohorts used in this thesis.
Paper
I
II
III
IV
INPH n=10
X
X
Healthy elderly n=44
X
X
Cognitively impaired n=28
X
Hydrocephalus n=40
X
INPHa n=26
X
INPH shunt surgery n=31
X
aOf
the 40 patients in the Hydrocephalus group in paper III, 26 were diagnosed with
INPH.
INPH patients of paper I and II
Ten INPH patients (two women) were included, mean age 72.4 years (range
55-78). MRI revealed communicating hydrocephalus, with dilated ventricles
(all had Evens Index >0.3), without cortical atrophy or severe white matter
lesions. Diagnostics included clinical characteristics, routine laboratory tests,
MMSE (77), and video taped gait tests.
19
Overnight intra-parenchymal pressure monitoring was performed. A lumbar
CSF infusion test and tap test was performed in the morning the day after
implantation of the catheter. While the infusion test was running, TCD of the
MCA and lumbar pressure monitoring was performed. All patients
subsequently received shunts. The studies were part of a larger project with
several previous publications (24,36,111).
Cognitively impaired patients of paper III
A research nurse screened the neurological ward for patients. Inclusion
criterion was a MMSE between 20 and 30. The only exclusion criterion was
impaired motor function (e.g. palsy). Thirty patients agreed to participate
and two patients were excluded because of inability to perform either
conventional or computerized testing, resulting in twenty-eight included
patients. See table 3 for demographics.
Hydrocephalus patients of paper III
Forty consecutive patients referred to our department because of suspected
INPH were included. All patients had MRI-verified communicating
hydrocephalus, see table 3 for demographics. A review of the charts revealed
that 26 patients fulfilled the criteria for INPH according to American and
European guidelines (50).
Healthy elderly of papers III and IV
Forty-five healthy elderly were recruited via an ad in the local newspaper in
Umeå. Inclusion criteria were an age between 60 and 82 years old. The
participants were confirmed healthy regarding medical history and clinical
examination including on-going medication, physical and neurological
examinations, electrocardiography, blood pressure, body mass index, MMSE
and MRI. Exclusion criteria were: MMSE<28, disease of the nervous system,
medication affecting the nervous system (such as benzodiazepine or
antidepressants), anticoagulants, ischemic heart disease and diabetes. More
than one of: hypertension, smoking or hyperlipidemia, was also an
indication for exclusion. One participant was excluded due to technical
20
problems with the touch screen, resulting in 44 included subjects.
Demographics are displayed in table 3.
Table 3. Characteristics of the study populations of paper III
Healthy
elderly
Cognitively
imaired
patients
Hydrocephalus
patients
Age, y Median (range)
69 (60-79)
71 (56-86)
72 (50-85)
Numbers (n)
44a
28
40
Male/Female, %
41/59
50/50
63/37
Education y, Median (range) 11.5 (6-22)
10 (6-15.5)
8 (6-20)
Computer knowledgeb % Yes 60
50
53
Color blind %
0
11
13
MMSE, Median (range)
>28
26 (20-30)
26 (18-30)
GDS, Median (range)
0 (0-6)
3 (0-10)
4 (0-19)
MMSE = Mini Mental State Exam, GDS = Geriatric Depression Scale, aTwo tests
(Four-finger tapping and Ten word list learning) were redesigned during the study
and only 26 of the 44 participants of the healthy elderly took the slightly modified
battery. bThe subjects were asked, "Do you have computer knowledge, yes or no".
INPH patients of paper IV
Thirty-two INPH patients were recruited from three centres (Umeå,
Linköping and Aalborg). Patients were included if older than 60 years,
having INPH according to international guidelines (50), and were planned
for surgery. Exclusion criteria were secondary NPH or medical condition
preventing neuropsychological testing (e.g. blindness, deafness).
Demographics of included patients are displayed in table 4. One patient was
excluded due to non-functioning shunt, resulting in 31 included patients.
21
There was an overlap in recruitment, and four patients were included in both
papers III and IV.
Table 4. Characteristics of INPH patients in paper IV.
INPH
n=31
Age, y Median (range)
74 (64-85)
Male/Female, %
84/16
Education y, Median (range)
8 (6-17)
Computer knowledge % Yes
35
Color blind %
6
MMSE, Median (range)
26 (19-30)
GDS, Median (range)
4 (0-13)
MMSE = Mini Mental State Exam, GDS = Geriatric Depression Scale.
Simultaneous infusion test, transcranial Doppler
and pressure measurements
In papers I and II, the patients performed simultaneous measurements of
MCA blood flow velocity, blood pressure, ICP and lumbar CSF pressure,
while ICP was regulated by lumbar infusion/withdrawal of CSF. An
TM
intraparenchymal catheter tip sensor (Codman MicroSensor
Johnson &
Johnson Professional, Raynham, MA, USA) was inserted into brain
parenchyma close to the frontal horn of the right ventricle, and an overnight
registration of ICP was performed. On day two, with the intracranial
pressure sensor still in place, two needles were inserted in the lumbar
subarachnoid space. One needle was used for pressure monitoring via a fluid
catheter system (Becton Dickinson, Franklin Lakes, NJ, USA) and the other
for CSF volume alternation. The CSF pressure was measured at the midpoint
22
between the highest and lowest points of the head, with the patient in the
supine position.
For CSF volume alternation, an in house developed infusion apparatus was
used. A constant pressure method was used, meaning that the infusion was
regulated to keep mean ICP stable at pre-set levels. First a baseline recording
was performed, and then intracranial pressure was increased in two or
threes steps up to 45 mm Hg, before pressure was released back to baseline
again. The recording was ended with CSF withdrawal to obtain an ICP of
zero mm Hg, see figure 5. On each pressure level a 30-90 seconds TCD
recording from the left middle cerebral artery was performed.
Figure 5.1 Constant pressure infusion in one patient. The pressure was recorded at
baseline. Then pressure was increased in three steps (1st - 3rd excess), and finally
released to baseline again (Relax). The recording ended with CSF withdrawal to zero
ICP (Drain). The magnification shows intracranial and CSF pressure waves.
®
All data was exported and analyzed offline in Matlab (Mathworks Inc.,
Natick, MA, USA). A program for automatic waveform analysis was
developed. The program identified diastolic minima and systolic maxima for
blood pressure, TCD velocity, as well as intracranial and lumbar CSF
pressure waves, see figure 6. After pulse wave identification, the amplitude of
each pulse was calculated as systolic maxima minus previous diastolic
1 Reprinted from paper II with permission: Behrens, A., Lenfeldt, N., Qvarlander, S., Koskinen, L.-O., Malm,
J., & Eklund, A. (2013). Are intracranial pressure wave amplitudes measurable through lumbar puncture?
Acta Neurologica Scandinavica, 127(4), 233–241.
23
minima. The delay between intracranial and CSF pulses was calculated as
time difference between diastolic minima. Also the fast Fourier transform
(FFT) was applied to the data. The amplitude of the fundamental frequency,
and the two successive over-tones was calculated for consecutive 6-second
time windows.
Blood pressure
Pressure [mm Hg]
200
150
100
50
0
1
2
3
4
5
4
5
Intracranial pressure
Pressure [mm Hg]
26
24
22
20
18
16
0
1
2
3
Time [s]
Figure 6. In the figure, waveforms from intracranial pressure and blood pressure
are displayed. The program calculated the fundamental frequency for blood pressure
data and looked for successive local minima within this interval. Corresponding
minima for ICP was sought close these points. Systolic maximum was calculated as
the maximum value between two minima. Asterisks indicate the limits of the pulse
wave found by the program.
Mathematical model
In paper I a mathematical model was invoked (112-­‐114). The model is a
complex physiological model describing interaction of a number of
mechanisms that each are described by a differential equation. The model
incorporates the CSF circulation, intracranial pressure volume relationship,
and cerebral hemodynamics, including auto regulatory mechanisms. The
model has previously been invoked to study TCD waveforms (115).
Input to the model was a blood pressure waveform from one of the patients.
The differential equations were solved with numerical methods in Matlab.
Results were also verified in a Simulink model. In the model, the in-vivo
experiment was reproduced, simulating a constant pressure infusion
investigation, see figure 7. The simulations were done with different values
for physiological parameters and the resulting waveform of the MCA was
24
analyzed with the program described above. Diastolic minima and systolic
maxima were identified and the pulsatility index for each pressure level was
calculated.
Pressure [mm Hg]
40
30
20
10
0
0
100
200
300
400
500
Time [s]
600
700
800
900
1000
100
200
300
400
500
Time [s]
600
700
800
900
1000
Infusion rate [ml/s]
0.1
0.08
0.06
0.04
0.02
0
0
Figure 7. Simulated infusion test with the mathematical model. Above is the
resulting intracranial pressure and below the infusion rate as function of time.
Bench test
In paper II a bench test comparing the Codman catheter tip sensor with the
Becton fluid catheter system was performed. The systems were set up in the
same way as the in-vivo measurements, but connected to a pressure
waveform generator (Model 601A Blood pressure system calibrator, Biotech
Instruments, Inc. Burlington, VT, USA) for simultaneous measurements. A
pressure waveform typical for the patients was modified and fed to the
systems with different amplitudes, heart rates and mean pressures.
25
Neuropsychological test battery
In papers III and IV a computerized neuropsychological test battery was
developed and evaluated. A Pubmed literature search was performed,
finding the most frequently used neuropsychological tests that were also
sensitive to the cognitive profile of INPH patients. The tests were selected for
brevity, feasibility for implementation on a touch screen and to cover the
cognitive domains affected by INPH. Also, tests had to be in the public
domain, or the implemented version had to be sufficiently different not to
infringe any copyright. This was in effect solved by the computer adaptation
itself.
Software design
Tests were implemented using the computer languages JAVA (116) and
Adobe Flash (117). The Flash language was used for making the user
interface, with graphics, buttons and animations with pre-recorded sound
presenting the tests. JAVA was used to implement storage of test results and
sound from the microphone to the hard drive and the generation of a test
report. A pilot test battery was developed and tested on personnel at our
department. After this, bugs were corrected and the tests that did not work
well in practice were redesigned (e.g. Four-finger tapping on the touch
screen was replaced by a numeric keyboard).
Computerized General Neuropsychological INPH Test (CoGNIT)
The CoGNIT battery includes tests of memory, attention, psychomotor
speed, executive functions, and manual dexterity. Here follows a description
of implemented tests. Screen shots from some tests are displayed in figure 8.
1. Two choice reaction test (attention). The subject is instructed to
keep the stylus over a cross in the middle of the screen, and then press one of
two buttons indicated by an arrow (see figure 8A). The arrow appears after a
random interval of 5 to 15 seconds. Median reaction time over 20 trials is
used as the test score.
26
2. Stroop congruent colors (psychomotor speed) (118). Two buttons
of different colors (red, green, yellow or blue) are displayed. One
corresponds to the name of the color presented in black text (see figure 8B).
The subject is asked to press the button with the color corresponding to the
text. A new text and buttons then appears after a 2 seconds delay. Median
reaction time of 50 words is measured.
3. Stroop incongruent colors (executive functions) (118). Two
buttons of different colors (red, green, yellow or blue) and colored text are
displayed (see figure 8C). The subject is asked to press the button with color
corresponding to the color (not the text) of the text. A new text and buttons
then appears after a 2 seconds delay. If instructions are not followed while
the test is running, the observer has instructions to verbally correct the
patient at a maximum of two times. Median reaction time of 50 words is
measured. In paper III, if the patient made more than 50 % errors, the test
was regarded as failed. This criterion was loosened for paper IV, where 10
correct answers were sufficient to complete the test.
4. Trail making test A (psychomotor speed) (119). The subject is asked
to press numbers in consecutive order (1-25) displayed on the screen. Time
to completion is measured.
5. Trail making test B (executive functions) (119). The screen displays
letters (A-L) and numbers (1-13) (see figure 8D). The subject is asked to
press buttons by alternating between letters and digits (1-A-2-B-3-C...). Time
to completion is measured.
6. Ten-word-list (memory) (120). The subject is asked to remember 10
consecutive words presented with text and pre-recorded announcer speech.
The words are randomly drawn from a pool of the 50 most common Swedish
nouns. The subject is then asked to repeat the wordlist into a microphone
(see figure 8E). The task is repeated three times. The score is the sum of
correctly remembered words, summed over all three trials.
7. Delayed recall (memory) (120). After approximately 20 minutes of
distracter tasks, the subject is asked to repeat the words from the listlearning task. The number of correctly recalled words is used as the score.
8. Delayed recognition (memory) (120). The subject is asked to
recognize the 10 words from the list-learning task among 10 distracter
words. Words are presented consecutively. The test score is calculated as the
number of correct responses minus errors.
27
9. Figure copy task (visuospatial skill) (120). The subject is asked to
copy a cube displayed on the screen. The drawing is manually scored as
correct or incorrect. The figure is regarded as correct if the size is correct and
all lines are present.
A
B
C
D
E
F
Figure 8. Screen shots from some of the included tests. A, Two-choice reaction test,
B Stroop congruent colors test, C Stroop incongruent colors test, D, Trail making test
B, E Ten-word list test. F Four-finger tapping test.
28
10. Four-finger tapping (manual dexterity) (121). The subject is asked
to tap on a small numeric keyboard with fingers 2-5 of the dominant hand
(see figure 8F). The correct order of tapping is (digits): 2-3-4-5-4-3-2-3-4-54-3.... The number of correct taps during 10 seconds is measured. The task is
repeated for 5 times and the sum of taps over all trials is the score.
11. Geriatric Depression Scale (depression) (122). The Geriatric
depression scale is a 20 questions instrument intended to measure
symptoms of depression in the elderly. The questions are presented on the
screen in consecutive order and the subject is required to give answers by
pressing buttons marked yes or no.
Statistics
Significance threshold for all statistical tests was set to 0.05. Statistical
softwares used were R version 2.13 in paper I, and SPSS (SPSS Inc,
Chicago, IL, USA) in papers II-IV (version 18 in paper II and version 21 in
papers III-IV).
In paper I correlations and linear regressions between ICP and PI was
performed both on lumped data and on data stratified by individual patients.
Confidence and prediction intervals were calculated.
In paper II the relation between intracranial and lumbar pressure
amplitudes, as well as bench test data from the two sensors, were explored
by Bland Altman plots (123). Spearman correlations were consistently used.
Differences between intracranial and lumbar amplitudes (∆ICP - ∆LP) were
explored in a general linear model with mean intracranial pressure
(k*ICPmean) as covariate and a patient dependent factor (mpat). The model
was invoked to explain amplitude difference in time domain amplitudes
(ktime), amplitudes calculated from frequency domain fundamental frequency
and two over tones (kfund, kOT1 and kOT2). Also the same model was used to
explain the delay between intracranial and lumbar pressure waves (kdelay).
In paper III reliability was explored by Pearson correlation (r) between test
and retest scores of healthy elderly. The standard error of measurement SEm
was calculated for applicable tests, as SEm=SD*sqrt(1-r), where SD is the
standard deviation of the test scores. The SEm gives an error band around a
single given score, and the true score is approximately within ±2 SEm with
95% confidence. Practice effects between test and retest were explored with
29
paired T-tests when the normality assumption was met otherwise the
Wilcoxon signed rank test was used. Association between corresponding
scores from computerized and conventional, and between different
computerized tests was explored by Spearman correlations.
In paper IV improvement after shunt operation was explored by the
Wilcoxon signed rank test. The individual change (improved or worse) in a
test was defined as change in ability to complete the test, or scores changes
exceeding "SEm of the score difference" (SEmdiff), which was calculated from
values in paper III, as:
𝑆𝐸𝑚!"## = 2 ∗ 𝑆𝐸𝑚
Overall cognitive level was explored by summing Z-scores of all tests.
Standard deviations and means were obtained from INPH patients providing
data to each test. In tests where patients failed, the Z-score was set to the
lowest value of those patients who managed to complete the test.
30
RESULTS
Transcranial Doppler
intracranial pressure
pulsatility
index
and
ICP [mmHg]
0
20
40
60
The primary aim of paper I was to evaluate if TCD is a feasible method to
assess ICP. The overall regression line relating pulsatility index (PI) to ICP
was ICP=23*PI+14 (R2=0.22, p<0.01). The 95% prediction interval for ICP
given a PI measurement, was in the order of ±25 mm Hg, see Figure 9.
Linear regressions for data stratified by patient showed great intraindividual variability. For instance patient 3 (ICP=82*PI-32, R2=0.86) and
patient 4 (ICP=48*PI+3, R2=0.85) had a factor 2 slope difference.
Confidence interval
Prediction interval
0.0
0.5
1.0
1.5
PI
Figure 9.2 The regression line between PI and ICP. The 95% confidence and
prediction intervals are displayed.
Simulating the mathematical model showed that, with no changes in
physiological parameters, PI truly reflected ICP.
However, altered
physiological variables such as mean arterial pressure, arterial pressure
amplitude, autoregulation and compliance of the middle cerebral artery
distorted the relationship, see Figure 10.
2 Reprinted from paper III with permission: Behrens, A., Lenfeldt, N., Ambarki, K., Malm, J., Eklund, A., &
Koskinen, L.-O. (2010). Transcranial Doppler pulsatility index: not an accurate method to assess intracranial
pressure. Neurosurgery, 66(6), 1050–1057. Wolters Kluwer Health Lippincott Williams & Wilkins©
31
1.6
0.8
1.0
1.2
C
D
1.4
1.6
1.1
1.2
30
35
1.0
25
ICP [mmHg]
15
25
20
KMCA=4.5
KMCA=5.0
KMCA=5.5
10
10
Gain=1
Gain=0.5
Gain=0
0.9
20
30
35
40
PI
40
PI
15
ICP [mmHg]
30
35
1.4
25
15
ICP [mmHg]
1.2
Amplitude=40 mmHg
Amplitude=50 mmHg
Amplitude=60 mmHg
10
10
MAP=84.6 mmHg
MAP=94.6 mmHg
MAP=104.6 mmHg
1.0
20
30
25
20
15
ICP [mmHg]
35
40
B
40
A
1.3
1.0
PI
1.1
1.2
1.3
PI
Figure 10.3 Simulations of the ICP-PI relations’ sensitivity to alternations in
physiological parameters. Altered parameters are (A) mean arterial pressure (MAP),
(B) arterial pulse pressure amplitude, (C) autoregulatory gain, and (D) middle
cerebral artery (MCA) compliance (KMCA). An autoregulatory gain parameter equal
to 1 means normal autoregulation, whereas zero implies non-functioning
autoregulation. A low KMCA value indicates a more compliant vessel.
Intracranial and lumbar pressure waves
The aim of paper II was to determine if ICP pressure wave amplitudes were
accurately assessed via lumbar puncture. To account for hardware
differences in intracranial and lumbar measurement systems, also a bench
test was performed.
3 Reprinted from paper III with permission: Behrens, A., Lenfeldt, N., Ambarki, K., Malm, J., Eklund, A., &
Koskinen, L.-O. (2010). Transcranial Doppler pulsatility index: not an accurate method to assess intracranial
pressure. Neurosurgery, 66(6), 1050–1057. Wolters Kluwer Health Lippincott Williams & Wilkins©
32
Baseline
At baseline pressure (mean ICP = 18.7 mm Hg, mean ICP amplitude = 6.1
mm Hg) the intracranial amplitudes preceded and exceeded amplitudes
measured the lumbar route (∆Amp = 0.9 mm Hg, SD=1.0, p<0.05; Delay=39
ms). Bench test data showed small measurement related differences at
amplitudes corresponding to baseline readings (∆Ampsensors < 0.2 mm Hg).
Elevated ICP
The correlation between intracranial and lumbar pressure amplitudes was
high (r=0.98, p<0.001), see figure 10. At higher pressures the lumbar
readings increased more, and even exceeded intracranial amplitudes at the
highest pressure level (mean ICP = 44.3 mm Hg, mean ICP amplitude = 21.7
mm Hg, ∆Amp = -0.2). The lumbar delay were still present but smaller
(Delay = 18 ms). These results were supported by the GLMs (ktime = -0.033,
P < 0.01; kdelay=-0.63, P < 0.001). The GLMs also revealed variablity in the
patient dependent factor mpat.
A
B
Figure 11.4 In (A) the agreement between ∆ICP and ∆LP is displayed. There was an
excellent correlation (r=0.98, r<0.001). (B) shows the amplitude difference as
function of mean amplitude (r=-0.47, p<0.01). Numbers represent patients.
Bench test data revealed a relative overestimation by the lumbar sensor at
the higher-pressure amplitudes (Figure 11 A and B). Higher heart rate gave
4 Reprinted from paper II with permission: Behrens, A., Lenfeldt, N., Qvarlander, S., Koskinen, L.-O., Malm,
J., & Eklund, A. (2013). Are intracranial pressure wave amplitudes measurable through lumbar puncture?
Acta Neurologica Scandinavica, 127(4), 233–241.
33
larger sensor discrepancy (Figure 11 B). Differences in mean pressure did not
affect sensor readings.
B
Difference in amplitude between
Codman and Becton sensors (mean+/−SD) (mmHg)
Difference in amplitude between
Codman and Becton sensors (mean+/−SD) (mmHg)
A
0.4
0.2
0
−0.2
−0.4
−0.6
0
Time domain amplitude
Frequency domain amplitude
5
10
15
20
25
Mean amplitude of all sensors (mmHg)
30
0.4
0.2
0
−0.2
−0.4
−0.6
−0.8
−1
−1.2
−1.4
Time domain amplitude
Frequency domain amplitude
−1.6
0
5
10
15
20
25
30
Mean amplitude of all sensors (mmHg)
35
Figure 12.5 Bench test amplitude difference as function of mean pressure amplitude
calculated from time and frequency domain data. Data represents means for six
measurements with different sensors. Figure (A) displays results for a heart rate of 60
bpm (r = -0.73, p<0.001 and r=0.73, p<0.001) and (B) 120 bpm (r = -0.94, p < 0.001
and r = 0.13, p = 0.11).
Time versus frequency analysis
In the in vivo experiment, amplitudes calculated from the fundamental
frequency, consistently underestimated time domain analysis. Looking at
ICP amplitudes, the difference was 2 mm Hg at baseline and 3.8 mm Hg at
the highest-pressure level. A frequency analysis revealed that not the
fundamental, but the first two overtones increased with higher mean
pressures in lumbar compared to intracranial readings (kfund = -0.002, P =
0.85; kOT1 = -0.017, P < 0.01 and kOT2 = -0.015, P < 0.01). Bench test data
also revealed that amplitudes calculated from the fundamental frequency
were less affected by sensor differences.
5 Reprinted from paper II with permission: Behrens, A., Lenfeldt, N., Qvarlander, S., Koskinen, L.-O., Malm,
J., & Eklund, A. (2013). Are intracranial pressure wave amplitudes measurable through lumbar puncture?
Acta Neurologica Scandinavica, 127(4), 233–241.
34
CoGNIT
The aim of papers III and IV was to implement commonly used
neuropsychological tests for INPH into a computerized format, and to
evaluate the reliability, validity, responsiveness to shunt surgery and
completion rate for INPH patients. Also, scores for healthy elderly were
collected for comparison.
Validity
Validity was explored by correlations between corresponding conventional
and computerized test scores in patients with cognitive impairment. Results
are displayed in table 5. There were significant correlations between all
computerized and corresponding conventional tests (r=0.49-0.87).
Table 5. Correlation between conventional and computerized tests.
Testa
Correlation computerconventional test
Significance
of correlation
Stroop congruent
0.82
<0.001
Stroop incongruent
0.76
<0.001
Ten word list
0.66
<0.001
Delayed recall
0.72
<0.001
Delayed recognition
0.49
<0.01
Trail making test A
0.85
<0.001
Trail making test B
0.83
<0.001
Figure copy task
0.54
<0.01
aNo
corresponding paper and pencil test exist for the Two-choice reaction and Four
finger tapping tests.
35
Divergent validity was explored by correlation between different
computerized tests. Results are displayed in table 6. Significant correlations
exist between tests assessing the same domain. Also, tests with a strong
motor component correlate e.g. the Stroop tests, the Trail making tests and
the Four-finger tapping test. A correlation was also seen between the
Delayed recognition and Figure copy tests.
Table 6. Correlation matrix of healthy elderly performance at first computer
test session.
NS = Non significant.
Reliability
Reliability was explored by correlations between repeated computerized
testing in healthy subjects. Test-retest correlations and tests of improvement
between test and retest are displayed in table 7. The mean test-retest interval
was 21 days and correlations between test and retest were >0.7 in 8 out of 10
tests. The Figure copy test had a correlation of 0.57 and the Ten-word list
test had a correlation of 0.67. There were significant improvements at
retesting in 5 tests.
36
Table 7. Results from the reliability study.
Computer test
Improvement
test-retest
Correlation
test-retest
Two choice reaction
NS
0.75
Stroop congruent
NS
0.74
Stroop incongruent
<0.01
0.83
Ten word list
<0.001
0.67
Delayed recall
NS
0.74
Delayed recognition
NS
0.70
Trail making test A
< 0.05
0.87
Trail making test B
< 0.05
0.83
Figure copy task
NS
0.57
Four finger tapping
<0.001
0.90
NS = Non significant
Ability to complete the test
In paper III 78% of consecutively evaluated hydrocephalus patients
completed the battery with one or zero failed tests. The corresponding
number from paper IV during the preoperative investigation of INPH
patients was 81%. The most frequently non-completed tests were Trail
making test B, Stroop incongruent colors test and the Four-finger tapping
test.
37
Performance of untreated INPH patients
INPH patients recruited in papers III and IV performed significantly
worse compared to healthy on all administered tests. Interquartile ranges
between INPH patients and healthy were non-overlapping in all tests but a
small overlap in the Trail making test B in paper III (see figure 13). The
overall profile of cognitive impairment was similar between the two INPH
cohorts. In paper IV it was shown that vascular risk factors affected
preoperative scores negatively.
140%#
120%#
100%#
80%#
Healthy#
INPH#Study#III#
INPH#Study#IV#
60%#
40%#
20%#
0%#
Two#choice#
reac2on#
Stroop#
congruent#
Stroop#
Ten#word#list# Delayed#recall#
Delayed#
incongruent#
recogni2on#
Trail#making#
test#A#
Trail#making#
test#B#
Four#finger#
tappinge#
Figure 13. Performance of INPH patients recruited in papers III and IV. Results are
normed so that 100% represents median scores of healthy controls. Bars indicate
interquartile ranges.
Neuropsychological improvement after shunt
The aim of paper IV was to evaluate the effects of shunt surgery on CoGNIT
performance in INPH patients. Mean delay from cognitive testing to shunt
38
surgery was 5.6 months. Mean follow up was 4.2 months after surgery.
Improvement in tests on group and individual level are displayed in table 8.
After shunt surgery, patients were still impaired compared to healthy with
the exception of the Four finger-tapping test. Significant improvement was
seen in all cognitive domains: Memory (Ten-word list); Attention and
psychomotor speed (Two choice reaction test, Stroop congruent colors);
Executive functions (Stroop incongruent colors) and Manual dexterity (Fourfinger tapping). Improvement at trend level was also seen in the Delayed
recall test (p=0.16). Improvement in summed Z-scores was seen in 14
patients (45%).
Table 8. Neuropsychological improvement after shunt surgery.
Improvement
Individual changea
P-value
I/U/W
Two choice reaction
<0.01
16/11/4
Stroop congruent
<0.05
14/10/7
Stroop incongruent (n=19)
<0.05
14/11/6
Ten word list
<0.01
13/14/4
Delayed recall
NS
11/13/7
Delayed recognition
NS
9/18/4
Trail making test A (n=30)
NS
11/9/11
Trail making test B (n=22)
NS
10/11/10
Four finger tapping (n=12)
<0.01
11/13/7
Summed Z-scores
<0.01
14/12/5
Test
I = numbers improved, U = Unchanged, W = Worse, NS = Non significant
39
DISCUSSION
Paper I validated the non-invasive transcranial Doppler pulsatility index
(PI) method of estimating ICP. The experiment showed that PI was an
unreliable predictor of ICP. In a mathematical model, it was shown that PI
was affected by various physiological parameters apart from ICP.
In paper II, pressure wave amplitudes were simultaneously measured in
the brain parenchyma and via a lumbar puncture in INPH patients. Main
finding was that at resting pressure, amplitudes were slightly attenuated in
lumbar CSF compared to in the brain parenchyma. In summary, lumbar
pressure amplitude measurement with a fluid catheter system is an
alternative to intraparenchymal ICP measurement, but the thresholds for
what should be interpreted as elevated amplitudes need to be adjusted.
In papers III and IV a computerized neuropsychological test battery for
INPH
was
developed
and
evaluated
(COmputerized
General
Neuropsychological INPH Test = CoGNIT). Tests were chosen after a
literature review and correlated significantly with corresponding paper-andpen versions. In paper III it was shown that reliability was good, and the
completion rates in INPH satisfying. In paper IV it was established that
CoGNIT showed impairment in untreated INPH patients compared to
healthy controls and improvement after shunt surgery was detected in all
assessed cognitive domains. Consequently CoGNIT is a new instrument in
the INPH field, for research but also for daily patient care.
Pulsatility index and ICP
The results of paper I refutes the pulsatility index method of predicting ICP.
The prediction interval for ICP given a PI reading was in the order of ±25
mm Hg, and henceforth too wide to be of clinical utility. The large prediction
interval of lumped data was explained by a large variability in the PI/ICPrelation between patients. The mathematical simulations indicated that
differences may be due to physiological variation. Prior to publication of
paper I, there were many positive publications regarding the potential
usefulness of the pulsatility method based largely on data from patients with
traumatic brain injury (14-­‐17), even though regression lines for the ICP/PIrelation differed largely between studies (15,17).
40
A limitation with the TCD technique is that it is affected by operator
dependent factors, which may explain differences between studies (124).
Also differences in constitution of the temporal bone window make TCD data
inaccessible in approximately 10-15 % of patients(125).
The negative result from our study may be due to the clinical material, where
hydrocephalus patients may show a less perfect PI-ICP correlation. However,
after our publication, one large study with 290 TBI patients with continuous
ICP monitoring also looked at prediction intervals of ICP from TCD PI data.
The prediction interval was between ±15-35 mm Hg, depending magnitude
of PI (126). Also, our results from the mathematical simulations has been
confirmed showing that many other variables than ICP affect PI, e.g. arterial
pulse pressure, heart rate and cerebral perfusion pressure (127).
The strength of paper I is that PI was measured over a wide range of ICPs in
every patient. From this data it could be concluded that the wide prediction
interval of lumped data stem from differences between individuals. The
differences were so large that it is not possible to transfer a general PI-ICP
relation to the individual level.
One obvious limitation with paper I is the low number of participants.
However, even including more patients, the variation from the first patients
would remain, but of course give better estimations of the prediction interval
of ICPs.
A tool for non-invasive ICP measurement would be very useful in clinical
practice. However, with such a complex system, with a multitude of
interactions such as the cerebrovascular hemodynamic system of the brain, it
would be surprising if a simple index such as PI from the middle cerebral
artery would predict ICP with high accuracy. Extraordinary claims require
extraordinary evidence and even more so when the utility is with critically ill
patients. The more enthusiastic reports on this topic should be interpreted
carefully, and future studies should include prediction intervals for derived
ICP.
Pulse pressure amplitudes of the ICP
The pulsatile stress from each heartbeat will eventually exhaust
predominantly the non-regenerating materials of the body. Microfractures
and degeneration in the walls of aorta is seen in normal ageing, leading to
dilation and stiffening (128). There is then a breakdown of the arterial
41
Windkessel that leads to larger flow pulsations, transmitted predominantly
to large flow organs such as the brain and the kidney (128). Also, the
compliance of the craniospinal compartment decreases with age (129,130).
An increased pulsatility transmitted from the larger arteries lead to micro
vascular damage with smaller arterial lumen and less mean flow (131).
Larger arterial pulsatile flow, together with decreased compliance of the
brain will result in larger ICP pulsatility that may indirectly affect the brain
(132). Recent studies in healthy elderly has linked larger arterial and
intracranial pulsations to decreased cognitive performance, smaller total and
regional brain volumes, larger ventricles and more white matter lucencies on
MRI (132-134). INPH patients have larger ICP amplitudes and the
assessment of those is proposed as a predictive test for shunt surgery (72,74).
ICP amplitudes would be far more accessible if measurable the lumbar route.
In paper II, there was a good correlation between intracranial and lumbar
pressure amplitudes. Relying on lumbar measurement will underestimate
true ICP amplitudes by approximately 0.9±1 mm Hg at baseline ICP. This
underestimation holds for the clinically most relevant range, and is probably
due to dampening of the pulse waves traveling down the spinal canal. The
standard deviation of amplitude differences may to a smaller degree be due
to measurement error, and are most likely due to individual differences.
These limitations is important to take note of e.g. using lumbar
measurements in the test for selecting INPH-shunt candidates proposed by
Eide (74). Maybe, recording of resting lumbar CSF pressure amplitudes can
only reliably be used in a prognostic test if they are very large. Accordingly
Eide et al. found that lumbar CSF pressure amplitude at baseline was <2 mm
Hg in 17/42 of patients who had mean ICP amplitude ≥4 mm Hg in overnight intraparenchymal registrations (72). Interestingly, Eide also found that
prediction of patients with large ICP amplitudes in an over-night registration
was increased if lumbar pulse amplitude analysis was done while performing
lumbar infusion (72,135). Regardless of measurement route, the predictive
value of pressure amplitudes in INPH needs to be confirmed in a
randomized study.
The underestimation of ICP amplitudes was less at higher pressures. This
may in part be explained by difference in measurement systems, where
lumbar measurements with a fluid catheter system amplify overtones of the
frequency spectra, and more so at larger amplitudes. However, relying on
amplitudes derived from the fundamental frequency will underestimate true
amplitudes. Even though measurement system discrepancy increased at
higher amplitudes, the differences were small compared to the magnitude of
the measurement, and may not have clinical relevance.
42
The limitations and strengths of paper I also hold for paper II, i.e. low
number of patients, but many measurements over a wide range of ICPs in
each patient. One limitation may be that the baseline ICP and ICP
amplitudes of paper II were large compared to other studies on INPH
(49,72), and difference between lumbar and ICP amplitudes may be larger at
lower ICPs. A general limitation with the method of amplitude
measurements the lumbar route, may be influence from physiological factors
such as spinal stenosis. However, this needs to be researched.
Lumbar puncture is routine in preoperative assessment of INPH patients.
Intracranial pressure pulse amplitudes are reflected in lumbar space, which
opens up for studies of the predictive test described by Eide (72,74). Also,
following the results of paper II, our research group has described another
interesting application of lumbar CSF pressure amplitude measurements. If
large preoperative intracranial pressure amplitudes are predictive of
outcome after shunt surgery in INPH, and the curative effect of the shunt is
by lowering amplitudes, then there should be an association between
amplitude reduction and outcome. Accordingly it was demonstrated that
amplitude reduction measured through the lumbar route was significantly
larger among patients who improved in gait (responders) after shunt
surgery, than non-responders (49). However, this needs to be confirmed in a
larger study. In the cognitive domain Foss et al. found that INPH shunt
responders had larger preoperative ICP amplitudes (98). It remains to be
examined if amplitude reductions are associated to improvements also in the
cognitive domain. The effect may be even more pronounced by combined
measures.
The potential for amplitude reduction from shunt surgery differ between
patients (21). The idea of preoperatively assessing the potential for
amplitude reduction opens up for interesting predictive tests. Especially
important in a field, where the largest challenge is excluding patients who
will not benefit from shunt surgery (65).
The prospect of lumbar pulse pressure measurements in INPH seems
promising. But to study their clinical impact, accessible and sensitive
outcome measures are needed.
CoGNIT development
When setting out developing the computer test the concerns were manifold.
Would elderly, potentially non computer-skilled persons be testable? How to
43
select appropriate tests? Also, the purely technical aspects were of concern,
e.g. what computer language is suitable?
The CoGNIT battery was evaluated with regards to completion rates of tests,
validity, reliability, discriminative ability and ability to detect cognitive
improvement after shunt surgery, discussed in the following paragraphs.
Test evaluation
The technical solution of using Java and Flash programming languages was
good. Java is a widely used programing language and programs run on any
computer. In the Flash language development of user interfaces and
animations are very easy. When starting the neuropsychological test battery,
the Java program, starts the Flash application; receive data from the tests
(and the microphone), and store results to a file. This technical solution
makes for easy development, fast translation of tests into other languages
but also portability to different computer platforms (e.g. Apple OS X or,
Microsoft Windows). It is also clear that the touch screen interface was a
good solution, providing a natural way of interacting with the computer. It
has also been proposed as the best input modality for testing elderly (108).
Completion rates
Completion rates were good. The healthy elderly completed all tests.
Hydrocephalus patients showed a preoperative completion rate around 80
%, and for individual tests the completion rates were in parity with
conventional neuropsychological testing. The tests that were not always
completed were the Trail making test B, Stroop incongruent colors and the
Four-finger tapping test. Trail making test B was regarded as failed if the
patient gave up, whereas in the Stroop test, the test was failed if more than
50 % errors. In paper IV, this criterion was loosened, and only 10 correct
answers were enough to gain a score on the test. This was in concordance
with other studies using the Stroop test in INPH (82). Still, completion of the
Stroop test was around 70 %, which is in line with other studies on INPH
(82). Trail making test B was completed by 75 % which is far more than the
38 % completion rate seen in the largest study of neuropsychology in INPH
to date (79). It should be noted that also a failed test provides information
about patient performance on a test. This may be more so in the executive
function domain where there is a component of "getting it". In the Four-
44
finger tapping test many patients started out correctly but soon reverted to
tapping with only the index finger. This might be relived by adding an extra
instruction, reminding that four fingers should be used. However, no data
confirming this are present at the moment. Still, completion rates for
CoGNIT can be regarded as satisfactory. Importantly, completion rates were
unaffected by reported computer knowledge.
Validity
The implemented tests were valid, in the sense that they correlated
significantly with their conventional paper-and-pen versions. The
magnitudes of correlations are dependent on range, but also on reliability
(measurement error) in both the computerized and conventional tests.
Hence, because of the poor reliability, the Figure-copy test showed a weak
correlation in paper III and was removed in paper IV.
Divergent validity was established between most of the tests. Not
surprisingly there were correlations between tests in the same cognitive
domain e.g. Trail making test A and Stroop congruent (psychomotor speed),
Trail making test B and Stroop incongruent (executive functions) and Tenword list sum of three trials and Delayed recall and recognition (memory).
Also, there were correlations between tests with a strong motor component.
Neuropsychological tests are seldom a pure measure of a cognitive domain
or ability. In all test with touch screen interface there will be some "motor
contamination".
A similar problem exist in many conventional
neuropsychological tests e.g. in the Trail making tests. Though correlation
between tests, it was noted that median improvement after surgery in the
Stoop incongruent test was double that of the Stroop congruent test (tests
with the same motor component), indicating that functions beyond motor
performance improve. It has been argued that divergent validity is more
important to establish when evaluating computerized tests (105), as e.g.
computer knowledge can influence all scores. Reported computer knowledge
was only associated with the baseline score of INPH patients in Trail making
test A in paper IV. A finding that was not replicated in either healthy or
INPH patients of paper III.
45
Reliability
In paper III it was shown that 8 of 10 the implemented tests had good testretest correlations (r>0.7). The Ten word list (r=0.67) and the Figure copy
test (r=0.57) were the tests with moderate performance with this regard.
Reliability is a measure of error in a score and it can be shown that
correlations between test and retest estimate this error (123). However, as
any correlation, this number is largely dependent on the range of
measurements and test-retest correlations cannot be compared between
groups. An absolute reliability measure such as the standard error of
measurement (SEm) is less affected by the population under study (136). In
this study the SEm was calculated as:
𝑆𝐸𝑚 = 𝑆! (1 − 𝑟)
where 𝑆! is the standard deviation of the observed scores and 𝑟 the testretest correlation. The SEm allows for making statements about the
confidence level of individual scores. An obtained score is approximately
within one SEm with 68 % confidence. It is clear that the SEm is less
influenced by the variation of scores and thus populations. An increased
variability of scores will increase 𝑆! but will be compensated by an increased
r.
After paper III, it was clear that the Figure-copy test showed poor
reliability. Probably performance was hampered because of a dichotomous
score, where a difference from pass to fail has a large impact. Also, in this
test the result has to be judged by a human, which makes for the traditional
errors when human judgments are involved namely intra and interrater
variability. This test was therefore left out of CoGNIT. Of the remaining tests
it was shown that the test with least test-retest reliability (Ten-word list,
r=0.67), with great confidence separated INPH from healthy, even when
measurement error was accounted for.
This is an example of relating
measurement errors to "analytical goals", i.e. the implication of the
measurement errors on inferences from a test (136). The subtests of CoGNIT
were reliable enough to be evaluated for clinical usefulness.
46
Baseline impairment and cognitive change
Pathology in INPH is related to dysfunction in subcortical white matter with
the largest deficit closest to the ventricles (137). Thus, impairments in the
rich connectivity from basal ganglia to frontal brain areas, connected via a so
called basal ganglia-thalamocortical loops, has been hypothesized to cause
the observed pattern of cognitive deficits (99,121,138-140). In the context of
INPH their disruption would be possible at several sites, at the striatum,
most likely at the caudate nucleus close to the lateral ventricles, the
thalamus, adjacent to the third ventricle and the periventricular white
matter. A finding in INPH, potentially affecting several cognitive domains is
that many INPH patients show impaired and reversible level of wakefulness,
that may be due to disturbance of the brain arousal system (141). A
radiological finding in INPH is reduced hippocampal volume and regional
cerebral blood flow, though not to the extent seen in Alzheimer’s (141,142).
This would explain the relatively mild memory deficiency in INPH (87).
CoGNIT detected significant impairments in all tests in untreated INPH
patients in both papers III and IV, see figure 13. Looking at individual
data the interquartile ranges of healthy and INPH scores were nonoverlapping in all tests but a slight overlap in Trail making test B of paper
III. Overall cognitive impairment reflected by the summed Z-scores in
paper IV showed that only one INPH patient scored just above the 5th
percentile of healthy. The ability for CoGNIT to detect cognitive impairment
of INPH patients must be regarded as good. Patients performed worst in the
Delayed recall and Stroop incongruent tests and best in the Delayed
recognition test, corroborating previous studies on conventional
neuropsychological testing in INPH (62,81). CoGNIT can thus be regarded to
capture the "finger print" of cognitive impairments present in INPH patients.
Also, the effect of vascular risk factors on neuropsychological performance in
INPH has been described before (81,99).
In paper IV, it was shown that CoGNIT was sensitive to detect post shunt
improvements in attention/psychomotor speed, manual dexterity, memory
and executive functions. No ceiling effects were observed in any test. Some
tests did not show any improvement on the group level (Trail making tests
and, the Delayed recognition test). Notably the Trail making test B was not
found to improve on a group level in the largest INPH cohort studied to date
(79). Post-shunt improvement in the domains of CoGNIT has been described
before (62,79,82). The magnitude and number of patients who improve after
shunt surgery differ between studies for a number of reasons. Firstly,
47
different neuropsychological tests and criteria for cognitive change are used
in different studies. Secondly, the time between neuropsychological
evaluation and shunt surgery likely vary between cohorts. Andrén et al.
showed that neuropsychological performance merely reached to baseline
values when surgery was delayed by more than 6 months (61). This delay is
rarely reported in studies and may be a large factor affecting results. In
paper IV, the mean pre-operative test delay was 5.6 months, probably
hampering results to a bit. Thirdly, centres have different criteria for
shunting, and may have different success rates with regards to this.
Regardless, the sensitivity of CoGNIT to detect cognitive impairment in
INPH patients seems good.
There are several intricacies with regards to measurement of cognitive
change. Ideally each patient should be compared with an untreated clone. As
this is not possible other methods must be used. In this thesis patients were
compared to the measurement error of the test. This is at least theoretically
stable across groups. Others have related baseline impairment and
improvements after shunt surgery to the standard deviation of a control
group, which makes for a more floating interpretation, as biases can be
introduced by i.e. different ways of recruiting controls.
Limitations
Limitations in the evaluation of CoGNIT include a relatively small number of
healthy controls that may hamper generalizability. Also, the small number of
INPH patients in paper IV, is limiting the power to detect post shunt
improvement.
The general criticism of computerized neuropsychological testing is that of
failure to show validity, and often lacking normative and error data. Also the
field is suffering from poorly designed computer-patient interfaces and
failure to show influence by e.g. computer knowledge (105,107). In the
development and evaluation of CoGNIT these issues were addressed.
It must be stated that testing with CoGNIT only provides data collection (and
a written report immediately after ending the testing session). Skilled
personnel must perform interpretation and diagnostics. Also, computerized
neuropsychological testing does not provide the qualitative data from an
evaluation by a neuropsychologist. Of course, before drawing conclusions
about higher cognitive functions, lower functions must be assured (e.g.
arousal). However, this is not unique for computerized testing.
48
Standardization in INPH
A great leap forward in the field of INPH was the international guidelines for
diagnosing the disease published in 2005 (143). A standard definition of the
disease makes for comparability of patient material between studies. In the
measurement of outcome however, there are a plethora of methods. Several
INPH specific scales have been proposed, but none widely adopted (25,144146). A promising development in this field, is the new INPH scale put
forward by Hellström et al. (100). The neuropsychological part includes an
average of normed scores of the Stroop tests, RAVLT and the Grooved
pegboard tests. CoGNIT includes tests tapping in to the same cognitive
domains, but is more extensive, and does not require special training to
administer. The MMSE is often included in many studies on INPH and
makes for comparison. It is practical as various health workers can
administer it, but lacks sensitivity in INPH (76).
Future directions
Having a standardized cognitive test that can be administered by e.g. nurses
will be practical and if adopted aid in comparability and cooperation
between centres managing INPH patients. A perquisite is that CoGNIT is
translated to other languages. Today, there are translations into Danish,
American English and Swedish. As of today German seems to be the next
language of translation. There is a bit of effort attached to this, as norms may
differ between countries. Also, translation may to some degree alter validity
and reliability of the tests.
The technical solution of CoGNIT test development makes for easy adding
and studying new neuropsychological tests, potentially advancing the field of
neuropsychological testing in INPH.
A next step could be porting CoGNIT to iPad. An iPad version would be
easily distributed and potentially increasing adoption.
The accessibility of CoGNIT may aid the longitudinal management of INPH
patients, where a sudden impairment may indicate other evaluations for e.g.
shunt dysfunction.
49
There are few studies that have looked at the utility of cognitive testing
before and after CSF drainage tests (69,86,147), and results are mixed. It
would probably not be practical to administer the full CoGNIT battery pre
and post e.g. a spinal tap. But a subset with the most responsive tests may be
useful. These tests may provide data in patients who are not able to walk,
where a standard test protocol with gait tests would give no information.
With retesting over such a short interval as 24 or 72 hours, practice effects
need to be explored and accounted for. A new research protocol is already
under way with regards to this.
An interesting research path lead from the finding that gait improvers show
larger CSF-pressure amplitude reductions after shunt surgery. Can the
findings be replicated in the cognitive domain or combined measures?
Comorbidities are common in INPH (53). CoGNIT is a focused battery for
INPH, and the usefulness for differential diagnostics is probably limited.
However, there might be clues, as e.g. a very large memory impairment
including a impaired Delayed recognition test might give a suspicion of a
concomitant Alzheimer's disease (AD). Also the potential for cognitive
recovery may be hampered by comorbidities such as AD (86). However, this
is yet to be researched.
50
CONCLUSIONS
Ia. The pulsatility index (PI) derived from the middle cerebral artery was
associated with intracranial pressure (ICP).
Ib. The prediction interval for ICP from PI data was too wide to be of clinical
use. The large difference was explained by individual differences in the PI ICP relation.
IIa. Intracranial pressure waves were measurable through lumbar puncture.
There was a difference in amplitudes from ICP and lumbar measurements at
resting pressures, likely due to dampening of pulse waves in the spinal canal.
IIb. Lumbar pressure amplitude measurement with fluid catheter is an
alternative to direct ICP measurement with a catheter tip sensor, but the
thresholds for what should be interpreted as elevated amplitudes need to be
adjusted.
III. The computerized neuropsychological test battery (CoGNIT) shows
satisfactory test-retest reliability in healthy elderly and validity for cognitive
impaired patients. Additionally, health workers without special training can
administer it.
IV. CoGNIT show good completion rates in INPH patients. CoGNIT is
sensitive to detect the preoperative cognitive impairment and changes after
shunt surgery in INPH patients.
51
ACKNOWLEDGEMENTS
I want to express my sincere gratitude to all who have contributed to this
thesis. Without you, and modern technology this thesis would never have
come true.
Jan Malm, my main supervisor. You have been supportive, patient, available
and helpful.
Anders Eklund, co-supervisor. For your dedicated mentorship. You taught
me to really look at the data.
Niklas Lenfeldt, co-supervisor. For your help and sincerity. You got me
started in research, for which I am truly grateful.
Eva Elgh, co-supervisor in the field of neuropsychology. For your help,
friendly and encouraging way.
The always-friendly Lars-Owe Koskinen for good cooperation.
Colleagues in the group, PhD candidates and co-authors, for interesting
discussions and lots of fun, Sara Qvarlander, Anders Wåhlin, Khalid
Ambarki, Kenneth Anderson and Nina Sundström. Special thanks to Hanna
Israelsson for valuable comments on the thesis.
I would like to thank Michael Williams and Cynthia Smith for great
hospitality in my visit to Baltimore and for good cooperation with CoGNIT.
Friendly partners cooperating with the CoGNIT in Sweden and Denmark,
Göran Leijon, Katarina Laurell, Bo Traberg-Kristenssen and Helle Vibe Krog
for providing the Danish translation of the test.
Sonja Edvinsson, Jörgen Andersson, Hanna Ackelind, Carina Gillewård,
Marie Bäckström and Kerstin Gruffman at the neurological department in
Umeå. Special thanks to always-cheerful Kristin Nyman, a great working
companion. Without you, this would be a thesis without any data.
My employer at Department of internal medicine at Karlskrona Hospital for
patience and funding. Collegues at the Department for internal medicine.
Especially my skilled colleagues in the neurology department: Marcus
Svennerud, Håkan Johansson, Malgorzata Losiak, Björn Servin, Evelina
52
Kask-Ogenblad, Lars-Erik Sjöberg, Lars Brattström, Maria P Idbrant and
Carina Gunnarson. Also, the skilled staff at 58:an.
Vetenskapliga rådet and Kompetenscentrum at Karlsrona Hospital, for
funding, a place to work and friendly colleagues. Alzheimerfonden,
Kempefonden, Forskningsfonden för Klinisk Neurovetenskap vid Norrlands
Universitetssjukhus for funding.
My friends. Especially Daniel Jinnelöv for valuable help on Java
programming and Magnus Ekström for valuable comments on the thesis.
My parents for love and encouragement. You are great. My brother and
sister. You are great too.
Fanny, my future wife. For patience, love and our wonderful children Lovisa
and Astrid.
53
REFERENCES
1.
Smith M. Monitoring intracranial pressure in traumatic brain
injury. Anesth Analg. 2008 Jan;106(1):240–8.
2.
Raboel PH, Bartek J, Andresen M, Bellander BM, Romner B.
Intracranial Pressure Monitoring: Invasive versus Non-Invasive
Methods-A Review. Crit Care Res Pract. 2012.
3.
Rangel-Castilla L, Rangel-Castillo L, Gopinath S, Robertson CS.
Management of intracranial hypertension. Neurol Clin. 2008
May;26(2):521–41.
4.
Binz DD, Toussaint LG, Friedman JA. Hemorrhagic complications
of ventriculostomy placement: a meta-analysis. Neurocrit Care.
2009;10(2):253–6.
5.
Lozier AP, Sciacca RR, Romagnoli MF, Connolly ES.
Ventriculostomy-related infections: a critical review of the
literature. Neurosurgery. 2008 Feb;62 Suppl 2:688–700.
6.
Zhang L-F, Hargens AR. Intraocular/Intracranial pressure
mismatch hypothesis for visual impairment syndrome in space.
Aviat Space Environ Med. 2014 Jan;85(1):78–80.
7.
Shimbles S, Dodd C, Banister K, Mendelow AD, Chambers IR.
Clinical comparison of tympanic membrane displacement with
invasive intracranial pressure measurements. Physiol Meas. 2005
Dec;26(6):1085–92.
8.
Soldatos T, Chatzimichail K, Papathanasiou M, Gouliamos A. Optic
nerve sonography: a new window for the non-invasive evaluation of
intracranial pressure in brain injury. Emerg Med J. 2009
Sep;26(9):630–4.
9.
Alperin NJ, Lee SH, Loth F, Raksin PB, Lichtor T. MR-Intracranial
pressure (ICP): a method to measure intracranial elastance and
pressure noninvasively by means of MR imaging: baboon and
human study. Radiology. 2000 Dec;217(3):877–85.
10.
Aaslid R, Markwalder TM, Nornes H. Noninvasive transcranial
Doppler ultrasound recording of flow velocity in basal cerebral
arteries. J Neurosurg. 1982 Dec 1;57(6):769–74.
11.
Moppett IK, Mahajan RP. Transcranial Doppler ultrasonography in
anaesthesia and intensive care. Br J Anaesth. 2004 Nov
54
1;93(5):710–24.
12.
Kirsch JD, Mathur M, Johnson MH, Gowthaman G, Scoutt LM.
Advances in Transcranial Doppler US: Imaging Ahead.
RadioGraphics. 2013 Jan;33(1):1–14.
13.
Gosling RG, King DH. Arterial assessment by Doppler-shift
ultrasound. Proc R Soc Med. 1974 Jun 1;67(6 Pt 1):447–9.
14.
Homburg AM, Jakobsen M, Enevoldsen E. Transcranial Doppler
recordings in raised intracranial pressure. Acta Neurol Scand. 1993
Jun;87(6):488–93.
15.
Voulgaris SG, Partheni M, Kaliora H, Haftouras N, Pessach IS,
Polyzoidis KS. Early cerebral monitoring using the transcranial
Doppler pulsatility index in patients with severe brain trauma. Med
Sci Monit. 2005 Feb 1;11(2):49–52.
16.
Moreno JA, Mesalles E, Gener J, Tomasa A, Ley A, Roca J, et al.
Evaluating the outcome of severe head injury with transcranial
Doppler ultrasonography. Neurosurgical focus. 2000;8(1):e8.
17.
Bellner J, Romner B, Reinstrup P, Kristiansson K-A, Ryding E,
Brandt L. Transcranial Doppler sonography pulsatility index (PI)
reflects intracranial pressure (ICP). Surgical neurology. 2004
Jul;62(1):45–51.
18.
Mokri B. The Monro-Kellie hypothesis: Applications in CSF volume
depletion. Neurology. 2001 Jun 26;56(12):1746–8.
19.
Marmarou A, Shulman K, Rosende RM. A nonlinear analysis of the
cerebrospinal fluid system and intracranial pressure dynamics. J
Neurosurg. 1978 Mar;48(3):332–44.
20.
Avezaat CJ, van Eijndhoven JH. Clinical observations on the
relationship between cerebrospinal fluid pulse pressure and
intracranial pressure. Acta neurochir. 1986;79(1):13–29.
21.
Qvarlander S, Malm J, Eklund A. The pulsatility curve-the
relationship between mean intracranial pressure and pulsation
amplitude. Physiol Meas. 2010 Nov;31(11):1517–28.
22.
Williams B. Simultaneous cerebral and spinal fluid pressure
recordings. 2. Cerebrospinal dissociation with lesions at the
foramen magnum. Acta neurochir. 1981;59(1-2):123–42.
23.
Williams B. Simultaneous cerebral and spinal fluid pressure
recordings. I. Technique, physiology, and normal results. Acta
55
neurochir. 1981;58(3-4):167–85.
24.
Lenfeldt N, Koskinen L-OD, Bergenheim AT, Malm J, Eklund A.
CSF pressure assessed by lumbar puncture agrees with intracranial
pressure. Neurology. 2007 Jan 9;68(2):155–8.
25.
Eide PK. Intracranial pressure parameters in idiopathic normal
pressure hydrocephalus patients treated with ventriculo-peritoneal
shunts. Acta neurochir. 2006 Jan;148(1):21–9.
26.
Aschoff A, Kremer P, Hashemi B, Kunze S. The scientific history of
hydrocephalus and its treatment. Neurosurg Rev. 1999 Oct;22(23):67–93.
27.
Rekate HL. The definition and classification of hydrocephalus: a
personal recommendation to stimulate debate. Cerebrospinal Fluid
Res. 2008;5(1):2.
28.
Irani DN. Cerebrospinal Fluid in Clinical Practice. Philadelphia:
Elsevier Health Sciences; 2008.
29.
Williams MA, Relkin NR. Diagnosis and management of idiopathic
normal-pressure hydrocephalus. Neurol Clin Pract. 2013 Oct
14;3(5):375–85.
30.
Hakim S, Adams RD. The special clinical problem of symptomatic
hydrocephalus with normal cerebrospinal fluid pressure.
Observations on cerebrospinal fluid hydrodynamics. J Neurol Sci.
1965 Jul;2(4):307–27.
31.
Malm J, Eklund A. Idiopathic normal pressure hydrocephalus.
Pract Neurol. 2006;6:14–27.
32.
Jaraj D, Rabiei K, Marlow T, Jensen C, Skoog I, Wikkelsø C.
Prevalence of idiopathic normal-pressure hydrocephalus.
Neurology. 2014 Apr 22;82(16):1449–54.
33.
Israelsson H, Birgander R, Ambarki K, Eklund A, Malm J.
Ventriculomegaly and balance disturbances in patients with TIA.
Acta Neurol Scand. 2011 May 7;125(3):163–70.
34.
Brean A, Eide PK. Prevalence of probable idiopathic normal
pressure hydrocephalus in a Norwegian population. Acta Neurol
Scand. 2008 Jul 1;118(1):48–53.
35.
Tisell M, Höglund M, Wikkelso C. National and regional incidence
of surgery for adult hydrocephalus in Sweden. Acta Neurol Scand.
2005 Aug;112(2):72–5.
56
36.
Agren-Wilsson A, Eklund A, Koskinen L-OD, Bergenheim AT,
Malm J. Brain energy metabolism and intracranial pressure in
idiopathic adult hydrocephalus syndrome. J Neurol Neurosurg
Psychiatr. 2005 Aug;76(8):1088–93.
37.
Malm J, Jacobsson J, Birgander R, Eklund A. Reference values for
CSF outflow resistance and intracranial pressure in healthy elderly.
Neurology. 2011 Mar 8;76(10):903–9.
38.
Børgesen SE. Conductance to outflow of CSF in normal pressure
hydrocephalus. Acta neurochir. 1984;71(1-2):1–45.
39.
Boon AJ, Tans JT, Delwel EJ, Egeler-Peerdeman SM, Hanlo PW,
Wurzer HA, et al. Dutch normal-pressure hydrocephalus study:
prediction of outcome after shunting by resistance to outflow of
cerebrospinal fluid. J Neurosurg. 1997 Nov;87(5):687–93.
40.
Egnor MR, Zheng L, Rosiello A, Gutman F, Davis R. A model of
pulsations in communicating hydrocephalus. Pediatr Neurosurg.
2002 Jun;36(6):281–303.
41.
Greitz D. Radiological assessment of hydrocephalus: new theories
and implications for therapy. Neurosurg Rev. 2004 Jul
1;27(3):145–65.
42.
Bateman GA. Pulse-wave encephalopathy: a comparative study of
the hydrodynamics of leukoaraiosis and normal-pressure
hydrocephalus. Neuroradiology. 2002 Sep;44(9):740–8.
43.
Bateman GA. Vascular compliance in normal pressure
hydrocephalus. AJNR Am J Neuroradiol. 2000 Oct;21(9):1574–85.
44.
Bateman GA, Levi CR, Schofield P, Wang Y, Lovett EC. The venous
manifestations of pulse wave encephalopathy: windkessel
dysfunction in normal aging and senile dementia. Neuroradiology.
2008 Jun 1;50(6):491–7.
45.
Graff-Radford NR, Knopman DS, Penman AD, Coker LH, Mosley
TH. Do systolic BP and pulse pressure relate to ventricular
enlargement? Eur J Neurol. 2013 Apr;20(4):720–4.
46.
Krauss JK, Regel JP, Vach W, Droste DW, Borremans JJ, Mergner
T. Vascular risk factors and arteriosclerotic disease in idiopathic
normal-pressure hydrocephalus of the elderly. Stroke.
1996;27(1):24–9.
47.
Eide PK, Sorteberg W. Preoperative spinal hydrodynamics versus
clinical change 1 year after shunt treatment in idiopathic normal
57
pressure hydrocephalus patients. Brit J Neurosurg. 2005
Dec;19(6):475–83.
48.
Eide PK, Sorteberg W. Simultaneous measurements of intracranial
pressure parameters in the epidural space and in brain parenchyma
in patients with hydrocephalus. J Neurosurg. 2010
Dec;113(6):1317–25.
49.
Qvarlander S, Lundkvist B, Koskinen L-OD, Malm J, Eklund A.
Pulsatility in CSF dynamics: pathophysiology of idiopathic normal
pressure hydrocephalus. J Neurol Neurosurg Psychiatr. 2013
Jul;84(7):735–41.
50.
Relkin N, Marmarou A, Klinge P, Bergsneider M, Black PM.
Diagnosing idiopathic normal-pressure hydrocephalus.
Neurosurgery. 2005 Sep;57(3 Suppl):4–16.
51.
Marmarou A, Bergsneider M, Klinge P, Relkin N, Black PM. The
value of supplemental prognostic tests for the preoperative
assessment of idiopathic normal-pressure hydrocephalus.
Neurosurgery. 2005 Sep;57(3 Suppl):17–28.
52.
Bech-Azeddine R, Waldemar G, Knudsen GM, Høgh P, Bruhn P,
Wildschiødtz G, et al. Idiopathic normal‐pressure hydrocephalus:
evaluation and findings in a multidisciplinary memory clinic. Eur J
Neurol. 2001 Nov 14;8(6):601–11.
53.
Malm J, Graff-Radford NR, Ishikawa M, Kristensen B, Leinonen V,
Mori E, et al. Influence of comorbidities in idiopathic normal
pressure hydrocephalus - research and clinical care. A report of the
ISHCSF task force on comorbidities in INPH. Fluids Barriers CNS.
2013;10(1):22.
54.
Bech-Azeddine R, Høgh P, Juhler M, Gjerris F, Waldemar G.
Idiopathic normal-pressure hydrocephalus: clinical comorbidity
correlated with cerebral biopsy findings and outcome of
cerebrospinal fluid shunting. J Neurol Neurosurg Psychiatr. 2007
Feb 1;78(2):157–61.
55.
Koivisto AM, Alafuzoff I, Savolainen S, Sutela A, Rummukainen J,
Kurki M, et al. Poor cognitive outcome in shunt-responsive
idiopathic normal pressure hydrocephalus. Neurosurgery. 2013
Jan;72(1):1–8.
56.
Eide PK, Sorteberg W. Changes in intracranial pulse pressure
amplitudes after shunt implantation and adjustment of shunt valve
opening pressure in normal pressure hydrocephalus. Acta
neurochir. 2008 Nov;150(11):1141.
58
57.
Tisell M, Tullberg M, Hellström P, Edsbagge M, Högfeldt M,
Wikkelsø C. Shunt surgery in patients with hydrocephalus and
white matter changes. J Neurosurg. 2011 May;114(5):1432–8.
58.
Klinge P, Hellström P, Tans J, Wikkelsö C, On behalf of the
European iNPH Multicentre Study Group. One-year outcome in the
European multicentre study on iNPH. Acta Neurol Scand. 2012
May 10;126(3):145–53.
59.
Malm J, Kristensen B, Karlsson T, Fagerlund M, Elfverson J,
Ekstedt J. The predictive value of cerebrospinal fluid dynamic tests
in patients with th idiopathic adult hydrocephalus syndrome. Arch
Neurol. 1995 Aug;52(8):783–9.
60.
Vanneste J, Augustijn P, Tan WF, Dirven C. Shunting normal
pressure hydrocephalus: the predictive value of combined clinical
and CT data. J Neurol Neurosurg Psychiatr. 1993 Mar 1;56(3):251–
6.
61.
Andrén K, Wikkelsø C, Tisell M, Hellström P. Natural course of
idiopathic normal pressure hydrocephalus. J Neurol Neurosurg
Psychiatr. 2014 Jul;85(7):806–10.
62.
Hellström P, Edsbagge M, Blomsterwall E, Archer T, Tisell M,
Tullberg M, et al. Neuropsychological effects of shunt treatment in
idiopathic normal pressure hydrocephalus. Neurosurgery. 2008
Sep 1;63(3):27–35.
63.
Bergsneider M, Black PM, Klinge P, Marmarou A, Relkin N.
Surgical management of idiopathic normal-pressure
hydrocephalus. Neurosurgery. 2005 Sep;57(3 Suppl):29–39.
64.
Wikkelsö C, Andersson H, Blomstrand C, Lindqvist G, Svendsen P.
Normal pressure hydrocephalus. Predictive value of the
cerebrospinal fluid tap-test. Acta Neurol Scand. 1986 Jun
1;73(6):566–73.
65.
Wikkelso C, Hellstrom P, Klinge PM, Tans JTJ, on behalf of the
European iNPH Multicentre Study Group. The European iNPH
Multicentre Study on the predictive values of resistance to CSF
outflow and the CSF Tap Test in patients with idiopathic normal
pressure hydrocephalus. J Neurol Neurosurg Psychiatr. 2013
May;84(5):62–8.
66.
Walchenbach R, Geiger E, Thomeer RT, Vanneste J. The value of
temporary external lumbar CSF drainage in predicting the outcome
of shunting on normal pressure hydrocephalus. J Neurol
Neurosurg Psychiatr. 2002 Apr;72(4):503–6.
59
67.
Ishikawa M, Hashimoto M, Mori E, Kuwana N, Kazui H. The value
of the cerebrospinal fluid tap test for predicting shunt effectiveness
in idiopathic normal pressure hydrocephalus. Fluids Barriers CNS.
2012;9(1):1.
68.
Haan J, Thomeer RT. Predictive value of temporary external
lumbar drainage in normal pressure hydrocephalus. Neurosurgery.
1988 Feb;22(2):388–91.
69.
Marmarou A, Young HF, Aygok GA, Sawauchi S, Tsuji O,
Yamamoto T, et al. Diagnosis and management of idiopathic
normal-pressure hydrocephalus: a prospective study in 151 patients.
J Neurosurg. 2005 Jun;102(6):987–97.
70.
Williams MA, Razumovsky AY, Hanley DF. Comparison of Pcsf
monitoring and controlled CSF drainage diagnose normal pressure
hydrocephalus. Acta Neurochir Suppl. 1998;71:328–30.
71.
Lenfeldt N, Hansson W, Larsson A, Birgander R, Eklund A, Malm J.
Three-day CSF drainage barely reduces ventricular size in normal
pressure hydrocephalus. Neurology. 2012 Jul 17;79(3):237–42.
72.
Eide PK, Brean A. Intracranial pulse pressure amplitude levels
determined during preoperative assessment of subjects with
possible idiopathic normal pressure hydrocephalus. Acta neurochir.
2006 Nov;148(11):1151–6.
73.
Wagshul ME, Eide PK, Madsen JR. The pulsating brain: A review of
experimental and clinical studies of intracranial pulsatility. Fluids
Barriers CNS. 2011;8(1):5.
74.
Eide PK, Sorteberg W. Diagnostic Intracranial Pressure Monitoring
and Surgical Management in Idiopathic Normal Pressure
Hydrocephalus. Neurosurgery. 2010 Jan;66(1):80–91.
75.
Prince M, Bryce R, Albanese E, Wimo A, Ribeiro W, Ferri CP. The
global prevalence of dementia: a systematic review and
metaanalysis. Alzheimers Dement. 2013 Jan;9(1):63–75.
76.
Vanneste J. Diagnosis and management of normal-pressure
hydrocephalus. J Neurol. 2000;247(1):5–14.
77.
Folstein MF, Folstein SE, McHugh PR. “Mini-mental state.” A
practical method for grading the cognitive state of patients for the
clinician. J Psychiatr Res. 1975 Nov;12(3):189–98.
78.
Kukull WA, Larson EB, Teri L, Bowen J, McCormick W,
Pfanschmidt ML. The Mini-Mental State Examination score and the
60
clinical diagnosis of dementia. J Clin Epidemiol. 1994
Sep;47(9):1061–7.
79.
Solana E, Sahuquillo J, Junqué C, Quintana M, Poca MA. Cognitive
disturbances and neuropsychological changes after surgical
treatment in a cohort of 185 patients with idiopathic normal
pressure hydrocephalus. Arch Clin Neuropsychol. 2012
May;27(3):304–17.
80.
Hoops S, Nazem S, Siderowf AD, Duda JE, Xie SX, Stern MB, et al.
Validity of the MoCA and MMSE in the detection of MCI and
dementia in Parkinson disease. Neurology. 2009 Nov
24;73(21):1738–45.
81.
Hellström P, Edsbagge M, Archer T, Tisell M, Tullberg M, Wikkelsø
C. The neuropsychology of patients with clinically diagnosed
idiopathic normal pressure hydrocephalus. Neurosurgery. 2007
Dec;61(6):1219–26.
82.
Hellström P, Klinge P, Tans J, Wikkelsø C. The neuropsychology of
iNPH: findings and evaluation of tests in the European multicentre
study. Clin Neurol Neurosurg. 2012 Feb;114(2):130–4.
83.
Hart RP, Kwentus JA. Psychomotor slowing and subcortical-type
dysfunction in depression. J Neurol Neurosurg Psychiatr. 1987
Oct;50(10):1263–6.
84.
Cummings JL. Subcortical dementia. Neuropsychology,
neuropsychiatry, and pathophysiology. Br J Psychiatry. 1986 Dec
1;149(6):682–97.
85.
Filley CM. The behavioral neurology of cerebral white matter.
Neurology. 1998;50(6):1535–40.
86.
Thomas G, Mcgirt MJ, Woodworth G, Heidler J, Rigamonti D,
Hillis AE, et al. Baseline neuropsychological profile and cognitive
response to cerebrospinal fluid shunting for idiopathic normal
pressure hydrocephalus. Dement Geriatr Cogn Disord. 2005;20(23):163–8.
87.
Ogino A, Kazui H, Miyoshi N, Hashimoto M, Ohkawa S, Tokunaga
H, et al. Cognitive impairment in patients with idiopathic normal
pressure hydrocephalus. Dement Geriatr Cogn Disord.
2006;21(2):113–9.
88.
Iddon JL, Morgan DJR, Loveday C, Sahakian BJ, Pickard JD.
Neuropsychological profile of young adults with spina bifida with or
without hydrocephalus. J Neurol Neurosurg Psychiatr. 2004 Aug
61
1;75(8):1112–8.
89.
Mataró M, Poca MA, Del Mar Matarín M, Catalan R, Sahuquillo J,
Galard R. CSF galanin and cognition after shunt surgery in normal
pressure hydrocephalus. J Neurol Neurosurg Psychiatr. 2003
Sep;74(9):1272–7.
90.
Miyoshi N, Kazui H, Ogino A, Ishikawa M, Miyake H, Tokunaga H,
et al. Association between cognitive impairment and gait
disturbance in patients with idiopathic normal pressure
hydrocephalus. Dement Geriatr Cogn Disord. 2005;20(2-3):71–6.
91.
Donnet A, Schmitt A, Dufour H, Giorgi R, Grisoli F. Differential
patterns of cognitive impairment in patients with aqueductal
stenosis and normal pressure hydrocephalus. Acta neurochir. 2004
Dec 1;146(12):1301–8.
92.
Gustafson L, Hagberg B. Recovery in hydrocephalic dementia after
shunt operation. J Neurol Neurosurg Psychiatr. 1978 Oct
1;41(10):940–7.
93.
Tsakanikas D, Katzen H, Ravdin LD, Relkin NR. Upper extremity
motor measures of Tap Test response in Normal Pressure
Hydrocephalus. Clin Neurol Neurosurg. 2009 Nov;111(9):752–7.
94.
Nowak DA, Topka HR. Broadening a classic clinical triad: The
hypokinetic motor disorder of normal pressure hydrocephalus also
affects the hand. Exp Neurol. 2006 Mar 1;198(1):81–7.
95.
Merten T, Merten T. [Neuropsychology of normal pressure
hydrocephalus]. Nervenarzt. 1999 Jun;70(6):496–503.
96.
Tsakanikas D, Relkin N. Normal pressure hydrocephalus. Semin
Neurol. 2007 Feb;27(1):58–65.
97.
Duinkerke A, Williams MA, Rigamonti D, Hillis AE. Cognitive
recovery in idiopathic normal pressure hydrocephalus after shunt.
Cogn Behav Neurol. 2004 Sep;17(3):179–84.
98.
Foss T, Eide PK, Finset A. Intracranial pressure parameters in
idiopathic normal pressure hydrocephalus patients with or without
improvement of cognitive function after shunt treatment. Dement
Geriatr Cogn Disord. 2007;23(1):47–54.
99.
Hellström P. The neuropsychology of idiopathic normal pressure
hydrocephalus. Institute of Neuroscience and Physiology.
Department of Clinical Neuroscience and Rehabilitation; 2011.
62
100.
Hellstrom P, Klinge P, Tans J, Wikkelso C. A new scale for
assessment of severity and outcome in iNPH. Acta Neurol Scand.
2012 Oct;126(4):229–37.
101.
Dikmen S, Heaton RK, Grant I, Temkin NR. Test-retest reliability
and practice effects of expanded Halstead-Reitan
Neuropsychological Test Battery. J Int Neuropsychol Soc. 1999 May
1;5(4):346–56.
102.
Katzen H, Ravdin LD, Assuras S, Heros R, Kaplitt M, Schwartz TH,
et al. Postshunt cognitive and functional improvement in idiopathic
normal pressure hydrocephalus. Neurosurgery. 2011
Feb;68(2):416–9.
103.
Solana E, Poca MA, Sahuquillo J, Benejam B, Junqué C, Dronavalli
M. Cognitive and motor improvement after retesting in normalpressure hydrocephalus: a real change or merely a learning effect? J
Neurosurg. 2010 Feb;112(2):399–409.
104.
Schlegel RE, Gilliland K. Development and quality assurance of
computer-based assessment batteries. Arch Clin Neuropsych. 2007
Feb;22 Suppl 1(Suppl 1):49–61.
105.
Zygouris S, Tsolaki M. Computerized Cognitive Testing for Older
Adults: A review. Am J Alzheimers Dis Other Demen. 2014
Feb;20:1–16.
106.
Lezak M, Howieson D, Bigler E, Tranel D. Neuropsychological
assessment. New York: Oxford University Press; 2012 .
107.
Wild K, Howieson D, Webbe F, Seelye A, Kaye J. Status of
computerized cognitive testing in aging: a systematic review.
Alzheimers Dement. 2008 Nov;4(6):428–37.
108.
Canini M, Battista P, Rosa Della PA, Catricalà E, Salvatore C,
Gilardi MC, et al. Computerized Neuropsychological Assessment in
Aging: Testing Efficacy and Clinical Ecology of Different Interfaces.
Comput Math Methods Med. 2014;2014(1):1–13.
109.
Terwee CB, Mokkink LB. Measurement in Medicine. New York:
Cambridge University Press; 2011.
110.
Chaytor N, Schmitter-Edgecombe M. The ecological validity of
neuropsychological tests: a review of the literature on everyday
cognitive skills. Neuropsychol Rev. 2003 Dec;13(4):181–97.
111.
Agren-Wilsson A, Roslin M, Eklund A, Koskinen L-OD, Bergenheim
AT, Malm J. Intracerebral microdialysis and CSF hydrodynamics in
63
idiopathic adult hydrocephalus syndrome. J Neurol Neurosurg
Psychiatr. 2003 Feb 1;74(2):217–21.
112.
Ursino M. A mathematical study of human intracranial
hydrodynamics. Part 1--The cerebrospinal fluid pulse pressure. Ann
Biomed Eng. 1988;16(4):379–401.
113.
Ursino M. A mathematical study of human intracranial
hydrodynamics. Part 2--Simulation of clinical tests. Ann Biomed
Eng. 1988;16(4):403–16.
114.
Ursino M, Lodi CA. Interaction among autoregulation, CO2
reactivity, and intracranial pressure: a mathematical model. Am J
Physiol. 1998 May 1;274(5 Pt 2):1715–28.
115.
Ursino M, Giulioni M, Lodi CA. Relationships among cerebral
perfusion pressure, autoregulation, and transcranial Doppler
waveform: a modeling study. J Neurosurg. 1998 Aug 1;89(2):255–
66.
116.
Farrell J. Java Programming. Boston: Cengage Learning; 2013.
117.
Yam C-S. Using macromedia flash for electronic presentations: a
new alternative. AJR Am J Roentgenol. 2006 Aug;187(2):209–17.
118.
Smith G, Eberhard Nyman G, Hentschel U. Manual till CWT serialt färgordtest. Stockholm: Psykologiförlaget; 1986.
119.
Reitan R. Validity of the Trail Making Test as an indicator of
organic brain damage. Percept Mot Skills. 1958;8:271–6.
120.
Morris J, Heyman A, Mohs R, Hughes J, Van G. The Consortium to
Establish a Registry for Alzheimer's Disease (CERAD). Part I.
Clinical and Neuropsychological assessment of Alsheimer's disease.
Neurology. 1989;39:1159-65.
121.
Lenfeldt N, Larsson A, Nyberg L, Andersson M, Birgander R,
Eklund A, et al. Idiopathic normal pressure hydrocephalus:
increased supplementary motor activity accounts for improvement
after CSF drainage. Brain. 2008 Nov;131(Pt 11):2904–12.
122.
Yesavage JA, Brink TL, Rose TL, Lum O, Huang V, Adey M, et al.
Development and validation of a geriatric depression screening
scale: a preliminary report. J Psychiatr Res. 1982;17(1):37–49.
123.
Altman DG, Bland JM. Measurement in medicine: the analysis of
method comparison studies. The statistician. 1983;(32):307–17.
64
124.
McMahon CJ, McDermott P, Horsfall D, Selvarajah JR, King AT,
Vail A. The reproducibility of transcranial Doppler middle cerebral
artery velocity measurements: implications for clinical practice.
Brit J Neurosurg. 2007 Feb;21(1):21–7.
125.
Kristiansson H, Nissborg E, Bartek J, Andresen M, Reinstrup P,
Romner B. Measuring elevated intracranial pressure through
noninvasive methods: a review of the literature. J Neurosurg
Anesthesiol. 2013 Oct;25(4):372–85.
126.
Zweifel C, Czosnyka M, Carrera E, de Riva N, Pickard JD,
Smielewski P. Reliability of the Blood Flow Velocity Pulsatility
Index for Assessment of Intracranial and Cerebral Perfusion
Pressures in Head-Injured Patients. Neurosurgery. 2012
Oct;71(4):853–61.
127.
de Riva N, Budohoski KP, Smielewski P, Kasprowicz M, Zweifel C,
Steiner LA, et al. Transcranial Doppler Pulsatility Index: What it is
and What it Isn’t. Neurocrit Care. 2012 Aug 1;17(1):58–66.
128.
O'Rourke MF, Hashimoto J. Mechanical factors in arterial aging: a
clinical perspective. J Am Coll Cardiol. 2007 Jul 3;50(1):1–13.
129.
Wåhlin A, Ambarki K, Birgander R, Alperin N, Malm J, Eklund A.
Assessment of craniospinal pressure-volume indices. Am J
Neuroradiol. 2010 Oct;31(9):1645–50.
130.
Czosnyka M, Czosnyka ZH, Whitfield PC, Donovan T, Pickard JD.
Age dependence of cerebrospinal pressure-volume compensation in
patients with hydrocephalus. J Neurosurg. 2001 Mar;94(3):482–6.
131.
Scuteri A, Nilsson PM, Tzourio C, Redon J, Laurent S.
Microvascular brain damage with aging and hypertension. J
Hypertens. 2011 Aug;29(8):1469–77.
132.
Wåhlin A, Ambarki K, Birgander R, Malm J, Eklund A. Intracranial
pulsatility is associated with regional brain volume in elderly
individuals. Neurobiol Aging. 2014 Feb;35(2):365–72.
133.
Mitchell GF, van Buchem MA, Sigurdsson S, Gotal JD, Jonsdottir
MK, Kjartansson O, et al. Arterial stiffness, pressure and flow
pulsatility and brain structure and function: the Age,
Gene/Environment Susceptibility - Reykjavik Study. Brain. 2011
Nov 1;134(11):3398–407.
134.
Mok V, Ding D, Fu J, Xiong Y, Chu WWC, Wang D, et al.
Transcranial Doppler ultrasound for screening cerebral small vessel
disease: a community study. Stroke. 2012 Oct;43(10):2791–3.
65
135.
Eide PK, Brean A. Lumbar cerebrospinal fluid pressure waves
versus intracranial pressure waves in idiopathic normal pressure
hydrocephalus. Brit J Neurosurg. 2006 Dec;20(6):407–14.
136.
Atkinson G, Nevill AM. Statistical methods for assessing
measurement error (reliability) in variables relevant to sports
medicine. Sports Med. 1998 Oct;26(4):217–38.
137.
Momjian S, Owler BK, Czosnyka ZH, Czosnyka M, Pena A, Pickard
JD. Pattern of white matter regional cerebral blood flow and
autoregulation in normal pressure hydrocephalus. Brain. 2004 May
1;127(Pt 5):965–72.
138.
Lenfeldt N, Larsson A, Nyberg L, Birgander R, Eklund A, Malm J.
Diffusion tensor imaging reveals supplementary lesions to frontal
white matter in idiopathic normal pressure hydrocephalus.
Neurosurgery. 2011 Jun;68(6):1586–93.
139.
Alexander GE, DeLong MR, Strick PL. Parallel organization of
functionally segregated circuits linking basal ganglia and cortex.
Annu Rev Neurosci. 1986;9:357–81.
140.
Middleton FA, Strick PL. Basal Ganglia Output and Cognition:
Evidence from Anatomical, Behavioral, and Clinical Studies. Brain
Cogn. 2000 Mar;42(2):183–200.
141.
Tullberg M, Hellström P, Piechnik SK, Starmark J-E, Wikkelsö C.
Impaired wakefulness is associated with reduced anterior cingulate
CBF in patients with normal pressure hydrocephalus. Acta Neurol
Scand. 2004 Nov 1;110(5):322–30.
142.
Savolainen S, Laakso MP, Paljärvi L, Alafuzoff I, Hurskainen H,
Partanen K, et al. MR imaging of the hippocampus in normal
pressure hydrocephalus: correlations with cortical Alzheimer's
disease confirmed by pathologic analysis. AJNR Am J Neuroradiol.
2000 Feb;21(2):409–14.
143.
Marmarou A, Bergsneider M, Relkin N, Klinge P, Black PM.
Development of guidelines for idiopathic normal-pressure
hydrocephalus: introduction. Neurosurgery. 2005 Sep;57(3
Suppl):1–3.
144.
Boon AJ, Tans JT, Delwel EJ, Egeler-Peerdeman SM, Hanlo PW,
Wurzer JA, et al. Dutch normal pressure hydrocephalus study:
baseline characteristics with emphasis on clinical findings. Eur J
Neurol. 1997 Jan;4(1):39–47.
145.
Krauss JK, Droste DW, Vach W, Regel JP, Orszagh M, Borremans
66
JJ, et al. Cerebrospinal fluid shunting in idiopathic normalpressure hydrocephalus of the elderly: effect of periventricular and
deep white matter lesions. Neurosurgery. 1996 Aug;39(2):292–9.
146.
Owler BK, Momjian S, Czosnyka Z, Czosnyka M, Péna A, Harris NG,
et al. Normal pressure hydrocephalus and cerebral blood flow: a
PET study of baseline values. J Cereb Blood Flow Metab. 2004
Jan;24(1):17–23.
147.
Wikkelsö C, Andersson H, Blomstrand C, Lindqvist G. The clinical
effect of lumbar puncture in normal pressure hydrocephalus. J
Neurol Neurosurg Psychiatr. 1982 Jan;45(1):64–9.
67