Cessation of Diastolic Cerebral Blood Flow Velocity: The Role of
Critical Closing Pressure
Georgios V. Varsos • Hugh K. Richards • Magdalena Kasprowicz • Matthias Reinhard • Peter Smielewski • Ken
M. Brady • John D. Pickard •
Marek Czosnyka
Published online: 19 November 2013
Springer Science+Business Media New York 2013
Abstract
Background Reducing cerebral perfusion pressure
(CPP) below the lower limit of autoregulation
(LLA) causes cerebral blood flow (CBF) to become
pressure passive. Further reductions in CPP can
cause cessation of CBF during diastole. We
hypothesized that zero diastolic flow velocity (FV)
occurs when diastolic blood pressure becomes less
than the critical closing pressure (CrCP). Methods
We retrospectively analyzed studies of 34 rabbits
with CPP below the LLA, induced with
pharmacologic sympathectomy (N = 23) or
cerebrospinal fluid infusion (N = 11). Basilar artery
blood FV and cortical Laser Doppler Flow (LDF)
were monitored. CrCP was trended using a model
of cerebrovascular impedance. The diastolic closing
margin (DCM) was monitored as the difference
123
G. V. Varsos (&) H. K. Richards P. Smielewski J. D.
Pickard M. Czosnyka
Division of Neurosurgery, Department of Clinical
Neurosciences, University of Cambridge Biomedical Campus,
Cambridge CB20QQ, UK e-mail: gv249@cam.ac.uk
M. Kasprowicz
Institute of Biomedical Engineering and Instrumentation,
Wroclaw University of Technology, Wrocław, Poland
M. Reinhard
Department of Neurology, University of Freiburg, Freiburg,
Germany
K. M. Brady
Baylor College of Medicine, Texas Children’s Hospital,
Houston, TX, USA
M. Czosnyka
Institute of Electronic Systems, Warsaw University of
Technology, Warsaw, Poland
between diastolic blood pressure and CrCP. LDF
was recorded for DCM values greater than and less
than zero. Results Arterial hypotension caused a
reduction of CrCP (p < 0.001), consistent with
decreased wall tension (p < 0.001) and a drop in
intracranial pressure (ICP; p = 0.004).
Cerebrospinal infusion caused an increase of CrCP
(p = 0.002) accounted for by increasing ICP (p <
0.001). The DCM was compromised by either
arterial hypotension or intracranial hypertension (p
< 0.001 for both). When the DCM reached zero,
diastolic FV ceased for a short period during each
heart cycle (R = 0.426, p < 0.001). CBF pressure
passivity accelerated when DCM decreased below
zero (from 1.51 ± 0.51 to 2.17 ± 1.17 %
DLDF/DmmHg; mean ± SD; p = 0.010).
Conclusions The disappearance of diastolic CBF
below LLA can be explained by DCM reaching
zero or negative values. Below this point the
decrease in CBF accelerates with further
decrements of CPP.
41
Keywords Brain ischemia Cerebral autoregulation
Vanishing diastolic flow velocity
Critical closing pressure Wall tension
Introduction
The brain is protected against ischemic injury
during low cerebral perfusion pressure (CPP) by
cerebrovascular pressure autoregulation [1–4]. The
lower limit of autoregulation (LLA) is the CPP
below which cerebrovascular dilatation fails to
maintain cerebral blood flow (CBF), which then
becomes passive to changes in CPP [2, 5].
With further reduction of CPP an additional
threshold can be observed, below which CBF
occurs only during systole, and is absent during
diastole, despite the diastolic CPP being >0. This
indicates a discontinuity of arterial blood flow
during a single heart cycle and can be observed
with transcranial Doppler ultrasonography (TCD)
as cessation of diastolic blood flow velocity (FVd);
an example depicting this phenomenon in a case of
arterial hypotension is demonstrated in Fig. 1.
In clinical terms, vanishing FVd in the TCD
signal has been associated with imminent cerebral
circulatory arrest and brain death in both adults [6,
7] and children [8]. Moreover, a low magnitude of
FVd obtained with TCD on admission has been
demonstrated to be an accurate detector of patients
at risk for a secondary neurological deterioration
after mild to moderate traumatic brain injury [9].
The phenomenon of disappearing FVd has also
been described in infants after hypothermic
123
Fig. 1 a Waveforms of ABP and cerebral FV, depicting an event of arterial hypotension from baseline to low ABP. b Waveform of FV during the
‘baseline’ part of a. c Waveform of FV during the ‘low ABP’ part of a, demonstrating the vanishing diastolic FV during this case of arterial
hypotension
Neurocrit Care (2014) 20:40–48
circulatory arrest for cardiac surgery, observed to
persist up to 4 h after surgery [10–12]. All the
above described scenarios signify the important role
of FVd in clinical reality, prompting further
investigations for enhancing the understanding of
the underlying mechanisms affecting diastolic
blood flow. We hypothesized that the disappearance
of FVd reflects a fall in arterial blood pressure
(ABP) during diastole below the critical closing
pressure (CrCP).
CrCP was first introduced and theoretically
described by Burton, who stated that small brain
vessels can collapse when local blood pressure is
reduced to a critical value [13]. According to
Burton’s model, CrCP equals the sum of
intracranial pressure (ICP) and vascular wall
tension (WT) [13, 14]. Therefore, in addition to the
obvious clinical potential of knowing a critical
threshold of ABP, CrCP may also provide insight
into the behavior of vascular tone in pathological
states [15].
Classic methods for estimating CrCP from TCD
are burdened with assumptions of linearity between
ABP and CBF [16–18]. These assumptions can
cause underestimation of CrCP, and occasionally
render negative, noninterpretable values in various
pathologies, limiting the clinical worth of the
threshold [19]. We recently introduced a method for
estimating CrCP based on a model of
cerebrovascular impedance, which eliminates the
negative values’ drawback of past methods [15].
In this study, we used this method of CrCP
determination in experimental models of low CPP
Neurocrit Care (2014) 20:40–48 42
to evaluate the behavior of the cerebrovascular bed
below the LLA. Based on the fact that mean CBF
ceases when mean ABP goes below the CrCP
threshold, we hypothesized that diastolic ABP
(ABPd) less than CrCP would result in a cessation
of diastolic CBF observed as zero FVd on the TCD
monitor. For this reason, we introduce a new term
called diastolic closing margin (DCM) depicting
the difference between ABPd and CrCP. By
exploring what is happening below the LLA, we
investigated whether there is another threshold in
the CPP–CBF relationship, below which the

123
43
decrease in CBF past the LLA would become faster,
accelerating then the risk for ischemia. We further
hypothesized that below this new threshold the
systoledependent CBF would be associated with
further increased passivity of TCD mean flow
velocity (FV) and Laser Doppler Flow (LDF) to
changes in mean ABP. The exact mechanisms of
this phenomenon and whether it differs in
circulatory failure because of low systemic ABP or
high ICP are unclear. Knowledge of these
mechanisms can enhance the understanding of TCD
findings while providing a physiological
explanation of this dramatic hemodynamic
phenomenon. In that respect, we tested the above
mentioned hypotheses in rabbits with
experimentally induced reduction of CPP, either
with arterial hypotension caused by chemical
sympathectomy, or with intracranial hypertension
caused by cerebrospinal fluid infusion.
Materials and Methods
The animal experiments in this study were carried
out under UK Home Office license, with
permission from the Institutional Animal Care and
Use Committee at Cambridge University and in
accordance with the standards set by the UK
Animals Scientific Procedures Act of 1986.
Recordings from 34 experiments of male New
Zealand White rabbits (weighting from 2.7 to 3.7
kg) were performed [20] including high-resolution
sampling of ABP, ICP, and FV. The nature of this
study is retrospective, and the experimental
protocols have been published in detail in the past
[15, 21].
ABP was monitored through a cannula in a
femoral artery with a catheter advancing to lie high
in the dorsal aorta (GaelTec, Dunvegan, UK). CBF
was approximated by monitoring FV with an 8
MHz pulsed Doppler ultrasound probe (PcDop 842,
SciMed, Bristol, UK) positioned over the basilar
artery. Access to the artery was made through a
posterior frontal burr hole (7 mm diameter) at the
bregma to the left of the midline. A second burr
hole of 1.5 mm in diameter was made over the
contralateral cerebral hemisphere close to the
bregma, and a Laser Doppler probe of 1 mm in
diameter (Moor Instruments, Axbridge, Devon,
UK) was placed epidurally in order to measure
cerebral LDF and validate the TCD measurements.
ICP was monitored through an intraparenchymal
123
Neurocrit Care (2014) 20:40–48
microsensor (Codman and Shurtleff, Raynham,
MA, USA). As part of multi-purpose experiments
of cerebral hemodynamics, lumbar laminectomy
was also performed in the group of arterial
hypotension, with a catheter inserted into the
lumbar subarachnoid space, sealed with
cyanoacrylate. The laminectomy was used in order
to perform two lumbar infusions of Hartmann’s
solution at baseline and decreased ABP, thus before
and after the maneuver of arterial hypotension was
performed. Sufficient time was given after the first
lumbar infusion in order for the ABP baseline to
return to normal, in order not to affect then the
subsequent arterial hypotension maneuver. These
two maneuvers of lumbar infusion were not studied
in the group of arterial hypotension, as we used the
second group of rabbits specifically for intracranial
hypertension purposes.
After recovery from surgery and stabilization of
ABP, temperature and arterial blood gas tensions,
recordings of ABP, ICP, basilar artery FV and LDF
were used to establish a baseline prior to decreasing
ABP. Sympathetic blockade to induce hypotension
was studied, by administrating the short-acting
ganglion blocking drug trimetaphan (Arfonad,
Roche, Welwyn, UK) at a dose of 0.5 mg/kg i.v.
When necessary, additional doses of trimetaphan
were administered to maintain the desired ABP
level.
In the group of rabbits with solely increasing ICP
(N = 11), the same principles of the experimental
protocols were used. ICP was increased by infusion
of
Hartmann’s solution into the lumbar subarachnoid
space at a stepwise increasing rate from 0.1 to 2
mL/min producing gradual rise in ICP to 70–80
mmHg. The experimental protocol for studying
intracranial hypertension was designed for a better
understanding of the cerebral circulation in cases of
high ICP [22]. For this reason, periods of already
elevated ICP (with a previous infusion resulting in
baseline ICP around 40 mmHg) to terminal rise in
ICP were analyzed for the sake of integrity of LDF
readings (optical couplings of fiberoptic were often
shifted during recording resulting into a change in
the output signal level). This also explains why the
first measurements of baseline ICP of the second
group where considerably higher than the ones in
the arterial hypotension group (see Table 1).
Neurocrit Care (2014) 20:40–48 44
Data Acquisition and Analysis Calculation of CrCP, WT, and DCM

ABP, ICP, FV, and LDF were converted into digital Impedance-modeled CrCP was calculated as [15,
samples using an analog-to-digital converter with a 23]:
sampling frequency set to 50 Hz. The recorded CPP
signals were then analyzed using our own software CrCP ¼ ABP
for clinical data processing (ICM+;
http://www.neurosurg.cam.ac.uk/icmplus). qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiff
Heart rate was calculated using the spectral iffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
position of the peak associated with the first ðmmHgÞ
ffiffiffiffiffiffiffiffiffiffiffi2
harmonic of ABP. Mean values of ABP, ICP, FV,
ðCVR Ca HR 2pÞ þ1
and CPP were calculated in ICM+ by averaging
values in a 10 s time window. The minimal values
where CPP is mean cerebral perfusion pressure
of ABP and FV from consecutive 2 s epochs were
(estimated as CPP = ABP - ICP); CVR is
calculated and treated as end-diastolic components.
cerebrovascular resistance; Ca is compliance of
These components were then averaged over 10 s to
cerebral arterial bed, while HR denotes heart rate
give the mean values for end-ABPd and FV, termed
(beats/s).
as ABPd and FVd, respectively. Individual LDF
Although CVR and Ca cannot be measured
recordings were normalized and expressed as
directly, they can be estimated using TCD blood FV
fraction of baselines (highest ABP in the
and ABP or CPP waveforms according to an
hypotensive group or baseline CPP in
Table 1 Mean values and SD’s of measured and calculated variables during arterial hypotension and intracranial hypertension
Arterial N = 23) Intracranial hypertension (N =
hypotension ( 11)
Baseline ABP Low ABP Baseline ICP High ICP Level of
Level of significance significance
(p value) (p value)
ABP (mmHg) 87.00 ± 21.57 51.53 ± 14.18 p < 0.001 97.26 ± 15.56 108.17 ± 26.19 p = 0.252 (NS)
ABPd (mmHg) 74.30 ± 20.25 41.72 ± 12.22 p < 0.001 84.87 ± 14.85 95.51 ± 22.78 p = 0.212 (NS)
ICP (mmHg) 13.35 ± 7.11 11.90 ± 5.82 p = 0.004 43.23 ± 16.94 78.10 ± 20.33 p < 0.001
CPP (mmHg) 73.64 ± 22.72 39.63 ± 15.65 p < 0.001 54.03 ± 17.18 30.07 ± 14.68 P = 0.002
FV (cm/s) 16.83 ± 5.72 10.50 ± 3.68 p < 0.001 30.38 ± 11.26 8.25 ± 4.87 p < 0.001
FVs (cm/s) 37.62 ± 11.44 34.96 ± 11.19 p = 0.086 (NS) 43.08 ± 14.77 22.61 ± 13.26 p = 0.003
FVd (cm/s) 7.15 ± 3.55 0.40 ± 1.11 p < 0.001 22.16 ± 9.35 0.42 ± 0.84 p < 0.001
CrCP (mmHg) 57.79 ± 15.40 43.53 ± 11.88 p < 0.001 61.80 ± 18.21 95.52 ± 24.05 p = 0.002
WT (mmHg) 63.39 ± 20.15 37.06 ± 14.85 p < 0.001 36.30 ± 19.36 25.05 ± 14.63 p = 0.141 (NS)
DCM (mmHg) 16.51 ± 10.80 -1.81 ± 1.77 p < 0.001 23.07 ± 11.32 0.00 ± 3.99 p < 0.001
LDF (%) 78.77 ± 35.21 49.38 ± 30.10 p < 0.001 67.95 ± 48.61 41.98 ± 40.95 p = 0.010
ABP mean arterial blood pressure, ABPd diastolic ABP, ICP intracranial pressure, CPP cerebral perfusion pressure, FV mean flow velocity, FVs
systolic FV, FVd diastolic FV, CrCP critical closing pressure, WT vascular wall tension, DCM diastolic closing margin, LDF Laser Doppler
Flow, normalized in % units
Results are presented in mean value ± standard deviation (SD) format; p values for univariate tests were used; NS non-significant result
the spinal infusion group) after subtracting the algorithm described in previous studies [24, 25].
values obtained at demise (biologic zero). In that The vascular WT was calculated as the
way, changes in LDF can expressed then non- difference CrCP - ICP [14].
unitized as percentages in the range of 0–100 %. DCM was calculated as:
DCM ¼ ABPd CrCP mmHgð Þ

123
45 Neurocrit Care (2014) 20:40–48
Statistical Methods < 0.001) due to dilation of vessels compensating for
the decrease in ABP prior to loss of autoregulation
Statistical analysis of the data was performed with and continuing past the LLA, although with a lower
IBM SPSS Statistics 20 package. The analysis rate.
consisted of comparing changes in an array of
parameters, averaged from steady periods of Group II
baseline ABP to averaged periods of low and stable
ABP following arterial hypotension and in the same After lumbar spinal infusion, ICP was significantly
manner from baseline ICP to high and stable increased by 80.66 % (p < 0.001), leading to a
plateau ICP following intracranial hypertension significant decrease in CPP of 44.35 % (p < 0.001).
(Table 1). Results are presented in mean value ± ABP and ABPd showed a slight upward trend that
standard deviation (SD) format. Normal was not significant (p = 0.252 and 0.212,
distribution was established with the Shapiro–Wilk respectively). Exhausted autoregulation was
test and T tests were used to examine the confirmed by observing a significant reduction in
significance of each sample comparison with the CBF: mean FV was reduced by 72.8 % (p < 0.001),
level of significance being set at 0.05. The criterion FVs was reduced by 47.52 % (p = 0.003), and LDF
for vanishing diastolic FV was FVd to be equal or decreased by 38.22 % (p = 0.010). FVd was ablated
lower to 1.0 cm/s. at high ICP, after a decrease of 98.10 % (p < 0.001).
CrCP increased by 54.56 % (p < 0.001), which in
absence of change in ABPd resulted in a decrement
and exhaustion of DCM (p < 0.001). Examples of
Results
the behavior of CrCP, ABPd, FV, and FVd during
intracranial hypertension are demonstrated in Fig.
Measurements and calculations for both
2b. The vascular WT showed a trend for decrease
experimental groups, with hypotension and with
that was not significant (p = 0.141).
elevated ICP, are shown in Table 1, and considered
in turn below.
Exhaustion of DCM Accelerates the Decrease in
Group I
CBF
Inducing trimetaphan reduced mean ABP by 40.77
% and CPP by 46.18 %. Exhausted autoregulation
(CPP going below LLA) was confirmed by a
decrease in CBF: mean FV decreased by 37.61 %
(p < 0.001) and LDF decreased by 37.31 % (p <
0.001). Further, also ICP was decreased during
hypotension by 10.86 %, (p < 0.001) suggesting
passive transmission of ABP changes to ICP.
ABPd was also decreased by 43.85 %, whereas
CrCP was decreased by only 24.67 % (p < 0.001 for
both). This disparity led to a significant DCM
reduction of 110.96 % (p < 0.001) and a negative
DCM was observed at this degree of hypotension.
Synchronous to the reduction of the DCM, FVd was
reduced by 94.41 % and was completely ablated at
low ABP. An example of behaviors of CrCP, ABPd,
mean FV, and FVd is demonstrated in the Fig. 2a.
Decreases in DCM were significantly positively
correlated to decrement of CPP (R = 0.806, p <
0.001). There was a trend for decreased systolic FV
(FVs) that was not significant (p = 0.086). The
vascular WT decreased significantly by 41.54 % (p

123
Neurocrit Care (2014) 20:40–48 46
Overall, with decreasing
DCM, FVd decreased (R =
0.426, p < 0.001, pooled data
from both groups: N = 34) and
became zero with depletion of
the DCM (Figs. 3a, b). In
addition to being associated
with a zero FVd state, a DCM
B 0 was associated with a
change in the rate of pressure
autoregulation and decrease in
CBF. This is seen when
plotting normalized LDF as a
function of CPP and
comparing the slope below the
point when DCM B 0, and
between zero DCM and the
LLA. With positive DCM, the
slope of this static
autoregulation curve was 1.51
± 0.51 (% change LDF)/
Fig. 2 a During arterial
hypotension both CrCP and
diastolic values of ABPd
decrease. Mean FV also
decreases with its diastolic
values (FVd) following the
same pattern and then vanishing
when ABPd falls below the
CrCP threshold. b During
intracranial hypertension CrCP
is increasing while ABPd
remains static at the same levels.
Following FV, FVd is
decreasing and then vanishes
when CrCP overcomes ABPd
(mmHg), while with DCM B 0 the slope increased Fig. 3 a Scatter plot depicting the phenomenon of decreasing
to 2.17 ± 1.17 (% change LDF)/(mmHg), a DCM being linked to decreasing FVd. When DCM is getting
lower than zero (vertical line), FVd vanishes, taking zero
difference that was statistically significant (p = (ceased flow-horizontal line) or negative values (reversed
0.010; paired samples T test). This increase in the flow). b Decrement of mean FVd (averaged from the full
rate of change of LDF across CPP is indicative of cohort of rabbits) with decreasing DCM. Error bars present
95 % confidence limit
increased pressure passivity associated with
exhaustion of the DCM, illustrated from an
individual rabbit in Fig. 4a. Static autoregulation
plots were only analyzed in the hypotensive group, Using pooled data from both groups, static
as limited data below the DCM was available in the autoregulation curves of mean FV as a function of
group with elevated ICP. CPP also demonstrated a similar inflection with
DCM. Above DCM = 0, FV plotted across CPP had
a slope of 1.35(% change FV)/(mmHg) (1.48–8.30;
median, IQR). Below the DCM = 0 point FV
plotted across CPP had a slope of 4.61(% change
FV)/(mmHg) (0.77–2.02; median, IQR; p = 0.018,
N = 34). A static plot of pressure autoregulation

123
47
with FV plotted across CPP is shown with the DCM
highlighted in Fig. 4b.
Discussion
The main finding of this study is a demonstration
that ABPd less than CrCP is associated with loss of
measureable blood FV during diastole when TCD is
used. Loss of CBF during diastole is associated
with a measurable increase in pressure passivity. By
monitoring CrCP, the DCM can also be trended and
may represent an important clinical threshold in
patients with hypotension, elevated ICP, or with
pathologically increased vascular WT.
When CPP decreases below LLA, CBF is known
to become pressure passive as cerebrovascular
dilation is no longer able to compensate for
reductions in CPP [1, 5]. However, the pressure
passivity of both LDF and FV measurements in this
study increased with exhaustion of the DCM. The
point of DCM exhaustion therefore appears to be an
important landmark of CPP in addition to the LLA,
likely to be associated with significant cerebral
ischemia.
Vascular Wall Tension Below the LLA
With intact autoregulation, a decrease in ABP
causes compensatory arterial dilation [1, 5] and a
decrease in active vasomotor tone, manifested as a
decrease in WT [15]. Below LLA, WT was seen to
decrease further, correlated with decrements of the
DCM, evidence that the LLA is not necessarily the
threshold of maximal arteriolar vasodilatation.
When CPP is further reduced below the point where
DCM = 0, the correlation between DCM and WT is
inversed, and WT increases then with decrements
of DCM (Fig. 5). In theory, this behavior of the
vasomotor tone reflects the collapse of cerebral
microvasculature during systole as well as diastole
[5].
Possible Clinical Consequences
The role of the DCM mechanism could be to
provide supplementary information on the patients’
hemodynamics based on the concept of CrCP,
which can follow a diagnosis made with TCD
measurements. In this instance, when for example
zero diastolic flow is seen with TCD, it may
indicate an imminent cerebral circulatory arrest,
123
Neurocrit Care (2014) 20:40–48
which in some situations may be reversible [10,
26]. Therapeutics would then require an increase in
mean arterial pressure or treatment of increased
ICP. In a neurointensive environment where DCM
is also being monitored, the partial cessation of
CBF in some patients would be indicated as a
decreased DCM reaching exhaustion. The
therapeutic procedures mentioned above can then
also been explained based on the DCM mechanism,
as both of them can increase DCM in order to reach
positive values and reverse the partial CBF
cessation phenomenon. In that direction, increasing
DCM could also be achieved with an induction of
vasodilatation, leading to decreasing CrCP, due to
decreased WT, even though the use of vasodilators
would be debatable in this setting.
Further studies on DCM may be able to explore
its timeevolution in a clinical scenario where CPP is
getting reduced considerably. Figure 2,
demonstrating both arterial
Fig. 4 Cases of arterial hypotension with brain ischemia
getting accelerated below the LLA. a Relationship of
normalized LDF with decreasing CPP for an individual
rabbit. Below the point where the diastolic closing margin
equals zero (DCM = 0, depicted as solid line), the slope
between LDF and CPP is increased, denoting the
Fig. 5 Relationship of vascular WT with decreasing DCM for
the full cohort of rabbits. Behavior of WT changes before
and after DCM becomes exhausted (DCM = 0). Error bars
present 95 % confidence
limit
hypotension and intracranial hypertension, is
representative of the implemented changes in time,
where diastolic FV is vanishing. In this particular
example it can be seen that a big reduction of ABP
resulted in exhaustion of DCM and disappearance
Neurocrit Care (2014) 20:40–48 48

of diastolic FV in about 30 s, whereas rising ICP basilar artery in comparison, i.e., to the middle
had had the same effect later, after 2 min and 30 s. cerebral artery is its minimal insonation depth, near
the take-off of the posterior cerebral arteries, which
Limitations makes the insonation more convenient on the small
skull of rabbits. In that sense, the use of basilar
Even though these experimental data have been artery could provide more accurate continuous
published previously [15, 20–22], reanalysis of recordings in terms of the quality of measurements.
existing experimental material for mathematical In some animals the extracranial section of the
modeling studies is scientifically sound and artery may have been insonated as the basilar artery
ethically justifiable (http://www.nc3rs.org.uk/). has both an intra- and extracranial course. In this
decrease in LDF is accelerating. b Relationship of mean FV model both cranial and lumbar spaces are treated as
with decreasing CPP. Below the point where DCM = 0
(depicted as solid line), the decrease in FV is accelerating.
a single craniospinal system. It has been
Example taken from another single rabbit demonstrated that infusion of mock CSF into the
lumbar space interacts with the pulsatility of FV in
the basilar artery [15], justifying the assumption of
this craniospinal coupling.
The recordings were analyzed from an experimental
The estimation of changes in global CBF may be
data set originally intended for multiple studies of
confounded by the fact the disappearance of FVd is
cerebral hemodynamics. The presented result of
related to the collapse of small brain vessels distal
estimated CrCP and DCM is a unique analysis of
to the basilar artery. We assumed that the basilar
the dataset.
artery TCD recording is reflective of the state of
The experimental design may be a limitation for
flow in the collapsing distal microvasculature. For
this study, potentially limiting a further
this reason, the basilar artery TCD results were also
understanding of the exact relationship among the
validated by synchronous use of LDF recordings at
measured variables. However, the original
the contralateral cerebral hemisphere close to the
experimental protocol was intended for having
bregma. LDF measurements in these studies
prospective data from different experiments in
provided a direct examination of cortical
regards to brain hemodynamics. Even though
microvasculature for comparison, and rendered
further maneuvers could have been implemented in
similar results to TCD recordings in the
the experimental protocol, exploiting these existent
autoregulation analyses above and below DCM
data in the best manner possible reduces the need
[27–
for new animals to be killed in compliance with the
29].
3Rs principles of replacement, reduction, and
In rabbits the pulsatility of the blood flow is
refinement.
high, as can be seen in Fig. 1, which is different to
The use of basilar artery TCD FV as an
what is usually found in humans. This could limit
approximation of CBF can pose a limitation for this
the extrapolability of the results in the human
study as it does not reflect global CBF. However,
setting. However, the purpose of the study was
without focal brain lesions, changes in blood flow
associated with changes being made in the cerebral
in the basilar artery can be considered as
circulation through the two maneuvers of reducing
proportional to changes of CBF. In that case, the
CPP. As such, even if the blood flow pulsatility is
interest of the study is associated on induced
higher than that seen on humans, the changes on the
changes on blood flow and not on an approximation
measured parameters would be proportional from
of absolute values CBF. The advantage of using the

123
49
baseline to the end of each maneuvers. Therefore,
the knowledge of understanding then these changes
in rabbits could be beneficial in the human setting.
Pharmacologically induced hypotension can
affect most of the variables used in the estimation
of CrCP, which are fully delineated in our previous
reports of this method [15, 30]. It is possible then,
that the behavior of CrCP shown in this study might
be different in other clinical causes of hypotension.
However, estimation of CrCP has been done with
both trimetaphan-induced hypotension and
hemorrhagic hypotension, and no qualitative
difference between the two groups was observed
[30].
Animals in the trimetaphan infusion group had
antecedent elevations of ICP by cerebrospinal
infusion as part of a separate study of cerebral
hemodynamics. Although the animals were allowed
to stabilize before taking baseline ABP values, it is
possible that the preceding acute intracranial
hypertension could have interacted with our
measurements of CrCP and WT.
In clinical reality the actual value of DCM might
vary considerably within an injured brain,
depending for example on which artery we are
measuring the CBF, as it is known that typical TCD
insonated arteries like ACA, MCA, or PCA differ in
magnitude of flow [31]. This could make the actual
translation to a clinical application difficult.
However, regardless of which big artery is
insonated, we hypothesize that reaching the
threshold of zero DCM would be indicative of a
distally collapsing microvasculature.
Conclusion
A decrease in diastolic closing margin below zero
reflects vascular collapse and zero CBF during
diastole. The CPP at which DCM is equal to zero
denotes an additional breakpoint for CPP positioned
below LLA, below which pressure passivity of
CBF accelerates. Monitoring of CrCP, DCM, and
vascular WT may be clinically informative in the
effort to maintain ABP at a level of adequate
cerebral perfusion.
Acknowledgments This study was supported by the National
Institute of Health Research, Biomedical Research Center
(Neuroscience Theme), the Medical Research Council
(Grants G0600986 and G9439390), Ministry of Science and
Higher Education (MK), and NIHR Senior Investigator
Awards (JDP). The experimental laboratory work was
123
Neurocrit Care (2014) 20:40–48
supported with internal funding (Neurosurgical Unit,
University of Cambridge, UK).
Disclosure ICM+ is licensed by the Cambridge Enterprise
Ltd. (Cambridge, UK, www.neurosurg.cam.ac.uk/icmplus).
PS and MC have financial interest in a fraction of the
licensing fee.
References
1. Fog M. Cerebral circulation. The reaction of the pial
arteries to a fall in blood pressure. Arch Neurol Psychiatry.
1937;37:351–64. 2. Lassen NA. Autoregulation of cerebral
blood flow. Circ Res. 1964;15(SUPPL):201–4.
3. Aaslid R, Lindegaard KF, Sorteberg W, Nornes H.
Cerebralautoregulation dynamics in humans. Stroke.
1989;20:45–52.
4. Paulson OB, Olesen J, Christensen MS. Restoration of
autoregulation of cerebral blood flow by hypocapnia. J
Neurol. 1972;22: 286–93.
5. Kontos HA, Wei EP, Navari RM, Levasseur JE,
Rosenblum WI,Patterson JL Jr. Responses of cerebral
arteries and arterioles to acute hypotension and
hypertension. Am J Physiol. 1978;234: H371–83.
6. Wijdicks FME. Determining brain death in adults. J
Neurol.1995;45:1003–11.
7. Lovrencˇic´-Huzjan A, Vukovic´ V, Jergovic´ K,
Demarin V. Transcranial Doppler as a confirmatory test
in brain death. Acta Clin Croat. 2006;45:365–73.
8. Kirkham FJ, Levin SD, Padayachee TS, Kyme MC,
Neville BGR,Gosling RG. Transcranial pulsed Doppler
ultrasound findings in brain stem death. J Neurol
Neurosurg Psychiatry. 1987;50: 1504–13.
9. Bouzat P, Francony G, Declety P, Genty C, Kaddour A,
BessouP, Brun J, Jacquot C, Chabardes S, Bosson JL,
Payen JF.
Transcranial Doppler to screen on admission patients
with mild to moderate traumatic brain injury.
Neurosurgery. 2011;68:1603–9.
10. Astudillo R, van der Linden J, Ekroth R, Wesslen O,
Hallhagen S,Scallan M, Shore D, Lincoln C. Absent
diastolic cerebral blood flow velocity after circulatory
arrest but not after low flow in infants. Ann Thorac Surg.
1993;56:515–9.
11. Jonassen AE, Quaegebeur JM, Young WL. Cerebral
blood flowvelocity in pediatric patients is reduced after
cardiopulmonary bypass with profound hypothermia. J
Thorac Cardiovasc Surg. 1995;110:934–43.
12. O’Hare B, Bissonnette B, Bohn D, Cox P, Williams W.
Persistentlow cerebral blood flow velocity following
profound hypothermic circulatory arrest in infants. Can J
Anaesth. 1995;42:964–71.
13. Burton AC. Fundamental instability of the small blood
vesselsand critical closing pressure in vascular beds. Am
J Physiol. 1951;164:330–1.
14. Dewey RC, Pierer HP, Hunt WE. Experimental cerebral
hemodynamics-vasomotor tone, critical closing
pressure, and vascular bed resistance. J Neurosurg.
1974;41(5):597–606.
15. Varsos GV, Richards H, Kasprowicz M, Budohoski KP,
BradyKM, Reinhard M, Avolio M, Smielewski P,
Pickard JD, Czosnyka M. Critical closing pressure
Neurocrit Care (2014) 20:40–48 50
determined with a model of cerebrovascular impedance. constant: dependence on cerebral perfusion pressure and
J Cereb Blood Flow Metab. end-tidal carbon dioxide concentration. Neurol Res.
2012;33:235–43. 2012;34:17–24.
16. Aaslid R, Lash SR, Bardy GH, Gild WH, Newell DW. 31. Aaslid R, Markwalder TM, Nornes H. Noninvasive
Dynamicpressure–flow velocity relationships in the transcranialDoppler recording of flow velocity in basal
human cerebral circulation. Stroke. 2003;34:1645–9. cerebral arteries. J Neurosurg. 1982;57(6):769–74.
17. Panerai RB. The critical closing pressure of the cerebral
circulation. Med Eng Phys. 2003;25:621–32.
18. Michel E, Hillebrand S, von Twickel J, Zernikow B,
Jorch G.Frequency dependence of cerebrovascular
impedance in preterm neonates: a different view on
critical closing pressure. J Cereb Blood Flow Metab.
1997;17:1127–31.
19. Puppo C, Camacho J, Yelicich B, Moraes L, Biestro A,
Gomez H.Bedside study of cerebral critical closing
pressure in patients with severe traumatic brain injury: a
transcranial Doppler study. Acta Neurosurg Suppl.
2012;114:283–8.
20. Richards HK, Czosnyka M, Pickard JD. Assessment of
criticalclosing pressure in the cerebral circulation as a
measure of cerebrovascular tone. Acta Neurochir
(Wien). 1999;141:1221–7.
21. Nelson RJ, Czosnyka M, Pickard JD, Maksymowicz W,
Perry S,Martin JL, et al. Experimental aspects of
cerebrospinal haemodynamics: the relationship between
blood flow velocity waveform and cerebral
autoregulation. J Neurosurg. 1992;31: 705–10.
22. Harland S, Richards HK, Czosnyka M, Piechnik P,
Pickard JD.Dissociation of cerebral autoregulation and
CO2 reactivity following carotid occlusion in rabbits. J
Cereb Blood Flow Metab. 1999;19:S636.
23. de Riva N, Budohoski KP, Smielewski P, Kasprowicz
M, ZweifelC, Luzius A, Reinhard M, Fabregas N,
Pickard JD, Czosnyka M. Transcranial Doppler
pulsatility index: what it is and what it isn’t. Neurocrit
Care. 2012;17:58–66.
24. Carrera E, Kim DJ, Castellani G, Zweifel C, Smielewski
P,Pickard JD, Kirkpatrick PJ, Czosnyka M. Cerebral
arterial compliance in patients with internal carotid
artery disease. Eur J Neurol. 2010;18:711–8.
25. Kasprowicz M, Czosnyka M, Soehle M, Smielewski P,
Kirkpatrick PJ, Pickard JD, Budohoski KP. Vasospasm
shortens cerebral arterial time constant. Neurocrit Care.
2011;16:213–8.
26. Reinhard M, Petrick M, Steinfurth G, Ziyeh S, Hetzel A.
Acuteincrease in intracranial pressure revealed by
transcranial Doppler sonography. J Clin Ultrasound.
2003;31:324–7.
27. Kempski O, Heimann A, Strecker U. On the number of
measurements necessary to assess regional cerebral
blood flow by local laser Doppler recordings: a
simulation study with data from 45 rabbits. Int J
Microcirc Clin Exp. 1995;15:37–42.
28. Ungersbo¨ck K, Heimann A, Kempski O. Monitoring of
cortical blood flow: clinical relevance of experimental
laser Doppler studies. Neurol Res. 1996;18:273–6.
29. Heimann A, Kroppenstedt S, Ulrich P, Kempski OS.
Cerebralblood flow autoregulation during hypobaric
hypotension assessed by laser Doppler scanning. J Cereb
Blood Flow Metab.
1994;14:1100–5.
30. Czosnyka M, Richards HK, Reinhard M, Steiner AL,
BudohoskiK, Smielewski P, et al. Cerebrovascular time

123