stroke - brain schemia - collateral circulation - tomography - angiography - perfusion
acidente vascular cerebral - isquemia encefálica - circulação colateral - tomografia
- angiografia - perfusão
Stroke is the second most common cause of death and was responsible for approximately
6.7 million deaths worldwide in 2012[1 ]. Ischemia, or restricted blood flow, is the main cause of stroke and is typically
due to abrupt occlusion of a cerebral artery as a result of progressive atherosclerosis
or embolism[2 ]. Acute ischemic stroke (AIS) can result in severe neurologic disability or death[3 ].
Since the late 1990s intravenous thrombolysis has been a recommended treatment for
AIS[4 ]
,
[5 ]. Additionally, in the last decade, several clinical trials have investigated the
effects of endovascular treatment (EVT) in the setting of an intracranial or extracranial
large artery occlusion. Several studies, including MR CLEAN[6 ], ESCAPE[3 ], EXTEND-IA[7 ] and SWIFT-PRIME[8 ], recently proved EVT to be more effective than standard medical care, with or without
intravenous thrombolysis, using stentrievers in the majority of the patients in the
EVT arms.
However, although EVT has been shown to be generally effective, the trials have documented
erratic individual and overall patient outcomes. These differences may not be solely
the result of the various methods used but may also be related to patient-specific
characteristics[9 ]. Among these patient-specific characteristics, collateral status has emerged as
an independent factor that is associated with angiographic and clinical outcomes in
AIS patients[10 ]
,
[11 ].
The collateral circulation is a physiologic pathway of specialized endogenous bypass
vessels that is present in most tissues and protects against ischemic injury during
initial oligemic status[12 ]. In the setting of AIS, the extent of collateral circulation influences the size
of the final infarct and the growth of the penumbra. Hence, the relationship between
the collateralization grade and the predictability of infarct evolution has been a
primary focus in recent years[11 ]
,
[13 ]
,
[14 ].
The refinement of diagnostic techniques for evaluation of collateral circulation may
contribute to improved anatomical and pathophysiological characterization of this
vascular network and its potential therapeutic and prognostic implications[15 ]. The ability to define physiologic parameters through diagnostic techniques is particularly
useful when the time of the stroke is unknown or a wider nonconventional window for
treatment is being considered.
In this review, we summarize the basic anatomy and physiology of the collateral circulation
and its potential as an endogenous therapeutic target in AIS. The relevance of multidetector
computed tomography (MDCT) in clinical settings to support medical decisions regarding
AIS is highlighted, and recent evidence indicating that good collateral circulation
can prevent or delay permanent neural damage is presented.
THE ANATOMY OF COLLATERAL CIRCULATION
THE ANATOMY OF COLLATERAL CIRCULATION
Two main routes underlie collateral perfusion of the brain parenchyma. The anatomy
of this arterial circulation includes extracranial sources of blood flow that can
supply intracranial vessels, as well as intracranial routes that can supplement other
intracranial areas when pathophysiologic mechanisms become actived[15 ].
The extracranial sources consist of large connections between the extracranial and
intracranial arteries. The external carotid artery gives rise to many branches in
the neck that are potential sources of collateral blood flow, particularly when chronic
stenosis or occlusion has developed in the internal carotid artery[16 ]. The facial, maxillary, middle meningeal, and occipital arteries are the main branches
that can shunt flow via anastomoses to the intracranial arteries. Apart from these
branches, common anastomotic routes include the ophthalmic artery, which may fill
in a retrograde direction, as well as smaller and unnamed dural arteries[17 ].
Intracranial collateral routes can be further subdivided into primary and secondary
routes. The primary pathways include the permanently active components of the circle
of Willis, and the secondary pathways include less direct routes that develop over
time. The blood supply to the brain is unique because four major arteries coalesce
to form an equalizing distributor, i.e., the circle of Willis, which, despite its
variability and asymmetry, can redistribute blood flow in the event of sudden occlusion
of a parent vessel.
The secondary pathway comprises leptomeningeal anastomoses that link distal sections
of the major cerebral arteries. It has been reported that some small arteriolar connections
(~50–400 μm) allow retrograde perfusion of adjacent territories[18 ]. It is assumed that these connections are important routes for collateral blood
flow, especially when an acute arterial occlusion occurs. These arteriolar anastomoses
mimic the circle of arteries but connect a much larger extension of the microvasculature,
joining territories of the middle cerebral artery (MCA) with both the anterior cerebral
artery (ACA) and the posterior cerebral artery (PCA)[19 ].
The development of native pial collateral circulation (collaterogenesis), which begins
in the embryo, has been shown to determine the extent of the collaterals in adulthood[20 ]. Acute obstructions induce blood flow across the collateral network (recruitment)
followed by remodeling and, potentially, formation of additional collaterals in chronic
obstructive disease (neocollateral formation) [17 ].
There is wide variation in collateral status among healthy adults, and recent animal
studies indicate that genetic background may be a major factor[21 ]. {Zhang, 2010 #110} A single polymorphic locus on chromosome 7 in mice, i.e., the
determinant of collateral extent 1 (Dce1), has been shown to influence the extent
of collateralization, blood flow and infarct volume after middle cerebral artery occlusion[22 ]. Whether human Dce1 or related loci are responsible for the wide variation in collateral
status in humans is still under investigation.
A number of other factors, including environmental and clinical features, have also
been shown to affect the quality and quantity of collaterals (rarefaction) at the
time of presentation in AIS. Of these, the strongest predictor by far is age[23 ]. Other clinical features include elevated glucose at the time of presentation, uric
acid level, history of hypertension, and history of smoking[24 ]
,
[25 ].
PHYSIOLOGY OF COLLATERAL BLOOD FLOW REGULATION
PHYSIOLOGY OF COLLATERAL BLOOD FLOW REGULATION
The importance of the collateral circulation in brain physiology may be demonstrated
with the concept of the collaterome, which is an extension of the connectome concept[26 ]. The collaterome provides a physiologically relevant approach to the management
of stroke and the influential balance of collateral perfusion that determines both
stroke evolution and related clinical sequelae[27 ].
Beyond structural assessments of collateral circulation, advances in perfusion-based
imaging have allowed for functional evaluations of the quality of collateral blood
flow (effective parenchymal perfusion). Cerebral blood flow (CBF) is regulated by
the metabolic demands of the brain itself, which vary regionally and with neuronal
activity. Although the precise mechanisms underlying cerebral autoregulation are not
fully understood, the process seems to be mediated at several levels and involves
neurons, neuropil, and cerebral blood vessels[28 ].
Normal CBF ranges between 50 and 60 mL/100 g/min and is tightly controlled by cerebral
autoregulation[29 ]. The pace of cellular death in the brain after an arterial occlusion is closely
linked to the severity of the decrease in blood flow within the local environment.
When blood flow is less than 10 mL/100 g/min, damage is rapid, and most cells die
within minutes of the insult[30 ]
,
[31 ]. When CBF is between 10 and 20 mL/100 g/min (hypoperfusion), neurons cease to function
but remain structurally intact and are potentially revivable if normal blood flow
is restored[31 ]. Therefore, neuronal damage is not uniform when an intracranial artery is occluded,
especially in the first few hours after an insult. Depending on the extent of collateral
perfusion, infarction may not be complete for hours or even days[32 ].
In thrombotic and embolic strokes, the intravascular pressure distal to the occlusion
falls immediately. Concurrently, the pressure within the pial vessels is relatively
well preserved, resulting in a gradient that is able to promote flow through anastomoses[18 ]. The effectiveness of collateral vessel flow can be assessed only with measurements
of tissue perfusion, which reflect the statuses of both the microcirculation and the
macrocirculation. Computed tomography (CT) and magnetic resonance imaging (MRI) perfusion
techniques and other methods, such as positron emission tomography (PET) and single-emission
computed tomography (SPECT), can provide insight into the collateral flow in patients
with cerebrovascular disease[17 ]
,
[33 ]
,
[34 ]. Physiologically effective collateral perfusion is evident when CBF and cerebral
blood volume (CBV) are maintained within the territory of an occluded artery.
IMAGING OF COLLATERAL VESSELS
IMAGING OF COLLATERAL VESSELS
Digital subtraction angiography (DSA) remains the gold standard for the anatomic evaluation
of the collateral circulation. This technique allows for the dynamic visualization
of blood flow through pial collaterals or other secondary collaterals[17 ]
,
[35 ]. The main limitations of DSA are its invasive nature, its reliance on iodinated
contrast and ionizing radiation, and its inability to evaluate brain parenchyma. Furthermore,
performing DSA in AIS when intra-arterial therapy is not considered may generate an
additional delay to treatment[36 ].
Several noninvasive approaches have been proposed to evaluate intracranial collateral
blood flow and the network, but none of these techniques have been shown to be as
effective as a reference standard for quantifying collateral flow[37 ].
Computed tomography angiography (CTA) is fast, reproducible, and widely available,
and its reasonable cost-to-effectiveness ratio makes this technique one of the most
widely used methods of evaluating the locations of vascular occlusions and the collateral
system. Analysis of CTA source images (CTA-SI) has a higher sensitivity for demonstrating
the infarct core than non-contrast computed tomography (NCCT)[38 ]. Post-processed CTA data involving maximal intensity projections (MIP) and multiplanar
reconstruction (MPR) allow for better visualization of the occluded vessel and the
extent of leptomeningeal flow[39 ]. The main limitation of CTA is that it is a snapshot of arterial contrast enhancement,
providing limited information about flow dynamics. Some studies have attempted to
circumvent this limitation with dual-phase CTA[40 ]. Such variations in acquisition protocols and differences in classification have
led to low levels of agreement among the relevant results even among experienced observers
(K-alpha 0.3-0.6)[41 ].
Dynamic CTA has become available for clinical practice in recent years and merges
the noninvasive nature of CTA and the dynamic acquisition of DSA. This technique,
also referred to as 4D-CTA, enables the noninvasive evaluation of the flow dynamics
of the intracranial vasculature by multiple subsequent CT acquisitions or continuous
volume CT acquisition over a period of time[42 ].
Several protocols of acquisition have been proposed, including a toggling-table technique,
shuttle mode scanning, and volume mode. The volume mode is considered the most versatile
option and allows for complete or partial coverage of the whole brain during 1 rotation
of the scanner[42 ]. Dynamic acquisitions in volume mode can be performed discontinuously or continuously,
depending on the required temporal resolution. When collateral flow, as in the case
of an arterial occlusion, needs to be evaluated, a lower temporal resolution is necessary.
In the setting of AIS, 4D-CTA better estimates thrombus burden and the presence of
collateral vessels than conventional CTA[42 ]. The challenge of the radiation dose level remains, although recently available
noise-reduction filters have dramatically reduced radiation exposure[43 ].
Magnetic resonance angiography (MRA) using time-of-flight technique (TOF) is one of
the most used MR techniques for accessing collateral circulation, but it remains controversial.
MRA provides structural information based on flow-sensitive images but is less effective
for collateral evaluations than CTA, especially when the objective is to estimate
the occlusion of distal branches[36 ]
,
[44 ]. Contrast-enhanced MRA (CE-MRA) allows better delineation of slow-moving blood in
the distal branches and is a better predictor of infarct outcome, but it provides
lower spatial resolution[44 ].
Transcranial Doppler (TCD) also provides information regarding cerebral autoregulation
and cerebral circulation. Flow direction changes, such as those found in the ophthalmic
artery, and increased velocity in vessels ipsilateral to a stenosis are correlated
with the presence of leptomeningeal collaterals[45 ]. However considerable variability has been found in TCD performance and interpretation[35 ]
,
[45 ].
IMAGING STUDIES OF COLLATERAL CIRCULATION IN THE CLINICAL SETTING OF ACUTE ISCHEMIA
IMAGING STUDIES OF COLLATERAL CIRCULATION IN THE CLINICAL SETTING OF ACUTE ISCHEMIA
The concept that a vascular network can potentially bypass the effects of a blocked
cerebral artery and influence ischemic lesion size and growth[37 ]
,
[46 ] has recently had an increased impact on the management of stroke patients[4 ].
Assessments of collaterals in angiographic studies have proven be widely variable.
A recent meta-analysis by Leng et al.[9 ] found 12 studies that used the American Society of Interventional and Therapeutic
Neuroradiology/Society of Interventional Radiology (ASITN/SIR) collateral flow grading
system by DSA and primarily defined grades 3-4 and 0-2 as good and poor collateral
statuses, respectively. Eleven studies have used other grading methods for DSA, 9
studies have used different grading methods for CTA, and others have used CTP or combined
grading methods with different imaging modalities.
Two DSA classifications stand out: Higashida[47 ] used the ASITN/SIR assessment, which was based on the extent and delay of retrograde
filling, whereas Christoforidis[48 ] proposed a distinct model that is based solely on the extent of the feedback.
The ASITN/SIR classification[47 ] seems to be the most appropriate because it has higher reproducibility and accounts
for both the parameters of the extent and the delay of the feedback via collaterals.
The ASITN/SIR classification involves a five-point system ([Table 1 ]) that has been used in several endovascular trials[9 ]
,
[10 ]
,
[49 ]
,
[50 ]
,
[51 ]
,
[52 ]. Grades 0 and 1 indicate only marginal flow, grade 2 indicates only partial filling
of the ischemic territory, and grades 3 and 4 indicate varying rates of complete filling
of the occluded arterial territory.
Table 1
American Society of Interventional and Therapeutic Neuroradiology / Society of interventional
Radiology (ASITN/SIR) collateral grade scale.
Grade
Angiographic collaterals (Digital subtraction angiography)
0
No collaterals visible to the ischemic site
1
Slow collaterals to the periphery of the ischemic site with persistence of some of
the defect
2
Rapid collaterals to the periphery of ischemic site with persistence of some of the
defect and to only a portion of the ischemic territory
3
Collaterals with slow but complete angiographic blood flow of the ischemic bed by
the late venous phase
4
Complete and rapid collateral blood flow to the vascular bed in the entire ischemic
territory by retrograde perfusion
Additional non-invasive grading systems for assessing collateral blood flow circulation
have also been described[53 ]. Several strategies have been used to describe CTA-visualized vessels by either
comparing them with the contralateral brain hemisphere or estimating the percentage
of MCA branches that become filled by contrast media during examination[54 ]
,
[55 ]. Souza et al.[56 ] proposed a simple and reliable grading system that correlates collateral scores
and diffusion-weighted imaging (DWI) lesion volumes on admission.
The best method of evaluating and grading collateral flow remains controversial. The
currently proposed methods of assessing collaterals are largely qualitative or semiquantitative,
and there are no clear indications of the superiority of any of the available techniques[57 ]. This lack of consensus may contribute to an overall under appreciation of the fundamental
role of collateral circulation in outcomes following AIS.
Nevertheless, Shet and Liebeskind[58 ], when revising the status of collateral circulation in endovascular therapy for
stroke, concluded that collateral blood supply is pivotal in determining clinical
outcomes. In the setting of vascular occlusion, patients with more robust collaterals
have smaller infarcts ([Figure 1 ])[17 ]
,
[59 ]. Regardless of the method used to determine the collateral score, extremely poor
outcomes are predicted when collateral blood flow is reduced or absent ([Figures 2 ] and [3 ])[60 ]
,
[61 ]
,
[62 ]
,
[63 ].
Figure 1 Proximal occlusion with suitable reconstitution of the distal middle cerebral artery
(MCA) branches (Mittef grade 3). (A) Computed tomography angiografhy (CTA) identified
right MCA occlusion. (B) Note collateral circulation throughout the MCA territory,
filled from leptomeningeal branches. (C) Follow-up non-contrast compute tomography
(NCCT) 24 hours later demonstrating a small final infarct volume confined to the lenticulostriate
territory (proximal MCA); this patient exhibited a good clinical outcome.
Figure 2 Proximal occlusion with vessel filling restricted to the Sylvian fissure (Mittef
grade 2). (A) A left Mittef grade 1 (M1) occlusion was demonstrated on computed tomography
angiografhy (CTA), whereas only minimal collateral filling was noted in the middle
cerebral artery (MCA) territory (B). (C) Follow-up non-contrast compute tomography
(NCCT) 48 hours later illustrating a large infarct volume involving the lenticulostriate
territory, insula and temporal lobe.
Figure 3 Proximal occlusion with faint contrast opacification restricted to the distal superficial
branches (Mittef grade 1 – M1). (A) A left M1 occlusion was found on computed tomography
angiografhy (CTA), and the collateral circulation in the middle cerebral artery (MCA)
territory was considered absent (B). (C) Follow-up non-contrast compute tomography
(NCCT) 24 hours later demonstrating a malignant infarct. This patient died 3 days
later.
Physiology is now considered to be more relevant than time in AIS because it is the
degree of collateral flow and not simply the time elapsed since the stroke that is
the main factor determining core infarct volume within the first 6 hours of stroke
onset[46 ]. Collateral status is recognized as having an influence on stroke prognosis, particularly
in terms of recanalization, reperfusion, hemorrhagic transformation and subsequent
neurological outcomes[59 ]. Collateral grading ([Table 2 ]) may represent a currently available opportunity to predict possible future outcomes,
whereas elapsed time solely reflects the past and does not have a causal connection
with the future of infracted brain tissue.
Table 2
Collateral grade by Miteff system.
Grade
Computed tomography angiography collaterals
1
The contrast opacification is merely seen in the distal superficial branches
2
Vessels can be seen at the Sylvian fissure
3
If the vessels are reconstituted distal to the occlusion
RELATIONSHIP WITH PARENCHYMAL PERFUSION
RELATIONSHIP WITH PARENCHYMAL PERFUSION
Collaterals are usually evaluated by examining arterial flow with angiography techniques,
and parenchymal perfusion is profoundly influenced by downstream microcirculation[59 ]. The microcirculation is crucial to the restoration of the blood supply to the brain,
and collateral circulation may increase ischemic tolerance by enhancing microvascular
perfusion[35 ].
A relatively simple method of evaluating brain perfusion and viable tissue is the
assessment of capillary blush using DSA. The capillary index score (CIS) can define
an ischemic area in two ways: either by a lack of anterograde flow (in an area that
receives blood supply in a retrograde fashion through pial collaterals) or by a significant
delay in anterograde flow due to a proximal partially recanalized clot[64 ]. This angiographic index has been shown to be a good predictor of outcomes and a
powerful strategy for improving outcomes in endovascular treatment[65 ].
Beyond structural assessments, perfusion-based images have allowed functional evaluations
to differentiate critically hypoperfused areas (infarct cores), penumbral areas (potentially
savable areas) and benign oligemic tissues. Current evidence indicates that the goals
of acute stroke treatment should be to determine tissue viability by noninvasive techniques,
use this information to individualize thrombolytic therapy, extend the therapeutic
time window and rescue penumbral tissue[46 ]. Therefore, knowing the factors that influence the loss of penumbral tissue is crucial,
and collaterals have emerged as a major feature that is relevant to this knowledge.
The presence of robust collaterals both markedly reduces and slows down penumbra loss[14 ]. This can be seen in patients without significant reperfusion after treatment, reinforcing
the fact that poor collaterals alone are able to predict larger infarcts. In contrast,
smaller infarcts, at least in part, result from good collaterals. Reperfusion can
occur not only via successful recanalization of the primary occlusion and restoration
of downstream flow but also via viable collateral blood flow[66 ]. However, it is important to note that mismatch (the difference between the penumbra
and the infarct core) is also an independent prognostic factor with a strong association
with better outcomes in target mismatch patients, adding information to the study
of collateral profiles[67 ].
Some studies have demonstrated the use of ischemia-induced vascular damage estimates
in AIS in combination with collateral scores and brain perfusion analysis to predict
hemorrhagic complications[17 ]
,
[68 ]
,
[69 ]. In the setting of poor collaterals, a finding of hyperperfusion may indicate a
higher risk of hemorrhagic transformation. When revascularization is achieved, symptomatic
hemorrhagic transformation may occur more frequently in patients who have presented
with poor collaterals[49 ]. Higher frequencies of infarct growth and symptomatic hemorrhagic transformation
in patients with poor collaterals in whom therapeutic recanalization has been achieved
may support the concept of reperfusion injury[70 ].
TEACHING POINTS FOR CLINICAL USE
TEACHING POINTS FOR CLINICAL USE
Larger recent trials have established the use of intracranial vascular studies in
the setting of AIS to detect proximal obstructions, define the treatment subtype and
select an intraarterial approach[3 ]
,
[6 ]
,
[7 ]. Therefore, the collateral profile should be determined in all of these patients.
Some critical findings may be established with the knowledge of collateral status.
Abundant native (preexisting) collateral circulation is directly correlated with better
clinical status and smaller volumes of infarcted brain[10 ]. Improved collateral circulation also predicts higher rates of recanalization, favorable
outcomes and lower rates of mortality[9 ]
,
[10 ]
,
[50 ].
The absence or relative paucity of a collateral network is a major predictor of extensive
infarct on admission[56 ], and a proximal thrombus associated with such an absence has been termed a “malignant
profile”[71 ]. Recent studies have also demonstrated that worse collaterals are associated with
increased hemorrhagic complications, reinforcing the relevance of specific approaches
for these patients[49 ]
,
[70 ].
FINAL REMARKS
This literature review supports the view that noninvasive vascular studies should
be used to identify proximal arterial occlusions and to estimate collateral grading
in the setting of AIS. A personalized approach that is not solely restricted by time
should be provided to maximize the effect of therapy, including appropriate patient
selection for EVT.
The currently available knowledge has increased the pathophysiologic understanding
of intrinsic compensatory vascular mechanisms, supports the use of MDCT techniques
to rapidly evaluate hyperacute AIS, and provides evidence for therapeutic decisions.
Modern techniques for reducing radiation exposure should be employed to ensure that
diagnostic tests preserve patient safety.
