Characterizing Hemodynamic Compensation in Patients With Intracranial Stenosis

The overall aim of this work is to assess the relationship between stroke risk and
hemodynamic compensation strategies, as measured using a novel 3.0 Tesla (T) MRI protocol, in
patients with symptomatic intracranial (IC) steno-occlusive disease. Recent studies have
shown high two-year ischemic stroke rates of approximately 20% in symptomatic IC stenosis
patients. Aggressive therapy for IC stenosis patients includes revascularization with
angioplasty, IC stenting, or bypass, however identification of patients most likely to
benefit from these more aggressive interventions, rather than medical management alone, has
been problematic. Accurate measurements of hemodynamic compromise are likely required to
better define stroke risk and guide treatment decisions. Specifically, in IC stenosis
patients with compromised cerebral perfusion pressure (CPP), the extent of hemodynamic
compromise reflects the autoregulatory capacity of vasculature to increase arterial cerebral
blood volume (aCBV) and/or develop collaterals to supplement cerebral blood flow (CBF). The
prevalence of CBF collateralization and aCBV autoregulation has been hypothesized to
correlate uniquely with stroke risk, however the extent of this correlation is debated.

The critical barrier to stratifying stroke risk rests with a lack of (i) methodology for
measuring multiple hemodynamic factors with high specificity and (ii) noninvasive approaches
capable of monitoring longitudinal progression of impairment. The investigators have
demonstrated the clinical utility of relatively new, noninvasive MRI approaches for assessing
cerebrovascular reactivity (CVR), aCBV, and collateral CBF. These approaches have been
optimized and feasibility assessed in healthy volunteers or isolated clinical studies,
however comparatively little information is available regarding the relationship between
these collective measures and stroke risk. Therefore, the investigators propose to complement
established angiographic and structural imaging with more novel, validated hemodynamic
measurements to assess tissue-level impairment and compensation strategies in patients with
IC stenosis. Using a collective approach combining measurements of collateral CBF, aCBV and
CVR in multiple brain regions, in conjunction with a statistical model incorporating the
above parameters as possible prognostic factors, the investigators will quantify the extent
to which two-year stroke risk is associated with hemodynamic compensation mechanisms in a
population of symptomatic IC stenosis patients. The noninvasive and multi-faceted scope of
this investigation is intended to expand the diagnostic stroke infrastructure and elucidate
new hemodynamic prognostic indicators of stroke in this high-risk population.

Hypothesis (1). Changes in collateral CBF with hypercarbic stimulation positively correlate
with two-year stroke incidence in symptomatic IC stenosis patients.

Aim (1). By separately magnetically labeling blood water in different feeding arteries (8),
the investigators will use a tested vessel-selective arterial spin labeling (VS-ASL) approach
and a calculated perfusion asymmetry index to noninvasively assess changes in collateral CBF
patterns in patients with IC stenosis. IC stenosis (n=90) patients will be monitored for two
years during which their known stroke incidence is 20%. Stroke incidence will be recorded and
separate correlations between perfusion asymmetry and stroke risk will be assessed.

Hypothesis (2). Regionally reduced CVR, indicative of vascular steal phenomena, and elevated
aCBV, will positively correlate with two-year stroke incidence in symptomatic IC stenosis
patients.

Aim (2). CVR, as measured using the blood oxygenation level-dependent (BOLD) MRI signal
change in response to hypercarbia, will be used to assess cerebrovascular reserve. The inflow
vascular-space-occupancy with dynamic subtraction (iVASO-DS) approach, which we have
developed and demonstrated in patients, will be applied to assess baseline aCBV. For the same
patients as in Aim 1, CVR dynamics and aCBV will be statistically assessed for unique
relationships with two-year stroke incidence.

Hypothesis (3). Collateral CBF, autoregulatory aCBV, and CVR will correlate unequally with
two-year stroke incidence, the extent of which can be quantified using a multi-component
prediction model.

Aim (3). The investigators will build a statistical prediction model, using collateral CBF,
aCBV and CVR as prognostic factors, to quantify the extent to which trends in the above
parameters collectively contribute to stroke risk. The high two-year stroke rate (20%) and
large number of admitted patients with symptomatic IC stenosis, will enable this study to
detect hypothesized group differences with high statistical power.

This work is an extension of recent methodological work whereby a novel, collective
compensatory hemodynamic protocol is applied to a specific clinical population to understand
prognostic potential. Successful completion should provide new information on tissue
hemodynamics and stroke risk in patients that can be used to guide management decisions,
ultimately reducing stroke incidence in this high-risk population.

Inclusion Criteria:

1. Adult patients (age 18-85, inclusive)

2. Symptomatic (TIA or ischemic stroke) in the hemispheric carotid territory of vascular
stenosis

3. Vascular imaging demonstrating large vessel IC stenosis>50% or occlusion of IC carotid
or MCA

4. Sub-acute stroke patients who have received intravenous or intra-arterial treatments

5. Sub-acute stroke patients ineligible for conventional acute stroke intervention.
Language comprehension intact, motor aphasia mild or absent, competent to give
informed consent

6. Most recent qualifying TIA or stroke within 60 days prior to performance date of
hemodynamic MRI

Exclusion Criteria:

1. ECA stenosis > 70% determined by MRA, CTA, or DSA.

2. Acute stroke patients presenting with anterior circulation stroke onset eligible for
intervention with intra-arterial thrombolysis or mechanical thrombectomy

3. MRI contraindications (e.g. non-compatible implants, pregnancy, etc.)

4. Non-atherosclerotic cervical or intracranial stenosis

5. Heart disease likely to cause cerebral ischemia, including cardiomyopathy with
ejection fraction<25%, prosthetic valve, infective endocarditis, sick sinus syndrome,
myxoma, left atrial or ventricular thrombus

6. Existing condition likely to lead to death within 2 years