Stress CT Perfusion in Patients With Chest Pain

Background. In the United States, more than 8 million patients require emergency department
evaluation for acute chest pain every year. The estimated cost for these evaluations is > $10
billion and the loss of economic productivity is likely far greater. Multidetector computed
tomography (MDCT) is a popular alternative to diagnostic invasive coronary angiography (ICA)
(Deetjen, et al. 2007, Schroeder, et al. 2008, de Roos 2010) and is gaining wide clinical
acceptance for its ability to rule out significant coronary artery disease (CAD) (Hulten,
Bittencourt, Ghoshhajra and Blankstein 2012). However, in patients with acute chest pain,
abnormal CT coronary angiography (CTCA) findings frequently result in additional nuclear
myocardial perfusion imaging (MPI) to determine the hemodynamic significance of coronary
stenosis (Garcia, Lessick and Hoffmann 2006, Deetjen, et al. 2007, Miller, et al. 2008,
Schroeder, et al. 2008, de Roos 2010). This is because the presence and extent of myocardial
ischemia is more important than the severity of a coronary stenosis for identifying patients
who would benefit from coronary revascularization. Accordingly, with the growing interest in
simultaneous evaluation of coronary anatomy and the hemodynamic significance of CAD in a
single test, studies have focused on the potential of MDCT to assess myocardial perfusion
(Techasith and Cury 2011). One hurdle this approach needs to overcome is that it relies on
visual assessment of manually selected 2D slices, rather than 3D analysis of the entire
myocardium, and requires manual adjustment of contrast windows, both carrying the risk of
missing subendocardial perfusion defects.

Prior data and hypothesis. To overcome these limitations, we recently developed a technique
for quantitative 3D analysis of myocardial perfusion, which uses the distribution of x-ray
attenuation to calculate for each myocardial segment an index of severity and extent of
perfusion abnormality (Kachenoura, et al. 2010). Having confirmed the ability of this
analysis to detect regadenoson-induced perfusion abnormalities in consecutive patients
referred for CTCA (Mor-Avi, et al. 2012), we propose a new study aimed at determining the
value of this methodology in patients presenting in the emergency department with acute chest
pain. We selected this cohort to further validate CT perfusion analysis and to determine
whether it provides additive utility over CTCA alone, because these patients are increasingly
referred by emergency departments for CTCA in large numbers, instead of nuclear vasodilator
stress testing, thus losing valuable physiologic information. Because our perfusion index was
specifically designed to take into account the fact that stress-induced perfusion defects are
subendocardial, we hypothesize that our quantitative 3D analysis can accurately identify
patients presenting at the emergency department with acute chest pain due to underlying
hemodynamically significant coronary stenosis, aid in the identification of individuals most
likely to benefit from revascularization, and thus improve the ability to predict patient
outcomes. Indeed, our previous study showed that with regadenoson, our perfusion index is 2-3
times higher in myocardial segments supplied by arteries with significant stenosis (Patel, et
al. 2011), and thus improves the diagnosis of hemodynamically significant CAD.

Aims. This study was designed to achieve the following goals: (1) to test the above
hypothesis by comparing stress MDCT perfusion data with invasive fractional flow reserve
(FFR) data in patients with significant stenosis who undergo ICA; and (2) to determine the
added value of MDCT perfusion as an adjunct to CTCA for predicting patient outcomes.

Study design. In this study, MDCT imaging will be performed during regadenoson stress in
approximately 150 consecutive patients with chest pain referred for CTCA, who agree to
undergo additional MDCT imaging during vasodilator stress. Patients with contraindications to
CTCA, including known allergies to iodine, renal dysfunction (creatinine >1.6 mg/dL),
inability to perform a 10 sec breath-hold, and contraindications to beta-blockers or
regadenoson, such as chronic obstructive pulmonary disease, advanced heart block or systolic
blood pressure <90 mmHg, will be excluded from the study. In addition, patients with a
history of cardiothoracic surgery or pacemaker or coronary stent implantation will be
excluded. Myocardial perfusion will be assessed using quantitative volumetric analysis, but
will not be reported to the referring physician in order to avoid referral bias. In each
patient, we will collect the following information at the time of enrollment: age, gender,
height, weight, blood pressure, tobacco use, history of heart disease, hypertension, stroke
and diabetes. In addition, a blood sample will be obtained to allow analysis of lipids and
cardiac enzymes. In a subgroup of patients who also undergo ICA, MDCT perfusion will be
compared with the ICA findings, including the degree of stenosis and FFR. Enrollment will be
stopped when 30 patients with ICA data are enrolled (based on power analysis). All study
patients will be followed up to determine the predictive value of MDCT perfusion for major
cardiovascular events at 1 moth and 1 year after presentation.

MDCT imaging protocol. Beta-blocker metoprolol will be given orally (50 mg, 1 hr prior to
imaging) and/or intravenously (5 to 15 mg immediately prior to imaging), as necessary to
achieve a target heart rate of <70 bpm. Images will be acquired during suspended respiration
(256-channel iCT scanner, Philips). Initially, CTCA will be performed at rest according to
the standard clinical protocol. Then, regadenoson (Lexiscan, Astellas) will be administered
(0.4mg, i.v. bolus) at least 15 minutes later to ensure contrast clearance. An additional set
of images will be acquired 1 minute after the administration of regadenoson to ensure imaging
during peak vasodilator effect. Stress images will be acquired following injection of 50 ml
of iodinated contrast at a rate of 4 ml/sec, using prospective gating, in order to minimize
radiation exposure, resulting in an average of 2 mSv for the additional stress scan.

MDCT image analysis. Myocardial perfusion will be analyzed both at rest and during
vasodilator stress. Following manual initialization of endo- and epicardial boundaries in 5
to 6 slices, the endo- and epicardial 3D surfaces will be automatically estimated using the
level-set technique (Corsi, et al. 2005). The 3D region of interest confined between the
endocardial and epicardial surfaces will be identified as LV myocardium and divided into 3D
wedge-shaped myocardial segments, according to standard AHA segmentation (Cerqueira, et al.
2002): 6 basal, 6 mid-ventricular, and 4 apical segments. For each myocardial segment, mean
x-ray attenuation value will be automatically measured in Hounsfeld units (HU). Then x-ray
attenuation in all LV slices from base to apex will be used as described previously
(Kachenoura, et al. 2010) to generate a bull's eye display of myocardial attenuation
normalized by mean LV cavity attenuation, resembling MPI bull's eyes. Unlike our previous
study of fixed perfusion defects (Kachenoura, et al. 2010), in which the bull's eyes
displayed transmural attenuation, in this study we will also create bull's eyes of
subendocardial attenuation to optimize the visualization of stress-induced subendocardial
perfusion defects. In addition to the bull's eye displays of subendocardial attenuation, this
information will be mapped onto the reconstructed 3D-rendered endocardial surface to
facilitate the appreciation of the location and extent of perfusion abnormality. This 3D
display allows manually rotating the model to allow visualization from any angle, and
provides a more realistic view of the hypoenhanced area with less distortion than the bull's
eyes. Finally, for each myocardial segment, a quantitative index of extent and severity of
perfusion abnormality, Qh, will be calculated as a mathematical product of the number of
voxels with low attenuation in percent of the total volume of the segment, reflecting the
extent of the defect, and the difference between the attenuation in these voxels and the
previously determined normal attenuation in the same anatomic location, reflecting the
severity of the defect (Kachenoura, et al. 2010).

Objective detection of perfusion abnormalities. To allow objective detection of perfusion
abnormalities, data collected in our earlier pilot study were used to determine the
abnormality cutoff values for perfusion index Qh. Myocardial segments were divided into two
groups according to CTCA findings: segments supplied by coronary arteries with stenosis
located proximally to the specific segment and causing >50% luminal narrowing on CTCA, and
segments supplied by arteries without significant stenosis or with stenosis located distally
to the segment. Receiver operator curve (ROC) analysis was used to determine optimal
abnormality thresholds for index Qh for both rest and stress. We will use these thresholds in
the planned prospective study to objectively classify segmental myocardial perfusion as
normal or abnormal. Segments with Qh above the corresponding threshold will be considered
abnormal. A territory of an individual coronary artery will be considered abnormal when the
perfusion index is abnormal in at least one segment. For the patient-by-patient analysis,
abnormal perfusion will be diagnosed when at least one territory is abnormal.

Inter-technique comparisons. In order to allow comparisons between MDCT perfusion and ICA
findings, coronary anatomy depicted on each patient's MDCT volume rendering of the heart will
be used to determine the perfusion territory of each artery and its major branches, i.e. to
assign each myocardial segment to the territory of a specific coronary artery. This will
allow us to predict normal or abnormal perfusion in each segment based on the ICA
determinations of the presence, location and severity of stenosis. Significant stenosis will
be defined by a 50% and separately 70% decrease in luminal cross-sectional area. Comparisons
with ICA will be performed for perfusion index Qh by itself, and also as an adjunct to CTCA
findings, i.e. percent luminal narrowing and FFR. All inter-technique comparisons will be
performed on a segment-by-segment, vascular territory and patient-by-patient basis.
Inter-technique agreement will be assessed by calculating sensitivity, specificity, positive
and negative predictive values (PPV, NPV) and overall accuracy against the respective
reference standard. The accuracy of MDCT perfusion combined with CTCA findings will be
compared with that of CTCA alone to determine the added clinical value of MDCT perfusion
data.

Patient follow-up. Patients enrolled in the study will be followed up for a minimum of 1 year
from the stress test for major adverse cardiovascular events, including coronary
revascularization, hospitalization for unstable angina, myocardial infarction or death. This
information will be obtained from the patients' medical records, Social Security Death Index,
and by phone calls to the patients' homes. The collected follow-up data will be used to
determine the prevalence of adverse events in patients with coronary stenosis >50% and
abnormal perfusion, which will be compared to the remaining patients (chi-square test). The
follow-up data will also be used to study the ability of regadenoson-induced perfusion
abnormalities detected on MDCT images to predict event-free survival. This will be achieved
by univariate logistic regression analysis, followed by multivariate regression for variables
found to predict outcomes by the univariate regression. The predictive value of the
combination of stenosis on CTCA with perfusion defects will be compared with that of the
conventional markers, such as percent coronary luminal narrowing on CTCA alone, cardiac
enzymes, and TIMI risk score. Finally, we will determine whether the combination of these
conventional markers with the MDCT-derived perfusion data would predict patient outcomes
better than either one of them alone.

Anticipated results. We anticipate that perfusion abnormalities detected on MDCT images will
correlate with the findings of ICA, and when added to CTCA findings, will improve the
accuracy of MDCT evaluation of CAD compared to these reference standards. In addition, we
anticipate that when added to the conventional markers, such as cardiac enzymes and TIMI risk
score, vasodilator stress induced MDCT perfusion abnormalities will predict adverse
cardiovascular events better than these traditional indices alone.

Innovation and significance. To our knowledge, this will be the first study to validate
against ICA reference quantitative 3D evaluation of myocardial perfusion from MDCT images
obtained during vasodilator stress in patients referred for CTCA evaluation of acute chest
pain. These patients are increasingly referred for CTCA in large numbers, instead of nuclear
vasodilator stress testing, thus losing valuable physiologic information. Because the
addition of stress perfusion will allow elucidating the impact of coronary lesions in the
same test, such addition promises to improve the diagnostic performance of cardiac CT in the
evaluation of acute chest pain. This methodology may prove as a single-stop alternative to
costly serial testing. Moreover, if this comprehensive approach is proven to better predict
patient outcomes compared to current clinical algorithms, it may become a new tool for risk
stratification and planning individual patient's treatment. We anticipate that the results of
our study will support the use of this methodology in every patient with chest pain referred
for CTCA, similar to the routine use of vasodilator stress with MPI.

Inclusion Criteria:

- patients with chest pain referred for CT coronary angiography

Exclusion Criteria:

- allergy to iodine,

- renal dysfunction (creatinine >1.6 mg/dL)

- chronic obstructive pulmonary disease

- advanced heart block

- or systolic blood pressure <90 mmHg