The aim of this study was to compare the efficacy of myocardial perfusion (MP) and wall motion (WM) analysis obtained with real-time myocardial contrast echocardiography (RTMCE) and two widely used contrast agents in detecting coronary artery disease after injection of the vasodilator regadenoson.
One hundred fifty patients were studied at two academic centers using regadenoson (400-μg intravenous bolus) vasodilator stress RTMCE (7.5% Optison infusion [ n = 50] or 1.5% Definity infusion [ n = 100]). Both MP and WM with RTMCE were analyzed at rest and after regadenoson bolus. Comparisons of WM and MP sensitivity, specificity, and accuracy were made. Quantitative angiography was performed in all patients within 1 month of the regadenoson stress study (>50% and >70% diameter stenosis was considered significant). Reviewers were blinded to all clinical and quantitative angiographic data.
Rate-pressure product after regadenoson was higher in Optison than Definity patients ( P = .004). Using a 50% diameter stenosis on quantitative angiography as a reference standard, overall sensitivity, specificity, and accuracy for combined WM and MP analysis were not different for both agents (Optison, 77%, 64%, and 73%; Definity, 80%, 74%, and 78%; P = NS). The sensitivity, specificity, and accuracy of WM analysis alone for Optison were 68%, 71%, and 69% compared with 60%, 72%, and 66% for Definity ( P = NS). Adding MP analysis improved the sensitivity and accuracy of Definity for detecting both >50% and >70% stenoses ( P < .001 vs WM), while MP analysis did not improve the sensitivity of Optison for detecting either >50 or >70% stenoses.
RTMCE during regadenoson stress using either Optison or Definity is a rapid and effective method for the detection of coronary artery disease. The ability of MP imaging to improve WM accuracy may depend on the rate-pressure product achieved.
The overall sensitivity and specificity of Definity and Optison regadenoson stress were comparable.
No significant side effects were seen with either agent during regadenoson stress.
Perfusion added to WM analysis only during Definity regadenoson stress.
Perfusion imaging improved the detection of multivessel CAD when using Definity.
The development of a noninvasive test for the evaluation of myocardial perfusion (MP) is important for early diagnosis and appropriate management of patients being evaluated for coronary artery disease (CAD). Real-time myocardial contrast echocardiography (RTMCE) uses compressible microbubbles similar to the rheology of red blood cells. Destruction of microbubbles and observation of replenishment into the myocardium enables the evaluation of both wall motion (WM) and MP in real time.
Previous studies have shown that qualitative RTMCE with very low mechanical index (MI) imaging accurately detects abnormal MP at rest and during adenosine or dipyridamole vasodilator stress. Regadenoson (Lexiscan; Astellas Pharma, Tokyo, Japan) is the first selective adenosine A 2A receptor agonist (given as a fixed-dose intravenous bolus) to be approved by the US Food and Drug Administration and is currently used clinically for radionuclide MP studies. When administered as an intravenous bolus, regadenoson has a rapid onset and short duration of action. Our previous observations confirmed the feasibility of both WM and MP analysis with RTMCE during regadenoson stress to detect CAD. However, despite numerous clinical trials demonstrating detection of MP abnormalities at rest and during stress, ultrasound contrast agents (UCAs) are not currently approved by the Food and Drug Administration for MP imaging with either of the commercially available agents. MP abnormalities are often induced in the absence of WM abnormalities with vasodilator stress agents such as regadenoson because of capillary blood flow and volume reserve abnormalities that are not always associated with ischemia. Therefore, using UCAs to detect MP in addition to WM after regadenoson administration would have significant clinical impact, because the study could have the added sensitivity of MP imaging and also be performed rapidly at the bedside without radiation risk or costs associated with radiolabeled tracers. The objective of the present study was to compare the feasibility and diagnostic accuracy (sensitivity, specificity, and accuracy) of two different UCAs to detect CAD during regadenoson stress.
The study was approved by the institutional review boards at both the Mayo Clinic and the University of Nebraska Medical Center (UNMC). An approved regadenoson investigational new drug application to the Food and Drug Administration (IND No. 104,710) was obtained.
From July 2009 to April 2014, 150 consenting patients referred for cardiac catheterization at both institutions, for clinically indicated coronary angiography (because of suspected CAD, abnormal stress test results, or the presence of multiple risk factors for CAD) were screened for eligibility for this study. These included 100 patients in whom the feasibility, safety, and accuracy of RTMCE with Definity (Lantheus Medical Imaging, North Billerica, MA) at different time intervals after regadenoson bolus injections have already been published during an enrollment period from July 2009 through February 2011. Once Optison (GE Healthcare, Little Chalfont, United Kingdom) became available again in the United States in 2011 and was added to the investigational new drug application, the study sponsor (Astellas Pharma) funded a study examining regadenoson Optison sensitivity and specificity using the same ultrasound systems and settings and the same recruitment criteria. Therefore, the studies with Optison were performed from January 2012 through April 2014. As a result, there was a comparison made between two prospectively acquired sets of data. Criteria for the exclusion of patients from the study were the following: ejection fraction < 40% by biplane echocardiography, New York Heart Association functional class III or IV symptoms, history of high-grade atrioventricular block without pacemaker, use of dipyridamole or caffeine within 24 hours of the stress test, severe reactive airway disease, and reduced life expectancy.
RTMCE was performed using ultrasound machines (iE33 [Philips Medical Systems, Andover, MA] or Acuson Sequoia [Siemens Medical Solutions, Mountain View, CA]) equipped with broadband transducers and low-MI contrast-specific presets. Apical four-, three-, and two-chamber views and short-axis views were acquired with MIs ≤ 0.20, frame rates of 20 to 25 Hz, and a focus at the mitral valve level or, in cases of a possible apical “pseudodefect,” the apical level. Time gain compensation and two-dimensional gain settings were adjusted and kept constant throughout the study, with potentiometers moved slightly higher in the near field ( Figure 1 ). A real-time destruction-replenishment technique was used in each view, with a transient, high-MI (1.3) imaging flash applied to destroy myocardial microbubbles and replenishment observed over a time period of five to 10 cardiac cycles. Images were stored digitally for offline analysis. A continuous infusion of perflutren lipid microspheres (1.5% diluted Definity) or perflutren albumin microspheres (7.5% diluted Optison) at an average rate of 4 mL/min was administered through an intravenous line. This translated into an average infusion of 7.2 × 10 8 microbubbles/min for Definity and 2.4 × 10 8 microbubbles/min for Optison. These dilutions were determined on the basis of what produced optimal left ventricular cavity and myocardial opacification without attenuation in the basal segments at the 4 mL/min infusion rate. The microbubble number for Optison was smaller most likely because this agent has a slightly higher reported size distribution in its package insert description (2.5–4.5 μm) than Definity (1.1–3.3 μm). Contrast infusion was started 1 min before real-time myocardial contrast echocardiographic acquisition at rest and was kept constant (achieving optimal myocardial enhancement without left ventricular cavity attenuation) until the end of imaging. Stress real-time myocardial contrast echocardiographic images were acquired in the same apical planes again in the 4-min period after the regadenoson bolus administration ( Figure 2 ). This time period was determined in the previously published feasibility study to be the optimal time to assess MP with RTMCE following the bolus injection of regadenoson.
Digitized real-time myocardial contrast echocardiographic images were analyzed offline by one experienced observer (F.X.), who was blinded to all clinical history and coronary angiographic data. The 17-segment model was used to evaluate left ventricular WM and MP simultaneously. Segments were ascribed to coronary territories where the four apical segments—apex, base–mid anterior wall, base–mid anteroseptum, and mid inferior septum—were assigned to the anterior circulation (left anterior descending coronary artery [LAD]). The base–mid anterolateral and base–mid inferolateral walls were assigned to the left circumflex coronary artery (LCx), and the base–mid inferior wall and the base–inferior septum were assigned to the right coronary artery (RCA). Posterior circulation included the LCx and RCA.
WM was evaluated with the established American Society of Echocardiography grading criteria: 1 = normal, 2 = hypokinetic, 3 = akinetic, and 4 = dyskinetic. Biplane Simpson quantitation was used to quantify resting left ventricular ejection fraction.
Comparison of rest and stress real-time myocardial contrast echocardiographic contrast enhancement images was done, and perfusion was graded as abnormal or normal. A contrast perfusion abnormality was defined as visually decreased contrast replenishment and enhancement in the frames near end-systole after a brief high-MI (flash) impulse at the following time points after the impulse: 4 sec (during rest imaging) and 2 sec (during regadenoson stress imaging). A delay in replenishment in any segment was considered abnormal, once attenuation had been ruled out. Attenuation was defined as decreased replenishment and overall enhancement within a segment that was present before and after high-MI impulses. This was also associated with decreased signal intensity in adjacent nonmyocardial regions surrounding the segment of interest.
MP or WM abnormalities were further classified as reversible defects (regional perfusion defects after regadenoson bolus not observed under resting conditions) within at least two contiguous segments, while a fixed defect was defined as defects seen at rest and stress, without change in distribution following regadenoson.
Side effects at any time within 1 hour of the regadenoson bolus injection were recorded and included a checklist of those previously reported with regadenoson stress during radionuclide imaging.
Coronary angiography was evaluated in all patients within 1 month of RTMCE. Quantitative analysis was independently performed by an experienced operator (E.O.) using prior methodology. Significant CAD was defined by quantitative angiography (QA) using both a >50% and a 70% diameter cutoff in a major coronary artery or branch vessel, using the smallest diameter of the stenosis on multiple projections and using as a reference the diameter of the vessel immediately proximal to the stenosis. Diagnostic accuracies of WM and MP, independently and combined, were determined using QA as the reference standard for the cutoffs described (50% and 70%) on territorial and patient levels.
Continuous data are reported as mean ± SD and were compared using paired and unpaired t test. Frequencies are used to report categorical variables, which were compared using χ 2 tests. The sample size justification was based on previous analyses in this same setting using an assumed disease prevalence of 60% for a >50% diameter cutoff on QA. The 100 patients receiving Definity were recruited first (50 at the Mayo Clinic, 50 at the UNMC), and on the basis of disease prevalence and accuracies observed with this cohort, a smaller sample of 50 patients receiving Optison (25 at the Mayo Clinic, 25 at the UNMC) was expected to allow adequate statistical comparisons of WM and MP between contrast agents. These statistical comparisons (overall and across groups of contrast agents) included sensitivity, specificity, and accuracy. All comparisons (WM and MP) were made on a patient and coronary artery territory (CAT) basis with 95% CIs. A true-positive result, on a patient basis, was defined as the presence of a perfusion defect (fixed or reversible) or WM abnormality (fixed or reversible) in any territory and the presence of a >50% or >70% coronary stenosis in any vessel on QA. Analysis was done with JMP version 7.0 software (SAS Institute, Cary, NC) and SAS/STAT version 9.3 (SAS Institute). Significance levels were set at a two-sided P value of .05.
The number of patients screened at both participating institutions in both cohorts and those meeting the inclusion criteria are illustrated in Figure 3 . A total of 150 patients in both cohorts (Definity and Optison) were consented for simultaneous echocardiographic imaging with the administration of UCAs and regadenoson stress. Clinical characteristics are summarized in Table 1 . The indications for coronary angiography in both cohorts were similar: for the Definity cohort, abnormal dobutamine stress echocardiographic results in 14%, abnormal adenosine radionuclide findings in 7%, abnormal treadmill stress test results in 19%, and high suspicion for CAD (as part of a workup for chest pain, preoperative noncardiac surgery, or valvular heart disease) in 60%; for the Optison cohort, abnormal dobutamine stress echocardiographic results in 16%, abnormal adenosine radionuclide findings in 6%, abnormal treadmill stress test results in 16%, and high suspicion for CAD (as part of a workup for chest pain, preoperative noncardiac surgery, or valvular heart disease) in 62%. The mean imaging duration (for rest and stress imaging) for RTMCE was longer for the Optison cohort (10.2 ± 6.3 min) than the Definity cohort (8.6 ± 2.8 min) ( P = .033).
|Variable||Definity cohort ( n = 100)||Optison cohort ( n = 50)||P ∗|
|Age (y)||62 ± 11||61 ± 11||.601|
|Body mass index (kg/m 2 )||32 ± 7||30 ± 7||.101|
|Resting ejection fraction (%)||57 ± 7||56 ± 8||.433|
|Resting HR (beats/min)||64 ± 12||67 ± 12||.151|
|Peak HR (beats/min)||84 ± 12||90 ± 14||.007|
|Resting SBP (mm Hg)||130 ± 22||132 ± 22||.601|
|Peak SBP (mm Hg)||124 ± 21||130 ± 25||.124|
|Resting DBP (mm Hg)||73 ± 13||75 ± 12||.364|
|Peak DBP (mm Hg)||67 ± 9||71 ± 13||.029|
|Resting RPP||8,838 ± 2,174||8,754 ± 1,846||.815|
|Peak RPP||10,431 ± 2,473||11,738 ± 2,862||.004|
No significant adverse side effects were noted during the injection of either Definity or Optison. Nonserious adverse effects during regadenoson and Definity administration included chest pain in 7% of patients, shortness of breath in 48%, headache in 23%, and back pain in 4%. Similarly, during regadenoson and Optison administration, 14% of patients experienced chest pain, 52% shortness of breath, and 36% headache. All symptoms persisted no longer than 5 min after completion of stress imaging. Because of machine availability at the UNMC, the Siemens system was used 23 times for the Definity regadenoson study and the Philips system 27 times. For the Optison regadenoson study, the Philips system was used for all 25 studies. At the Mayo Clinic, the Siemens system was used for all Definity and Optison studies.
Regadenoson administration resulted in significant increases (rest vs stress) in heart rate ( P < .0001) and rate-pressure product ( P < .0001) at both cohorts. Peak heart rate, diastolic blood pressure, and rate-pressure product achieved were slightly, but significantly, higher in the Optison cohort than the Definity cohort ( P = .007, P = .029, and P = .004, respectively). No differences in oxygen saturation were noted between the Optison and Definity patients.
Coronary Angiographic Data
Coronary angiographic results were available in 147 patients (98%); all three with unavailable data were in the Optison cohort. A total of 85 (53% of Definity patients and 64% of Optison patients) of the 147 patients (58% overall) had >50% diameter stenoses in at least one CAT, whereas 65 patients (38% of Definity patients and 54% of Optison patients) (44% overall) had >70% diameter stenoses. Multivessel coronary stenoses (with >50% diameter cutoff) were present in 52 patients (32% of Definity patients and 40% of Optison patients), while multivessel CAD using a >70% diameter cutoff was present in 28 patients (19% of Definity patients and 18% of Optison patients). On the CAT level, 156 territories were supplied by >50% diameter stenoses (59 in the LAD, 44 in the LCx, and 53 in the RCA). There were no significant differences in disease prevalence between the two cohorts (53% for Definity, 64% for Optison; P = .268).
MP and WM Comparisons
Real-time myocardial contrast echocardiographic data were feasible for analysis in 146 of 150 patients (97%). Suboptimal images precluded perfusion assessment in four patients (two in each cohort). On a CAT basis, 433 of 450 territories (96%) were analyzable for MP. MP in 17 CATs (nine in Definity patients and eight in Optison patients) was not analyzable (four in the LAD, four in the LCx, and nine in the RCA).
MP defects after regadenoson stress were observed in 81 patients (55%) (53% with Definity and 56% with Optison). Of the analyzable CATs by RTMCE, 112 (26%) were abnormal for MP (75 with Definity and 37 with Optison). The LAD was abnormal in 49 (33 reversible, 10 fixed, and six mixed), the LCx was abnormal in 24 (15 reversible, six fixed, and three mixed), and the RCA was abnormal in 39 (16 reversible, 17 fixed, and six mixed). Perfusion defects after regadenoson stress in more than one CAT were observed in 27 patients (19% with Definity and 16% with Optison).
WM was abnormal in 70 patients (47%) (45% with Definity and 50% with Optison) and 91 CATs: 36 in the LAD (19 reversible, 12 fixed, and five mixed), 22 in the LCx (14 reversible, seven fixed, and one mixed), and 33 in the RCA (10 reversible, 19 fixed, and four mixed).
Compared with QA, overall analysis of both WM and MP had 77% sensitivity (95% CI, 67%–85%) and 73% specificity (95% CI, 57%–84%) for detecting CAD of >50% severity, on a patient level. MP sensitivity and accuracy were significantly better than WM sensitivity and accuracy in the Definity cohort ( P < .001 for both), whereas specificities were similar ( Table 2 ). MP sensitivity was not better than WM sensitivity in the Optison cohort ( P = .34). Similar findings were observed for CAD stenosis detection of >70% severity on a patient level ( Table 2 ). In the combined analysis of WM and RTMCE, sensitivity, specificity, and accuracy for the Definity cohort were 80% (95% CI, 67%–88%), 74% (95% CI, 57%–83%), and 78% (95% CI, 63%–85%), respectively. For the Optison cohort, these values were 77% (95% CI, 60%–89%), 64% (95% CI, 39%–84%), and 73% (95% CI, 56%–86%), respectively ( P = NS for any parameter).