Dipyridamole is the most common vasodilator used with positron emission tomography for the evaluation of patients with hypertrophic cardiomyopathy (HC). The aim of this study was to evaluate whether positron emission tomographic quantification of regional myocardial perfusion (rMP), myocardial blood flow (MBF), and coronary flow reserve are comparable between dipyridamole and the newer vasodilator regadenoson in HC. An additional aim was to evaluate the association between vasodilator-induced ST-segment depression on electrocardiography and myocardial flow in HC. Nitrogen-13 ammonia positron emission tomography was performed in 57 patients with symptomatic HC at rest and during vasodilator stress (peak) with either dipyridamole (0.56 mg/kg during 4-minute infusion) or regadenoson (0.4 mg fixed bolus dose) for assessment of electrocardiographic findings, rMP (17-segment American Heart Association summed difference score), MBF, and coronary flow reserve. The dipyridamole and regadenoson groups consisted of 28 and 29 patients respectively. Baseline characteristics, including rest MBF (0.92 ± 0.22 vs 0.89 ± 0.23 ml/min/g, p = 0.60), were similar between the 2 groups. During stress, the presence and severity of abnormal rMP (summed difference score 5.5 ± 5.5 vs 5.8 ± 6.7, p = 0.80), peak MBF (1.81 ± 0.44 vs 1.82 ± 0.50 ml/min/g, p = 0.90), and coronary flow reserve (2.02 ± 0.53 vs 2.12 ± 0.12, p = 0.50) were comparable between the dipyridamole and regadenoson groups. Fewer patients exhibited side effects with regadenoson (2 vs 7, p = 0.06). Vasodilator-induced ST-segment depression showed high specificity (about 92%) but low sensitivity (about 34%) to predict abnormal rMP (summed difference score ≥2). In conclusion, measurement of rMP and quantitative flow with positron emission tomography is similar between regadenoson and dipyridamole in patients with symptomatic HC. Regadenoson is tolerated better than dipyridamole and is easier to administer. Vasodilator-induced ST-segment depression is a specific but nonsensitive marker for the prediction of abnormal rMP in patients with HC.
Regadenoson is a new United States Food and Drug Administration–approved vasodilator agent administered as a single fixed dose (0.4 mg) that binds more selectively to the coronary A 2a adenosine receptors than dipyridamole or adenosine. At present, there are no reported studies investigating the clinical utility of regadenoson as a vasodilator stress agent for cardiac single photon emission computed tomography or positron emission tomographic (PET) imaging in patients with hypertrophic cardiomyopathy (HC). The main aim of the present study was to compare regadenoson to the standard agent, dipyridamole, for regional myocardial perfusion (rMP) and quantification of myocardial blood flow (MBF) with PET imaging in patients with HC. In addition, we evaluated the association between vasodilator-induced electrocardiographic changes and myocardial perfusion in patients with HC.
Methods
We performed a retrospective analysis of consecutive patients with histories of HC diagnosed by echocardiography, who were referred for cardiac PET imaging for clinical indications from June 2009 to February 2012. Subjects with histories of coronary artery disease, previous surgical myectomy, or alcohol septal ablation were excluded. The diagnosis of HC was based on echocardiographic criteria by demonstrating left ventricular (LV) hypertrophy with wall thickness ≥15 mm. LV outflow tract gradients were identified by Doppler echocardiography under basal conditions and after the Valsalva maneuver and amyl nitrite (inhaled) challenge to elicit latent obstruction during provocation.
All patients underwent imaging using a GE Discovery VCT PET/computed tomographic system (GE Healthcare, Waukesha, Wisconsin). Patients were positioned with the assistance of a computed tomographic topogram, and a low-dose computed tomographic scan was performed for attenuation correction of PET emission data. Subsequently, PET images were acquired using a same-day rest/stress protocol as follows: nitrogen-13 ammonia (approximately 10 mCi) was injected intravenously as a bolus, and a 2-dimensional list-mode PET scan was obtained over 20 minutes. Approximately 60 minutes after injection of the rest dose, dipyridamole or regadenoson was administered for vasodilator stress. Dipyridamole was the drug of choice at the beginning of our PET protocol for HC in 2009; however, beginning in June 2011, all patients with HC were stressed with regadenoson. Dipyridamole (0.56 mg/kg) was infused over a period of 4 minutes, followed by a second dose of nitrogen-13 ammonia (approximately 10 mCi) 4 minutes after the end of dipyridamole infusion. Regadenoson (0.4 mg) was injected as a bolus (about 15 to 20 seconds), followed by a 5-ml saline flush, and nitrogen-13 ammonia was administered 30 seconds later. Stress acquisition began concomitantly with the second nitrogen-13 ammonia injection, and all other parameters were the same as during rest. Heart rate, blood pressure, and a 12-lead electrocardiogram were recorded before, during, and after the completion of the stress protocol. List-mode data were resampled to create various images, including static (4-minutes prescan delay), electrocardiographically gated (8 bins per cardiac cycle), and 36-frame-dynamic (20 × 6 seconds, 5 × 12 seconds, 4 × 30 seconds, 5 × 60 seconds, and 2 × 300 seconds).
For the electrocardiographic analysis, patients with left bundle branch block and electronically paced ventricular rhythms were excluded. All electrocardiograms were evaluated at baseline for the presence of LV hypertrophy (by the Sokolow-Lyon criterion or Cornell product) and ST-T abnormalities. Vasodilator-induced ST-segment changes were evaluated in all leads. Any ST-segment deviation from the J point at baseline was subtracted from that observed during pharmacologic stress to determine the overall ST-segment shift in each lead. Negative shift values signified depression and positive represented elevation of the ST segment. ST changes were summed and averaged out in leads II, III, and aVF to assess mean ST-segment depression in the inferior leads, and this was also performed in leads V 3 , V 4 , and V 5 to estimate mean ST-segment depression in the precordial leads. For each patient, the final ST-segment depression reported consisted of the largest (maximum) ST shift between the mean inferior and mean precordial leads. Reciprocal ST-segment elevation in leads aVR and V 1 was also evaluated.
The CardIQ Physio package (GE Healthcare) was used for analysis of the LV ejection fraction at rest and during stress. PET rMP was semiquantitatively assessed for each set of rest/stress images using the summed difference score (SDS) and the standard 17-segment 5-point scale, which is the American Heart Association model. An SDS ≥2 was considered abnormal in this study. Quantitative Gated SPECT software (Cedars-Sinai Medical Center, Los Angeles, California) was used for regional wall motion assessment using a 5-point, 17-segment LV model scoring system. This method derived rest and stress wall motion scores. Pharmacologically induced wall motion segmental abnormalities were considered present when the stress-rest wall motion score difference was >1. The Munich Heart package was used for absolute flow quantification. Software computation of MBF is based on a well-established 2-tissue compartment tracer kinetic model, as previously described. Global MBF during vasodilator-stress and rest was measured in milliliters per minute per gram. Coronary flow reserve (CFR) was determined as the ratio of stress MBF to rest MBF (unitless). Rest MBF and CFR were also corrected for the rate-pressure product, a product of rest heart rate and systolic blood pressure, using the following equation : corrected rest MBF = observed rest MBF × 8,500 (mean rate-pressure product of the study cohort)/rate-pressure product of each subject. Corrected CFR was calculated as observed stress MBF/corrected rest MBF.
Statistical analyses were performed using SPSS version 19.0 (SPSS, Inc., Chicago, Illinois). Continuous variables are presented as mean ± SD. An independent-measures Student’s t test was used to assess the differences between the parametric subgroups, and the Mann-Whitney U test was used for the nonparametric groups. Categorical variables were compared between groups using chi-square tests and are presented as percentages. Receiver-operating characteristic curves were used to determine the optimal maximum ST-segment depression cut point that identifies abnormal rMP by PET imaging. A p value <0.05 was considered statistically significant.
Results
A total of 57 patients with HC were included in this cohort: 28 were stressed with dipyridamole (group 1) and 29 with regadenoson (group 2). Baseline and echocardiographic characteristics, including maximal wall thickness, were highly comparable between the 2 groups ( Table 1 ). Heart rate and mean arterial blood pressure were statistically similar between the groups before the administration of either stress agent, although there was a trend toward higher rate-pressure products in the regadenoson group ( Table 2 ). During vasodilator stress, peak heart rate showed a trend for higher values with regadenoson compared to dipyridamole, but mean blood pressure decreased to a similar nadir with the 2 agents ( Table 2 ). From a safety profile perspective, 7 patients (26%) experienced side effects with dipyridamole, including chest tightness or pain (n = 4), nausea (n = 3), and hypotension (n = 3), requiring aminophylline in all instances. In contrast, only 2 patients (7%) (p = 0.06) experienced side effects after regadenoson administration (headache and chest pain). Aminophylline was given to 1 patient with chest pain. The prevalence of abnormal rMP (71% vs 83%, p = 0.30) and the severity of reversible perfusion defects (SDS 5.5 ± 5.5 vs 5.8 ± 6.7, p = 0.80) were similar in patients who underwent dipyridamole and regadenoson vasodilator stress. At baseline, global MBF was similar between the dipyridamole and regadenoson group ( Figure 1 ). After pharmacologic administration, the hyperemic MBF achieved in the entire left ventricle with dipyridamole was similar to that obtained with regadenoson ( Figure 1 ). As a result, global CFR was not significantly different between patients with HC stressed with dipyridamole or regadenoson (2.02 ± 0.53 vs 2.12 ± 0.12, p = 0.50). After correction for baseline differences in heart rate and systolic blood pressure, rest MBF (0.99 ± 0.26 vs 0.88 ± 0.22 ml/min/g, p = 0.10) and CFR (1.90 ± 0.52 vs 2.19 ± 0.74, p = 0.10) remained comparable between the dipyridamole and regadenoson groups. Regionally, no significant differences between dipyridamole and regadenoson in hyperemic MBF were observed in any myocardial wall. In the lateral wall, there was a slight trend toward higher peak flow values with regadenoson ( Figure 2 ). Figure 3 shows PET images of 1 patient with HC who underwent pharmacologic stress with dipyridamole and regadenoson 9 months apart.
| Characteristic | Dipyridamole Group | Regadenoson Group | p Value |
|---|---|---|---|
| (n = 28) | (n = 29) | ||
| Age (years) | 51 ± 16 | 51 ± 12 | 0.90 |
| Men | 16 (57%) | 17 (59%) | 0.90 |
| Chest pain and/or dyspnea | 25 (89%) | 27 (93%) | 0.60 |
| Syncope | 4 (14%) | 4 (14%) | 0.90 |
| Hypertension | 12 (43%) | 13 (45%) | 0.90 |
| Diabetes mellitus | 3 (11%) | 2 (7%) | 0.60 |
| β blockers | 25 (89%) | 26 (90%) | 0.90 |
| Calcium channel blockers | 5 (18%) | 4 (14%) | 0.70 |
| Family history of sudden cardiac death | 7 (25%) | 8 (28%) | 0.80 |
| Family history of HC | 5 (18%) | 6 (21%) | 0.80 |
| Maximal myocardial wall thickness (cm) | 2.16 ± 0.57 | 2.17 ± 0.49 | 0.90 |
| LV posterior wall (cm) | 1.20 ± 0.24 | 1.19 ± 0.30 | 0.90 |
| LV ejection fraction (%) | 60 ± 7 | 58 ± 9 | 0.40 |
| Rest outflow tract gradient (mm Hg) | 27 ± 24 | 25 ± 31 | 0.70 |
| Provoked outflow tract gradient (mm Hg) | 69 ± 53 | 47 ± 47 | 0.10 |
| Nonobstructive outflow tract gradients | 9 (32%) | 17 (58%) | |
| Obstructive outflow tract gradients | 10 (36%) | 6 (21%) | 0.10 |
| Latent outflow tract obstruction | 9 (32%) | 6 (21%) |
| Characteristic | Dipyridamole Group | Regadenoson Group | p Value |
|---|---|---|---|
| (n = 28) | (n = 29) | ||
| Baseline heart rate (beats/min) | 61 ± 10 | 64 ± 10 | 0.30 |
| Peak heart rate (beats/min) | 86 ± 15 | 92 ± 14 | 0.10 |
| Heart rate difference (beats/min) | 25 ± 13 | 28 ± 10 | 0.30 |
| Baseline systolic blood pressure (mm Hg) | 133 ± 16 | 138 ± 23 | 0.40 |
| Baseline diastolic blood pressure (mm Hg) | 67 ± 12 | 70 ± 11 | 0.40 |
| Baseline mean blood pressure (mm Hg) | 89 ± 12 | 92 ± 14 | 0.40 |
| Mean blood pressure nadir (mm Hg) | 82 ± 14 | 86 ± 14 | 0.30 |
| Mean blood pressure difference (mm Hg) | −8 ± 8 | −7 ± 7 | 0.70 |
| Baseline rate-pressure product (beats/min · mm Hg) | 8,114 ± 1,576 | 8,818 ± 2,019 | 0.10 |
A total of 53 patients had evaluable electrocardiograms at baseline (26 with dipyridamole and 27 in the regadenoson group). Patients were divided into quartiles on the basis of the maximum vasodilator-induced ST-segment shift ( Table 3 ). Significant ST-segment depression (≥1 mm) occurred in 14 patients in a diffuse pattern (leads II, III, aVF, and V 3 to V 5 ) and was observed equally with dipyridamole and regadenoson. Reciprocal ST-segment elevation (≥1 mm) was observed in leads aVR, aVL, and V 1 . This was seen in 3 patients (11%) who received dipyridamole and 1 subject (4%) (p = 0.30) stressed with regadenoson. Using the standard cut point of a ≥1-mm shift, vasodilator-induced ST-segment depression showed high specificity (about 92%) but low sensitivity (about 34%) to predict abnormal rMP. When maximal ST-segment shift during stress electrocardiography was plotted against abnormal rMP on a receiver-operating characteristics curve, ST-segment deviation >0.67 mm from baseline showed a somewhat better cut point to predict abnormal rMP (sensitivity about 51%, specificity about 92%). Figure 4 illustrates the high specificity and Figure 5 the low sensitivity of ST-segment shifts to predict abnormal rMP. Patients with greater ST-segment deviations had greater LV outflow tract gradients, LV ejection fractions, and ST-T abnormalities on electrocardiography at baseline and a higher incidence of vasodilator-induced tachycardia, chest pain, LV systolic dysfunction, and regional wall motion abnormalities compared to patients with lesser ST-segment shifts ( Table 3 ). Maximal wall thickness, corrected resting MBF, stress MBF, and CFR were similar among patients with varying degrees of ST shifts. Abnormal rMP showed a trend toward higher incidence in patients with greater ST shifts ( Table 3 ). Patients with vasodilator-induced chest pain (n = 5) had significantly greater wall thickness (2.54 ± 1.00 vs 2.08 ± 0.40 cm, p = 0.045) and lower corrected CFR (1.50 ± 0.44 vs 2.13 ± 0.65, p = 0.04). All patients with vasodilator-induced chest pain had abnormal rMP and wall motion abnormalities on PET; 4 of 5 patients had ST depression ≥1 mm on stress electrocardiography.
| Characteristic | Vasodilator-Induced ST-Segment Deviation Groups (n = 53) | p Value | |||
|---|---|---|---|---|---|
| <0.1 mm (n = 12) | 0.1–0.4 mm (n = 14) | 0.5–0.9 mm (n = 13) | ≥1.0 mm (n = 14) | ||
| Echocardiography | |||||
| Maximal myocardial wall thickness (cm) | 2.28 ± 0.44 | 2.09 ± 0.56 | 1.97 ± 0.36 | 2.15 ± 0.57 | 0.50 |
| LV posterior wall (cm) | 1.18 ± 0.25 | 1.16 ± 0.28 | 1.17 ± 0.21 | 1.18 ± 0.22 | 0.90 |
| Rest outflow tract gradient (mm Hg) | 11 ± 9 | 21 ± 26 | 23 ± 18 | 38 ± 32 | 0.04 |
| Provoked outflow tract gradient (mm Hg) | 52 ± 50 | 43 ± 42 | 44 ± 27 | 88 ± 67 | 0.06 |
| Nonobstructive outflow tract gradients | 6 (50%) | 8 (57%) | 5 (38.5%) | 5 (36%) | 0.20 |
| Obstructive outflow tract gradients | 1 (8%) | 3 (21.5%) | 3 (23%) | 7 (50%) | |
| Latent outflow tract obstruction | 5 (42%) | 3 (21.5%) | 5 (38.5%) | 2 (14%) | |
| Baseline electrocardiography | |||||
| Strain pattern and/or T-wave inversion | 3 (25%) | 7 (50%) | 11 (85%) | 11 (79%) | 0.007 |
| Early repolarization changes | 3 (25%) | 3 (21%) | 1 (8%) | 1 (7%) | 0.50 |
| Any ST-T abnormality | 6 (50%) | 10 (71%) | 12 (92%) | 12 (86%) | 0.07 |
| LV hypertrophy | 3 (25%) | 7 (50%) | 7 (54%) | 8 (57%) | 0.40 |
| Hemodynamics | |||||
| Rest heart rate (beats/min) | 58 ± 10 | 66 ± 11 | 61 ± 8 | 67 ± 8 | 0.053 |
| Rest systolic blood pressure (mm Hg) | 139 ± 25 | 130 ± 18 | 133 ± 18 | 136 ± 19 | 0.70 |
| Rate-pressure product (beats/min · mm Hg) | 8,106 ± 2,368 | 8,666 ± 1,907 | 8,123 ± 1,507 | 9,045 ± 1,741 | 0.50 |
| Peak heart rate (beats/min) | 84 ± 12 | 90 ± 9 | 85 ± 11 | 101 ± 16 | 0.002 |
| Heart rate difference (beats/min) | 26 ± 10 | 24 ± 11 | 24 ± 7 | 35 ± 13 | 0.03 |
| Regadenoson as vasodilator stress agent | 9 (75%) | 6 (43%) | 5 (38.5%) | 7 (50%) | 0.30 |
| Vasodilator-induced chest pain | 0 | 1 (7%) | 0 | 4 (29%) | 0.03 |
| Stress electrocardiography | |||||
| Mean ST-segment shift, leads II, III, and aVF (mm) | −0.00 ± 0.03 | −0.18 ± 0.13 | −0.59 ± 0.20 | −1.05 ± 0.66 | <0.0001 |
| Mean ST-segment shift, leads V 3 to V 5 (mm) | −0.00 ± 0.02 | −0.15 ± 0.15 | −0.63 ± 0.20 | −1.58 ± 0.63 | <0.0001 |
| Maximum ST-segment shift (mm) | −0.01 ± 0.01 | −0.21 ± 0.12 | −0.68 ± 0.16 | −1.60 ± 0.62 | <0.0001 |
| PET imaging | |||||
| SDS | 3.4 ± 3.2 | 7.6 ± 9.4 | 5.0 ± 3.1 | 7.4 ± 6.1 | 0.30 |
| Abnormal rMP | 8 (67%) | 9 (64%) | 12 (92%) | 13 (93%) | 0.10 |
| Rest LV ejection fraction (%) | 53 ± 8 | 59 ± 9 | 61 ± 8 | 62 ± 7 | 0.04 |
| Stress ejection fraction (%) | 48 ± 12 | 51 ± 13 | 44 ± 11 | 43 ± 11 | 0.30 |
| Ejection fraction reserve (%) | −5 ± 12 | −8 ± 11 | −17 ± 12 | −19 ± 13 | 0.01 |
| Negative LV ejection fraction reserve | 7 (58%) | 10 (71%) | 11 (85%) | 14 (100%) | 0.055 |
| Stress-induced wall motion abnormalities | 3 (25%) | 8 (57%) | 10 (77%) | 12 (86%) | 0.008 |
| Rest MBF (ml/min/g) | 0.75 ± 0.14 | 0.95 ± 0.22 | 0.98 ± 0.24 | 0.95 ± 0.22 | 0.04 |
| Stress MBF (ml/min/g) | 1.75 ± 0.41 | 1.81 ± 0.56 | 1.92 ± 0.44 | 1.88 ± 0.42 | 0.80 |
| CFR | 2.37 ± 0.49 | 1.97 ± 0.60 | 2.04 ± 0.60 | 2.02 ± 0.44 | 0.20 |
| Corrected rest myocardial flow (ml/min/g) | 0.83 ± 0.22 | 0.96 ± 0.27 | 1.05 ± 0.28 | 0.91 ± 0.20 | 0.20 |
| Corrected CFR | 2.22 ± 0.66 | 1.99 ± 0.75 | 1.95 ± 0.65 | 2.15 ± 0.60 | 0.70 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
