Mechanism of decreased sensitivity of dobutamine associated left ventricular wall motion analyses for appreciating inducible ischemia in older adults

The study was approved by the Institutional Review Board of Wake Forest Health Sciences, and each participant provided witnessed written informed consent. This study was registered within the United States on Clinicaltrials.gov (NCT00542503) and was performed in accordance with the National Institutes of Health grants R01HL076438 and P30AG21332. In addition, this study complied with the Declaration of Helsinki. The purpose of this initiative was to utilize advanced DCMR techniques to identify rest and stress-induced cardiac and vascular abnormalities in older individuals that are associated with CV disease. In this study, each participant underwent an interview to collect and record information pertinent to CV disease (e.g., demographics, CV risk factors, etc.) and then underwent DCMR in which rate pressure product (RPP), aortic pulse wave velocity (PWV), LV myocardial oxygen demand (pressure volume area, PVA, stroke work, SW), mass, volumes, concentricity, and the presence of WMA and first pass gadolinium enhanced perfusion defects (PDs) indicative of ischemia were measured and recorded.

The study included participants from the rural counties of Central and Western North Carolina that exhibited a >5 years presence of risk factors (e.g., hypertension, diabetes, or CAD) for a cardiac event, but no history of myocardial infarction within 6 months of enrollment, no contraindication to intravenous dobutamine or gadolinium based contrast nor the performance of a CMR exam (such as the presence of incompatible bio-metallic implants or claustrophobia). None of the subjects had angina or symptoms of heart failure in the 12 months prior to enrollment. Recruitment of study participants was achieved through newspaper and television advertisements, or mailings to randomly selected individuals over the age of 55, but <85 years within Forsyth, Davie, and Davidson counties of Northwest North Carolina [4,5,11].

The DCMR protocol was accomplished according to previously published techniques [4,5] with images that were acquired on a 1.5-T (Siemens Avanto) whole-body imaging system using a phased-array cardiothoracic surface coil placed on the chest. Dobutamine was infused incrementally from low (7.5 mcg/kg/min) to peak dose (20 to 40 mcg/kg/min), along with atropine (up to 1.5 mg), to achieve 80% of the maximum predicted heart rate response (MPHRR) for age. Cine images of the left ventricle were obtained in multiple contiguous short axis slices (apex to base) and in 3 long axis views (2, 3, and 4 chamber) at rest, low and peak dose dobutamine, and then after 10 minutes of recovery. Measurements of brachial artery systolic (SBP) and diastolic blood pressure (DBP) were performed using a CMR compatible sphygmomanometer.

According to previously published techniques [11,12], LV volumes were measured from the short-axis series of cine white blood imaging sequences using a modified Simpson’s rule method [12]. LV concentricity was measured as the ratio of the LV mass to the LV end-diastolic volume as described by Chuang et al. [13]. Image acquisition parameters included a 45 msec repetition time (TR), a 1 msec echo time (TE), a 78° flip angle (FA), a 400 × 324 mm field of view (FOV), a 192 × 109 matrix, and an 8 mm thick slice with a 2 mm gap and an acceleration factor of 2 [5].

The LV wall motion at rest, peak dose and in recovery was assessed with a visual scoring system in which 1 = normal wall motion, 2 = hypokinesis, 3 = akinesis, and 4 = dyskinesis. An LV inducible WMA was defined as an increase in score of ≥1 (e.g., normal to hypokinetic) in 1 or more myocardial segments. Segments with an LV wall motion score of 2 or 3 at rest with no worsening of wall motion were considered negative for ischemia.

Ten minutes after the administration of gadobenate dimeglumine (Multihance) contrast, late gadolinium enhanced (LGE) inversion recovery images with steady state free precession readout were collected in the same short axis planes used to assess LV volumes. The sequence parameters included a 6 mm thick slice with a 2 mm gap, an 800 msec TR, a 1.7 msec TE, a 40° FA, a 360 × 270 mm FOV, and a 192 × 109 matrix with an inversion time adjusted to null the myocardium. Enhanced (>3 standard deviations in mean signal intensity above background non-enhanced) regions were identified.

First pass perfusion imaging with gadobenate dimeglumine (0.1 mmol/kg) was performed when 80% MPHRR was achieved. Eighty percent (80%) of the maximum predicted heart rate response for age was selected for this study because [5] this heart rate response was previously shown to be accurate for identifying ischemia and forecasting cardiac prognosis [5,14]. A potential benefit of the mild decrease in peak stress heart rate (80% relative to 85%), is that it allowed for acquisition of 2 slices for assessing myocardial first pass perfusion at peak stress. These perfusion images were collected in the short axis orientation in the middle and apical segments (2 slice positions due to the rapid HR). Image parameters included an 8 mm thick slice, a 169 msec TR, a 1.1 msec TE, a FA of 12°, a FOV of 360 × 270 mm and a 192 × 108 matrix. Rest first-pass perfusion imaging was not performed.

For each LV myocardial segment, a two-step process was utilized to identify perfusion defects indicative of ischemia. First, regions of first-pass hypo-perfusion were measured as a percentage of the corresponding myocardial wall thickness for that LV myocardial segment in the same imaging plane (Figure 1). The radial length of the PD was expressed as a percentage of the total LV myocardial wall thickness. In addition, the duration (or number of frames) for each PD was calculated from onset of LV myocardial enhancement until complete resolution of the defect. Any PD that persisted for more than 5 frames from onset of myocardial enhancement and encompassed >25% of the thickness of the wall was further evaluated for classification as ischemic [15]. This criterion of 25% transmural involvement was used successfully by other investigators to exclude dark rim artifacts [16,17].

Further evaluation was performed to exclude PDs related to prior infarcted territories. The extent of these PDs suspected of ischemia was compared to the extent of infarction as assessed in the LGE images within the same myocardial segments. Participants with PDs that extended beyond the areas of LGE or occurred in a territory or segment that differed from the segmental territory in which the LGE was noted were classified as ischemic. Any PD that matched the area of LGE was not classified as ischemic.

According to previously published techniques [18], PWV was assessed using phase-contrast cardiovascular magnetic resonance (PC-CMR). Images of the proximal thoracic aorta were obtained in an axial cross-sectional plane placed at the apex of the main pulmonary artery (identified with a sagittal localizer). PC-CMR imaging parameters included an 8 mm thick slice, a 10 msec TR, a 3–5 msec TE, a 15-20°FA, 20 ms temporal resolution, a 340 to 360 mm FOV, a 256 × 192 matrix and a through plane velocity encoding of 150 cm/sec. Pulse wave velocity (PWV) was calculated by dividing the distance between the ascending and descending thoracic aorta by the transit time of the flow wave [18].

In addition to the changes in HR, blood pressure (BP) and the RPP, we also assessed the LV stroke work (SW) and the systolic PVA, a measure indicating the global LV myocardial oxygen demand [19,20]. The systolic PVA is defined as both the mechanical LV SW and the mechanical potential energy which is expended during systole [21,22]. This is the area enclosed by the slopes of the end-systolic and the end-diastolic pressure volume relationship [22]. Invasive studies have shown an excellent correlation between the PVA and the myocardial oxygen consumption under different loading conditions and with dobutamine. The SW was defined as the mean arterial pressure * LV stroke volume. The systolic PVA was defined as: LV SW+ ½* LV end-systolic pressure * end-systolic volume. LV end systolic pressure was calculated as 0.85* brachial systolic blood pressure (SBP) [23].

In accordance with the American Heart Association Scientific Statement [24], the 10 middle and apical LV myocardial segments were assessed for LV first pass PDs, LGE, and inducible WMA. Participants in this study were characterized into one of 3 groups including those: 1) without a PD or WMA (Group I), 2) with a PD but without a WMA (Group II), and 3) with both a PD and a WMA indicative of ischemia (Group III). The differences in demographic, hemodynamic, CMR volumetric parameters, and indices of myocardial oxygen demand between the 3 Groups were assessed by an analysis of variance test of equality (ANOVA). In Tables 1 and 2, the p-values are displayed for the overall equality of the three groups. In the Figures, the comparisons between specific groups were accomplished using pairwise comparisons. A p-value of <0.05 was considered significant for either forms of testing. The differences in the LV myocardial SW and the PVA were adjusted for preload (resting and peak dose LV end-diastolic volume index or LVEDVi), afterload (peak dose PWV), contractility (LV ejection fraction or LVEF), and LV concentricity using analysis of covariance. Multiple regression models adjusting for SBP, LVEDV baseline and LVEDV peak dose were selected and reported by stepwise regression. The sensitivity and specificity of dobutamine related wall motion abnormalities for detecting obstructive CAD was assessed with the results of the perfusion component of the DCMR protocol serving as the reference standard. Results were expressed as means ± standard error of the estimate unless stated otherwise.