- Email: [email protected]
Determinants and Functional Significance of Myocardial Perfusion Reserve in Severe Aortic Stenosis
JACC: CARDIOVASCULAR IMAGING
VOL. 5, NO. 2, 2012
© 2012 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION PUBLISHED BY ELSEVIER INC.
ISSN 1936-878X/$36.00 DOI:10.1016/j.jcmg.2011.09.022
Determinants and Functional Significance of Myocardial Perfusion Reserve in Severe Aortic Stenosis Christopher D. Steadman, MB, CHB,*† Michael Jerosch-Herold, PHD,‡ Benjamin Grundy, BSC,* Suzanne Rafelt, MSC,† Leong L. Ng, MD,† Iain B. Squire, MD,† Nilesh J. Samani, MD,*† Gerry P. McCann, MD*† Leicester, United Kingdom; and Boston, Massachusetts
O B J E C T I V E S The purpose of this study was to assess the functional significance of cardiac
magnetic resonance (CMR) measures of left ventricular (LV) remodeling and myocardial perfusion reserve (MPR) in patients with severe aortic stenosis (AS), without obstructive coronary artery disease. B A C K G R O U N D Measures of stenosis severity do not correlate well with exercise intolerance in AS. LV remodeling in AS is associated with myocardial fibrosis and impaired MPR. The functional significance and determinants of MPR in AS are unclear. M E T H O D S Forty-six patients with isolated severe AS were prospectively studied before aortic valve replacement. The following investigations were undertaken: cardiopulmonary exercise testing to measure aerobic exercise capacity (peak VO2); CMR to assess left ventricular mass index (LVMI), myocardial fibrosis with late gadolinium enhancement (LGE), myocardial blood flow (MBF), and MPR; and transthoracic echocardiography to assess stenosis severity and diastolic function. R E S U L T S Peak VO2 was associated with sex ( ⫽ ⫺0.41), age ( ⫽ ⫺0.32), MPR ( ⫽ 0.45), resting
MBF ( ⫽ ⫺0.53), and septal transmitral flow velocity to annular velocity ratio (E/E=) ( ⫽ ⫺0.34), but not with LVMI, LGE, or echocardiographic measures of AS severity. On stepwise regression analysis, only MPR was independently associated with age- and sex-corrected peak VO2 ( ⫽ 0.46, p ⫽ 0.001). MPR was also inversely related to New York Heart Association functional class (p ⫽ 0.001). Univariate associations with MPR were sex ( ⫽ 0.38, p ⫽ 0.02), septal E/E= ( ⫽ ⫺0.30, p ⫽ 0.03), peak aortic valve velocity ( ⫽ ⫺0.34, p ⫽ 0.02), LVMI ( ⫽ ⫺0.51, p ⬍ 0.001), and LGE category ( ⫽ ⫺0.46, p ⫽ 0.002). On multivariate analysis, LVMI and LGE were independently associated with MPR. C O N C L U S I O N S CMR-quantified MPR is independently associated with aerobic exercise capacity in severe AS. LV remodeling appears to be a more important determinant of impaired MPR than stenosis severity per se. Further work is required to determine how CMR assessment of MPR can aid clinical management of patients with AS. (J Am Coll Cardiol Img 2012;5:182–9) © 2012 by the American College of Cardiology Foundation
From the *National Institute for Health Research (NIHR) Leicester Cardiovascular Biomedical Research Unit, Leicester, United Kingdom; †Department of Cardiovascular Sciences, University Hospitals of Leicester, Leicester, United Kingdom; and the ‡Department of Radiology, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts. This work was supported by the British Heart Foundation (PG/07/068/2334), the NIHR Comprehensive Local Research Network, and the NIHR Leicester Cardiovascular Biomedical Research Unit. All authors have reported they have no relationships relevant to the contents of this paper to disclose. Manuscript received June 30, 2011; revised manuscript received August 31, 2011, accepted September 1, 2011.
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
T
he development of left ventricular hypertrophy (LVH) in aortic stenosis (AS) has been regarded as a necessary physiological adaptation to maintain wall stress and preserve cardiac function. Left ventricular (LV) remodeling in AS is associated with a number of detrimental pathophysiological sequelae: interstitial fibrosis (1,2), diastolic dysfunction (3), and reduced myocardial perfusion reserve (MPR) (4). There is marked variation in the extent of LVH See page 190
in patients with severe AS (4,5). Experimental models of pressure overload have shown that animals that do not have LVH are better able to maintain ejection fraction (6). Furthermore, patients with critical AS, but without LVH, are less likely to have heart failure than patients with LVH (5). Cardiac magnetic resonance (CMR) is the ideal noninvasive technique to study LV remodeling. Left ventricular mass (LVM) and volumes, late gadolinium enhancement (LGE) for the detection of focal myocardial fibrosis, and MPR can be quantified accurately in a single examination. Myocardial fibrosis in AS is detected on LGE images in 27% to 62% of patients (1,7,8). The extent of LGE correlates with, although underestimates the extent of, interstitial fibrosis on myocardial biopsy (1,2) and increases with LVM (1,7,8). Peak oxygen consumption (VO2) is an objective measure of exercise capacity and an independent predictor of prognosis in chronic heart failure (9). There is a paucity of studies in AS, although patients with moderate to severe asymptomatic AS who are limited by symptoms on exercise testing do have lower peak VO2 and cardiac index than those who are limited by fatigue (10). Resting echocardiographic measures of stenosis severity and ejection fraction do not predict exercise capacity, although LVH and diastolic dysfunction may be important (3,11). The aim of the current study was to assess the importance of CMR measured LVM, LGE, and MPR in relation to peak VO2 in patients with isolated, severe AS. METHODS Patient selection. Patients with severe AS listed for
aortic valve replacement (AVR) were prospectively enrolled from a single tertiary center between October 2008 and April 2010. Inclusion criteria were ages 18 to 85 years and isolated severe AS; defined as one of the following: aortic valve area (AVA) ⬍1
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
183
cm2, peak aortic velocity ⱖ4 m/s, or mean pressure gradient ⬎40 mm Hg. Exclusion criteria were syncope; moderate/severe valve disease other than AS, previous valve surgery, obstructive coronary artery disease (⬎50% luminal stenosis on angiography), chronic obstructive pulmonary disease, atrial fibrillation, inability to exercise, a CMR contraindication, and estimated glomerular filtration rate ⬍30 ml/min. The local research and ethics committee approved the study. All subjects gave written informed consent before investigations, which were performed on the same day, 48 h after discontinuing -blockers. Echocardiography. Transthoracic echocardiography was performed using a Vivid 7 (GE Healthcare, Waukesha, Wisconsin) according to ABBREVIATIONS American Society of Echocardiography AND ACRONYMS guidelines (12). Tissue Doppler imaging allowed calculation of the transmitral flow AS ⴝ aortic stenosis velocity to annular velocity ratio (E/E=), a AVR ⴝ aortic valve replacement measure of LV filling pressure (13). AnalCMR ⴝ cardiac magnetic ysis was performed offline blinded to paresonance tient details using EchoPAC software DPT ⴝ diastolic perfusion time (GE Healthcare). The mean of 3 readings E/E’ ⴝ transmitral flow velocity to annular velocity ratio was used for each parameter. Diastolic perfusion time and LV rate pressure product. Resting diastolic perfusion time
LGE ⴝ late gadolinium enhancement
LV ⴝ left ventricular (DPT) was calculated as before (14). MyocarLVEF ⴝ left ventricular ejection dial work was estimated by calculating the fraction resting left ventricular rate pressure product LVH ⴝ left ventricular hypertrophy (LVRPP): (peak aortic pressure gradient ⫹ LVMI ⴝ left ventricular mass index systolic blood pressure) ⫻ heart rate (HR). MBF ⴝ myocardial blood flow Cardiopulmonary exercise testing. Symptomlimited exercise testing was performed on a MPR ⴝ myocardial perfusion reserve bicycle ergometer using a 1-min ramp proNYHA ⴝ New York Heart tocol with age- and sex-specific workloads. Association A 12-lead electrocardiogram was monitored VO2 ⴝ oxygen consumption continuously and blood pressure recorded every 2 min. Expired ventilatory gases were analyzed using an ErgoCard CPEX Test Station (Medisoft, Dinant, Belgium) to determine peak VO2. All exercise tests were physician supervised with indications for termination, as previously published (15). Cardiac magnetic resonance. CMR was performed using a 1.5-T scanner (Siemens, Avanto, Erlangen, Germany) with retrospective electrocardiographic triggering and a 6-channel phased array cardiac coil. Steady-state free precession cine images were acquired in 4-, 3-, and 2-chamber views. Perfusion images were acquired after pharmacological vasodilation with adenosine, 140 g/kg/min, for 3 min and during data acquisition. A gadolinium-based contrast agent (Magnevist, Bayer Healthcare, Leverkusen,
184
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
Germany) was administered intravenously (0.05 mmol/kg) at 5 ml/s, followed by a 20-ml saline flush. First-pass perfusion was assessed for 3 slices (basal, mid, and apical) acquiring every heartbeat using a saturation recovery gradient echo sequence. Rest imaging was performed approximately 10 min after stress imaging with a further 0.05 mmol/kg of contrast. In the intervening time, a stack of short-axis slices were obtained using cine imaging covering the entire left ventricle. A further 0.1 mmol/kg of contrast was given to bring the total dose to 0.2 mmol/kg. At least 10 min after this, LGE images were obtained with the use of an inversion-recovery prepared, segmented gradient echo sequence, as previously described (16). CMR analysis. Analysis was performed offline blinded to patient details using QMass software, version 7.1 (Medis, Leiden, the Netherlands). The LV contours were drawn manually and left ventricular end-diastolic volume (LVEDV), LV end-systolic volume, stroke volume, left ventricular ejection fraction (LVEF), and end-diastolic LVM were calculated. Values were indexed by body surface area, denoted by the suffix ‘I,’ for example, LVMI. Interobserver and intraobserver variability was calculated on 10 random datasets by 2 experienced observers (C.D.S., G.P.M.). The epicardium and endocardium were contoured on the perfusion images, along with a region of interest in the LV blood pool, to generate signal intensity curves. The measured arterial input, measured in arbitrary signal intensity units, was converted to a curve of percent enhancement, by dividing by the baseline signal-intensity, namely, the signal intensity in the blood pool before any contrast enhancement. A calibration curve of effective contrast
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
enhancement versus the contrast enhancement extrapolated from the low R1 (contrast concentration) range was calculated by numerical simulation, using the sequence parameters, and a mean pre-contrast T1 value for blood of 1,450 ms. This calibration curve was then inverted, and used for correction of the observed percent contrast enhancement in the blood pool. Saturation correction resulted, on average, in a 20% to 30% increase of the peak contrast enhancement of the arterial input function. The arterial input function corrected for signal saturation was used for myocardial blood flow (MBF) quantification by model-independent deconvolution (17). Transmural MPR was calculated by dividing hyperemic MBF by resting MBF. All MPR results refer to uncorrected values, although resting MBF results are also displayed corrected for LVRPP. Two experienced observers (C.D.S., G.P.M.) qualitatively assessed the perfusion images for subendocardial perfusion defects (a visible defect for ⱖ5 heartbeats) and the LGE images for focal fibrosis, categorized as present or absent. Additionally, computer-assisted planimetry was used to determine the mass and percentage of enhanced myocardium ⬎5 SD above the mean signal intensity of remote normal myocardium. Statistical analysis. The distributions of continuous variables were assessed using graphical displays and the Shapiro-Wilk test for normality. Values are reported as mean ⫾ SD or median (interquartile range) if the distribution was not normal. Septal E/E=, lateral E/E=, and LGE (%LV mass) were skewed and logarithmically transformed before regression analysis. Statistical analysis was undertaken using SPSS, version 16.0 (SPSS, Chicago, Illinois) and SAS (SAS Institute, Cary, North Carolina). Comparison of means of 2 groups was done with a t test and ⬎2 groups with 1-way analysis of variance. Linear regression investigated possible associations of continuous outcome measures. Stepwise selection methods were used to determine the most important associations. Hyperemic MBF was adjusted for resting MBF (thereby assessing MPR) by forcing resting MBF into the model before selection methods were applied. The same method was used when adjusting peak VO2 for age and sex. All p values ⬍0.05 were considered significant.
RESULTS Figure 1. Recruitment Data The number of patients approached, excluded, and recruited. AR ⫽ aortic regurgitation; CMR ⫽ cardiac magnetic resonance; COPD ⫽ chronic obstructive pulmonary disease; EF ⫽ ejection fraction; LGE ⫽ late gadolinium enhancement.
Population characteristics. Details of recruitment are outlined in Figure 1. Demographic data for the 46 patients included in this report are presented in Table 1.
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
0.86 ⫾ 0.22/0.45 ⫾ 0.13
VO2 are summarized in Table 3. Peak VO2 was higher in men compared with women: 16.2 ⫾ 4.1 ml/kg/min versus 12.3 ⫾ 3.1 ml/kg/min, respectively (p ⫽ 0.005), and decreased with age (Fig. 3A). Peak VO2 was inversely related to septal E/E= and resting MBF, and correlated with MPR ( ⫽ 0.45, r2 ⫽ 0.2) (Fig. 3B), but not with resting MBF/LVRPP or hyperemic MBF. Only MPR was independently associated with age- and sex-corrected peak VO2 on stepwise analysis. Correction for -blocker usage did not affect the significance of the models. Symptomatic status. NYHA functional class was associated with peak VO2, even after adjustment for age and sex ( ⫽ ⫺0.4, p ⫽ 0.002). Patients with higher NYHA functional class had lower MPR (p ⫽ 0.001) (Fig. 4), despite no significant differences in AS severity between groups. Adding NYHA functional class to the stepwise regression for associations with peak VO2 had no effect on the significance of MPR as the only independent variable.
1.46 ⫾ 0.29
Associations with resting MBF and perfusion reserve.
Table 1. Patient Characteristics Age, yrs Male/female
66.2 ⫾ 9.3 34/12
Body surface area, m2
1.92 ⫾ 0.21
Systolic blood pressure, mm Hg
131 ⫾ 20
Diastolic blood pressure, mm Hg
74 ⫾ 10
Resting heart rate, beats/min
69 ⫾ 12
ACEI/ARB
15 (33%)
Beta-blocker
18 (39%)
Statin
28 (61%)
Diabetes mellitus
10 (22%)
Hypertension
27 (59%)
Smoking Former Current NYHA functional class, (n)
29 (63%) 3 (6%) I (9); II (32); III (5)
Echocardiography Peak AV velocity, m/s Mean PG, mm Hg AVA/I, (cm2)/(cm2/m2) LVRPP, mm Hg·beats/min·10⫺4 Resting DPT, s/min
4.4 ⫾ 0.6 48.5 ⫾ 14.1
37.9 ⫾ 3.1
Septal E/E=
13.7 (11.1–18.6)
Lateral E/E=
10.2 (7.2–14.0)
Values are mean ⫾ SD, n, or n(%). ACEI ⫽ angiotensin-converting enzyme inhibitor; ARB ⫽ angiotensinreceptor blocker; AV ⫽ aortic valve; AVA/I ⫽ aortic valve area/index; DPT ⫽ diastolic perfusion time; E/E= ⫽ transmitral flow velocity to annular velocity ratio; LVRPP ⫽ left ventricular rate pressure product; NYHA ⫽ New York Heart Association; PG ⫽ pressure gradient.
Resting and hyperemic MBF were higher in women than men (resting, 1.07 ⫾ 0.24 vs. 0.84 ⫾ 0.15, p ⫽ 0.001; hyperemic, 2.16 ⫾ 0.53 vs. 1.64 ⫾ 0.38, p ⫽ Table 2. Cardiac Magnetic Resonance and Cardiopulmonary Exercise Data Cardiac magnetic resonance LVMI, g/m2
69.6 ⫾ 17.7
Cardiopulmonary exercise data. One exercise test was
LVEDVI, ml/m2
96.9 ⫾ 14.6
medically terminated because of angina associated with a ⬎20 mm Hg drop in systolic blood pressure. There were no complications. Twenty-two patients (48%) were symptom limited, and the remainder were limited by fatigue, including all 9 patients in New York Heart Association (NYHA) functional class I. Exercise data are presented in Table 2. Cardiac magnetic resonance. Volumetric data and LGE images were available for all patients. Results are displayed in Table 2. Interobserver and intraobserver variability were as follows: LVM 5.2% and 4.9%, LVEDV 3.1% and 3%, LVEF 6.6% and 4.7%. Thirty-four patients had global subendocardial perfusion defects. There was a wide range of LVMI and MPR, but all patients had normal or near normal systolic function (n ⫽ 5, LVEF 45% to 50%; n ⫽ 1, LVEF 40% to 45%). LGE was present in 26 (56%) patients, and except in the patient excluded, there were no clear myocardial infarctions. Examples of patients with variable LVM, LGE, and MPR are shown in Figure 2. Associations with exercise capacity. Univariate analyses and results from the stepwise model selection for peak
LVESVI, ml/m2
42.5 ⫾ 10.6
LVM/LVEDV, g/ml
0.72 ⫾ 0.15
LVEF, %
56.5 ⫾ 6.5
Stroke volume index, ml/m2
53.6 ⫾ 7.7
Maximum wall thickness, mm
13.2 ⫾ 2.5
Resting MBF, ml/min/g
0.90 ⫾ 0.20
Resting MBF/LVRPP, (ml/min/g)/ (mm Hg.beats/min.10⫺4)
0.60 ⫾ 0.14
Hyperemic MBF, ml/min/g
1.77 ⫾ 0.47
Hyperemic heart rate, beats/min
88 ⫾ 12
MPR (n ⫽ 41)
2.03 ⫾ 0.55
LGE positive
26 (56%)
LGE, % (⬎5 SD above remote) LGE mass, g (⬎5 SD above remote)
1 (0–28) 1.8 (0.8–4.8)
Cardiopulmonary exercise testing Peak heart rate, % predicted Rise in systolic blood pressure, mm Hg
85 ⫾ 12 39 ⫾ 23
Respiratory exchange ratio
1.09 ⫾ 0.1
Peak VO2, ml/kg/min
15.2 ⫾ 4.2
Values are mean ⫾ SD, n (%), or median interquartile range. LGE ⫽ late gadolinium enhancement; LVEDV ⫽ left ventricular enddiastolic volume; LVEDVI ⫽ left ventricular end-diastolic volume index; LVEF ⫽ left ventricular ejection fraction; LVESVI ⫽ left ventricular end-systolic volume index; LVM ⫽ left ventricular mass; LVMI ⫽ left ventricular mass index; MBF ⫽ myocardial blood flow; MPR ⫽ myocardial perfusion reserve, VO2 ⫽ volume oxygen consumption; other abbreviation as in Table 1.
185
186
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
Figure 2. Examples of Patients With Varying LV Remodeling Cardiac magnetic resonance images: (i) short-axis cine end-diastole at middle level (A, B, and C) and basal level (D); (ii) late gadolinium enhancement (LGE) at same level as “i” (LGE marked with white arrows); (iii) perfusion imaging in middle left ventricular (LV) slice during hyperemia; and (iv) during rest perfusion. Four male patients displayed: (A) Significantly raised left ventricular mass (LVM), clear insertion point LGE, and reduced myocardial perfusion reserve (MPR [age 63 years, left ventricular mass index (LVMI) 126 g/m2, aortic valve area (AVA) 0.92 cm2, MPR 1.07, peak volume oxygen consumption (VO2) 12.7 ml/kg/min). (B) Mildly elevated LVM with mild insertion point LGE (age 65 years, LVMI 76 g/m2, AVA 0.79 cm2, MPR 1.59, peak VO2 16.4 ml/kg/min). (C) Normal LVM with no LGE (age 65 years, LVMI 55 g/m2, AVA 0.99 cm2, MPR 2.30, peak VO2 18.5 ml/kg/min). (D) Normal LVM with mild insertion point LGE but more significant remote LGE (age 63 years, LVMI 66 g/m2, AVA 0.73 cm2, MPR 2.4, peak VO2 20.6 ml/kg/min).
0.002). Resting MBF correlated with resting LVRPP ( ⫽ 0.47, p ⫽ 0.02) and was inversely related to DPT ( ⫽ ⫺0.35, p ⫽ 0.025). Sex and resting LVRPP were independent associations with resting MBF. MPR was inversely related to peak aortic valve velocity, mean pressure gradient, CMR measures of LVH, LGE category, and septal E/E=. Results are summarized in Table 4. MPR was not related to history of smoking, diabetes, or hypertension. With all univariable associations, the best stepwise multivariate regression model contained LVMI and LGE category as independent associations. MPR was reduced in patients with LGE (1.87 ⫾ 0.47 vs. 2.24 ⫾ 0.58, p ⫽ 0.031), and there was also an inverse correlation between MPR and LGE scar mass at 5 SD ( ⫽ ⫺0.4, p ⫽ 0.008). DISCUSSION Determinants of resting MBF and MPR. Resting and hyperemic MBF were higher in women, as has been found previously (18). As reported by Rajappan et al. (4), we did not find that measures of AS severity or LVMI were related to resting MBF. Contrary to that
study, there was an association between resting MBF and estimated myocardial work (LVRPP), as has also been shown in hypertrophic cardiomyopathy (18). The results of MPR in this study are remarkably consistent with those from Rajappan’s similar cohort of patients with severe AS before AVR (2.03 ⫾ 0.55 vs. 1.90 ⫾ 0.60, respectively), considering that positron emission tomography was used in that report (4). We have shown that aortic valve velocities/pressure gradients, measures of LVM, and, for the first time, female sex, LV filling pressure (septal E/E=), and myocardial fibrosis (LGE), have univariate associations with MPR. Of these, LVMI and LGE appear to be the most important. These findings are somewhat different from those of Rajappan et al. (4), whose results suggested that AS severity and DPT were more important determinants of MPR in severe AS. The discrepancies are likely due to the smaller numbers (20 vs. 41), that positron emission tomography was used, that no LGE for fibrosis was performed, and that the majority of patients in the Rajappan study were asymptomatic (4). Given the modest correlations found in our study, we cannot exclude that the
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
Table 3. Associations With Peak VO2 Univariate Associations

p Value
Sex
⫺0.41
0.005
Age
⫺0.32
0.03
Peak AV velocity
⫺0.18
0.24
Mean PG
⫺0.18
0.24
0.04
0.79
Septal E/E=
AVAI
⫺0.34
0.02
Lateral E/E=
⫺0.23
0.13
0.02
0.89
LVMI LVM/LVEDV ratio
⫺0.04
0.79
LV maximum wall thickness
⫺0.2
0.90
LVEF
0.01
0.97
LGE
0.05
0.73
and sex-corrected exercise capacity. This clinical study has identified imaging and clinical parameters that have a clear and independent association with exercise capacity in AS. Previous studies have suggested a range of parameters that include aortic valve compliance (23–24), reduced longitudinal strain (25), LV filling pressure (13) and diastolic dysfunction (3, 11). However, those studies have looked at exerciseinduced symptoms, which may be subjective, as in our study, where 50% of symptomatic patients were limited by fatigue, and they have not accounted for age and sex as determinants of exercise capacity. Many other factors may influence peak VO2, including genetics, physical activity, and skeletal muscle changes, which may be particularly important in patients with cardiac dysfunction.
Resting MBF
⫺0.53
0.001
Resting MBF/LVRPP
⫺0.28
0.08
Hyperemic MBF
⫺0.01
Mechanisms linking reduced myocardial perfusion reserve and peak VO2. With exercise, cardiac output
0.94
rises incrementally with workload. Increased cardiac
MPR
0.45
0.004
Final Model (Stepwise Selection)

p Value
R2
Sex
⫺0.44
0.002
0.42
Age
⫺0.15
0.25
MPR
0.46
0.001
LV ⫽ left ventricular; other abbreviations as in Tables 1 and 2.
findings may be spurious; however, they are physiologically plausible. The lack of association between smoking, hypertension, and diabetes (19) and MPR in this study is almost certainly due to the smaller sample size and the large effect of LV remodeling in these patients with AS. This is the first study to demonstrate the importance of myocardial fibrosis (LGE) in relation to MPR in AS. This association may be explained by reduced arteriolar and capillary density seen with LV remodeling and fibrosis (20) and loss of the “suction” wave seen in patients with LVH (21). It is important to emphasize that diffuse myocardial fibrosis (22) was not assessed (as newer T1mapping techniques were not available to us at the time), and this may be an important determinant of MPR. However, the extent of LGE does correlate with diffuse fibrosis on myocardial biopsy (1). Associations with exercise capacity. We hypothesized that CMR measures of LV remodeling would be independently associated with aerobic exercise capacity. As previously demonstrated, measures of AS severity did not predict exercise capacity (15). Septal E/E=, resting MBF and MPR, but not measures of LVH, had univariate associations with peak VO2. Only MPR was independently associated with age-
Figure 3. Associations With Exercise Capacity Relationship between peak volume oxygen consumption (VO2) and (A) age and (B) myocardial perfusion reserve.
187
188
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
Figure 4. Perfusion Reserve and Symptomatic Status Myocardial perfusion reserve and New York Heart Association (NYHA) functional class. CI ⫽ confidence interval.
output in severe AS is largely dependent on increased heart rate (26), which is associated with the development of impaired MPR in both patients (27) and experimental models of AS (28). The inability to increase blood flow to the myocardium is limited because vasodilation may already be near maximal (29), DPT will be limited further with increased heart rate, and there is a rapid increase in LV end-diastolic pressure (26), which further reduces the effective Table 4. Associations With Perfusion Reserve Univariate Associations

p Value
Sex
0.38
0.02
Age
⫺0.09
0.54
Hyperemic heart rate
0.2
0.19
Beta-blockers
0.008
0.96
LVRPP
⫺0.1
0.58
Resting DPT
⫺0.28
0.08
Peak AV velocity
⫺0.34
0.02
Mean PG
⫺0.30
0.04
AVAI
0.21
0.17
LVMI
⫺0.51
⬍0.001
LVM/LVEDV
⫺0.43
0.003
LV maximum wall thickness
⫺0.41
0.005
LVEF
0.28
0.07
LGE
⫺0.46
0.002
Septal E/E=
⫺0.33
0.03
Lateral E/E=
⫺0.19
0.20
Final Model (Stepwise Selection)

p Value
R2
LVMI
⫺0.40
0.004
0.43
LGE
⫺0.31
0.02
Abbreviations as in Tables 1, 2, and 3.
pressure gradient for perfusion. Patients unable to increase blood flow to the myocardium on exercise will likely develop subendocardial myocardial dysfunction (25,28), which will limit cardiac output and contribute to exercise intolerance. Clinical implications. MPR is an important prognostic marker in hypertrophic cardiomyopathy, predicting adverse ventricular remodeling and clinical outcomes (30). Impaired MPR could be a therapeutic target in AS patients unable to have aortic valve intervention and in patients with persistent symptoms after AVR. Traditional therapies targeting LV remodeling such as inhibitors of the renin-angiotensin-aldosterone system and drugs with novel mechanisms of action, for example, perhexilene, may be useful in this context. The importance of impaired MPR to the development of symptoms and outcome in asymptomatic patients with AS warrants further investigation. Study limitations. The results of this mechanistic study may not be applicable to the wider AS population. Although we have touched on symptomatic status in this paper, the number of asymptomatic patients is small and, therefore, limited conclusions can be drawn about this group. The cross-sectional nature of our data does not allow one to draw conclusions as to whether myocardial fibrosis leads to a reduction in MPR or if the reduced MPR causes myocardial fibrosis. Longitudinal studies in patients with less severe AS with assessment of diffuse fibrosis may be able to answer this question. The reproducibility of MPR measurements in AS is unknown. Finally, because of the relatively small numbers, the number of variables that can be appropriately used in the multivariate regression analysis is necessarily limited. Therefore, we limited the number of univariate associations entered into the multivariate model to 5 and used the strongest univariate associations, avoiding overlapping variables. CONCLUSIONS
We have shown for the first time that MPR is independently associated with objectively measured exercise capacity in severe AS. Impaired MPR is determined by a combination of factors including AS severity, LV remodeling, and filling pressure. Further work is required to determine how CMR assessment of MPR can aid the clinical management of patients with AS. Reprint requests and correspondence: Dr. Gerry P. Mc-
Cann, Department of Cardiovascular Sciences, Glenfield Hospital, Groby Road, Leicester LE3 9QP, United Kingdom. E-mail: [email protected].
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
REFERENCES
1. Azevedo CF, Nigri M, Higuchi ML, et al. Prognostic significance of myocardial fibrosis quantification by histopathology and magnetic resonance imaging in patients with severe aortic valve disease. J Am Coll Cardiol 2010; 56:278 – 87. 2. Weidemann F, Herrmann S, Stork S, et al. Impact of myocardial fibrosis in patients with symptomatic severe aortic stenosis. Circulation 2009;120: 577– 84. 3. Hess OM, Ritter M, Schneider J, Grimm J, Turina M, Krayenbuehl HP. Diastolic stiffness and myocardial structure in aortic valve disease before and after valve replacement. Circulation 1984;69:855– 65. 4. Rajappan K, Rimoldi OE, Dutka DP, et al. Mechanisms of coronary microcirculatory dysfunction in patients with aortic stenosis and angiographically normal coronary arteries. Circulation 2002;105:470 – 6. 5. Kupari M, Turto H, Lommi J. Left ventricular hypertrophy in aortic valve stenosis: preventive or promotive of systolic dysfunction and heart failure? Eur Heart J 2005;26:1790 – 6. 6. Esposito G, Rapacciuolo A, Naga Prasad SV, et al. Genetic alterations that inhibit in vivo pressure-overload hypertrophy prevent cardiac dysfunction despite increased wall stress. Circulation 2002;105:85–92. 7. Debl K, Djavidani B, Buchner S, et al. Delayed hyperenhancement in magnetic resonance imaging of left ventricular hypertrophy caused by aortic stenosis and hypertrophic cardiomyopathy: visualisation of focal fibrosis. Heart 2006;92:1447–51. 8. Rudolph A, Abdel-Aty H, Bohl S, et al. Noninvasive detection of fibrosis applying contrast-enhanced cardiac magnetic resonance in different forms of left ventricular hypertrophy: relation to remodeling. J Am Coll Cardiol 2009;53:284 –91. 9. Francis DP, Shamim W, Ceri Davies L, et al. Cardiopulmonary exercise testing for prognosis in chronic heart failure: continuous and independent prognostic value from VE/VCO2 slope and peak VO2. Eur Heart J 2000;21:154 – 61. 10. Rajani R, Rimington H, Chambers JB. Treadmill exercise in apparently asymptomatic patients with moderate or severe aortic stenosis: relationship
between cardiac index and revealed symptoms. Heart 2010;96:689 –95. 11. Monrad ES, Hess OM, Murakami T, Nonogi H, Corin WJ, Krayenbuehl HP. Abnormal exercise hemodynamics in patients with normal systolic function late after aortic valve replacement. Circulation 1988;77:613–24. 12. Gottdiener JS, Bednarz J, Devereux R, et al. American Society of Echocardiography recommendations for use of echocardiography in clinical trials. J Am Soc Echocardiogr 2004; 17:1086 –119. 13. Dalsgaard M, Kjaergaard J, Pecini R, et al. Left ventricular filling pressure estimation at rest and during exercise in patients with severe aortic valve stenosis: comparison of echocardiographic and invasive measurements. J Am Soc Echocardiogr 2009;22:343–9. 14. Ferro G, Duilio C, Spinelli L, Liucci GA, Mazza F, Indolfi C. Relation between diastolic perfusion time and coronary artery stenosis during stressinduced myocardial ischemia. Circulation 1995;92:342–7. 15. Das P, Rimington H, Chambers J. Exercise testing to stratify risk in aortic stenosis. Eur Heart J 2005;26: 1309 –13. 16. Simonetti OP, Kim RJ, Fieno DS, et al. An improved MR imaging technique for the visualization of myocardial infarction. Radiology 2001;218:215–23. 17. Jerosch-Herold M, Swingen C, Seethamraju RT. Myocardial blood flow quantification with MRI by model-independent deconvolution. Med Phys 2002;29:886 –97. 18. Petersen SE, Jerosch-Herold M, Hudsmith LE, et al. Evidence for microvascular dysfunction in hypertrophic cardiomyopathy: new insights from multiparametric magnetic resonance imaging. Circulation 2007;115: 2418 –25. 19. Wang L, Jerosch-Herold M, Jacobs DR Jr., Shahar E, Detrano R, Folsom AR. Coronary artery calcification and myocardial perfusion in asymptomatic adults: the MESA (Multi-Ethnic Study of Atherosclerosis). J Am Coll Cardiol 2006;48:1018 –26. 20. Marcus ML, Harrison DG, Chilian WM, et al. Alterations in the coronary circulation in hypertrophied ventricles. Circulation 1987;75:I19 –I25. 21. Davies JE, Whinnett ZI, Francis DP, et al. Evidence of a dominant backwardpropagating “suction” wave responsible
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
for diastolic coronary filling in humans, attenuated in left ventricular hypertrophy. Circulation 2006;113:1768 –78. 22. Flett AS, Hayward MP, Ashworth MT, et al. Equilibrium contrast cardiovascular magnetic resonance for the measurement of diffuse myocardial fibrosis: preliminary validation in humans. Circulation 2010;122:138 – 44. 23. Das P, Rimington H, Smeeton N, Chambers J. Determinants of symptoms and exercise capacity in aortic stenosis: a comparison of resting haemodynamics and valve compliance during dobutamine stress. Eur Heart J 2003;24:1254 – 63. 24. Lancellotti P, Karsera D, Tumminello G, Lebois F, Pierard LA. Determinants of an abnormal response to exercise in patients with asymptomatic valvular aortic stenosis. Eur J Echocardiogr 2008;9:338 – 43. 25. Lafitte S, Perlant M, Reant P, et al. Impact of impaired myocardial deformations on exercise tolerance and prognosis in patients with asymptomatic aortic stenosis. Eur J Echocardiogr 2009;10:414 –9. 26. Lee SJ, Jonsson B, Bevegard S, Karlof I, Astrom MD. Hemodynamic changes at rest and during exercise in patients with aortic stenosis of varying severity. Am Heart J 1970;79:318 –31. 27. Trenouth RS, Phelps NC, Neill WA. Determinants of left ventricular hypertrophy and oxygen supply in chronic aortic valve disease. Circulation 1976;53:644 –50. 28. Nakano K, Corin WJ, Spann JF Jr., Biederman RW, Denslow S, Carabello BA. Abnormal subendocardial blood flow in pressure overload hypertrophy is associated with pacinginduced subendocardial dysfunction. Circ Res 1989;65:1555– 64. 29. Hoffman JI. Determinants and prediction of transmural myocardial perfusion. Circulation 1978;58:381–91. 30. Cecchi F, Olivotto I, Gistri R, Lorenzoni R, Chiriatti G, Camici PG. Coronary microvascular dysfunction and prognosis in hypertrophic cardiomyopathy. N Engl J Med 2009;349: 1027–35.
Key Words: aortic stenosis y cardiac magnetic resonance y cardiopulmonary exercise testing y myocardial perfusion reserve.
189
VOL. 5, NO. 2, 2012
© 2012 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION PUBLISHED BY ELSEVIER INC.
ISSN 1936-878X/$36.00 DOI:10.1016/j.jcmg.2011.09.022
Determinants and Functional Significance of Myocardial Perfusion Reserve in Severe Aortic Stenosis Christopher D. Steadman, MB, CHB,*† Michael Jerosch-Herold, PHD,‡ Benjamin Grundy, BSC,* Suzanne Rafelt, MSC,† Leong L. Ng, MD,† Iain B. Squire, MD,† Nilesh J. Samani, MD,*† Gerry P. McCann, MD*† Leicester, United Kingdom; and Boston, Massachusetts
O B J E C T I V E S The purpose of this study was to assess the functional significance of cardiac
magnetic resonance (CMR) measures of left ventricular (LV) remodeling and myocardial perfusion reserve (MPR) in patients with severe aortic stenosis (AS), without obstructive coronary artery disease. B A C K G R O U N D Measures of stenosis severity do not correlate well with exercise intolerance in AS. LV remodeling in AS is associated with myocardial fibrosis and impaired MPR. The functional significance and determinants of MPR in AS are unclear. M E T H O D S Forty-six patients with isolated severe AS were prospectively studied before aortic valve replacement. The following investigations were undertaken: cardiopulmonary exercise testing to measure aerobic exercise capacity (peak VO2); CMR to assess left ventricular mass index (LVMI), myocardial fibrosis with late gadolinium enhancement (LGE), myocardial blood flow (MBF), and MPR; and transthoracic echocardiography to assess stenosis severity and diastolic function. R E S U L T S Peak VO2 was associated with sex ( ⫽ ⫺0.41), age ( ⫽ ⫺0.32), MPR ( ⫽ 0.45), resting
MBF ( ⫽ ⫺0.53), and septal transmitral flow velocity to annular velocity ratio (E/E=) ( ⫽ ⫺0.34), but not with LVMI, LGE, or echocardiographic measures of AS severity. On stepwise regression analysis, only MPR was independently associated with age- and sex-corrected peak VO2 ( ⫽ 0.46, p ⫽ 0.001). MPR was also inversely related to New York Heart Association functional class (p ⫽ 0.001). Univariate associations with MPR were sex ( ⫽ 0.38, p ⫽ 0.02), septal E/E= ( ⫽ ⫺0.30, p ⫽ 0.03), peak aortic valve velocity ( ⫽ ⫺0.34, p ⫽ 0.02), LVMI ( ⫽ ⫺0.51, p ⬍ 0.001), and LGE category ( ⫽ ⫺0.46, p ⫽ 0.002). On multivariate analysis, LVMI and LGE were independently associated with MPR. C O N C L U S I O N S CMR-quantified MPR is independently associated with aerobic exercise capacity in severe AS. LV remodeling appears to be a more important determinant of impaired MPR than stenosis severity per se. Further work is required to determine how CMR assessment of MPR can aid clinical management of patients with AS. (J Am Coll Cardiol Img 2012;5:182–9) © 2012 by the American College of Cardiology Foundation
From the *National Institute for Health Research (NIHR) Leicester Cardiovascular Biomedical Research Unit, Leicester, United Kingdom; †Department of Cardiovascular Sciences, University Hospitals of Leicester, Leicester, United Kingdom; and the ‡Department of Radiology, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts. This work was supported by the British Heart Foundation (PG/07/068/2334), the NIHR Comprehensive Local Research Network, and the NIHR Leicester Cardiovascular Biomedical Research Unit. All authors have reported they have no relationships relevant to the contents of this paper to disclose. Manuscript received June 30, 2011; revised manuscript received August 31, 2011, accepted September 1, 2011.
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
T
he development of left ventricular hypertrophy (LVH) in aortic stenosis (AS) has been regarded as a necessary physiological adaptation to maintain wall stress and preserve cardiac function. Left ventricular (LV) remodeling in AS is associated with a number of detrimental pathophysiological sequelae: interstitial fibrosis (1,2), diastolic dysfunction (3), and reduced myocardial perfusion reserve (MPR) (4). There is marked variation in the extent of LVH See page 190
in patients with severe AS (4,5). Experimental models of pressure overload have shown that animals that do not have LVH are better able to maintain ejection fraction (6). Furthermore, patients with critical AS, but without LVH, are less likely to have heart failure than patients with LVH (5). Cardiac magnetic resonance (CMR) is the ideal noninvasive technique to study LV remodeling. Left ventricular mass (LVM) and volumes, late gadolinium enhancement (LGE) for the detection of focal myocardial fibrosis, and MPR can be quantified accurately in a single examination. Myocardial fibrosis in AS is detected on LGE images in 27% to 62% of patients (1,7,8). The extent of LGE correlates with, although underestimates the extent of, interstitial fibrosis on myocardial biopsy (1,2) and increases with LVM (1,7,8). Peak oxygen consumption (VO2) is an objective measure of exercise capacity and an independent predictor of prognosis in chronic heart failure (9). There is a paucity of studies in AS, although patients with moderate to severe asymptomatic AS who are limited by symptoms on exercise testing do have lower peak VO2 and cardiac index than those who are limited by fatigue (10). Resting echocardiographic measures of stenosis severity and ejection fraction do not predict exercise capacity, although LVH and diastolic dysfunction may be important (3,11). The aim of the current study was to assess the importance of CMR measured LVM, LGE, and MPR in relation to peak VO2 in patients with isolated, severe AS. METHODS Patient selection. Patients with severe AS listed for
aortic valve replacement (AVR) were prospectively enrolled from a single tertiary center between October 2008 and April 2010. Inclusion criteria were ages 18 to 85 years and isolated severe AS; defined as one of the following: aortic valve area (AVA) ⬍1
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
183
cm2, peak aortic velocity ⱖ4 m/s, or mean pressure gradient ⬎40 mm Hg. Exclusion criteria were syncope; moderate/severe valve disease other than AS, previous valve surgery, obstructive coronary artery disease (⬎50% luminal stenosis on angiography), chronic obstructive pulmonary disease, atrial fibrillation, inability to exercise, a CMR contraindication, and estimated glomerular filtration rate ⬍30 ml/min. The local research and ethics committee approved the study. All subjects gave written informed consent before investigations, which were performed on the same day, 48 h after discontinuing -blockers. Echocardiography. Transthoracic echocardiography was performed using a Vivid 7 (GE Healthcare, Waukesha, Wisconsin) according to ABBREVIATIONS American Society of Echocardiography AND ACRONYMS guidelines (12). Tissue Doppler imaging allowed calculation of the transmitral flow AS ⴝ aortic stenosis velocity to annular velocity ratio (E/E=), a AVR ⴝ aortic valve replacement measure of LV filling pressure (13). AnalCMR ⴝ cardiac magnetic ysis was performed offline blinded to paresonance tient details using EchoPAC software DPT ⴝ diastolic perfusion time (GE Healthcare). The mean of 3 readings E/E’ ⴝ transmitral flow velocity to annular velocity ratio was used for each parameter. Diastolic perfusion time and LV rate pressure product. Resting diastolic perfusion time
LGE ⴝ late gadolinium enhancement
LV ⴝ left ventricular (DPT) was calculated as before (14). MyocarLVEF ⴝ left ventricular ejection dial work was estimated by calculating the fraction resting left ventricular rate pressure product LVH ⴝ left ventricular hypertrophy (LVRPP): (peak aortic pressure gradient ⫹ LVMI ⴝ left ventricular mass index systolic blood pressure) ⫻ heart rate (HR). MBF ⴝ myocardial blood flow Cardiopulmonary exercise testing. Symptomlimited exercise testing was performed on a MPR ⴝ myocardial perfusion reserve bicycle ergometer using a 1-min ramp proNYHA ⴝ New York Heart tocol with age- and sex-specific workloads. Association A 12-lead electrocardiogram was monitored VO2 ⴝ oxygen consumption continuously and blood pressure recorded every 2 min. Expired ventilatory gases were analyzed using an ErgoCard CPEX Test Station (Medisoft, Dinant, Belgium) to determine peak VO2. All exercise tests were physician supervised with indications for termination, as previously published (15). Cardiac magnetic resonance. CMR was performed using a 1.5-T scanner (Siemens, Avanto, Erlangen, Germany) with retrospective electrocardiographic triggering and a 6-channel phased array cardiac coil. Steady-state free precession cine images were acquired in 4-, 3-, and 2-chamber views. Perfusion images were acquired after pharmacological vasodilation with adenosine, 140 g/kg/min, for 3 min and during data acquisition. A gadolinium-based contrast agent (Magnevist, Bayer Healthcare, Leverkusen,
184
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
Germany) was administered intravenously (0.05 mmol/kg) at 5 ml/s, followed by a 20-ml saline flush. First-pass perfusion was assessed for 3 slices (basal, mid, and apical) acquiring every heartbeat using a saturation recovery gradient echo sequence. Rest imaging was performed approximately 10 min after stress imaging with a further 0.05 mmol/kg of contrast. In the intervening time, a stack of short-axis slices were obtained using cine imaging covering the entire left ventricle. A further 0.1 mmol/kg of contrast was given to bring the total dose to 0.2 mmol/kg. At least 10 min after this, LGE images were obtained with the use of an inversion-recovery prepared, segmented gradient echo sequence, as previously described (16). CMR analysis. Analysis was performed offline blinded to patient details using QMass software, version 7.1 (Medis, Leiden, the Netherlands). The LV contours were drawn manually and left ventricular end-diastolic volume (LVEDV), LV end-systolic volume, stroke volume, left ventricular ejection fraction (LVEF), and end-diastolic LVM were calculated. Values were indexed by body surface area, denoted by the suffix ‘I,’ for example, LVMI. Interobserver and intraobserver variability was calculated on 10 random datasets by 2 experienced observers (C.D.S., G.P.M.). The epicardium and endocardium were contoured on the perfusion images, along with a region of interest in the LV blood pool, to generate signal intensity curves. The measured arterial input, measured in arbitrary signal intensity units, was converted to a curve of percent enhancement, by dividing by the baseline signal-intensity, namely, the signal intensity in the blood pool before any contrast enhancement. A calibration curve of effective contrast
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
enhancement versus the contrast enhancement extrapolated from the low R1 (contrast concentration) range was calculated by numerical simulation, using the sequence parameters, and a mean pre-contrast T1 value for blood of 1,450 ms. This calibration curve was then inverted, and used for correction of the observed percent contrast enhancement in the blood pool. Saturation correction resulted, on average, in a 20% to 30% increase of the peak contrast enhancement of the arterial input function. The arterial input function corrected for signal saturation was used for myocardial blood flow (MBF) quantification by model-independent deconvolution (17). Transmural MPR was calculated by dividing hyperemic MBF by resting MBF. All MPR results refer to uncorrected values, although resting MBF results are also displayed corrected for LVRPP. Two experienced observers (C.D.S., G.P.M.) qualitatively assessed the perfusion images for subendocardial perfusion defects (a visible defect for ⱖ5 heartbeats) and the LGE images for focal fibrosis, categorized as present or absent. Additionally, computer-assisted planimetry was used to determine the mass and percentage of enhanced myocardium ⬎5 SD above the mean signal intensity of remote normal myocardium. Statistical analysis. The distributions of continuous variables were assessed using graphical displays and the Shapiro-Wilk test for normality. Values are reported as mean ⫾ SD or median (interquartile range) if the distribution was not normal. Septal E/E=, lateral E/E=, and LGE (%LV mass) were skewed and logarithmically transformed before regression analysis. Statistical analysis was undertaken using SPSS, version 16.0 (SPSS, Chicago, Illinois) and SAS (SAS Institute, Cary, North Carolina). Comparison of means of 2 groups was done with a t test and ⬎2 groups with 1-way analysis of variance. Linear regression investigated possible associations of continuous outcome measures. Stepwise selection methods were used to determine the most important associations. Hyperemic MBF was adjusted for resting MBF (thereby assessing MPR) by forcing resting MBF into the model before selection methods were applied. The same method was used when adjusting peak VO2 for age and sex. All p values ⬍0.05 were considered significant.
RESULTS Figure 1. Recruitment Data The number of patients approached, excluded, and recruited. AR ⫽ aortic regurgitation; CMR ⫽ cardiac magnetic resonance; COPD ⫽ chronic obstructive pulmonary disease; EF ⫽ ejection fraction; LGE ⫽ late gadolinium enhancement.
Population characteristics. Details of recruitment are outlined in Figure 1. Demographic data for the 46 patients included in this report are presented in Table 1.
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
0.86 ⫾ 0.22/0.45 ⫾ 0.13
VO2 are summarized in Table 3. Peak VO2 was higher in men compared with women: 16.2 ⫾ 4.1 ml/kg/min versus 12.3 ⫾ 3.1 ml/kg/min, respectively (p ⫽ 0.005), and decreased with age (Fig. 3A). Peak VO2 was inversely related to septal E/E= and resting MBF, and correlated with MPR ( ⫽ 0.45, r2 ⫽ 0.2) (Fig. 3B), but not with resting MBF/LVRPP or hyperemic MBF. Only MPR was independently associated with age- and sex-corrected peak VO2 on stepwise analysis. Correction for -blocker usage did not affect the significance of the models. Symptomatic status. NYHA functional class was associated with peak VO2, even after adjustment for age and sex ( ⫽ ⫺0.4, p ⫽ 0.002). Patients with higher NYHA functional class had lower MPR (p ⫽ 0.001) (Fig. 4), despite no significant differences in AS severity between groups. Adding NYHA functional class to the stepwise regression for associations with peak VO2 had no effect on the significance of MPR as the only independent variable.
1.46 ⫾ 0.29
Associations with resting MBF and perfusion reserve.
Table 1. Patient Characteristics Age, yrs Male/female
66.2 ⫾ 9.3 34/12
Body surface area, m2
1.92 ⫾ 0.21
Systolic blood pressure, mm Hg
131 ⫾ 20
Diastolic blood pressure, mm Hg
74 ⫾ 10
Resting heart rate, beats/min
69 ⫾ 12
ACEI/ARB
15 (33%)
Beta-blocker
18 (39%)
Statin
28 (61%)
Diabetes mellitus
10 (22%)
Hypertension
27 (59%)
Smoking Former Current NYHA functional class, (n)
29 (63%) 3 (6%) I (9); II (32); III (5)
Echocardiography Peak AV velocity, m/s Mean PG, mm Hg AVA/I, (cm2)/(cm2/m2) LVRPP, mm Hg·beats/min·10⫺4 Resting DPT, s/min
4.4 ⫾ 0.6 48.5 ⫾ 14.1
37.9 ⫾ 3.1
Septal E/E=
13.7 (11.1–18.6)
Lateral E/E=
10.2 (7.2–14.0)
Values are mean ⫾ SD, n, or n(%). ACEI ⫽ angiotensin-converting enzyme inhibitor; ARB ⫽ angiotensinreceptor blocker; AV ⫽ aortic valve; AVA/I ⫽ aortic valve area/index; DPT ⫽ diastolic perfusion time; E/E= ⫽ transmitral flow velocity to annular velocity ratio; LVRPP ⫽ left ventricular rate pressure product; NYHA ⫽ New York Heart Association; PG ⫽ pressure gradient.
Resting and hyperemic MBF were higher in women than men (resting, 1.07 ⫾ 0.24 vs. 0.84 ⫾ 0.15, p ⫽ 0.001; hyperemic, 2.16 ⫾ 0.53 vs. 1.64 ⫾ 0.38, p ⫽ Table 2. Cardiac Magnetic Resonance and Cardiopulmonary Exercise Data Cardiac magnetic resonance LVMI, g/m2
69.6 ⫾ 17.7
Cardiopulmonary exercise data. One exercise test was
LVEDVI, ml/m2
96.9 ⫾ 14.6
medically terminated because of angina associated with a ⬎20 mm Hg drop in systolic blood pressure. There were no complications. Twenty-two patients (48%) were symptom limited, and the remainder were limited by fatigue, including all 9 patients in New York Heart Association (NYHA) functional class I. Exercise data are presented in Table 2. Cardiac magnetic resonance. Volumetric data and LGE images were available for all patients. Results are displayed in Table 2. Interobserver and intraobserver variability were as follows: LVM 5.2% and 4.9%, LVEDV 3.1% and 3%, LVEF 6.6% and 4.7%. Thirty-four patients had global subendocardial perfusion defects. There was a wide range of LVMI and MPR, but all patients had normal or near normal systolic function (n ⫽ 5, LVEF 45% to 50%; n ⫽ 1, LVEF 40% to 45%). LGE was present in 26 (56%) patients, and except in the patient excluded, there were no clear myocardial infarctions. Examples of patients with variable LVM, LGE, and MPR are shown in Figure 2. Associations with exercise capacity. Univariate analyses and results from the stepwise model selection for peak
LVESVI, ml/m2
42.5 ⫾ 10.6
LVM/LVEDV, g/ml
0.72 ⫾ 0.15
LVEF, %
56.5 ⫾ 6.5
Stroke volume index, ml/m2
53.6 ⫾ 7.7
Maximum wall thickness, mm
13.2 ⫾ 2.5
Resting MBF, ml/min/g
0.90 ⫾ 0.20
Resting MBF/LVRPP, (ml/min/g)/ (mm Hg.beats/min.10⫺4)
0.60 ⫾ 0.14
Hyperemic MBF, ml/min/g
1.77 ⫾ 0.47
Hyperemic heart rate, beats/min
88 ⫾ 12
MPR (n ⫽ 41)
2.03 ⫾ 0.55
LGE positive
26 (56%)
LGE, % (⬎5 SD above remote) LGE mass, g (⬎5 SD above remote)
1 (0–28) 1.8 (0.8–4.8)
Cardiopulmonary exercise testing Peak heart rate, % predicted Rise in systolic blood pressure, mm Hg
85 ⫾ 12 39 ⫾ 23
Respiratory exchange ratio
1.09 ⫾ 0.1
Peak VO2, ml/kg/min
15.2 ⫾ 4.2
Values are mean ⫾ SD, n (%), or median interquartile range. LGE ⫽ late gadolinium enhancement; LVEDV ⫽ left ventricular enddiastolic volume; LVEDVI ⫽ left ventricular end-diastolic volume index; LVEF ⫽ left ventricular ejection fraction; LVESVI ⫽ left ventricular end-systolic volume index; LVM ⫽ left ventricular mass; LVMI ⫽ left ventricular mass index; MBF ⫽ myocardial blood flow; MPR ⫽ myocardial perfusion reserve, VO2 ⫽ volume oxygen consumption; other abbreviation as in Table 1.
185
186
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
Figure 2. Examples of Patients With Varying LV Remodeling Cardiac magnetic resonance images: (i) short-axis cine end-diastole at middle level (A, B, and C) and basal level (D); (ii) late gadolinium enhancement (LGE) at same level as “i” (LGE marked with white arrows); (iii) perfusion imaging in middle left ventricular (LV) slice during hyperemia; and (iv) during rest perfusion. Four male patients displayed: (A) Significantly raised left ventricular mass (LVM), clear insertion point LGE, and reduced myocardial perfusion reserve (MPR [age 63 years, left ventricular mass index (LVMI) 126 g/m2, aortic valve area (AVA) 0.92 cm2, MPR 1.07, peak volume oxygen consumption (VO2) 12.7 ml/kg/min). (B) Mildly elevated LVM with mild insertion point LGE (age 65 years, LVMI 76 g/m2, AVA 0.79 cm2, MPR 1.59, peak VO2 16.4 ml/kg/min). (C) Normal LVM with no LGE (age 65 years, LVMI 55 g/m2, AVA 0.99 cm2, MPR 2.30, peak VO2 18.5 ml/kg/min). (D) Normal LVM with mild insertion point LGE but more significant remote LGE (age 63 years, LVMI 66 g/m2, AVA 0.73 cm2, MPR 2.4, peak VO2 20.6 ml/kg/min).
0.002). Resting MBF correlated with resting LVRPP ( ⫽ 0.47, p ⫽ 0.02) and was inversely related to DPT ( ⫽ ⫺0.35, p ⫽ 0.025). Sex and resting LVRPP were independent associations with resting MBF. MPR was inversely related to peak aortic valve velocity, mean pressure gradient, CMR measures of LVH, LGE category, and septal E/E=. Results are summarized in Table 4. MPR was not related to history of smoking, diabetes, or hypertension. With all univariable associations, the best stepwise multivariate regression model contained LVMI and LGE category as independent associations. MPR was reduced in patients with LGE (1.87 ⫾ 0.47 vs. 2.24 ⫾ 0.58, p ⫽ 0.031), and there was also an inverse correlation between MPR and LGE scar mass at 5 SD ( ⫽ ⫺0.4, p ⫽ 0.008). DISCUSSION Determinants of resting MBF and MPR. Resting and hyperemic MBF were higher in women, as has been found previously (18). As reported by Rajappan et al. (4), we did not find that measures of AS severity or LVMI were related to resting MBF. Contrary to that
study, there was an association between resting MBF and estimated myocardial work (LVRPP), as has also been shown in hypertrophic cardiomyopathy (18). The results of MPR in this study are remarkably consistent with those from Rajappan’s similar cohort of patients with severe AS before AVR (2.03 ⫾ 0.55 vs. 1.90 ⫾ 0.60, respectively), considering that positron emission tomography was used in that report (4). We have shown that aortic valve velocities/pressure gradients, measures of LVM, and, for the first time, female sex, LV filling pressure (septal E/E=), and myocardial fibrosis (LGE), have univariate associations with MPR. Of these, LVMI and LGE appear to be the most important. These findings are somewhat different from those of Rajappan et al. (4), whose results suggested that AS severity and DPT were more important determinants of MPR in severe AS. The discrepancies are likely due to the smaller numbers (20 vs. 41), that positron emission tomography was used, that no LGE for fibrosis was performed, and that the majority of patients in the Rajappan study were asymptomatic (4). Given the modest correlations found in our study, we cannot exclude that the
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
Table 3. Associations With Peak VO2 Univariate Associations

p Value
Sex
⫺0.41
0.005
Age
⫺0.32
0.03
Peak AV velocity
⫺0.18
0.24
Mean PG
⫺0.18
0.24
0.04
0.79
Septal E/E=
AVAI
⫺0.34
0.02
Lateral E/E=
⫺0.23
0.13
0.02
0.89
LVMI LVM/LVEDV ratio
⫺0.04
0.79
LV maximum wall thickness
⫺0.2
0.90
LVEF
0.01
0.97
LGE
0.05
0.73
and sex-corrected exercise capacity. This clinical study has identified imaging and clinical parameters that have a clear and independent association with exercise capacity in AS. Previous studies have suggested a range of parameters that include aortic valve compliance (23–24), reduced longitudinal strain (25), LV filling pressure (13) and diastolic dysfunction (3, 11). However, those studies have looked at exerciseinduced symptoms, which may be subjective, as in our study, where 50% of symptomatic patients were limited by fatigue, and they have not accounted for age and sex as determinants of exercise capacity. Many other factors may influence peak VO2, including genetics, physical activity, and skeletal muscle changes, which may be particularly important in patients with cardiac dysfunction.
Resting MBF
⫺0.53
0.001
Resting MBF/LVRPP
⫺0.28
0.08
Hyperemic MBF
⫺0.01
Mechanisms linking reduced myocardial perfusion reserve and peak VO2. With exercise, cardiac output
0.94
rises incrementally with workload. Increased cardiac
MPR
0.45
0.004
Final Model (Stepwise Selection)

p Value
R2
Sex
⫺0.44
0.002
0.42
Age
⫺0.15
0.25
MPR
0.46
0.001
LV ⫽ left ventricular; other abbreviations as in Tables 1 and 2.
findings may be spurious; however, they are physiologically plausible. The lack of association between smoking, hypertension, and diabetes (19) and MPR in this study is almost certainly due to the smaller sample size and the large effect of LV remodeling in these patients with AS. This is the first study to demonstrate the importance of myocardial fibrosis (LGE) in relation to MPR in AS. This association may be explained by reduced arteriolar and capillary density seen with LV remodeling and fibrosis (20) and loss of the “suction” wave seen in patients with LVH (21). It is important to emphasize that diffuse myocardial fibrosis (22) was not assessed (as newer T1mapping techniques were not available to us at the time), and this may be an important determinant of MPR. However, the extent of LGE does correlate with diffuse fibrosis on myocardial biopsy (1). Associations with exercise capacity. We hypothesized that CMR measures of LV remodeling would be independently associated with aerobic exercise capacity. As previously demonstrated, measures of AS severity did not predict exercise capacity (15). Septal E/E=, resting MBF and MPR, but not measures of LVH, had univariate associations with peak VO2. Only MPR was independently associated with age-
Figure 3. Associations With Exercise Capacity Relationship between peak volume oxygen consumption (VO2) and (A) age and (B) myocardial perfusion reserve.
187
188
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
Figure 4. Perfusion Reserve and Symptomatic Status Myocardial perfusion reserve and New York Heart Association (NYHA) functional class. CI ⫽ confidence interval.
output in severe AS is largely dependent on increased heart rate (26), which is associated with the development of impaired MPR in both patients (27) and experimental models of AS (28). The inability to increase blood flow to the myocardium is limited because vasodilation may already be near maximal (29), DPT will be limited further with increased heart rate, and there is a rapid increase in LV end-diastolic pressure (26), which further reduces the effective Table 4. Associations With Perfusion Reserve Univariate Associations

p Value
Sex
0.38
0.02
Age
⫺0.09
0.54
Hyperemic heart rate
0.2
0.19
Beta-blockers
0.008
0.96
LVRPP
⫺0.1
0.58
Resting DPT
⫺0.28
0.08
Peak AV velocity
⫺0.34
0.02
Mean PG
⫺0.30
0.04
AVAI
0.21
0.17
LVMI
⫺0.51
⬍0.001
LVM/LVEDV
⫺0.43
0.003
LV maximum wall thickness
⫺0.41
0.005
LVEF
0.28
0.07
LGE
⫺0.46
0.002
Septal E/E=
⫺0.33
0.03
Lateral E/E=
⫺0.19
0.20
Final Model (Stepwise Selection)

p Value
R2
LVMI
⫺0.40
0.004
0.43
LGE
⫺0.31
0.02
Abbreviations as in Tables 1, 2, and 3.
pressure gradient for perfusion. Patients unable to increase blood flow to the myocardium on exercise will likely develop subendocardial myocardial dysfunction (25,28), which will limit cardiac output and contribute to exercise intolerance. Clinical implications. MPR is an important prognostic marker in hypertrophic cardiomyopathy, predicting adverse ventricular remodeling and clinical outcomes (30). Impaired MPR could be a therapeutic target in AS patients unable to have aortic valve intervention and in patients with persistent symptoms after AVR. Traditional therapies targeting LV remodeling such as inhibitors of the renin-angiotensin-aldosterone system and drugs with novel mechanisms of action, for example, perhexilene, may be useful in this context. The importance of impaired MPR to the development of symptoms and outcome in asymptomatic patients with AS warrants further investigation. Study limitations. The results of this mechanistic study may not be applicable to the wider AS population. Although we have touched on symptomatic status in this paper, the number of asymptomatic patients is small and, therefore, limited conclusions can be drawn about this group. The cross-sectional nature of our data does not allow one to draw conclusions as to whether myocardial fibrosis leads to a reduction in MPR or if the reduced MPR causes myocardial fibrosis. Longitudinal studies in patients with less severe AS with assessment of diffuse fibrosis may be able to answer this question. The reproducibility of MPR measurements in AS is unknown. Finally, because of the relatively small numbers, the number of variables that can be appropriately used in the multivariate regression analysis is necessarily limited. Therefore, we limited the number of univariate associations entered into the multivariate model to 5 and used the strongest univariate associations, avoiding overlapping variables. CONCLUSIONS
We have shown for the first time that MPR is independently associated with objectively measured exercise capacity in severe AS. Impaired MPR is determined by a combination of factors including AS severity, LV remodeling, and filling pressure. Further work is required to determine how CMR assessment of MPR can aid the clinical management of patients with AS. Reprint requests and correspondence: Dr. Gerry P. Mc-
Cann, Department of Cardiovascular Sciences, Glenfield Hospital, Groby Road, Leicester LE3 9QP, United Kingdom. E-mail: [email protected].
JACC: CARDIOVASCULAR IMAGING, VOL. 5, NO. 2, 2012 FEBRUARY 2012:182–9
REFERENCES
1. Azevedo CF, Nigri M, Higuchi ML, et al. Prognostic significance of myocardial fibrosis quantification by histopathology and magnetic resonance imaging in patients with severe aortic valve disease. J Am Coll Cardiol 2010; 56:278 – 87. 2. Weidemann F, Herrmann S, Stork S, et al. Impact of myocardial fibrosis in patients with symptomatic severe aortic stenosis. Circulation 2009;120: 577– 84. 3. Hess OM, Ritter M, Schneider J, Grimm J, Turina M, Krayenbuehl HP. Diastolic stiffness and myocardial structure in aortic valve disease before and after valve replacement. Circulation 1984;69:855– 65. 4. Rajappan K, Rimoldi OE, Dutka DP, et al. Mechanisms of coronary microcirculatory dysfunction in patients with aortic stenosis and angiographically normal coronary arteries. Circulation 2002;105:470 – 6. 5. Kupari M, Turto H, Lommi J. Left ventricular hypertrophy in aortic valve stenosis: preventive or promotive of systolic dysfunction and heart failure? Eur Heart J 2005;26:1790 – 6. 6. Esposito G, Rapacciuolo A, Naga Prasad SV, et al. Genetic alterations that inhibit in vivo pressure-overload hypertrophy prevent cardiac dysfunction despite increased wall stress. Circulation 2002;105:85–92. 7. Debl K, Djavidani B, Buchner S, et al. Delayed hyperenhancement in magnetic resonance imaging of left ventricular hypertrophy caused by aortic stenosis and hypertrophic cardiomyopathy: visualisation of focal fibrosis. Heart 2006;92:1447–51. 8. Rudolph A, Abdel-Aty H, Bohl S, et al. Noninvasive detection of fibrosis applying contrast-enhanced cardiac magnetic resonance in different forms of left ventricular hypertrophy: relation to remodeling. J Am Coll Cardiol 2009;53:284 –91. 9. Francis DP, Shamim W, Ceri Davies L, et al. Cardiopulmonary exercise testing for prognosis in chronic heart failure: continuous and independent prognostic value from VE/VCO2 slope and peak VO2. Eur Heart J 2000;21:154 – 61. 10. Rajani R, Rimington H, Chambers JB. Treadmill exercise in apparently asymptomatic patients with moderate or severe aortic stenosis: relationship
between cardiac index and revealed symptoms. Heart 2010;96:689 –95. 11. Monrad ES, Hess OM, Murakami T, Nonogi H, Corin WJ, Krayenbuehl HP. Abnormal exercise hemodynamics in patients with normal systolic function late after aortic valve replacement. Circulation 1988;77:613–24. 12. Gottdiener JS, Bednarz J, Devereux R, et al. American Society of Echocardiography recommendations for use of echocardiography in clinical trials. J Am Soc Echocardiogr 2004; 17:1086 –119. 13. Dalsgaard M, Kjaergaard J, Pecini R, et al. Left ventricular filling pressure estimation at rest and during exercise in patients with severe aortic valve stenosis: comparison of echocardiographic and invasive measurements. J Am Soc Echocardiogr 2009;22:343–9. 14. Ferro G, Duilio C, Spinelli L, Liucci GA, Mazza F, Indolfi C. Relation between diastolic perfusion time and coronary artery stenosis during stressinduced myocardial ischemia. Circulation 1995;92:342–7. 15. Das P, Rimington H, Chambers J. Exercise testing to stratify risk in aortic stenosis. Eur Heart J 2005;26: 1309 –13. 16. Simonetti OP, Kim RJ, Fieno DS, et al. An improved MR imaging technique for the visualization of myocardial infarction. Radiology 2001;218:215–23. 17. Jerosch-Herold M, Swingen C, Seethamraju RT. Myocardial blood flow quantification with MRI by model-independent deconvolution. Med Phys 2002;29:886 –97. 18. Petersen SE, Jerosch-Herold M, Hudsmith LE, et al. Evidence for microvascular dysfunction in hypertrophic cardiomyopathy: new insights from multiparametric magnetic resonance imaging. Circulation 2007;115: 2418 –25. 19. Wang L, Jerosch-Herold M, Jacobs DR Jr., Shahar E, Detrano R, Folsom AR. Coronary artery calcification and myocardial perfusion in asymptomatic adults: the MESA (Multi-Ethnic Study of Atherosclerosis). J Am Coll Cardiol 2006;48:1018 –26. 20. Marcus ML, Harrison DG, Chilian WM, et al. Alterations in the coronary circulation in hypertrophied ventricles. Circulation 1987;75:I19 –I25. 21. Davies JE, Whinnett ZI, Francis DP, et al. Evidence of a dominant backwardpropagating “suction” wave responsible
Steadman et al. Myocardial Perfusion Reserve in Aortic Stenosis
for diastolic coronary filling in humans, attenuated in left ventricular hypertrophy. Circulation 2006;113:1768 –78. 22. Flett AS, Hayward MP, Ashworth MT, et al. Equilibrium contrast cardiovascular magnetic resonance for the measurement of diffuse myocardial fibrosis: preliminary validation in humans. Circulation 2010;122:138 – 44. 23. Das P, Rimington H, Smeeton N, Chambers J. Determinants of symptoms and exercise capacity in aortic stenosis: a comparison of resting haemodynamics and valve compliance during dobutamine stress. Eur Heart J 2003;24:1254 – 63. 24. Lancellotti P, Karsera D, Tumminello G, Lebois F, Pierard LA. Determinants of an abnormal response to exercise in patients with asymptomatic valvular aortic stenosis. Eur J Echocardiogr 2008;9:338 – 43. 25. Lafitte S, Perlant M, Reant P, et al. Impact of impaired myocardial deformations on exercise tolerance and prognosis in patients with asymptomatic aortic stenosis. Eur J Echocardiogr 2009;10:414 –9. 26. Lee SJ, Jonsson B, Bevegard S, Karlof I, Astrom MD. Hemodynamic changes at rest and during exercise in patients with aortic stenosis of varying severity. Am Heart J 1970;79:318 –31. 27. Trenouth RS, Phelps NC, Neill WA. Determinants of left ventricular hypertrophy and oxygen supply in chronic aortic valve disease. Circulation 1976;53:644 –50. 28. Nakano K, Corin WJ, Spann JF Jr., Biederman RW, Denslow S, Carabello BA. Abnormal subendocardial blood flow in pressure overload hypertrophy is associated with pacinginduced subendocardial dysfunction. Circ Res 1989;65:1555– 64. 29. Hoffman JI. Determinants and prediction of transmural myocardial perfusion. Circulation 1978;58:381–91. 30. Cecchi F, Olivotto I, Gistri R, Lorenzoni R, Chiriatti G, Camici PG. Coronary microvascular dysfunction and prognosis in hypertrophic cardiomyopathy. N Engl J Med 2009;349: 1027–35.
Key Words: aortic stenosis y cardiac magnetic resonance y cardiopulmonary exercise testing y myocardial perfusion reserve.
189