Objective
To evaluate, by noninvasive coronary flow velocity reserve (CFVR), whether patients with asymmetric hypertrophic cardiomyopathy (HC), with or without left ventricular outflow tract obstruction, demonstrate significant regional differences of CFVR.
Methods
We evaluated 61 patients with HC (27 men; mean age 49 ± 16 years), including 20 patients with hypertrophic obstructive cardiomyopathy (HOCM) and 41 patients without obstruction (HCM). The control group included 20 age- and sex-matched subjects. Transthoracic Doppler echocardiography CFVR of the left anterior descending coronary artery (LAD) and the posterior descending coronary artery (PD) were performed, including calculation of relative CFVR as the ratio between CFVR LAD and CFVR PD.
Results
Compared with the controls, all the patients with HC had lower CFVR LAD (2.12 ± 0.53 vs 3.34 ± 0.67; P < .001) and CFVR PD (2.29 ± 0.49 vs 3.21 ± 0.65; P < .001). CFVR LAD in HOCM group in comparison with the HCM group was significantly lower (1.93 ± 0.42 vs 2.22 ± 0.55; P = .047), due to higher basal diastolic coronary flow velocities (0.40 ± 0.09 vs 0.33 ± 0.07 m/sec; P = .002), with similar hyperemic diastolic flow velocities (0.71 ± 0.16 vs 0.76 ± 0.19 m/sec; P = .330), respectively. There was no significant difference in CFVR PD between patients with HOCM and those with HCM (2.33 ± 0.46 vs 2.27 ± 0.50; P = .636), respectively. Relative CFVR was lower in the HOCM group compared with the HCM group (0.84 ± 0.16 vs 0.98 ± 0.14; P = .001). By multivariable regression analysis, left ventricular outflow tract gradient was the independent predictor of CFVR LAD (B = −0.24; P = .008) and relative CFVR (B = −0.34; P = .016).
Conclusions
CFVR LAD and relative CFVR were significantly lower in patients with HOCM compared with patients with HCM. Regional differences of CFVR are present only in patients with significant left ventricular outflow tract obstruction, which suggests that obstruction per se, by increasing wall stress in basal conditions, leads to higher basal diastolic coronary flow velocities and results in lower CFVR in LAD compared with PD.
In patients with hypertrophic cardiomyopathy (HC), evaluation of microcirculation and its influence on prognosis has been a matter of considerable interest in the past years. In fact, the degree of microvascular dysfunction in patients with HC was identified as a marker of clinical deterioration and potentially poor survival, which suggests quantification of coronary flow velocity reserve (CFVR) as clinically relevant and important. Mechanisms of microvascular dysfunction in HC involve vascular remodeling and reduced capillary density, but extravascular compression due to elevated left ventricular (LV) end-diastolic pressure and LV systolic pressure might also contribute to perfusion abnormalities, predominantly in subendocardial layers of myocardium. In regard to regional differences in microcirculation, there are conflicting data on coronary flow reserve (CFR) in HC evaluated by invasive techniques or positron emission tomography (PET). In addition, specific differences of microcirculation in patients with HC and with and without LV outflow obstruction have not previously been studied in detail. Transthoracic Doppler echocardiography (TDE) CFVR is a versatile, reproducible, and noninvasive tool for evaluation of microcirculation in the absence of epicardial stenosis, which could further stratify patients with HC to increased risk of clinical worsening during follow up. The aim of this study was to evaluate, by noninvasive CVFR, whether patients with asymmetric HC, with or without LV outflow tract obstruction, demonstrate significant regional differences of CFVR.
Methods
Study Population
We prospectively included 64 patients with asymmetric HC; but, due to a technical inability to assess CFVR in both arteries in 3 patients, the final group consisted of 61 patients (27 men, 34 women; mean age, 49 ± 16 years). The patients were evaluated at the Clinic of Cardiology, Clinical Center of Serbia, during the period October 2007 to March 2011. Patients’ clinical characteristics, New York Heart Association (NYHA) class, the presence of arrhythmia on 24-hour Holter monitoring, and medical therapy were recorded for each patient.
Inclusion criteria were as follows: echocardiographic evidence of myocardial hypertrophy, defined by a LV myocardial wall thickness ≥15 mm and septum to posterior wall ratio >1.3, in the absence of another cardiac or systemic cause of LV hypertrophy. Hypertrophic obstructive cardiomyopathy (HOCM) was defined if systolic gradient at rest was ≥30 mm Hg in the LV outflow tract. Angiography was performed in 31 patients, and none of them had any coronary stenosis. In addition, no epicardial coronary “bridging” was observed in a patient who underwent cardiac catheterization. The remaining patients had either a negative stress echocardiography test or <5% probability of having coronary artery disease.
Criteria for exclusion from the study were as follows: significant mitral regurgitation, the presence of diabetes, NYHA class III or IV, poor echocardiographic window (for CFVR assessment), chronic obstructive pulmonary disease, and atrioventricular block. The control group included 20 age- and sex-matched subjects (13 men; mean age, 43 ± 11 years; P = .086 for age; P = .107 for sex) with atypical chest pain referred to coronary angiography by the discretion of their physician. All control subjects had normal findings on coronary angiography, echocardiography, and physical examination. Informed consent was obtained for all the participants; the local ethical committee approved the study protocol.
Echocardiographic Examination
Echocardiographic studies were performed with an available digital ultrasound system (Acuson Sequoia C256; Siemens Medical Solutions USA, Inc, Mountain View, CA), with a 3V2C multifrequency transducer by using second-harmonic technology. Standard two-dimensional (2D), M-mode, pulsed Doppler measures were performed according to the American Society of Echocardiography guidelines. The average of 3 cardiac cycles was used for measurements of cardiac dimensions. After data were collected in M-mode in the parasternal long-axis view: LV end-diastolic diameter, LV end-systolic diameter, end-diastolic left atrial diameter, end-diastolic LV interventricular septum (IVS), and left posterior wall (PW) thickness. In addition, we calculated the IVS-PW ratio as a LV morphologic parameter. Inferior wall and PW thickness were also measured by 2D echocardiography from the short-axis parasternal view at the mitral valve level. Left atrial volume and ejection fraction were assessed by using the modified Simpson biplane method. LV outflow tract gradient (LVOTG) was examined by combined use of color-Doppler, pulsed-wave Doppler, and continuous-wave Doppler echocardiography. LV rate-pressure product (RPP) ([peak LVOTG + systolic blood pressure] × heart rate) was calculated, as previously described.
CFVR
After standard echocardiographic examination, TDE CFVR was performed. The distal portion of the left anterior descending coronary artery (LAD) and posterior descending coronary artery (PD) was evaluated by using a 4-MHz transducer. For color-Doppler flow mapping, the velocity range was set in the range of 16–24 cm/sec. Visualization of distal segment of LAD artery was performed in a modified 3-chamber view. Evaluation of PD artery coronary flow was accomplished in apical 2-chamber view. From this position, a probe was slightly rotated anticlockwise and tilted anteriorly, until the coronary blood flow in the posterior interventricular groove was identified by color-Doppler flow mapping. Blood flow velocity was measured by pulsed-wave Doppler echocardiography by using a sample volume of 3–5 mm wide. Alignment of the ultrasound beam direction with the distal LAD and PD flow was as parallel as possible, with the stable transducer position at rest and during maximal hyperemia. Peak diastolic coronary flow velocity was measured in basal conditions and during maximal hyperemia, which was induced with adenosine (0.14 mg/kg/min intravenously, during 2 minutes). Three optimal diastolic flow profiles at rest and during hyperemia were measured, and the results were averaged. Absolute CFVR was calculated as the ratio of hyperemic to basal peak diastolic flow velocities. Relative CFVR was calculated as the ratio between CFVR LAD and CFVR PD. Diastolic deceleration time (DDT) was measured from the peak diastolic velocity to the point of intercept of initial decay slope with baseline. All studies were recorded on VHS videotapes, and echocardiographic images and clips were stored on magneto-optical discs for offline analysis. Blood flow velocity measurements were done offline by using the integrated softer package of the ultrasound system, by 2 experienced investigators (A.D.-D. and M.T.). We previously reported an interobserver agreement for CFVR evaluation of 90%.
Statistical Analysis
All data were entered into a specially created database and then processed in the statistical program SPSS version 15.0 (SPSS Inc, Chicago, Illinois). All numeric data were expressed as mean (SD), and attributive as frequency or percentages. Differences in continuous variables were assessed with the Student’s t test or the Fisher one-way analysis of variance with Bonferroni adjustment in the course of multiple tests by following Kolmogorov-Smirnov test. The χ 2 test was used for categorical variables. A statistical correlation among echocardiographic variables was examined with the Pearson linear correlation coefficient. Multivariable linear regression was performed, which allowed all continuous variables (for greater statistical power) with P < .05 from univariable linear regression analysis to enter the model. Results are expressed as partial linear regression coefficients (B) and their 95% CIs for a 1 SD change in the variable. We used enter method of variable entry in multivariable linear regression analysis to detect independent predictors of CFVR for LAD, CFVR for PD, and relative CFVR for patients with HC. Statistical significance was defined as P < .05.
Results
Of 64 patients, 1 patient was excluded because of a technically inadequate coronary flow signal in LAD artery during hyperemia, whereas 2 patients were excluded due to an inability to visualize baseline coronary flow in PD artery. Thus, feasibility to assess CFVR for patients with HC was 98.4% for LAD, and 96.7% for PD. According to the presence of LV outflow tract gradient at rest (≥30 mm Hg), the patients were divided into 20 patients who had HOCM, and 41 patients with HCM.
Clinical characteristics of the whole HC group as well as patients with HOCM and those with HCM are presented in Table 1 . There were no statistical significant differences in age, sex, and body surface area, the presence of angina and syncope, cardiac arrhythmias, medical therapy, and baseline heart rate between patients with HOCM and patients with HCM. An NYHA functional class II was more prevalent in patients with HOCM. Baseline RPP was significantly higher in the patients with HOCM compared with the patients with HCM.
| Variables | Controls ( n = 20) | HCM ( n = 41) | HOCM ( n = 20) | P value; HCM vs HOCM |
|---|---|---|---|---|
| Mean (SD) age, y | 43 ± 11 | 46 ± 16 | 53 ± 15 | .080 |
| No. (%) men | 13 (65) | 19 (46) | 8 (40) | .640 |
| BSA, m 2 | 2.00 ± 0.23 | 1.84 ± 0.17 ∗ | 1.85 ± 0.16 | .885 |
| Angina, no. (%) | 10 (50) | 28 (68) | 14 (70) | .892 |
| Syncope, no. (%) | 0 (0) | 3 (7) | 5 (25) | .055 |
| NYHA functional class, no. (%) | † | † | .020 | |
| I | 20 (100) | 21 (51) | 4 (20) | |
| II | 0 (0) | 20 (49) | 16 (80) | |
| Unsustained ventricular tachycardia on Holter ECG, no. (%) | 0 (0) | 5 (12) | 3 (15) | .818 |
| Paroxysmal or chronic atrial fibrillation, no. (%) | 0 (0) | 10 (24) | 3 (15) | .372 |
| Medical therapy, no. (%) | 15 (75) | 37 (90) | 19 (95) | .525 |
| Beta blockers, no. (%) | 15 (75) | 32 (78) | 18 (90) | .254 |
| Verapamil, no. (%) | 1 (5) | 10 (24) | 8 (40) ∗ | .210 |
| Amiodarone, no. (%) | 0 (0) | 8 (20) ∗ | 4 (20) ∗ | .964 |
| Mean (SD) baseline heart rate, beats/min | 74 ± 19 | 70 ± 14 | 72 ± 11 | .632 |
| Mean (SD) peak heart rate during hyperemia, beats/min | 84 ± 18 | 75 ± 18 | 76 ± 12 | .854 |
| Diastolic blood pressure, mm Hg | 81 ± 7 | 77 ± 9 | 77 ± 8 | .968 |
| Systolic blood pressure, mm Hg | 124 ± 9 | 121 ± 17 | 120 ± 18 | .807 |
| Baseline RPP, mm Hg/min | 9862 ± 2601 | 9185 ± 2191 | 13103 ± 2805 † | <.001 |
∗ P < .05 for comparison of Control vs HCM/HOCM.
Echocardiographic Data
Echocardiographic data are presented in Table 2 . Patients with HOCM in comparison with patients with HCM had significantly thicker IVS and PW, whereas IVS to PW ratio was not statistically different, which suggests a asymmetric pattern in both groups of patients. There was no significant difference between inferior and PW thickness for HOCM (mean [SD], 13.7 ± 3.33 mm vs 13.3 ± 3.2 mm; P = ns) and HCM (mean [SD], 10.8 ± 2.8 mm vs 10.6 ± 2.4 mm; P = ns) group measured by 2D echocardiography from the short-axis parasternal view at the mitral valve level. LV ejection fraction and fractional shortening were also significantly higher in patients with HOCM. However, the LV end-diastolic dimension and the LV end-systolic dimension were larger in the HCM group.
| Variables | Controls ( n = 20) | HCM ( n = 41) | HOCM ( n = 20) | P value; HCM vs HOCM |
|---|---|---|---|---|
| LV end-diastolic dimension, mm | 49.4 ± 3.8 | 46.8 ± 5.0 | 44.3 ± 3.8 ∗ | .050 |
| LV end-systolic dimension, mm | 31.2 ± 4.1 | 27.5 ± 5.2 | 23.8 ± 4.4 † | .008 |
| IVS thickness, mm | 9.2 ± 2.2 | 18.8 ± 3.5 † | 21.2 ± 4.4 † | .027 |
| PW thickness, mm | 8.6 ± 1.1 | 10.1 ± 2.3 | 12.7 ± 3.6 † | .001 |
| Maximal LV thickness, mm | 9.7 ± 2.8 | 21.6 ± 3.8 † | 23.8 ± 4.2 † | .040 |
| IVS to PW ratio | 1.06 ± 0.14 | 1.95 ± 0.6 † | 1.79 ± 0.7 † | .338 |
| LV ejection fraction, % | 67 ± 5.7 | 67.8 ± 8.8 | 74.2 ± 5.4 | .001 |
| Left atrial volume/BSA, mL/m 2 | 27.0 ± 4.9 | 29.0 ± 11.4 | 38.4 ± 15.6 ∗ | .011 |
| LVOTG at rest, mm Hg | 6 ± 2.0 | 10.1 ± 5.1 | 62.1 ± 27.8 † | .001 |
∗ P < .05 for comparison of Control vs HCM/HOCM.
Coronary Flow Characteristics
Quantitative values of coronary flow of LAD and PD at rest, during hyperemia, CFVR, relative CFVR, and deceleration time for the whole HC group, controls, HOCM, and HCM are presented in Table 3 . Compared with the control group, all the patients with HC had a higher peak baseline diastolic flow velocity both in LAD and PD, and a lower CFVR LAD (mean [SD], 2.12 ± 0.53 vs 3.34 ± 0.67; P < .001) and CFVR PD (mean [SD], 2.29 ± 0.49 vs 3.21 ± 0.65; P < .001).
| Variables | All patients ( n = 61) | Controls ( n = 20) | P value | HCM ( n = 41) | HOCM ( n = 20) | P value |
|---|---|---|---|---|---|---|
| Baseline diastolic flow velocity (m/sec), LAD | 0.35 ± 0.08 | 0.25 ± 0.03 | <.001 | 0.33 ± 0.07 ∗ | 0.40 ± 0.09 ∗ | .002 |
| Deceleration time of baseline diastolic flow (msec), LAD | 1125 ± 375 | 860 ± 99 | <.001 | 1016 ± 297 | 1348 ± 425 ∗ | .001 |
| Hyperemic diastolic flow velocity (m/sec), LAD | 0.73 ± 0.17 | 0.83 ± 0.18 | .029 | 0.71 ± 0.16 † | 0.76 ± 0.19 | .330 |
| CFVR LAD | 2.12 ± 0.53 | 3.34 ± 0.67 | <.001 | 2.22 ± 0.55 ∗ | 1.93 ± 0.42 ∗ | .047 |
| Baseline diastolic flow velocity (m/sec), PD | 0.35 ± 0.06 | 0.26 ± 0.02 | <.001 | 0.34 ± 0.06 ∗ | 0.37 ± 0.05 ∗ | .080 |
| Deceleration time of baseline diastolic flow (msec), PD | 985 ± 213 | 865 ± 157 | .093 | 994 ± 228 | 967.7 ± 185 | .655 |
| Hyperemic diastolic flow velocity (m/sec), PD | 0.78 ± 0.16 | 0.84 ± 0.17 | .180 | 0.75 ± 0.15 † | 0.85 ± 0.16 | .029 |
| CFVR PD | 2.29 ± 0.49 | 3.21 ± 0.65 | <.001 | 2.27 ± 0.50 ∗ | 2.33 ± 0.46 ∗ | .636 |
| Relative CFVR | 0.93 ± 0.16 | 1.05 ± 0.14 | .005 | 0.98 ± 0.14 | 0.84 ± 0.16 ∗ | .001 |
∗ P < .01 for comparison of Control vs HCM/HOCM.
We found significantly higher baseline diastolic flow velocity in LAD in patients with HOCM in comparison with patients with HCM (0.40 ± 0.09 m/sec vs 0.33 ± 0.07 m/sec; P = .002), which contributes to significantly lower CFVR LAD in the HOCM group (1.93 ± 0.42 vs 2.22 ± 0.55; P = .047). However, in the HOCM group compared with the HCM group, the baseline diastolic flow velocity in PD (mean [SD], 0.37 ± 0.05 m/sec vs 0.34 ± 0.06 m/sec; P = .080) and CFVR PD (mean [SD], 2.33 ± 0.46 vs 2.27 ± 0.50; P = .636) were similar between the groups. The baseline deceleration time was significantly longer in patients with HOCM compared with patients with HCM in LAD ( P = .001) but not in PD ( P = .655). In patients with HOCM, CFVR LAD was significantly lower in comparison with CFR PD, whereas there was no significant difference between CFVR LAD and CFVR PD in the HCM group and the control group ( Figure 1 ). Therefore, relative CFVR was lower in the HOCM group compared with the HCM group (mean [SD], 0.84 ± 0.16 vs 0.98 ± 0.14; P = .001).
