Background
Similar to power-to-weight ratio and weight-to-power ratio, which are measurements of the actual performance of any engine, the ratios of peak power output to left ventricular (LV) mass (peak power/mass) and of peak LV mass to power output (peak mass/power) are indices of LV performance potentially useful in heart failure (HF). This Doppler echocardiographic study was designed to evaluate peak power/mass and peak mass/power in patients with advanced HF compared with healthy subjects and to assess their prognostic value.
Methods
Power output was measured at rest and at peak exercise in 75 subjects, 60 patients with advanced HF (LV ejection fraction ≤ 35%) and 15 controls. Peak LV power output (W) was calculated as the maximal product of (133 × 10 −6 ) × stroke volume (mL) × mean arterial pressure (mm Hg) × heart rate (beats/min). LV mass was assessed using a standard M-mode echocardiographic method.
Results
Peak power/mass was 1.84 ± 0.46 W/100 g and 0.76 ± 0.31 W/100 g, and peak mass/power was 32 ± 10 g/m 2 /W and 84 ± 38 g/m 2 /W in controls and in patients with HF, respectively (both P values < .0001). Peak power/mass was a powerful predictor of outcome on multivariate logistic regression analysis (hazard ratio, 0.907; P = .009). On receiver operating characteristic curve analysis, the areas under the curve for HF-related events were greater for peak power/mass ( P = .002) and peak mass/power ( P = .011) with respect to resting ejection fraction. Comparisons of Cox models showed that peak power/mass added prognostic value to a model that included age, New York Heart Association class, etiology, ejection fraction, and diastolic dysfunction ( P < .0001).
Conclusion
Peak power/mass is useful to discriminate and risk stratify patients with advanced HF with additional power with respect to ejection fraction.
Left ventricular (LV) power output is a descriptor of cardiac function, derived from blood pressure (BP) and cardiac output, that reflects the rate of external work (pressure-volume work) done by the left ventricle. LV pumping capability can be defined as the maximum LV power output, which is the level of power output achieved by the ventricle, and can be invasively and noninvasively assessed during maximal exercise or pharmacologic stress testing. Previous studies have shown that noninvasively derived LV power output is a powerful predictor of prognosis for patients with chronic heart failure.
Because normal myocardium grows to match the workload imposed on it, the degree of recruitable LV power output at peak exercise should reflect the number of active contractile units. Similarly to any mechanical engine, whose performance is proportional to the power generated with respect to weight, the left ventricle can be viewed as a power generator whose performance can be related to LV muscular mass. Therefore, indices that relate peak LV power output and LV mass, such as the ratios of peak power output to LV mass (peak power/mass) and of peak LV indexed mass to power output (peak mass/power), may be proposed as measures for estimating LV systolic function. These indices can be determined by Doppler echocardiography and may be potentially useful to risk stratify patients with LV systolic dysfunction, especially those with advanced heart failure, for whom the prognostic value of LV ejection fraction has been challenged. Hence, this study was designed to evaluate peak power/mass and peak mass/power during exercise stress echocardiography in patients with advanced heart failure compared with healthy subjects and to assess their prognostic value.
Methods
Patients
This study included a total of 75 subjects: 15 controls and 60 patients with advanced heart failure (LV ejection fraction ≤ 35%) enrolled at Cardiovascular Diseases Unit 2, Santa Chiara Hospital, Pisa, Italy (15 controls and 54 patients with heart failure), and the Division of Cardiology, Scuola Medica Salernitana University Hospital, Salerno, Italy (six patients with heart failure). The exclusion criteria were heart failure secondary to valvular heart disease, the presence of aortic regurgitation, peripheral artery disease limiting the capability to perform exercise stress tests, and reduced exercise tolerance attributable to myocardial ischemia. Of the 64 patients initially selected for the study, four were subsequently excluded because of an inability to perform exercise ( n = 2) or inadequate image quality during the exercise tests ( n = 2).The 60 study patients with advanced heart failure were clinically stable and under oral treatment with angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers (95%), loop diuretics (82%), antialdosterone drugs (53%), digoxin (8%), and β-blockers (78%). Beta-blockers were withheld for ≥48 hours before the test. For means of comparison, 15 age-matched and sex-matched subjects without detectable cardiovascular disease and risk factors were also studied.
Echocardiography
All patients and healthy subjects underwent ultrasound examinations at baseline and during semisupine bicycle exercise using an Acuson Sequoia C256 ultrasound instrument (Siemens Medical Solutions USA, Inc, Mountain View, CA) equipped with a 3.5-MHz transducer and harmonic imaging.
Before exercise, a complete echocardiographic and Doppler examination was performed with the subject in the left lateral supine position. The standard parasternal long-axis view as well as the three standard apical four-chamber, two-chamber, and long-axis views were acquired, optimizing gain setting, sector angle, and depth. LV end-diastolic and end-systolic volumes were calculated according to the biplane Simpson’s rule. LV mass index was determined using the M-mode method according to the recommendations of the American Society of Echocardiography. The LV outflow tract anteroposterior diameter was measured in the parasternal long-axis view, and the LV outflow tract area was calculated as π r 2 (square centimeters). The LV stroke distance (centimeters) was measured tracing the outer edge of the most dense (or brightest) portion of the spectral aortic tracing recorded from the apical five-chamber view, with the pulsed-wave Doppler sample volume positioned about 5 mm proximal to the aortic valve. Doppler tissue imaging longitudinal velocities were recorded with the sample volume placed at the junction between the septal and lateral LV wall and the mitral annulus in the four-chamber view, and peak early myocardial wave (e′) velocities were measured. The ratio of mitral E peak velocity and averaged e′ velocity (E/e′) was calculated. Patients with more than mild mitral regurgitation were selected according to the vena contracta method.
Exercise Echocardiography
Symptom-limited graded semisupine bicycle exercise was performed at an initial workload of 20 W lasting for 1 min; thereafter, the workload was increased stepwise by 10 W every minute. Twelve-lead electrocardiography and BP determination were performed at baseline and every minute thereafter.
At baseline and during each exercise level, Doppler-derived cardiac output at the LV outflow tract, heart rate, and arterial systolic BP and diastolic BP (by cuff sphygmomanometry) were measured. Mean BP was estimated as follows: diastolic BP + 1/3(systolic BP − diastolic BP). Stroke volume was calculated as stroke distance × LV outflow tract area and cardiac output as stroke volume × heart rate. Great care was taken to ensure that patients held their breath at each acquisition time and to acquire three consecutive Doppler tracings. All measures were averaged over three consecutive cycles.
LV power output was calculated as (133 x 10 −6 ) × stroke volume (mL) × mean BP (mm Hg) × heart rate (beats/min). Power/mass was calculated as LV power output per 100 g of LV mass: 100 × LV power output/LV mass (W/100 g). Conversely, mass/power was estimated by dividing LV mass index by LV power output (g/m 2 /W).
End Points of Event-Free Survival
For event-free survival analyses, observation began on the date of index exercise stress echocardiography. End points were cardiac mortality and cardiac events (cardiac mortality or hospitalization for worsening heart failure). Event-free survival data were obtained through follow-up of patients and verified through local authority registry and hospital records.
Statistical Analysis
Data are presented as mean ± SD for continuous variables and as percentages for categorical variables. Continuous variables were compared using paired-samples and independent-samples Student’s t test. Categorical variables were compared by means of the χ 2 test. Time-independent associations between variables and outcomes were assessed using receiver operating characteristic (ROC) curve analysis and logistic regression. Statistical comparison of ROC curves was performed using the method of paired ROC curves of Hanley and McNeil. Cox proportional-hazard regression analysis was used to identify predictors of outcome in time-dependent analysis. Kaplan-Meier curves were constructed and log-rank tests were used to test for differences between event-free survival curves. The likelihood ratio test was used to compare different multivariate Cox models by taking into account the different number of variables used in each model and to explore the incremental prognostic value of power/mass ratio over established clinical and echocardiographic prognostic variables. The log likelihood and the overall model χ 2 (i.e., that calculated by testing the hypothesis that all regression coefficients for the variables in the model are identically zero) were calculated for each model. The probability level was p < .05 for all data examined. Data management and analyses were performed using SPSS version 15.0 for Windows (SPSS, Inc, Chicago, IL).
Results
The study population comprised patients with heart failure who were in New York Heart Association (NYHA) class I ( n = 8), NYHA class II ( n = 32), and NYHA class III ( n = 20). Etiology was ischemic in 38% of patients. More than mild mitral regurgitation, as defined by a vena contracta width ≥ 0.5 cm, was present in 17% of the study patients. They were followed up for a mean period of 19 ± 10 months. Seven patients (12%) died from cardiac causes, and 13 patients (22%) were hospitalized for worsening heart failure. Patients’ characteristics are summarized in Table 1 .
| Variable | Patients with heart failure | Controls | P |
|---|---|---|---|
| Age (yrs) | 61 ± 9 | 55 ± 14 | .058 |
| Men | 77% | 80% | .78 |
| LV mass (g) | 297 ± 74 | 190 ± 57 | <.0001 |
| Body surface area (m 2 ) | 1.9 ± 0.2 | 1.8 ± 0.1 | .23 |
| Indexed LV mass (g/m 2 ) | 156 ± 29 | 103 ± 28 | <.0001 |
| Left atrial volume (mL) | 75 ± 47 | 47 ± 5 | <.0001 |
| Left atrial volume index (mL/m 2 ) | 39 ± 8 | 26 ± 7 | <.0001 |
| E/e′ ratio | 9 ± 4 | 5 ± 1 | .0024 |
| Resting variables | |||
| Heart rate (beats/min) | 78 ± 13 | 65 ± 12 | .0006 |
| Systolic arterial pressure (mm Hg) | 119 ± 17 | 127 ± 12 | .072 |
| Diastolic arterial pressure (mm Hg) | 78 ± 12 | 83 ± 8 | .12 |
| End-diastolic volume index (mL/m 2 ) | 115 ± 29 | 74 ± 17 | <.0001 |
| End-systolic volume index (mL/m 2 ) | 84 ± 26 | 29 ± 7 | <.0001 |
| Stroke volume (mL) | 55 ± 15 | 65 ± 21 | .041 |
| Stroke volume index (mL/m 2 ) | 29 ± 7 | 35 ± 11 | .0081 |
| Ejection fraction (%) | 28 ± 6 | 61 ± 3 | <.0001 |
| Cardiac output (L/min) | 4.2 ± 1.2 | 4.3 ± 1.3 | .71 |
| Cardiac index (L/min/m 2 ) | 2.3 ± 0.6 | 2.3 ± 0.6 | .97 |
| Power output (W) | 0.87 ± 0.27 | 0.89 ± 0.33 | .77 |
| Indexed mass/power output (g/m 2 /W) | 196 ± 75 | 130 ± 51 | .002 |
| Power output/mass (W/100 g) | 0.31 ± 0.11 | 0.48 ± 0.16 | <.0001 |
| Exercise variables | |||
| Heart rate (beats/min) | 126 ± 18 | 129 ± 14 | .56 |
| Systolic arterial pressure (mm Hg) | 159 ± 29 | 199 ± 30 | <.0001 |
| Diastolic arterial pressure (mm Hg) | 89 ± 13 | 91 ± 16 | .64 |
| End-diastolic volume index (mL/m 2 ) | 109 ± 28 | 70 ± 16 | <.0001 |
| End-systolic volume index (mL/m 2 ) | 75 ± 26 | 20 ± 7 | <.0001 |
| Stroke volume (mL) | 69 ± 24 | 92 ± 24 | .0017 |
| Stroke volume index (mL/m 2 ) | 37 ± 11 | 51 ± 13 | <.0001 |
| Ejection fraction (%) | 32 ± 8 | 72 ± 4 | <.0001 |
| Δ ejection fraction (%) | 16 ± 14 | 19 ± 5 | .65 |
| Cardiac output (L/min) | 8.7 ± 3.1 | 11.9 ± 3.3 | .0007 |
| Cardiac index (L/min/m 2 ) | 4.6 ± 1.4 | 6.5 ± 1.7 | <.0001 |
| Power output (W) | 2.2 ± 0.9 | 3.4 ± 1.1 | <.006 |
| Indexed mass/power output (g/m 2 /W) | 84 ± 38 | 32 ± 10 | <.0001 |
| Power output/mass (W/100 g) | 0.76 ± 0.31 | 1.84 ± 0.46 | <.0001 |
Mean exercise duration was 7 ± 3 min in patients with advanced heart failure and 14 ± 7 min in healthy controls. From baseline to peak exercise, heart rate significantly increased in both the advanced heart failure and the control groups (both P values < .0001). In patients with advanced heart failure, mean BP increased from 92 ± 12 to 112 ± 17 mm Hg; in controls, mean BP was 98 ± 8 mm Hg at baseline and 127 ± 15 mm Hg at peak exercise (both P values < .0001). No patients had chest pain, ischemic electrocardiographic changes, wall motion abnormalities or significant arrhythmias during the exercise tests.
Resting power/mass in patients with heart failure was between 0.12 and 0.59 W/100 g, and it was between 0.22 and 0.71 W/100 g in controls. Peak power/mass was between 0.29 and 1.87 W/100 g in patients with heart failure, and it was between 1.12 and 2.68 W/100 g in controls. Resting mass/power ranged from 82 to 483 g/m 2 /W in patients with heart failure and from 76 to 240 g/m 2 /W in the control group. In patients with advanced heart failure, peak mass/power was between 29 and 198 g/m 2 /W, while it was between 21 and 60 g/m 2 /W in healthy subjects. The comparisons of peak power/mass and peak mass/power in healthy subjects and in patients with heart failure are presented in Figure 1 .
Time-Independent Analyses of Event-Free Survival
The results of ROC analysis for outcome prediction are shown in Table 2 . Among all variables, peak power/mass showed the largest area under the curve. A value of ≤0.58 W/100 g was identified as the best cutoff value, showing 80% sensitivity and 90% specificity for the prediction of events. The best cutoff value of peak mass/power was >91 g/m 2 /W (sensitivity, 85%; specificity, 90%). Statistically significant differences between the areas under the curve were apparent between peak power/mass and peak mass/power with respect to LV ejection fraction ( p = .002 and p = .011, respectively). Figure 2 displays the comparison of ROC curves of peak power/mass and LV ejection fraction. Results of univariate logistic regression are shown in Table 3 . Although a number of variables were associated with the risk for events, stepwise multivariate analysis showed that peak power/mass (hazard ratio [HR], 0.907; 95% confidence interval [CI], 0.843–0.976; P = .009), cardiac index at peak exercise (HR, 1.013; 95% CI, 1.001–1.024; P = .036), and systolic BP at peak exercise (HR, 0.947; 95% CI, 0.896-1.001; P = .053) were independent predictors (overall model P < .0001).
| Variable | Area under the curve | 95% CI | P |
|---|---|---|---|
| Age | 0.56 | 0.40–0.72 | .42 |
| NYHA class | 0.68 | 0.54–0.83 | .02 |
| Indexed LV mass | 0.66 | 0.51–0.82 | .04 |
| E/e′ ratio | 0.65 | 0.50–0.80 | .061 |
| Resting | |||
| Heart rate | 0.50 | 0.33–0.67 | .97 |
| Systolic arterial pressure | 0.83 | 0.71–0.94 | <.0001 |
| End-diastolic volume index | 0.56 | 0.40–0.72 | .46 |
| End-systolic volume index | 0.59 | 0.44–0.74 | .27 |
| Ejection fraction | 0.63 | 0.48–0.79 | .096 |
| Cardiac index | 0.62 | 0.46–0.78 | .14 |
| Exercise | |||
| Heart rate | 0.62 | 0.46–0.77 | .15 |
| Systolic arterial pressure | 0.83 | 0.73–0.94 | <.0001 |
| End-systolic volume index | 0.68 | 0.53–0.82 | .028 |
| Ejection fraction | 0.72 | 0.58–0.86 | .007 |
| Δ ejection fraction | 0.78 | 0.64–0.91 | .001 |
| Cardiac index | 0.73 | 0.59–0.88 | .004 |
| Power output | 0.81 | 0.67–0.95 | <.0001 |
| Indexed mass/power output | 0.85 | 0.73–0.98 | <.0001 |
| Power output/mass | 0.87 | 0.77–0.98 | <.0001 |
| Variable | HR | 95% CI | P |
|---|---|---|---|
| Age | 1.003 | 0.945–1.065 | .912 |
| Male gender | 3.855 | 0.771–19.277 | .10 |
| Ischemic etiology | 4.500 | 1.431–14.150 | .01 |
| NYHA class ∗ | |||
| II | 2.333 | 0.248–21.981 | .46 |
| III | 8.556 | 0.881–83.057 | .07 |
| Indexed LV mass | 1.022 | 1.000–1.044 | .047 |
| E/e′ ratio | 1.161 | 1.002–1.346 | .047 |
| Resting | |||
| Heart rate | 1.003 | 0.961–1.046 | .90 |
| Systolic arterial pressure | 0.897 | 0.843–0.954 | .001 |
| End-diastolic volume index | 1.012 | 0.993–1.031 | .22 |
| End-systolic volume index | 1.016 | 0.994–1.036 | .15 |
| Ejection fraction | 0.924 | 0.840–1.015 | .099 |
| Cardiac index | 0.995 | 0.985–1.004 | .26 |
| Exercise | |||
| Systolic arterial pressure | 0.938 | 0.904–0.972 | .001 |
| End-systolic volume index | 1.022 | 1.000–1.044 | .048 |
| Ejection fraction | 0.804 | 0.706–0.915 | .007 |
| Δ ejection fraction | 0.911 | 0.855–0.971 | .004 |
| Cardiac index | 0.993 | 0.988–0.998 | .01 |
| Power output | 0.985 | 0.975–0.995 | .003 |
| Indexed mass/power output | 1.059 | 1.026–1.083 | <.0001 |
| Power output/mass | 0.930 | 0.896–0.965 | <.0001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
