Author + information
- Received September 14, 2016
- Revision received January 15, 2017
- Accepted March 2, 2017
- Published online July 19, 2017.
- Kenya Kusunose, MD, PhDa,∗∗ (, )
- Hirotsugu Yamada, MD, PhDa,∗ (, )
- Susumu Nishio, RMSb,
- Ayumi Ishii, RNc,
- Yukina Hirata, RMS, PhDb,
- Hiromitsu Seno, MDa,
- Yoshihito Saijo, MDa,
- Takayuki Ise, MD, PhDa,
- Koji Yamaguchi, MD, PhDa,
- Shusuke Yagi, MD, PhDa,
- Takeshi Soeki, MD, PhDa,
- Tetsuzo Wakatsuki, MD, PhDa and
- Masataka Sata, MD, PhDa
- aDepartment of Cardiovascular Medicine, Tokushima University Hospital, Tokushima, Japan
- bUltrasound Examination Center, Tokushima University Hospital, Tokushima, Japan
- cDepartment of Nursing, Tokushima University Hospital, Tokushima, Japan
Objectives The aim of this study was to assess the relationship between right ventricular (RV) function during pre-load augmentation and exercise tolerance.
Background Peak oxygen uptake (VO2) is a strong predictor of mortality in chronic heart failure. Cardiac function during pre-load augmentation is an important part of the phenomenon in the evaluation of exercise capacity.
Methods We prospectively performed echocardiographic studies in 68 chronic heart failure patients with cardiopulmonary exercise testing (mean age 60 ± 12 years; 69% male). After resting evaluations, echocardiographic parameters were repeated during leg positive pressure (LPP). Exercise capacity was assessed by peak VO2 in all patients (left ventricular ejection fraction: 43 ± 15%).
Results Patients with severely reduced exercise capacity (peak VO2 <14 ml/kg/min) had significantly lower stroke volume index, left ventricular global longitudinal strain and RV strain and higher filling pressure (E/e’ and pulmonary arterial systolic pressure) than the remainder. Stroke volume index (β = 0.49), global longitudinal strain (β = −0.61), E/e’ (β = −0.32), pulmonary arterial systolic pressure (β = −0.57), and RV strain (β = −0.66) during LPP were independently correlated to peak VO2 (all p < 0.01). RV strain during LPP was the most powerful predictor in identifying patients with severely reduced exercise capacity (cut off value: −17%; sensitivity: 81%; specificity: 88%; areas under the curve: 0.88, p < 0.001) compared with other variables including resting parameters.
Conclusions RV strain during pre-load augmentation correlated independently to peak VO2 and was a powerful predictor in identifying patients with severely reduced exercise capacity.
Peak oxygen uptake (VO2) using cardiopulmonary exercise testing (CPX) is a strong predictor of mortality in chronic heart failure (CHF) patients with a wide range of left ventricular (LV) function (1). There were many studies on the prognostic value of exercise capacity, and the VO2 is used in heart failure guidelines (2–4). Previous studies have found a relationship between exercise capacity and cardiac function including LV systolic function, diastolic function, and right ventricular (RV) function (5–8). However, echocardiographic indices, especially in left ventricular ejection fraction (LVEF), were not always associated with exercise capacity, and it remains unclear which cardiac function is the most contributing factor on exercise capacity in CHF patients (7). A better understanding of the association between cardiac function and exercise capacity can be clinically relevant to the management of CHF patients.
Cardiac function during pre-load augmentation is an important part of the phenomenon in the evaluation of exercise capacity. A previous study showed that patients with E/e′ >15 during leg lifting had low peak VO2 compared to those with persistent E/e′ <15 (9). This result suggested that patients with low exercise capacity had easily increased LV filling pressures without an effective increase in stroke volume during pre-load augmentation (10). Therefore, the LV function during pre-load augmentation reflects exercise capacity. On the other hand, a pre-load–dependent augmentation of stroke volume was consistent with an increase in both LV and RV contractility in normal individuals (11). Furthermore, RV function has been well established as functional and prognostic parameters in several cardiac diseases (12–14). Evaluation of cardiac function during pre-load augmentation in the absence of knowledge of RV function may be problematic.
Recently, our laboratory has developed pre-load stress echocardiography using leg positive pressure (LPP) to assess cardiac function during pre-load augmentation (15,16). RV function during pre-load augmentation may provide additional information beyond that measured by conventional Doppler echocardiographic parameters at rest in CHF patients. We hypothesized that exercise capacity is more closely related to the RV function during pre-load augmentation than to variables at rest in patients with CHF.
We designed a single-center, cross-sectional, prospective study to assess the exercise capacity and cardiac function in patients with CHF between April 2015 and March 2016. All patients fulfilled the following inclusion criteria: 1) sinus rhythm; 2) stable clinical condition at the time of echocardiography with optimal medical treatment; 3) absence of chronic lung disease; 4) absence of unstable angina; 5) absence of severe valvular disease; 6) absence of anemia; and 7) technically adequate 2-dimensional (2D) and Doppler echocardiograms. All patients were clinically examined and assessed by New York Heart Association (NYHA) functional classification. The Institutional Review Board of the Tokushima University Hospital approved the study protocol, and written informed consent was obtained from all subjects.
Echocardiography was performed on the same day as CPX using a commercially available ultrasound machine (iE33, Philips Healthcare, Amsterdam, the Netherlands; Vivid E9, GE Healthcare, Waukesha, Wisconsin; and SSA-770A, Toshiba Medical, Otawara, Japan). Measurements and recordings were obtained according to American Society of Echocardiography recommendations (17). Left ventricular end-diastolic volume (LVEDV), left ventricular end-systolic volume (LVESV), left atrial volume, and LVEF were calculated by the biplane method of disks using 2D images and indexed to body surface area. Stroke volume index (SVi) was calculated as the product of the cross-sectional area of the LV outflow tract and the time-velocity integral in the LV outflow tract and indexed to body surface area. The early diastolic (e′) mitral annular tissue velocity was also measured in the apical 4-chamber view with the sample volume positioned at the lateral mitral annulus. Pulmonary artery systolic pressure (PASP) was measured from the maximal continuous-wave Doppler velocity of the tricuspid regurgitant jet using the systolic transtricuspid pressure gradient calculated by the modified Bernoulli equation. Right atrial pressure was estimated from the inferior vena cava diameter and collapsibility (18).
2D strain echocardiography
Peak systolic longitudinal strain (LS) measurements were obtained from gray-scale images recorded in the apical 4-chamber, 2-chamber, and long-axis views. The frame rate was maintained at a level >50 frame/s. LV strain was analyzed offline using speckle tracking software (EchoInsight, Epsilon Imaging, Ann Arbor, Michigan). Global longitudinal strain (GLS) was obtained by averaging all segmental strain values from the apical 4-chamber, 2-chamber, and long-axis views. Peak strain for the 3 RV free wall segments was averaged to produce global RV LS, with exclusion of the interventricular septum to avoid LV interaction (Online Video 1). These offline analyses were independently performed in a blinded manner by 2 observers who were not involved in the image acquisition and had no knowledge of examination dates and other echocardiographic or clinical data. The reproducibility of strains, expressed as the coefficient of variation, has been well described by our group as 5% to 7% and 7% to 9%, respectively, for intraobserver and interobserver variations (19,20).
We have previously shown that the LPP maneuver is useful for pre-load stress echocardiography because it allows noninvasive pre-load augmentation during an echocardiographic examination. Briefly, we customized a commercially available leg massage machine (Dr. Medomer DM-5000EX, Medo Industries Co., Ltd., Tokyo, Japan) and used a setting of 90 mm Hg because this pressure did not significantly increase either heart rate or systolic blood pressure, based on findings from our studies (15,16). All echocardiographic variables were obtained at rest and during LPP (Figure 1). All patients tolerated 90 mm Hg LPP during the echocardiographic examination without any complications.
Cardiopulmonary exercise testing
All patients performed exercise testing with ventilator expired gas analysis for clinical indications, mainly for the determinants of exercise training program. Exercise testing was performed on an upright bicycle ergometer (STB-3200, CATEYE Co. Ltd. Osaka, Japan). The test started with 2 min of rest and 2 min of warm-up at 10 Watts followed by a 10-Watt ramp. The test ended when symptoms of exhaustion were exhibited. VO2, carbon dioxide production, and ventilation were measured and calculated by the gas analysis system (CPEX-1, Inter Reha Co. Ltd., Tokyo, Japan). We defined peak VO2 as the highest VO2 obtained during and adequately performed test (2). At peak exercise, we recorded heart rate, blood pressure, Watts and metabolic equivalents. We used 14 ml/kg/min as an optimal value of VO2 for severely reduced exercise capacity in patients with heart failure (HF) because several investigators have reported that patients who achieve a peak VO2 of 14 ml/kg/min appear to have a poor prognosis in CHF (21). We also calculated % peak VO2 as the difference between actual and predicted VO2 divided by predicted VO2.
Data are presented as mean ± SD if the Kolmogorov-Smirnov test showed a normal distribution. Otherwise, the median and interquartile ranges were used. Comparisons between the 2 groups were performed using a Student t test (or Wilcoxon signed rank tests for non-normal data). We checked for collinearity between the independent variables and exercise capacity by using Pearson’s correlation coefficients. Linear regression analysis was used to evaluate the associations between several variables and peak VO2. We found no evidence for collinearity problems in our model (variance inflation factor values <2). Age, sex, body mass index (BMI), and NYHA functional classes III/IV were forced into the multivariate models because these variables were well known as important markers of exercise capacity. Receiver-operating characteristic (ROC) curve analysis was used to identify parameters that were best to diagnose severely reduced exercise capacity. The best cutoff value was defined as the upper limit of the confidence interval (CI) of the Youden index. To evaluate the impact of RV function during pre-load augmentation for predicting reduced exercise capacity along with clinical factors (age, sex, BMI, and NYHA functional classes III/IV), 2 models were constructed. Model 1, the basic model, consisted of clinical factors. Model 2 included variables in Model 1 plus RV strain during pre-load augmentation. The DeLong method was used to compare the C-statistic (22). Statistical analysis was performed using standard statistical software packages (SPSS software 21.0, SPSS Inc., Chicago, Illinois; and MedCalc Software 17, Mariakerke, Belgium). Statistical significance was defined by p < 0.05.
Baseline characteristics of the study group are presented in Table 1. The study population consisted of 68 patients (mean age, 60 ± 12 years; 69% male) with CHF. The patient population consisted of 75% patients with ischemic cardiomyopathy, and 25% with nonischemic cardiomyopathy. All patients with ischemic cardiomyopathy were completely revascularized. Mean peak VO2 was 15 ± 6 ml/kg/min in the total population. Thirty-six patients (53%) had severely reduced exercise capacity defined by peak VO2 <14 ml/kg/min and there was no difference of etiology between the 2 groups. The averaged exercise time was 7 ± 2 min.
Echocardiographic parameters at baseline and during pre-load augmentation
Echocardiographic data at baseline and during LPP are shown in Table 1. In this cohort, LVEF (43 ± 15%) and LVGLS (−13 ± 6%) were reduced. Measures of RV function were also below normal reference values. At baseline, patients in the severely reduced exercise capacity group had significantly lower SVi, lower GLS, lower RV strain, higher E/e′, and higher peak PASP profiles than patients in the preserved exercise capacity group. During LPP, patients in the severely reduced exercise capacity group also had more significantly lower LVEF (40 ± 17% vs. 49 ± 16%; p = 0.04), lower SVi (24 ± 9 ml/m2 vs. 34 ± 8 ml/m2; p < 0.001), lower GLS (−12 ± 4% vs. −16 ± 5%; p = 0.002), lower RV strain (−15 ± 4% vs. −20 ± 4%; p < 0.001), higher E/e′ (13.3 ± 6.2 vs. 9.5 ± 5.5; p = 0.009), and higher peak PASP (41 ± 10 mm Hg vs. 33 ± 5 mm Hg; p < 0.001) profiles than patients in the preserved exercise capacity group. In patients with preserved exercise capacity, SVi increased from 31 ± 7 ml/m2 to 34 ± 8 ml/m2 during LPP (p < 0.001). In contrast, SVi remained unchanged from 23 ± 8 ml/m2 to 24 ± 9 ml/m2 (p = 0.16) in patients with severely reduced exercise capacity (Figure 2).
Correlates of exercise capacity
Parameters of myocardial systolic and diastolic function correlated to peak VO2 (Table 2). There is a weak correlation between RV strain and exercise time (r = 0.37; p = 0.001). After adjustment for clinical variables (age, sex, BMI, and NYHA functional classes III/IV), left-sided heart function (SVi: standardized β = 0.45, p < 0.001; GLS: standardized β = −0.52, p < 0.001; and E/e′: standardized β = −0.32, p = 0.005) and right-sided heart function (RV strain: standardized β = −0.59, p < 0.001; and PASP: standardized β = −0.45, p < 0.001) were associated with exercise capacity at baseline. During LPP, the relationships between echocardiographic variables and peak VO2 were more significant (SVi: standardized β = 0.49, p < 0.001; GLS: standardized β = −0.61, p < 0.001; E/e′: standardized β = −0.32, p = 0.006; RV strain: standardized β = −0.66, p < 0.001; and PASP: standardized β = −0.57, p < 0.001) (Figures 3A to 3D). The change of RV strain by LPP was associated with peak VO2 (r = 0.52; p < 0.001).
Results of the ROC curve analysis used to identify the optimal cutoff point for predicting the severely reduced exercise capacity are shown in Figure 4. ROC analyses revealed that RV strain during LPP had significantly better ability to detect reduced peak VO2 <14 ml/kg/min compared with GLS, PASP, and E/e′. This RV strain during LPP had the highest area under the curve (AUC) (0.88; p < 0.001) among echocardiographic variables. An RV strain value of −17% during LPP had good sensitivity of 0.81 (95% CI: 0.64 to 0.92) and specificity of 0.88 (95% CI: 0.71 to 0.97) to identify patients with a peak VO2 of <14 ml/kg/min. In addition, the RV strain during LPP had incremental diagnostic value over other clinical factors (C-statistic: 0.89 vs. 0.77; p = 0.003). The RV strain during LPP was also correlated with predicted peak VO2 (r = 0.69; p < 0.001). Therefore, the RV strain during pre-load augmentation was an important marker of exercise capacity in patients with CHF.
This is the first study to clearly show the relationship between RV strain during pre-load augmentation and exercise capacity in CHF. Our study brings several new insights into the understanding of the relationship between exercise capacity and cardiac functions: 1) both LV and RV longitudinal myocardial function during pre-load augmentation were correlated to exercise capacity; 2) in patients with severely reduced exercise capacity, SVi remained unchanged during pre-load augmentation; and 3) the RV strain during pre-load augmentation was shown to be the strongest echocardiographic correlate of exercise capacity even after adjusting clinical variables. Therefore, the RV assessment during pre-load augmentation may hold promise to screen exercise intolerance in CHF.
The results of this study are consistent with the previous work linking GLS with exercise capacity in CHF (7). Although LV strain parameter is correlated to LVEF, it may be more sensitive in detecting sub-clinical myocardial dysfunction (23). On the other hand, when we have added both LV and RV strains in the same statistical model, the correlation for RV strain is better than for LV strain (RV strain: standardized β = −0.62, p < 0.001; LV strain: standardized β = −0.31, p < 0.001). A likely reason is that we included the majority of patients with impaired LVEF (averaged LVEF was 43 ± 15%) in this study and RV strain may be more important especially in patients with LV systolic dysfunction. In addition, other investigators showed that the pressure markers (E/e′ and PASP) were also correlated with exercise capacity (24,25). Patients with reduced exercise capacity have elevated LA pressure and PASP in our cohort. More interestingly, RV function is a sensitive marker to assess exercise capacity. In the present study, several parameters of RV function (e.g., RV fractional area change and RV strain) showed an association with exercise capacity. The RV strain had the highest AUC among echocardiographic variables. Recently, RV function provided important prognostic information and had a central role of cardiac function in heart failure and pulmonary hypertension (12–14). Thus, the results suggest that RV dysfunction defined by RV strain may further help to assess the exercise capacity in CHF. In our subjects, RV strain during pre-load augmentation was an independent predictor of reduced exercise capacity. Recent recommendations for CPX define a reference value for peak VO2 as 14 ml/kg/min based on previous studies (21,26). RV strain during pre-load augmentation was the best parameter to identify patients with peak VO2 <14 ml/kg/min. Therefore, the RV strain might be useful to discriminate patients with preserved exercise capacity from severely reduced exercise capacity.
Effects of pre-load augmentation
The cause of RV dysfunction during pre-load augmentation in reduced exercise capacity is not fully understood given the complex interaction between left and right sides of the heart. In normal subjects, responses to pre-load augmentation is an increase in SV according to Frank-Starling’s law (27). In subjects with severe cardiac dysfunction, LV filling pressure easily increased without an effective increase in stroke volume during pre-load augmentation. Our results showed that in patients with reduced exercise capacity, E/e′ easily increased and stroke volume index did not significantly change during pre-load augmentation. On the other hand, patients with preserved exercise capacity had an effective increase in SVi during pre-load augmentation. In this phenomenon, elevated LV filling pressure during pre-load augmentation leads to increased pulmonary venous and arterial pressure, in turn increasing RV afterload, which is one source of RV dysfunction. In addition, pre-load augmentation itself also influences the changes of RV systolic function. If there is a sub-clinical RV failure in patients, the RV systolic function could not appropriately increase during pre-load augmentation according to RV Frank-Starling’s law. Therefore, exercise capacity can be reduced in patients with impaired RV strain during pre-load augmentation due to the increased afterload and sub-clinical myocardial dysfunction. This finding is well matched with the concept that cardiac function during pre-load augmentation is an important factor of exercise capacity in CHF. In the assessment of exercise capacity, speckle tracking imaging can be used for detailed RV analysis during pre-load augmentation.
To the best of our knowledge, this is the first report of RV function during pre-load augmentation and may help differentiate patients with reduced exercise capacity. Current guidelines for CPX define the assessment of peak VO2 as a primary prognostic marker (3). However, CPX is underused in the clinical setting due to the expense of equipment, time, and capability. Echocardiography is a routine examination in the evaluation of all CHF patients. RV assessment during pre-load augmentation could work as a first-line examination and could avoid unnecessary CPX studies. It might also be a useful tool in differentiating patients with poor prognosis.
Our study was a cross-sectional, single-center study, and we did not relate our findings to clinical outcomes. The sample size was relatively small, and the study population was heterogeneous. Another limitation is the lack of a validation cohort. The present study should be considered hypothesis generating, and we believe that larger multicenter studies are warranted. We used absolute peak VO2 values in our study because the guidelines recommend use of the absolute value to assess HF.
RV strain during pre-load increment correlated independently to peak VO2 and was a powerful predictor in identifying patients with severely reduced exercise capacity. Pre-load stress echocardiography can be helpful in identifying a sub-group with reduced exercise capacity. Further studies are needed to assess whether these parameters can be used to assess the prognostic value in CHF in long-term follow-up.
COMPETENCY IN MEDICAL KNOWLEDGE: The right ventricular strain obtained by 2D speckle tracking echocardiography during pre-load augmentation was a predictor of severely reduced exercise capacity. This information might be useful for clinical evaluation and follow-up in CHF.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Strain values should be carefully assessed in CHF. Our results suggest that the RV function during pre-load augmentation is a sensitive marker of exercise capacity in patients with CHF.
TRANSLATIONAL OUTLOOK: Although this study suggests an association between strain and exercise capacity, this hypothesis should be tested on a large cohort of patients with CHF.
The authors acknowledge Kathryn Brock, BA, for her work editing the manuscript.
For a supplemental video, please see the online version of this paper.
Dr. Kusunose has received a grant from JSPS Kakenhi (No. 15K19381), and the Japan Heart Foundation Research Grant. Dr. Sata has received grants from JSPS Kakenhi (Nos. 16H05299 and 26248050), and from MEXT KAKENHI (No. 21117007). All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- chronic heart failure
- cardiopulmonary exercise testing
- ejection fraction
- leg positive pressure
- longitudinal strain
- left ventricular
- pulmonary artery systolic pressure
- right ventricular
- stroke volume index
- oxygen uptake
- Received September 14, 2016.
- Revision received January 15, 2017.
- Accepted March 2, 2017.
- 2017 American College of Cardiology Foundation
- Balady G.J.,
- Arena R.,
- Sietsema K.,
- et al.
- Guazzi M.,
- Adams V.,
- Conraads V.,
- et al.
- Guazzi M.,
- Myers J.,
- Arena R.
- Hasselberg N.E.,
- Haugaa K.H.,
- Sarvari S.I.,
- et al.
- Kusunose K.,
- Motoki H.,
- Popovic Z.B.,
- Thomas J.D.,
- Klein A.L.,
- Marwick T.H.
- Nonogi H.,
- Hess O.M.,
- Ritter M.,
- Krayenbuehl H.P.
- Park J.H.,
- Kusunose K.,
- Kwon D.H.,
- et al.
- Khemasuwan D.,
- Yingchoncharoen T.,
- Tunsupon P.,
- et al.
- Kusunose K.,
- Yamada H.,
- Nishio S.,
- et al.
- Yamada H.,
- Kusunose K.,
- Nishio S.,
- et al.
- Lang R.M.,
- Badano L.P.,
- Mor-Avi V.,
- et al.
- Rudski L.G.,
- Lai W.W.,
- Afilalo J.,
- et al.
- Kusunose K.,
- Tsutsui R.S.,
- Bhatt K.,
- et al.
- Goodman A.,
- Kusunose K.,
- Popovic Z.B.,
- et al.
- Kusunose K.,
- Agarwal S.,
- Marwick T.H.,
- Griffin B.P.,
- Popovic Z.B.
- Naeije R.,
- Huez S.,
- Lamotte M.,
- et al.
- Pina I.L.