Is ventilatory efficiency (VE/VCO₂ slope) associated with right ventricular oxidative metabolism in patients with congestive heart failure?

2.1. Subjects

The study population consisted of 30 consecutive consenting patients with stable NYHA class II-III CHF who had a reduced LVEF by radionuclide angiography, an ability to exercise and no other significant co-morbidities that could affect exercise tolerance or outcome. Heart failure medication was stable for at least 1 month before the study. Patients were excluded if they were unable to achieve adequate exercise (respiratory exchange ratio, RER <1.0). Of the 30 patients screened, 22 achieved an RER >1. One additional patient was excluded because of LV thrombus. Thus, the study cohort consisted of 21 patients.

2.2. Study design

The relationship between myocardial energetics and exercise capacity in CHF patients was studied using [¹¹C]acetate PET imaging to measure oxidative metabolism; echocardiography to measure LV volumes and stroke work, and a 6-minute walk and cardiopulmonary exercise test to measure exercise capacity. On the study day, cardiac performance was measured by echocardiography and followed immediately by a [¹¹C]acetate PET study. The 6-minute walk test and cardiopulmonary exercise testing were performed within 11±13 days of the combined echo/PET studies. There was no change in clinical status or medication between testing. The protocol was approved by the Ethics Review Board of University of Ottawa Heart Institute. Informed consent was obtained in all subjects.

2.3. Positron emission tomography

Patients were positioned in a whole body ECAT ART PET scanner (Siemens/CTI, Knoxville, TN). Following the 5-minute transmission scan, 370 MBq of [¹¹C]acetate was administered intravenously and a dynamic PET acquisition was initiated (10×10 s; 2×30 s; 5×100 s; 3×180 s, 2×300 s).

2.4. PET data analysis

The reconstructed dynamic PET images were analyzed by applying five regions of interest (ROI) covering the septum, anterior wall, lateral wall, whole LV myocardium and free wall of the RV, respectively, in three to five midventricular transaxial planes 11. A monoexponential function was fit to the myocardial time-activity data and monoexponential clearance rates (k_mono) were determined for each ROI and averaged across planes (Fig. 1) 9,11. Due to known individual variation (physiological and pathophysiological) in both RV and LV myocardial oxidative metabolism, normalized RV oxidative metabolism was used for analysis, Normalized or corrected RV oxidative metabolism (RVOx) was calculated as RV k_mono/LV k_mono.

2.5. Echocardiography

Echocardiography examinations were performed using a Phillips HP 2500 or 5500 cardiac ultrasound system (Andover, Massachusetts). Stroke volume (SV) was measured using Doppler echocardiography 12 and stroke work index (LVSWI) calculated as 8,9,11:

(1)

where SVI is SV indexed to body surface area and SBP is systolic blood pressure. LV volumes were calculated using the method of discs 13. Mitral regurgitation (MR) was quantified by measuring the largest regurgitant jet area on Doppler colour flow imaging 14.

2.6. Work metabolic index determination

Myocardial efficiency was derived using the work metabolic index (WMI) 8,9,11:

(2)

where HR is heart rate. All PET and echocardiographic data were analyzed blinded to the patient's clinical and exercise data.

2.7. Cardiopulmonary exercise testing

All study patients underwent a symptom limited, treadmill test to volitional fatigue with respiratory gas exchange analysis. Metabolic measurements were made using a Medgraphics CPX-D system operating Breeze Suite 5.3 software (Minneapolis, Minn, USA). Patients continued medical therapy on the day of the exercise test. VO₂, carbon dioxide release (VCO₂), tidal volume, RER and respiratory rate, were measured. A slow ramp (n=18) or a regular ramp (n=3) protocol were performed using 1-minute stages, constant speed and an increased grade every minute. Protocol selection was designed to obtain a peak exercise capacity within 8 to 12 min. HR and rhythm were monitored by 12-lead electrocardiogram (GE Case 8000 stress testing system GE, WI, USA.) Blood pressure was determined manually using a mercury Baumanometer (W.A. Baum Co., Inc.; Copiague, NY 11726). BP and rating of perceived exertion (RPE) were obtained at the end of each stage. Gas exchange data was collected in a breath-by-breath manner and averaged into 30-second time periods. Peak VO₂ was calculated as the highest 30-second time period before volitional fatigue was reached. Anaerobic threshold (VO₂ AT) was measured by the V-slope method 15. The VE/VCO₂ slope was calculated by linear regression, excluding the nonlinear part of the data after the onset of ventilatory compensation for metabolic acidosis. A VE/VCO₂ slope of 34 was defined as a threshold value based on the published data on prognosis in patients with CHF 3,16,17.

2.8. Six-minute walk test

The 6-minute walk test was performed on a level hallway surface 28 m long according to the method of Guyatt et al. 18 and administered by a nurse blinded to the clinical and research data. The total distance walked in meters during the 6-minute walk test was recorded. HR, BP and symptoms were recorded before, and at 6 min 17. Cardiopulmonary exercise tests and 6-minute walk data were analyzed blinded to clinical and imaging data.

2.9. Statistical analysis

Data are expressed as mean±SD. Correlation coefficients were calculated using Pearson's correlation coefficient. Comparisons were performed using paired t-testing and unpaired t-testing where appropriate. A p value≤0.05 was considered significant. Statistical calculations were carried out using SPSS 10.0 software.

3.1. Subjects

The characteristics of the 21 patients are shown in Table 1.

3.2. Exercise parameters

Exercise data in the 21 patients are summarized in Table 2. In three patients, ST-segment depression was observed. Peak VO₂was reduced at 16.2±4.2 mL/kg/min and VE/VCO₂ slope was elevated (33.4±6.1) compared to previously published normal values 19.

3.3. Positron emission tomography

Representative time-activity curves are depicted in Fig. 1. k_mono was 0.046±0.009 in the LV and 0.037±0.007 for the RV (Table 3). RVOx was 0.83±0.17. WMI was 6.15±1.67×10⁶ mm Hg mL/m².

3.4. Relationship of cardiac work, oxidative metabolism and exercise (Table 4)

LVSWI did not correlate with peak VO₂ (r=−0.23, p=0.31), VE/VCO₂ slope (r=0.03, p=0.90) nor the distance achieved during the 6-minute walk test (r=−0.14, p=0.54). Similarly, LV k_mono and WMI did not correlate with the peak VO₂ (r=−0.28, p=0.22 and r=0.10, p=0.66), VE/VCO₂ slope (r=−0.32, p=0.15 and r=0.22, p=0.33) nor distance achieved during the 6-minute walk test (r=−0.084, p=0.72 and r=−0.21, p=0.36), respectively (Table 4).

RV k_mono tended to correlate with peak VO₂ (r=−0.42, p=0.055) and VE/VCO₂ slope (r=0.40, p=0.071). However, the correlation with peak VO₂ was lost when correcting RV oxidative metabolism for LV oxidative metabolism (RVOx) (r=−0.19, p=0.42). In contrast, a good correlation between RVOx and VE/VCO₂ slope was observed (r=0.61, p=0.0034) (Fig. 2). Importantly, RVOx was lower in patients with VE/VCO₂ slope <34 compared to the RVOx in patients with a slope ≥34 (0.77±0.16 vs. 0.93±0.16; p=0.047, Fig. 3), a threshold related to poor prognosis in CHF 3,16,17. FEV-1 (1.75±0.50 vs. 2.57±0.45, p=0.0035) and FVC (2.57±0.58 vs. 3.40±0.56 L, p=0.0084) were lower in our patients with VE/VCO₂ slope ≥34.

4.1. Oxidative metabolism in the left and right ventricle

Myocardial oxidative metabolism is the primary biochemical pathway for the conversion of substrates into energy that will be used for contraction and other cellular functions. It has been well demonstrated that oxidative metabolism can be measured using the kinetics of [¹¹C]acetate with dynamic PET imaging which reflects TCA cycle flux and myocardial oxygen consumption in animal models and humans 20 21 22 23 24.

In the normal myocardium, metabolic demand of the RV is less than the LV reflecting the lesser metabolic demand required of the RV. However in heart failure, RV oxidative metabolism increases relative to the LV 9, which may reflect a relative increase in metabolic demand for the RV compared to the LV. In the current study, this corrected RV oxidative metabolism was higher than published normal values 9. Forward efficiency and work are often compromised as cardiac function deteriorates without a concomitant decrease in myocardial oxygen consumption i.e. instead of useful forward work, the heart uses its' energy to overcome increased wall stress.

4.2. Previous studies on oxidative metabolism and exercise in heart failure

The level of resting LV oxidative metabolism, as measured by k_mono of [¹¹C]acetate is the same in mild to moderate congestive heart failure and healthy volunteers 7. However, in advanced heart failure LV oxidative metabolism is reduced 7. RV oxidative metabolism and RVOx are increased in decompensated heart failure 9. To our knowledge there are no published in vivo data on RV oxidative metabolism in patients with mild or compensated heart failure.

Stolen et al 25 studied 20 patients with dilated cardiomyopathy and mildly symptomatic heart failure (NYHA class I-II) and found that peak VO₂ had an inverse relationship with the ratio of RV:LV oxidative metabolism. In contrast, we did not observe such a relationship, although there was a trend for a relationship between VO₂ and RV k_mono. The RER was >1 in both studies. However, patients were more symptomatic in our study (NYHA class 2.5±0.5 vs. 1.5±0.5 (Stolen et al. 25)). In addition, we included patients with ischaemic heart disease. Severity of disease and aetiology may explain the observed differences in the two studies.

4.3. VE/VCO₂ slope and RV oxidative metabolism

The relationship between minute ventilation and the rate of CO₂ elimination (the VE/VCO₂ slope) is strongly associated with mortality in patients with CHF 4,16. The VE/VCO₂ slope is also closely associated with symptoms in advanced CHF and provides complimentary prognostic information to peak VO₂ 17,26.

A VE/VCO₂ slope >34 is associated with a poor prognosis 3,17. It appears that the determinants of the VE/VCO₂ slope are multifactorial and include the peripheral ergoreceptor response 27 the muscle mass utilized for the exercise 28, increased ventilation/perfusion mismatch 5 and pulmonary vascular resistance (PVR) 29.

It is reasonable to assume that there would be an increased demand on the RV metabolism (as reflected by RV oxidative metabolism) due to the increased ventilation/perfusion mismatch and enhanced pulmonary vascular resistance. Therefore, a high VE/VCO₂ slope may be associated with increasing RV oxidative metabolism. In our study, the ratio of RV:LV oxidative metabolism, or corrected RV oxidative metabolism, had a good correlation with the VE/VCO₂ slope. There are significant variations in RV and LV oxidative metabolism in different physiological and pathophysiological circumstances. Therefore, correction of RV oxidative metabolism to LV oxidative metabolism (RV k_mono/LV k_mono ratio) may better describe the relative level of RV oxidative metabolism. Previous studies have demonstrated that RV oxidative metabolism is increased compared to LV metabolism in patients with CHF 9,25. This metabolic discordance likely reflects - at least in part - increased wall stress due to higher afterload in the RV leading to increased metabolic demand.

RV function contributes to exercise tolerance in patients with heart failure and carries prognostic value in this patient population 30,31. Further, mechanical discordance between the RV and LV has been associated with poor prognosis 30. In the current study, we have demonstrated a relationship between the corrected RV oxidative metabolism and a known prognostic marker, VE/VCO₂ slope, which further emphasizes the importance of the RV in CHF. Increased RVOx may reflect the metabolic load that is present prior to exercise.

4.4. Limitations of the study

While the RVOx was related to a parameter known to have prognostic value, the prognostic value of RVOx itself was not specifically tested. This will require future clinical follow-up studies.

Ideally, exercise [¹¹C]acetate data would have been compared to exercise VE/VCO₂. However exercise PET imaging is not practical with [¹¹C]acetate due to the requirement for early dynamic data and adverse effect of motion for supine exercise. Noteworthy is the observation that the VE/VCO₂ ratio at rest and the VE/VCO₂ slope during exercise are correlated 29. Catheterization (right and left side) was not part of our study protocol due its invasiveness. Therefore, direct measurements of pulmonary artery pressure were not available for our patients.

In spite of these limitations, the study does support a mechanistic link between RV oxidative metabolism and ventilation-perfusion mismatching.