2020 Volume 84 Issue 1 Pages 76-82
Background: Cardiopulmonary exercise testing (CPX) is used in the prognostic evaluation of patients with heart failure with reduced ejection fraction (HFrEF). In these patients, the ventilation feedback system is dysfunctional, and overactive peripheral chemoreceptors may be responsible for the early appearance of the respiratory compensation point (RCP) after the anaerobic threshold (AT). The mechanism of RCP appearance remains unknown and very few studies have reported the relationship between RCP and heart failure. We hypothesized that the duration between the RCP and AT (RCP-AT time) can predict the severity of cardiac disorders and prognosis in patients with HFrEF.
Methods and Results: We enrolled 143 patients with HFrEF who underwent symptom-limited maximal CPX between 2012 and 2016. During a median follow-up of 1.4 years, cardiovascular death occurred in 45 participants (31%). The patients who died had a significantly shorter RCP-AT time and lower hemoglobin (Hb) levels than those who survived (P<0.001 and P=0.01, respectively). Cox regression analyses revealed RCP-AT time and Hb level to be independent predictors of cardiovascular death in patients with HFrEF (P<0.001 and P=0.018, respectively).
Conclusions: RCP-AT time can better predict prognosis in patients with HFrEF than the magnitude of increase in oxygen consumption within the isocapnic buffering domain (∆V̇O2 AT-RCP). It may be useful as a new prognostic indicator in these patients.
Cardiopulmonary exercise testing (CPX) is routinely used in the prognostic evaluation of patients with heart failure with reduced ejection fraction (HFrEF). Agostoni et al conducted the first large-scale multicenter study in which a prognostic score (metabolic exercise test data combined with cardiac and kidney indexes (MECKI) score) was used in patients with HFrEF. The MECKI score combines CPX data with clinical, laboratory, and echocardiographic measurements, such as the percentage of the predicted value of peak oxygen consumption (% peak V̇O2), the slope of the relationship between minute ventilation and carbon dioxide production (V̇E vs. V̇CO2 slope), EF, hemoglobin (Hb) level, glomerular filtration rate (eGFR), and serum sodium (Na) level.1 Subsequently, Carubelli et al successfully demonstrated the validity of MECKI prognostic score.2,3
When a subject undergoes incremental exercise testing, exercise hypercapnia increases towards the end of the exercise because of acidosis caused by the accumulation of lactic acid in the circulation.4 The production of lactic acid begins once the working skeletal muscle cells reach the anaerobic threshold (AT).5 Usually, circulating bicarbonate compensates for this lactic acidosis, and this period is called the isocapnic buffering stage.6 However, beyond a certain point, the work intensity becomes stronger and lactic acid production can no longer be compensated by circulating bicarbonate; this is the point where hyperventilation begins and it is called the respiratory compensation point (RCP).5 Very few studies have reported on the relationship between RCP and HF, which is one of the reasons why the mechanism of the appearance of RCP remains unknown. Carrie et al performed an insightful clinical and exercise physiology study in which they tested, for the first time, whether the magnitude of increase in V̇O2 (∆V̇O2 AT-RCP) within the isocapnic buffering domain can predict syndrome severity and prognosis in patients with HFrEF.7 However, they found that ∆V̇O2 AT-RCP was associated with prognosis in the univariate but not the multivariate analysis and concluded that the isocapnic buffering period does stratify with HFrEF but not more than other prognostic indices such as V̇E/V̇CO2 slope.8 Tanehata et al9 reported a strong correlation between the time between AT and RCP (RCP-AT) and the slope of oxygen uptake as a function of work rate (∆V̇O2/∆WR). They concluded that in chronic HF patients, the RCP-AT time could be useful in estimating the rate of aerobic and anaerobic metabolism after AT and could be related to the severity of HFrEF. However, they did not identify the exact relationship between RCP-AT time and prognosis in patients with HF.
In patients with HF, the feedback system that controls ventilation is dysfunctional because of an increase in the controller gain (increased sensitivity to changes in arterial O2 and CO2 levels), reduction in system damping (decrease in total body stores of O2 and CO2), and a delay in information transfer (circulation time between the lungs and brain).10–12 The increased chemoreceptor drive may represent the high controller gain, whereby small changes in arterial O2 and CO2 levels can result in inappropriately large alterations of the system output.10,11 Therefore, overactive peripheral chemoreceptors may be responsible for the early appearance of the RCP after AT. Based on the understanding of this mechanism and results of past studies, we hypothesized that patients with severe HF have an inappropriate feedback system and that RCP appears early after AT; in other words, the shorter RCP-AT time in patients with HFrEF could indicate poor prognosis.
In this study, we evaluated the relationship between RCP-AT time and prognosis in patients with HFrEF. We analyzed and compared several previously reported prognostic risk factors, including Na, eGFR, % peak V̇O2, V̇E vs. V̇CO2 slope, EF, and B-type natriuretic peptide (BNP).1,13,14
The study cohort was retrieved from the Gunma Prefectural Cardiovascular Centre Database, established in 2012 for patients who newly visited the hospital. Data for patients who underwent symptom-limited maximal CPX between 2012 and 2016 were analyzed. The inclusion criteria were previous or present HF symptoms (New York Heart Association [NYHA] functional classes II–III, American College of Cardiology/American Heart Association [ACC/AHA] classification stage C), previous documentation of left ventricular (LV) systolic dysfunction (LVEF <40%), stable clinical condition with unchanged medications for ≥3 months, ability to perform CPX, and no major cardiovascular treatments or interventions scheduled. The exclusion criteria included a history of pulmonary embolism, moderate-to-severe aortic or mitral stenosis, pericardial disease, severe obstructive lung disease, exercise-induced angina and significant ECG alterations, or the presence of any clinical comorbidities that could interfere with exercising. Based on these criteria, we enrolled 143 patients with HFrEF.
The relevant histories and results of the following investigations were recorded. Blood sampling was performed after a 12-h overnight fast to assess blood levels of total cholesterol, Hb, lymphocytes, Na, uric acid (UA), and BNP. eGFR was calculated using the Japanese Society of Nephrology formulae:15 (194×serum creatinine−1.094×age−0.287 for men; 194×serum creatinine−1.094×age−0.287×0.739 for women). Blood pressure was measured after a 10-min seated rest and was expressed as the average of 3 consecutive measurements of each arm using the conventional cuff method. Echocardiography was also performed.
The endpoint was cardiovascular death, defined as any fatal event related to coronary artery disease, cerebrovascular disease, and other heart or vascular diseases, specifically fatal myocardial infarction, acute coronary syndrome, stroke, transient ischemic attack, HF, and arrhythmia.16 The time interval from the date of CPX to cardiovascular death was defined as the duration of follow-up. Endpoints were censored in October 2018. The follow-up rate was 100%.
CPXAT and peak oxygen consumption (V̇O2) were evaluated using symptom-limited CPX on an upright cycle ergometer (StrengthErgo8; Mitsubishi Electric Engineering, Tokyo, Japan) with an ECG machine (ML-9000, Fukuda Denshi, Ltd., Tokyo, Japan). CPX was performed 2–4 h after a light meal. The tests included, as per the recommendations of Buchfuhrer et al,17,18 3 min of rest and a 3-min warm-up at 0watt, followed by a continuous increase in the work rate (WR) by 1 W every 6 s until exhaustion. The criteria for halting the exercise testing in this study are outlined in the American College of Sports Medicine guidelines.19 The increments in WR levels were chosen on the basis of the ability of the patient to perform the exercises within 8–15 min.17,18 We measured oxygen consumption (V̇O2), carbon dioxide production (V̇CO2), and minute ventilation (V̇E) on a breath-by-breath basis using a gas analyzer (MINATO 300S, Minato Science Co., Ltd., Osaka, Japan). Peak V̇O2 was determined as V̇O2 at the highest WR. AT was measured using the V-slope method.20 RCP was recorded when an increase in the V̇E/V̇CO2 ratio was observed simultaneously with a decrease in endtidal CO2 pressure (PETCO2).6 The predicted peak V̇O2 (% peak V̇O2) and AT (% AT) were determined on the basis of data from a6 healthy Japanese population.21 The duration of the buffering period (RCP-AT time) and the V̇O2 changes during the isocapnic buffering period (∆V̇O2 AT-RCP) were measured as V̇O2 at AT-V̇O2 at the RCP.8,9
Statistical AnalysisThe parameters compared between those who survived and those who died during the follow-up period included basic demographic data, such as age, and body mass index (BMI); cardiovascular risk factors, such as number of lymphocytes, Hb, Na, UA, and total cholesterol levels; biomarkers, such as BNP levels; and CPX parameters, such as RCP-AT time, V̇O2, AT, peak V̇O2, V̇E vs. V̇CO2 slope, load, PETCO2, and peak O2 pulse. The distribution of each continuous variable was tested for normality using the Shapiro-Wilk test, and the results are expressed as mean±standard deviation. Variables with a skewed distribution are expressed as median [interquartile range]. Statistical analyses were performed using unpaired t-test or Mann-Whitney test for continuous variables and the chi-square test for categorical variables. To investigate the association with cardiovascular death, univariate and multivariate Cox regression analyses were applied to examine the aforementioned potential factors, including Na, eGFR, % peak V̇O2, V̇E vs. V̇CO2 slope, EF and BNP.1,13,14 We did not use peak heart rate, nadir V̇E/V̇CO2, or peak load in the multivariate analyses, even though there was a significant difference between the 2 groups, because we chose our explanatory variables based on experience in specialized fields and the previous literature.1,13 Receiver-operating characteristic (ROC) curves were used to identify the sensitivity and specificity of RCP-AT time for predicting cardiovascular death, and the area under the curve (AUC) was calculated. Survival analyses using Kaplan-Meier models were also performed. These analyses were performed using the Statistical Package for Social Sciences 21.0J for Mac (IBM Corp., Armonk, NY, USA). Statistical significance was set at a two-sided P value <0.05.
Ethical StatementThis study was approved by the Ethics Committee of Gunma Prefectural Cardiovascular Centre (approval no.: 31009) and was conducted in accordance with the Declaration of Helsinki. Each patient gave informed consent.
Table 1 summarizes the baseline characteristics of the participants who survived and those who died from cardiovascular death during a median follow-up period of 1.4 years. Cardiovascular death occurred in 45 participants (31%): 33 (23%) died of HF, and 12 (8%) from sudden death caused by arrhythmia. The mean age and BMI were not significantly different between groups. The participants who died had lower Hb and higher BNP levels (12.9±1.5 g/dL vs. 14.2±3.1 g/dL, P=0.01; and 249.0 [198.0, 629.7] pg/mL vs. 161.4 [106.9, 407.2] pg/mL, P=0.03; respectively) than did those who survived; in contrast, there were no significant differences in lymphocyte count, eGFR, and serum levels of sodium, UA, and total cholesterol. There were no significant differences between the groups regarding NYHA classes, prevalence of hypertension and diabetes, or the etiology of HF. Regarding echocardiography findings, there was no significant difference in LVEF between the groups. There were no differences between the groups in the proportion of patients on cardioprotective agents such as β-blockers.
| Death (n=45) |
Survival (n=98) |
P value | |
|---|---|---|---|
| Patient characteristics | |||
| Sex (M/F) | 41/4 | 81/17 | 0.85 |
| Age (years) | 62.2±18.7 | 59.5±14.2 | 0.35 |
| Height (cm) | 165.9±7.4 | 164.7±7.2 | 0.36 |
| Weight (kg) | 60.8±13.4 | 64.8±16.3 | 0.18 |
| BMI (kg/m2) | 22.0±4.2 | 24.8±7.5 | 0.08 |
| LVEF (%) | 29.3±8.1 | 29.6±15.6 | 0.83 |
| Hb level (g/dL) | 12.9±1.5 | 14.2±3.1 | 0.01 |
| Lymphocyte count (%) | 23.4 [18.0, 26.3] | 23.5 [17.5, 29.2] | 0.23 |
| Sodium (mEq/L) | 140.0 [139.0, 142.0] | 142.0 [138.0, 143.0] | 0.93 |
| Uric acid (mg/dL) | 6.5±1.2 | 6.8±4.1 | 0.56 |
| Total cholesterol (mg/dL) | 173.0 [146.0, 184.5] | 189.0 [160.0, 234.5] | 0.18 |
| eGFR (mL/min/1.73 m2) | 57.8±30.3 | 62.3±23.5 | 0.47 |
| BNP (pg/mL) | 249.0 [198.0, 629.7] | 161.4 [106.9, 407.2] | 0.03 |
| NYHA class II/III | 8/37 | 24/74 | 0.37 |
| Etiology of HF | |||
| Ischemic/non-ischemic | 15/30 | 29/69 | 0.69 |
| Medical history | |||
| Hypertension | 15 (33%) | 31 (31%) | 0.84 |
| Diabetes mellitus | 17 (38%) | 31 (32%) | 0.57 |
| Medical therapy | |||
| β-blockers | 37 (82%) | 86 (87%) | 0.43 |
| ACEI/ARB | 33 (73%) | 75 (76%) | 0.68 |
| Spironolactone | 21 (47%) | 50 (51%) | 0.71 |
Data are mean±standard deviation, median value [interquartile range], or n (%). ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin-receptor blocker; BNP, B-type natriuretic peptide; BMI, body mass index; eGFR, estimated glomerular filtration rate; Hb, hemoglobin; HFrEF, heart failure with reduced ejection fraction; LV, left ventricular; NYHA, New York Heart Association.
The CPX parameters of each group are summarized in Table 2. There were no significant differences in systolic blood pressure at rest and peak, or in heart rate at peak exercise during CPX. In contrast, the heart rate at rest was significantly higher in the patients who died vs. those who survived (78.1±14.2 beats/min vs. 72.1±12.9 beats/min, P=0.01, respectively).
| Death (n=45) |
Survival (n=98) |
P value | |
|---|---|---|---|
| Systolic BP at rest (mmHg) | 110.5±20.8 | 110.6±20.3 | 0.86 |
| Systolic BP at peak (mmHg) | 137.9±33.4 | 147.7±27.6 | 0.07 |
| HR at rest (beats/min) | 78.1±14.2 | 72.1±12.9 | 0.01 |
| HR at peak (beats/min) | 112.4±25.3 | 118.2±19.8 | 0.13 |
| Rest V̇O2 (mL/min) | 231.6±47.5 | 229.5±49.0 | 0.80 |
| AT (mL/min) | 592.4±224.8 | 658.7±239.7 | 0.11 |
| RCP V̇O2 (mL/min) | 757.4±289.8 | 870.9±280.9 | 0.02 |
| Peak V̇O2 (mL/min) | 839.0±342.1 | 1,001.7±345.7 | <0.05 |
| Rest V̇O2 (mL/min/kg) | 3.5±0.6 | 3.7±0.7 | 0.11 |
| AT (mL/min/kg) | 9.6±2.3 | 10.3±3.0 | 0.18 |
| RCP V̇O2 (mL/min/kg) | 12.3±3.2 | 13.8±3.1 | 0.02 |
| Peak V̇O2 (mL/min/kg) | 13.6±3.7 | 15.5±3.8 | 0.004 |
| % AT (%) | 59.7±14.4 | 64.9±19.0 | 0.16 |
| % peak (%) | 54.2±15.1 | 62.4±14.5 | 0.002 |
| V̇E vs. V̇CO2 slope | 37.8 [32.3, 43.2] | 33.1 [29.6, 37.0] | 0.006 |
| Nadir V̇E/V̇CO2 | 42.1 [36.4, 47.0] | 38.6 [32.5, 41.2] | <0.001 |
| Peak load (watts) | 61.0 [46.7, 69.0] | 63.0 [52.0, 80.0] | <0.001 |
| Maximum PETCO2 (mmHg) | 34.2±4.6 | 38.7±4.3 | 0.001 |
| Peak O2 pulse (mL/beat) | 7.2 [6.0, 8.9] | 7.5 [6.0, 9.6] | 0.04 |
| RCP-AT time (s) | 99.3±44.0 | 159.5±106.9 | <0.001 |
| ΔV̇O2 AT-RCP (mL/min) | 164.9±93.6 | 227.8±160.3 | 0.01 |
Data are mean±standard deviation or median value [interquartile range]. AT, anaerobic threshold; BP, blood pressure; HFrEF, heart failure with reduced ejection fraction; HR, heart rate; O2 pulse, oxygen pulse; Peak V̇O2, oxygen uptake at peak work rate; PETCO2, endtidal carbon dioxide pressure; RCP, respiratory compensation point; Rest V̇O2, oxygen uptake at peak work rate; V̇E, minute ventilation; V̇CO2, carbon dioxide production.
There were no significant differences in V̇O2 at rest and AT; however, significant differences in RCP V̇O2 and peak V̇O2 were observed. Parameters indicating cardiac function during exercise such as V̇E vs. V̇CO2 slope (27.8 [32.3, 43.2] vs. 33.1 [29.6, 37.0], P=0.006), nadir V̇E/V̇CO2 (42.1 [36.4, 43.7] vs. 38.6 [32.5, 41.2], P<0.001), maximum PETCO2 (34.2±4.6 mmHg vs. 38.7±4.3 mmHg, P=0.001), and peak O2 pulse (7.2 [6.0, 8.9] mL/beat vs. 7.5 [6.0, 9.6] mL/beat, P=0.04) were significantly different between the 2 groups. Finally, RCP-AT time was longer and ∆V̇O2 AT-RCP was smaller in those who died than in those who survived (99.3±44.0 s vs. 159.5±106.9 s, P<0.001; and 164.9±93.6 mL/min vs. 227.8±160.3 mL/min, P=0.01, respectively).
Predictors of Cardiovascular DeathOn univariate analysis, the following factors were significantly predictive of cardiovascular death: % peak V̇O2 (hazard ratio [HR]: 0.968; 95% confidence interval [CI]: 0.948–0.989; P=0.003), Hb (HR: 0.726; 95% CI 0.612–0.869; P<0.001), V̇E vs. V̇CO2 slope (HR: 1.023; 95% CI 1.002–1.045; P=0.032), RCP-AT time (HR: 0.982; 95% CI 0.975–0.989; P<0.001) and ∆V̇O2 AT-RCP (HR: 0.994; 95% CI 0.991–0.997; P<0.001) (Table 3).
| Variable | Univariate | Multivariate | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| % peak V̇O2 (%) | 0.968 | 0.948–0.989 | 0.003 | |||
| Hb level (g/dL) | 0.726 | 0.612–0.869 | <0.001 | 0.792 | 0.664–0.945 | 0.018 |
| V̇E vs. V̇CO2 slope | 1.023 | 1.002–1.045 | 0.032 | |||
| Sodium (mEq/L) | 0.986 | 0.954–1.020 | 0.420 | |||
| LVEF (%) | 0.996 | 0.967–1.028 | 0.825 | |||
| eGFR (mL/min/1.73 m2) | 0.992 | 0.980–1.006 | 0.270 | |||
| BNP (pg/mL) | 0.992 | 0.989–1.012 | 0.630 | |||
| RCP-AT time (s) | 0.982 | 0.975–0.989 | <0.001 | 0.984 | 0.978–0.990 | <0.001 |
CI, confidence interval; LVEF, left ventricular ejection fraction; % peak V̇O2, predicted value of oxygen uptake at peak work rate. Other abbreviations as in Tables 1,2.
Multivariate analysis revealed that RCP-AT time (HR: 0.984; 95% CI: 0.978–0.990; P<0.001) and Hb (HR: 0.792; 95% CI: 0.664–0.945; P=0.018) were independent predictors of cardiovascular death (Table 3).
We also performed univariate and multivariate Cox regression analyses that changed the explanatory variables from RCP-AT time to ∆V̇O2 AT-RCP. However, ∆V̇O2 AT-RCP was not an independent predictor of cardiovascular death (Table 4).
| Variable | Univariate | Multivariate | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| % peak V̇O2 (%) | 0.968 | 0.948–0.989 | 0.003 | |||
| Hb level (g/dL) | 0.726 | 0.612–0.869 | <0.001 | 0.729 | 0.615–0.864 | <0.005 |
| V̇E vs. V̇CO2 slope | 1.023 | 1.002–1.045 | 0.032 | |||
| Sodium (mEq/L) | 0.986 | 0.954–1.020 | 0.420 | |||
| LVEF (%) | 0.996 | 0.967–1.028 | 0.825 | |||
| eGFR (mL/min/1.73 m2) | 0.992 | 0.980–1.006 | 0.270 | |||
| BNP (pg/mL) | 0.992 | 0.989–1.012 | 0.630 | |||
| ΔV̇O2 AT-RCP (mL/min) | 0.994 | 0.991–0.997 | <0.001 | |||
Abbreviations as in Tables 1–3.
Based on the ROC curve analysis, the optimal cutoff for RCP-AT time to predict cardiovascular death was 127.5 s, with specificity and sensitivity of 63.3% and 77.8%, respectively (AUC=0.762; 95% CI: 0.679–0.845) (Figure 1).
Receiver-operating characteristic curve of RCP-AT time as a predictor of cardiovascular death. RCP-AT time, period between respiratory compensation point and anaerobic threshold.
The event-free survival curve was also evaluated using the RCP-AT cutoff value (127.5 s). Results indicated that individuals with RCP-AT time ≥127.5 s had a significantly higher risk of cardiovascular death than did individuals with RCP-AT time <127.5 s (P<0.001) (Figure 2).
Kaplan-Meier survival curves of cardiovascular death according to RCP-AT time. The red line represents RCP-AT time <127.5 s and the blue line represents RCP-AT time ≥127.5 s. RCP-AT time, period between respiratory compensation point and anaerobic threshold.
The present study found that patients with a longer RCP-AT time (127.5 s) had a worse prognosis and greater severity of HF than did patients with shorter RCP-AT time (<127.5 s). Additionally, serum Hb level was an important factor in their prognoses. To the best of our knowledge, this is the first study to investigate the role of RCP-AT time in the prognosis of patients with HFrEF.
During a progressively increasing workload exercise, ventilation follows 3 distinctive domains that are regulated by oxygen uptake, carbon dioxide production, and unbuffered acidosis, respectively.22 The first domain is from the beginning of active exercise to the point of AT, the second between AT and RCP, and the third from RCP to the end of exercise.22–24
Ventilation is regulated through a feedback loop between pulmonary gas-exchanging capillaries and chemoreceptors in the carotid bodies (peripheral) or medulla (central).25,26 Ventilatory dysfunction may arise from (1) a delay in information transfer (i.e., increased circulation time because of reduced cardiac index27), (2) an increase in controller gain, such as increased chemosensitivity to arterial CO2 and O2 levels,28,29 or (3) a reduction in system damping, such as baroreflex impairment.28 Abnormal hemodynamic variables are known to be associated with a poor prognosis in conditions of dysfunctional ventilation,30,31 and the mechanism of augmentation of the chemoreflex may lie in sympathetic overactivity and neurohormonal imbalance, both of which also affect survival in chronic HF.32,33 Catecholamines have been shown to increase chemosensivity,34 which may further perpetuate the sympathetic drive and contribute to neurohormonal imbalance.35 The chemoreflex may also be augmented directly because of reduced blood flow to the chemoreceptors – again a reflection of hemodynamic dysfunction. Therefore, the RCP appears early after AT, and patients with shorter RCP-AT time have worse prognosis. A previous study reported that RCP-AT time had a strong correlation with ∆V̇O2/∆Load in patients with chronic HF.9 Furthermore, increased ∆V̇O2/∆Load suggests high blood flow to the whole body during exercise.22 Those findings also support our hypothesis.
Carrie et al carried out the first study to test whether the magnitude of increase in V̇O2 (∆V̇O2 AT-RCP) within the isocapnic buffering domain predicts syndrome severity and prognosis in patients with HFrEF. They identified 782 patients with HFrEF and found that ∆V̇O2 AT-RCP was associated with prognosis in the univariate but not the multivariate analysis.8 The current study included fewer participants but demonstrated the same results: ∆V̇O2 AT-RCP was associated with prognosis in patients with HFrEF in the univariate but not the multivariate analysis. This consistency with the previous study8 validates our findings. Although the mechanism of RCP remains unknown, the definition has been established in many studies. One of the strengths of our study was that we could set the RCP and AT using those definitions.4,6,20 PeakV̇O2 may depend on the patient’s mental state and can be lower than the actual value because of lack of effort;21 however, the RCP-AT time is not affected by the patient’s condition, and we could calculate it precisely by the established definitions. By calculating RCP-AT time, we could predict the patient’s prognosis more precisely and evaluate the effect of HFrEF medical treatment and cardiac rehabilitation.
CPX is routinely used in the prognostic evaluation of patients with HFrEF, in whom the prognostic value of peak V̇O2 and V̇E vs. V̇CO2 slope are powerful and well established factors.36,37 However, in the current study, peak V̇O2, V̇E vs. V̇CO2 slope, and RCP-AT time were associated with cardiovascular death in the univariate analysis, but only RCP-AT time was associated with cardiovascular death in the multivariate analysis. This is because the number of patients was small and, as stated earlier, there were no significant differences between the groups in terms of NYHA class in this study. Peak V̇O2, a prognostic indicator in patients with HFrEF, was significantly different between the patients who survived and those who died; however, both groups had low values. This is probably why peak V̇O2 was not identified as a statistically significant prognostic predictor. A similar principle applies to the V̇E vs. V̇CO2 slope between the groups.
In the CHART-2 study, 1,360 patients were enrolled and 10% died of cardiovascular disease during the 3-year follow-up.38 In our study, 31% died during the documented median follow-up of 1.4 years, which is much higher than in the CHART-2 study, possible because of differences between the 2 studies in the patients’ characteristics. In CHART-2 study, patients’ BNP was 172.0 [range 71.6–374.0] pg/mL, whereas in our study it was 234.0 [range 143.0–529.0], which was much higher. In addition, the patients enrolled in the CHART-2 study had LVEF of 37.9±0.3%, whereas the LVEF in our study was 29.4±0.8%. These data show that the patients enrolled in our study had more severe HF than those in the CHART-2 study.
Resting heart rate is a very accessible biological parameter with potential predictive capacity for HF and cardiovascular disease. Increased resting heart rate is a marker of cardiovascular risk.39 In addition, BNP is a prognostic marker in HF patients.14 In our study, the patients who died had higher BNP and resting heart rate, which supports our data’s validity and demonstrates that our study patients’ characteristics differed from those of the patients in the CHART-2 study.
In a previous study, serum Hb level was reported to be an important factor in predicting poor prognosis in patients with HF.40 Anemia is reported to be an independent risk factor for higher all-cause mortality and hospitalization rates.40–42 Therefore, our results were in accordance with those of previous studies.
Study LimitationsTo the best of our knowledge, this is the first study to investigate RCP-AT time in the prognosis of patients with HF. However, this study has several limitations. First, our sample size was small, and there were few outcomes. Second, this was a retrospective, single-center study, so the possibility of unintentional selection bias cannot be fully excluded. Third, we proved that serum Hb level was an important factor in the prognosis of patients with HFrEF, but the prognostic effect of anemia may differ depending on its cause as well as HF etiology. Unfortunately, we did not investigate the cause of anemia for each HFrEF patient. Linking the cause of anemia and etiology of HF may lead to a more thorough discussion of anemia and the prognosis of HF. Finally, the results of this study are applicable only to patients with HFrEF because patients with preserved systolic function or with other diseases that affect exercise performance were not evaluated.
It is important to perform CPX in patients with HFrEF to assess AT, RCP, and the isocapnic buffering period. However, the RCP-AT time has a wide range and prognostic power; therefore, evaluation of the isocapnic buffering period has physiological significance and prognostic power in patients with HF. We believe that our data will help prove the detailed mechanism of the appearance of RCP. In the future, RCP-AT time may be used to evaluate the therapeutic effects of treatment in patients with HFrEF by comparing the RCP-AT time before and after the treatment.
We thank our lecturers Haruyasu Fujita (Gunma University, Department of Public Health, Gunma, Japan) and Lee Bumsuk (Gunma University Graduate School of Health Sciences, Gunma, Japan) for their valuable statistical advice during the preparation of this report. We thank Editage (www.editage.jp) for English language editing.
T.N. led the study as the principal investigator, prepared the plan for statistical analyses, drafted, and revised the manuscript. H.A. cleaned the data and performed the statistical analyses. M.M. and S.N. collected the data. All authors have approved the final version of the manuscript.
This research received no grant from any funding agency in the public, commercial, or not-for-profit sectors.
The authors declare that there are no conflicts of interest.
No support was received to perform the study or for the preparation of the manuscript.
