Increased chemosensitivity has been observed in HF (heart failure) and, in order to clarify its pathophysiological and clinical relevance, the aim of the present study was to investigate its impact on neurohormonal balance, breathing pattern, response to exercise and arrhythmic profile. A total of 60 patients with chronic HF [age, 66±1 years; LVEF (left ventricular ejection fraction), 31±1%; values are means±S.E.M.] underwent assessment of HVR (hypoxic ventilatory response) and HCVR (hypercapnic ventilatory response), neurohormonal evaluation, cardiopulmonary test, 24-h ECG monitoring, and assessment of CSR (Cheyne–Stokes respiration) by diurnal and nocturnal polygraphy. A total of 60% of patients had enhanced chemosensitivity. Those with enhanced chemosensitivity to both hypoxia and hypercapnia (i.e. HVR and HCVR), compared with those with normal chemosensitivity, had significantly (all P<0.01) higher noradrenaline (norepinephrine) and BNP (B-type natriuretic peptide) levels, higher prevalence of daytime and night-time CSR, worse NYHA (New York Heart Association) class and ventilatory efficiency [higher V̇E (minute ventilation)/V̇CO2 (carbon dioxide output) slope], and a higher incidence of chronic atrial fibrillation and paroxysmal non-sustained ventricular tachycardia, but no difference in left ventricular volumes or LVEF. A direct correlation was found between HVR or HCVR and noradrenaline (R=0.40 and R=0.37 respectively; P<0.01), BNP (R=0.40, P<0.01), N-terminal pro-BNP (R=0.37 and R=0.41 respectively, P<0.01), apnoea/hypopnoea index (R=0.57 and R=0.59 respectively, P<0.001) and V̇E/V̇CO2 slope (R=0.42 and R=0.50 respectively, P<0.001). Finally, by multivariate analysis, HCVR was shown to be an independent predictor of both daytime and night-time CSR. In conclusion, increased chemosensitivity to hypoxia and hypercapnia, particularly when combined, is associated with neurohormonal impairment, worse ventilatory efficiency, CSR and a higher incidence of arrhythmias, and probably plays a central pathophysiological role in patients with HF.

INTRODUCTION

Despite the advances in the treatment of chronic HF (heart failure), its epidemiological relevance is still increasing [1]. Ventricular remodelling, symptoms and final outcome are highly influenced by neurohormonal derangement [2], characterized by the activation of the sympathetic [3] and the renin–angiotensin–aldosterone system [4], and by the enhanced secretion of natriuretic peptides, whose plasma levels have diagnostic and prognostic value [5]. Increased sympathetic activity, one of the major determinants of the evolution of the disease and of life-threatening events, is elicited by changes in autonomic afferent feedback via baroreceptor desensitization [6], and ergoreceptor [7] and chemoreceptor sensitization [8].

In particular, increased chemosensitivity has also been associated with several markers of worse clinical status and prognosis in patients with HF, such as the ventilatory response to exercise [9] or night-time CSR (Cheyne–Stokes respiration) [10,11]. Previously, increased peripheral chemosensitivity has been highlighted as an independent prognostic factor in HF [12]. However, the importance of increased sensitivity to either hypoxia or hypercapnia and the possible role of their combination in the pathophysiology of HF have not been fully investigated. Moreover, there are few reports on the impact of chemosensitivity on other neurohormonal axes than the adrenergic one, including renin–angiotensin–aldosterone and cardiac natriuretic peptide systems.

Our hypothesis is that chemosensitivity might play a central role in influencing neurohormonal activation and the overall clinical picture in patients with HF, even on optimal pharmacological treatment. Therefore the aim of the present study was to evaluate in a prospective cohort of patients with chronic HF (optimally treated by a complete neurohormonal antagonist drug approach) the actual prevalence of enhanced chemosensitivity to both hypoxia and hypercapnia and its significance with regard to neurohormonal activation, including cardiac endocrine function, occurrence of CSR, clinical status and response to exercise.

MATERIALS AND METHODS

Subjects and study design

From 2005 to 2006, we screened 84 consecutive patients with HF from our outpatient clinic with echocardiographic evidence of impaired LV (left ventricular) systolic function [LVEF (LV ejection fraction) <45%]. Exclusion criteria were NYHA (New York Heart Association) class IV, acute coronary syndrome within 6 months before examination, severe renal dysfunction (i.e. creatinine clearance <35 ml/min), pulmonary disease [vital capacity and total lung capacity <50% of predicted value; FEV1 (forced expiratory volume in 1 s) <50% of predicted value; and FEV1/FVC (forced vital capacity) <70%], obstructive sleep apnoea syndrome (as determined by a preliminary polysomnography), and treatment with morphine or derivates, theophylline, oxygen, benzodiazepines or acetazolamide. A total of 60 patients matched these criteria and were enrolled in the study (Table 1). All of the patients were on stable (i.e. >1 month) optimal pharmacological treatment, with restriction of water/sodium intake. The study design included a standard clinical evaluation and: (i) the detection of chemosensitivity to hypoxia and to hypercapnia by assessing the individual HVR (hypoxic ventilatory response) and HCVR (hypercapnic ventilatory response); (ii) neurohormonal evaluation; (iii) echocardiography; (iv) arterial blood gas analysis; (v) CPT (cardiopulmonary exercise testing); (vi) 24-h ECG recording; and (vii) 20-min daytime polygraphy and nocturnal polysonography for CSR assessment. The entire protocol was completed for each patient within 3 days.

Table 1
Characteristics of the control subjects and patients with HF

Values are means±S.E.M., except for categorical data which are expressed as percentages. *P<0.05 and †P<0.01 compared with controls. HVR and HCVR indicate the chemosensitivity to hypoxia and hypercapnia respectively. ARBs, angiotensin receptor blockers.

Controls (n=12)HF patients (n=60)
Age (years) 65±1 66±1 
Male (%) 84 88 
BMI (kg/m226.7±0.4 27.4±0.5 
Creatinine clearance (ml/min) 85.3±6.8 79.2±4.1 
Aetiology of HF (%)   
 Ischaemic − 38 
 Idiopathic  50 
 Secondary  12 
Atrial fibrillation (%) − 25 
NYHA class I/II/III (%) − 10/50/40 
LVEF (%) 60.2±2.4 30.7±0.9* 
HVR (litre·min−1·%SaO2−10.35±0.06 0.74±0.06† 
HCVR (litre·min−1·mmHg−10.31±0.07 0.83±0.07† 
Medication (%)   
 Furosemide − 90 
 β-Blockers − 92 
 ACE inhibitors − 62 
 ARBs − 22 
 Spironolactone − 62 
Controls (n=12)HF patients (n=60)
Age (years) 65±1 66±1 
Male (%) 84 88 
BMI (kg/m226.7±0.4 27.4±0.5 
Creatinine clearance (ml/min) 85.3±6.8 79.2±4.1 
Aetiology of HF (%)   
 Ischaemic − 38 
 Idiopathic  50 
 Secondary  12 
Atrial fibrillation (%) − 25 
NYHA class I/II/III (%) − 10/50/40 
LVEF (%) 60.2±2.4 30.7±0.9* 
HVR (litre·min−1·%SaO2−10.35±0.06 0.74±0.06† 
HCVR (litre·min−1·mmHg−10.31±0.07 0.83±0.07† 
Medication (%)   
 Furosemide − 90 
 β-Blockers − 92 
 ACE inhibitors − 62 
 ARBs − 22 
 Spironolactone − 62 

The investigation was carried out in accordance with the Declaration of Helsinki (2000) of the World Medical Association, and has been approved by the Institutional Ethics Committee. Informed consent was obtained from all subjects enrolled in the study.

Evaluation of chemosensitivity

Chemoreceptor sensitivity was assessed using the rebreathing technique [13,14]. Subjects were examined in standardized conditions, in a quiet room at a comfortable temperature, while seated and connected to a rebreathing circuit through a mouthpiece. They were not allowed to smoke, or drink alcohol or caffeine-containing beverages in the 12 h preceding the study. ECG, airway flow and respiratory gases were recorded continuously through a breath-by-breath gas analyser (Vmax; Sensormedics), and oxygen saturation was recorded through a pulse oximeter (SET® Radical; Masimo). A 4-min baseline recording was performed during spontaneous breathing. The mean SaO2 (arterial oxygen saturation) and end-tidal CO2 during this recording were assumed as subject resting values. During the progressive isocapnic hypoxia trial (from resting SaO2 values to 70–80%, according to individual tolerance), end-tidal CO2 was maintained at a baseline value by passing a portion of the expired air into a scrubbing circuit before returning it to a 5 litre rebreathing bag. Conversely, during the progressive normoxic hypercapnic trial (from resting end-tidal CO2 values until 50 mmHg or an increase ≥10 mmHg from the basal values, according to individual tolerance), inspired PaO2 (arterial partial pressure of oxygen) was kept at the baseline value by adding oxygen to the circuit. The two trials were performed in a random order. All signals were digitized online (500 sample/s; National Instruments) and were analysed to derive respiratory rate, breath-to-breath V̇T (tidal volume) and V̇E (minute ventilation), as well as SaO2 and end-tidal pressure of CO2.

HVR was expressed by the linear regression slope between V̇E and SaO2 during a hypoxic-normocapnic trial. HCVR was expressed by the linear regression slope between V̇E and the end-tidal pressure of CO2 during the hypercapnic-normoxic trial. The mean HVR and HCVR values obtained in a group of 12 age- and gender-matched healthy subjects are shown in Table 1. As a cut-off value for defining increased chemosensitivity, we considered 0.77 litre·min−1·%SaO2−1 for HVR and 0.79 litre·min−1·mmHg−1 for HCVR (i.e. 2-fold the S.D. value of the control group).

Neurohormonal assay, CPT, echocardiographic study and 24-h ECG recording

Plasma BNP (B-type natriuretic peptide), catecholamine and aldosterone levels, PRA (plasma renin activity) and thyroid profile were assayed as described in detail previously [15]. NT-proBNP (N-terminal pro-BNP) was measured with an automated electrochemiluminescent immunoassay (Roche Diagnostics).

Patients underwent a symptom-limited CPT on a bicycle ergometer according to a ramp protocol with increments of 10 W/min (Vmax; Sensormedics). Peak V̇O2 (oxygen uptake; the highest value at peak exercise over a 20-s average) and ventilatory efficiency [regression slope relating V̇E to V̇CO2 (carbon dioxide uptake)] were determined. All CPTs and echocardiographic studies were performed by the same physician blinded to the results of the chemosensitivity tests. A 24-h ECG recording was obtained using a three-lead (precordial, posterior and inferior leads) digital system (Elamedical). In patients with sinus rhythm from 24-h ECG recordings, we also computed 24-h average values of normal RR intervals, SDNN (S.D. of all RR intervals), SDANN (S.D. of 5 min mean values of RR intervals), RMSSD (square root of the mean of the sum of the squares of differences between adjacent RR intervals) and pNN50 (number of adjacent RR intervals differing by more than 50 ms, as a percentage of the total number of RR intervals).

Daytime cardiorespiratory recording and polysomnography

All subjects underwent a 20-min recording while awake and spontaneously breathing in a supine position, as described previously [16]. We recorded a two-lead ECG, chest wall and abdominal movements by electrical inductance, oronasal airflow by nasal pressure transducers, beat-to-beat blood pressure (Colin® tonometry), SaO2 (Pulse Oxymeter Pulsox-7; Minolta) and end-tidal pressure of the CO2 signal (Cosmoplus®; Novametrics). All patients also underwent nocturnal continuous polygraphic recording by conventional polysomnography (PSG, E-series 2; Compumedics) within the same day of the short-term recording. An episode of apnoea was defined as the cessation of inspiratory airflow for at least 10 s, whereas hypopnoea was defined as a reduction in airflow (>50% of V̇T) lasting 10 s or more and associated with at least a 4% decrease in arterial oxyhaemoglobin saturation [17]. Apnoea and hypopnoea were considered as central or obstructive by the absence or presence of ribcage and abdominal excursions respectively. As regards to night-time polygraphy, the severity of apnoeas was quantified by means of the AHI (apnoea/hypopnoea index; i.e. the number of episodes of apnoea and hypopnoea/h, with a cut-off of diagnosis of CSR=10).

Statistical analysis

Statistical analysis was performed using SPSS version 13.0. Noradrenaline (norepinephrine), aldosterone, PRA, BNP and NT-proBNP were logarithmically transformed to correct for a skewed distribution. Mean differences among the groups were evaluated using ANOVA. Discrete variables were compared using a χ2 test with Yate's correction or Fisher's exact test when appropriate. Post-hoc testing was performed using the Bonferroni correction. The Spearman rank correlation was used to determine the direct relationship between different numerical variables. The predictive power of a variable was quantified in terms of the area under the ROC (receiver operating characteristic) curve, and the statistical significance of the difference in the AUC (area under the curve) from that of the line of ‘no information’ was evaluated by a Mann–Whitney U test. Multiple logistic regression analysis was employed in order to evaluate the influence of different variables on the occurrence of CSR. Values are means±S.E.M, and a P value <0.05 was considered statistically significant.

RESULTS

On the whole, 60% of patients with HF had increased hypoxic and/or hypercapnic chemosensitivity when compared with controls (Table 2). Isolated increased HVR and HCVR were found in 13 and 20% of patients respectively, whereas a combined increase in both HVR and HCVR was present in 27% of the patients (Table 2).

Table 2
Clinical characteristics, and cardiac, renal and pulmonary function according to chemosensitivity

Values are means±S.E.M., except for categorical data which are expressed as numbers (percentage). *P<0.001 compared with normal HVR and HCVR; †P<0.05 compared with increased HCVR alone. HVR and HCVR indicate the chemosensitivity to hypoxia and hypercapnia respectively. LVEDVi, LV end-diastolic volume index; LVESVi, LV end-systolic volume index; PaCO2, arterial partial pressure of carbon dioxide.

Normal HVR and HCVRIncreased HVRIncreased HCVRIncreased HVR and HCVR
n 24 (40%) 8 (13%) 12 (20%) 16 (27%) 
Age (years) 64±2 64±2 68±5 69±3 
Male (n20 (83%) 7 (88%) 11 (92%) 15 (93%) 
BMI (kg/m227.9±0.9 28.8±1.2 26.2±0.9 27.7±0.6 
NYHA class III (%) 16 50* 50* 56* 
LVEF (%) 31.4±1.4 28.3±2.9 29.8±2.3 31.4±2.0 
LVEDVi (ml/m) 135.1±7.2 136.3±16.4 146.9±10.1 140.4±9.2 
LVESVi (ml/m) 86.4±6.8 96.6±13.1 94.1±11.9 89.1±6.9 
Creatinine clearance (ml/min) 84.0±6.6 84.8±10.1 80.8±7.7 74.1±5.3 
FEV1 (litres) 2.64±0.24 2.61±0.35 2.27±0.23 2.56±0.19 
FEV1/FVC 72.3±1.9 78.7±3.2 73.3±4.9 68.5±2.8 
pH 7.44±0.01 7.45±0.01 7.44±0.01 7.46±0.01 
PaO2 (mmHg) 80.2±1.5 82.8±1.9 78.0±3.3 78.9±2.1 
PaCO2 (mmHg) 36.5±1.0 35.1±0.3 35.8±0.9 33.3±0.9 
HVR (litre·min−1·%SaO2−10.40±0.04 1.12±0.04* 0.55±0.05 1.17±0.07* 
HCVR (litre·min−1·mmHg−10.47±0.04 0.66±0.06 0.99±0.04* 1.29±0.09*† 
Normal HVR and HCVRIncreased HVRIncreased HCVRIncreased HVR and HCVR
n 24 (40%) 8 (13%) 12 (20%) 16 (27%) 
Age (years) 64±2 64±2 68±5 69±3 
Male (n20 (83%) 7 (88%) 11 (92%) 15 (93%) 
BMI (kg/m227.9±0.9 28.8±1.2 26.2±0.9 27.7±0.6 
NYHA class III (%) 16 50* 50* 56* 
LVEF (%) 31.4±1.4 28.3±2.9 29.8±2.3 31.4±2.0 
LVEDVi (ml/m) 135.1±7.2 136.3±16.4 146.9±10.1 140.4±9.2 
LVESVi (ml/m) 86.4±6.8 96.6±13.1 94.1±11.9 89.1±6.9 
Creatinine clearance (ml/min) 84.0±6.6 84.8±10.1 80.8±7.7 74.1±5.3 
FEV1 (litres) 2.64±0.24 2.61±0.35 2.27±0.23 2.56±0.19 
FEV1/FVC 72.3±1.9 78.7±3.2 73.3±4.9 68.5±2.8 
pH 7.44±0.01 7.45±0.01 7.44±0.01 7.46±0.01 
PaO2 (mmHg) 80.2±1.5 82.8±1.9 78.0±3.3 78.9±2.1 
PaCO2 (mmHg) 36.5±1.0 35.1±0.3 35.8±0.9 33.3±0.9 
HVR (litre·min−1·%SaO2−10.40±0.04 1.12±0.04* 0.55±0.05 1.17±0.07* 
HCVR (litre·min−1·mmHg−10.47±0.04 0.66±0.06 0.99±0.04* 1.29±0.09*† 

Patients with increased HVR and/or HCVR (i.e. chemosensitivity to hypoxia and/or hypercapnia) did not differ with regards to age, gender, BMI (body mass index), LV dimensions and function, renal and pulmonary function, and arterial gas analysis values compared with patients with normal chemosensitivity (Table 2). Increased HVR and/or HCVR were associated with worse clinical severity, as expressed by the percentage of patients in NYHA class III (Table 2).

Chemosensitivity and neurohormonal activation

Patients with combined increases in HVR and HCVR, compared with patients with normal chemosensitivity, had a significant increase in plasma noradrenaline concentrations, and the highest plasma levels of BNP and NT-proBNP, despite a similar degree of LV systolic dysfunction. No significant differences were found with regard to cortisol, PRA, aldosterone and thyroid profile (Table 3).

Table 3
Neurohormonal profile, functional capacity and ventilatory efficiency, and arrhythmias according to chemosensitivity

Values are means±S.E.M., except for categorical data which are expressed as numbers (percentage). *P< 0.05, **P<0.01 and ***P<0.001 compared with normal HVR and HCVR; †P<0.05 compared with increased HVR alone; and ‡P<0.05 compared with increased HCVR alone. HVR and HCVR indicate the chemosensitivity to hypoxia and hypercapnia respectively. fT3, free triiodothyronine; fT4, free tetraiodothyronine; TSH, thyroid-stimulating hormone. SDNN and SDANN were computed only in patients in sinus rhythm (n=34).

Normal HVR and HCVRIncreased HVRIncreased HCVRIncreased HVR and HCVR
Noradrenaline (ng/l) 427.1±69.9 619.1±84.5 621.7±97.9 689.0±72.3** 
BNP (ng/l) 102.4±20.6 270.6±77.6 314.2±75.1* 413.2±92.7*** 
NT-proBNP (ng/l) 832.7±195.5 1620.7±381.9 2945.5±713.6* 3035.0±787.3** 
Cortisol (ng/l) 155.4±9.2 155.8±24.8 196.1±40.0 184.9±15.7 
PRA (ng·ml−1·h−13.3±1.5 4.3±1.6 2.7±0.9 5.7±2.4 
Aldosterone (ng/l) 149.2±24.3 152.8±37.5 178.1±36.8 272.7±73.1 
fT3 (ng/l) 2.5±0.1 2.6±0.1 2.3±0.1 2.3±0.1 
fT4 (ng/l) 10.9±0.5 11.2±0.8 11.3±0.7 12.8±1.5 
TSH (μIU/ml) 2.1±0.2 2.2±0.6 2.1±0.4 2.2±0.3 
Peak V̇O2/kg of body weight (ml·min−1·kg−113.4±1.1 12.0±1.5 12.6±2.3 10.5±0.6 
V̇E/V̇CO2 slope 35.0±1.3 42.8±2.1 38.8±2.8 46.3±2.6*** 
Workload (W) 97.1±9.0 82.5±8.4 89.4±21.05 69.3±5.4 
SDANN (ms) 74.5±4.3 87.7±20.6 82.1±8.1 33.5±11.7*†‡ 
SDNN (ms) 95.4±4.7 99.7±18.7 90.6±13.9 71.8±28.4 
Atrial fibrillation (%) 14 16 62***†‡ 
NSVT (%) 20 38 58† 63** 
Normal HVR and HCVRIncreased HVRIncreased HCVRIncreased HVR and HCVR
Noradrenaline (ng/l) 427.1±69.9 619.1±84.5 621.7±97.9 689.0±72.3** 
BNP (ng/l) 102.4±20.6 270.6±77.6 314.2±75.1* 413.2±92.7*** 
NT-proBNP (ng/l) 832.7±195.5 1620.7±381.9 2945.5±713.6* 3035.0±787.3** 
Cortisol (ng/l) 155.4±9.2 155.8±24.8 196.1±40.0 184.9±15.7 
PRA (ng·ml−1·h−13.3±1.5 4.3±1.6 2.7±0.9 5.7±2.4 
Aldosterone (ng/l) 149.2±24.3 152.8±37.5 178.1±36.8 272.7±73.1 
fT3 (ng/l) 2.5±0.1 2.6±0.1 2.3±0.1 2.3±0.1 
fT4 (ng/l) 10.9±0.5 11.2±0.8 11.3±0.7 12.8±1.5 
TSH (μIU/ml) 2.1±0.2 2.2±0.6 2.1±0.4 2.2±0.3 
Peak V̇O2/kg of body weight (ml·min−1·kg−113.4±1.1 12.0±1.5 12.6±2.3 10.5±0.6 
V̇E/V̇CO2 slope 35.0±1.3 42.8±2.1 38.8±2.8 46.3±2.6*** 
Workload (W) 97.1±9.0 82.5±8.4 89.4±21.05 69.3±5.4 
SDANN (ms) 74.5±4.3 87.7±20.6 82.1±8.1 33.5±11.7*†‡ 
SDNN (ms) 95.4±4.7 99.7±18.7 90.6±13.9 71.8±28.4 
Atrial fibrillation (%) 14 16 62***†‡ 
NSVT (%) 20 38 58† 63** 

A significant relationship was found between HVR or HCVR and plasma noradrenaline level (R=0.40 and R=0.37 respectively, both P<0.01; Figure 1) and BNP expression (both R=0.40, P<0.01; Figure 1) and NT-proBNP expression (R=0.37 and R=0.41 respectively, both P<0.01; results not shown).

Relationship between enhanced chemosensitivity to hypoxia (a and c) and hypercapnia (b and d) with noradrenaline (a and b) and BNP (c and d)

Figure 1
Relationship between enhanced chemosensitivity to hypoxia (a and c) and hypercapnia (b and d) with noradrenaline (a and b) and BNP (c and d)

Values from all of the 60 subjects studied are included. HVR and HCVR indicate the chemosensitivity to hypoxia and hypercapnia respectively. NEPI, noradrenaline.

Figure 1
Relationship between enhanced chemosensitivity to hypoxia (a and c) and hypercapnia (b and d) with noradrenaline (a and b) and BNP (c and d)

Values from all of the 60 subjects studied are included. HVR and HCVR indicate the chemosensitivity to hypoxia and hypercapnia respectively. NEPI, noradrenaline.

Chemosensitivity and CSR

Daytime CSR was found in 17 patients (28%). All patients with HF with preserved chemoreceptor sensitivity had a normal breathing pattern, whereas daytime CSR occurrence increased progressively when enhanced HVR or HCVR were present and significantly (P<0.001) when a combined enhancement was detected (Figure 2). Using multivariate analysis, HCVR and the plasma BNP level were the only independent predictors of diurnal CSR (HCVR, β=9.2; BNP, β=3.5; P<0.05) among all of the univariate predictors (HVR, HCVR, V̇E/V̇CO2 slope, noradrenaline, BNP and NT-proBNP). The strong ability of HCVR and BNP plasma level in predicting the CSR occurrence was confirmed by ROC analysis (P<0.001 for both; Figure 3).

Prevalence of diurnal CSR and nocturnal AHI among patients with preserved chemoreflex and with isolated or combined enhancement in chemosensitivity to hypoxia and hypercapnia

Figure 2
Prevalence of diurnal CSR and nocturnal AHI among patients with preserved chemoreflex and with isolated or combined enhancement in chemosensitivity to hypoxia and hypercapnia

*P<0.001 and †P<0.001 compared with a normal chemoreflex. HVR and HCVR indicate the chemosensitivity to hypoxia and hypercapnia respectively.

Figure 2
Prevalence of diurnal CSR and nocturnal AHI among patients with preserved chemoreflex and with isolated or combined enhancement in chemosensitivity to hypoxia and hypercapnia

*P<0.001 and †P<0.001 compared with a normal chemoreflex. HVR and HCVR indicate the chemosensitivity to hypoxia and hypercapnia respectively.

ROC curve of the chemosensitivity to hypercapnia (a) and BNP (b) for the prediction of daytime CSR

Figure 3
ROC curve of the chemosensitivity to hypercapnia (a) and BNP (b) for the prediction of daytime CSR

HCVR indicates the chemosensitivity to hypercapnia.

Figure 3
ROC curve of the chemosensitivity to hypercapnia (a) and BNP (b) for the prediction of daytime CSR

HCVR indicates the chemosensitivity to hypercapnia.

With regards to nocturnal CSR, patients with combined enhancement of HVR and HCVR had the highest AHI during polysomnography (Figure 2). Moreover, a significant positive correlation between HVR/HCVR and AHI was found (R=0.57 and R=0.58 respectively, both P<0.001). Finally, HCVR (β=4.4, P<0.05) was the only independent predictor of night-time CSR (defined as an AHI >10) among all of the univariate predictors (HVR, HCVR, V̇E/V̇CO2 slope, noradrenaline, BNP and NT-proBNP), with significant findings at ROC curve analysis [AUC, 0.87±0.04; cut-off value, 0.76 litre·min−1·mmHg−1; sensitivity, 81.1%; and specificity, 73.2% (P<0.001)].

Chemosensitivity and response to exercise

Patients with enhanced chemosensitivity both to hypoxia and hypercapnia had lower ventilatory efficiency, as expressed by the V̇E/V̇CO2 slope during CPT, with a non-significant trend towards lower peak V̇O2 and maximum exercise workload (Table 2). Moreover, the degree of ventilatory inefficiency was related to the level of both HVR (chemosensitivity to hypoxia; R=0.42, P<0.01) and HCVR (chemosensitivity to hypercapnia; R=0.50, P<0.001) (Figure 4).

Relationship between enhanced chemosensitivity to hypoxia (a and c) and hypercapnia (b and d) with peak V̇O2 (a and b) and V̇E/V̇CO2 slope (c and d)

Figure 4
Relationship between enhanced chemosensitivity to hypoxia (a and c) and hypercapnia (b and d) with peak V̇O2 (a and b) and V̇E/V̇CO2 slope (c and d)

Values from all of the 60 subjects studied are included. HVR and HCVR indicate the chemosensitivity to hypoxia and hypercapnia respectively. VO2/kg represents peak V̇O2. NS, not significant.

Figure 4
Relationship between enhanced chemosensitivity to hypoxia (a and c) and hypercapnia (b and d) with peak V̇O2 (a and b) and V̇E/V̇CO2 slope (c and d)

Values from all of the 60 subjects studied are included. HVR and HCVR indicate the chemosensitivity to hypoxia and hypercapnia respectively. VO2/kg represents peak V̇O2. NS, not significant.

Chemosensitivity, arrhythmias and heart rate variability

A total of 15 patients (25%) had chronic atrial fibrillation during a 24-h ECG recording, where 25 (42%) had at least one episode of NSVT (non-sustained ventricular tachycardia; ≥3 consecutive ventricular complexes at a rate of more than 100 beats/min, lasting for less than 30 s). Patients with a combined enhancement of HVR and HCVR (i.e. chemosensitivity to hypoxia and hypercapnia) had a higher incidence of both atrial fibrillation (P<0.01) and NSVT (P<0.001) when compared with patients with normal chemosensitivity (Table 3), in spite of similar pharmacological treatment.

Furthermore, we also analysed the heart rate variability in patients in sinus rhythm (n=34). Patients with combined enhancement of HVR and HCVR had a significant decrease in SDANN, among time domain heart rate variability, with a non-significant trend for SDNN, probably due to the small sample size (as shown in Table 3). There were no differences between the groups with regard to both RMSSD and pNN50 (results not shown).

DISCUSSION

The results of the present study show that increased chemosensitivity is frequent in mild-to-moderate chronic HF, regardless of the optimized medical treatment [including β-blockade, anti-aldosterone drugs and ACE (angiotensin-converting enzyme)/angiotensin II inhibition]. Moreover, increased chemosensitivity (particularly when combined with both hypoxia and hypercapnia) enhances sympathetic activation and is associated with depressed heart rate variability and increased plasma concentrations of BNP and NT-proBNP. Finally, it is associated with the occurrence of respiratory abnormalities, such as altered ventilatory response to exercise and either daytime or night-time CSR, with an increased incidence of atrial fibrillation and ventricular arrhythmias and with worse clinical status, independently of the degree of LV systolic dysfunction.

Isolated or combined enhancement of chemosensitivity to hypoxia and hypercapnia was present in 60% of our patients, in spite of optimal treatment. When considered separately, the overall prevalence of increased chemosensitivity to hypoxia (40%) is similar to that reported previously [12]. On the other hand, increased chemosensitivity to hypercapnia occurred less frequently in our present study (47%) when compared with previous observations (76%) in patients with mild systolic HF [18]. This might be explained by a blunting effect of β-blocker administration (more frequent in our present study, 92 compared with 17% of patients) on the sympathetic efferent arm, with a possible prevalent effect on the central chemoreflex.

With regard to neurohormonal derangement, our present findings support the hypothesis that chronic enhancement of chemosensitivity might play a central role in sustaining adrenergic activation, possibly by the direct stimulation of sympathetic centres from chemoreceptors [8,19] or by inducing CSR-related periodic hypoxaemia [20,21]. A role of the chemoreflex in eliciting adrenergic activation in HF, beyond altered haemodynamics [22], is confirmed by its persistent action on the sympathetic axis in patients following cardiac transplantation [23]. Furthermore, enhanced chemosensitivity also appears to be associated with abnormal heart rate variability, indicating an autonomic imbalance at the sinus node level.

To our knowledge, the present study is the first in which the effect of the chemoreflex on cardiac endocrine function has been observed in humans. We found that patients with combined enhancement of chemosensitivity were characterized by higher plasma levels of BNP and NT-proBNP compared with patients with normal chemosensitivity, despite a similar degree of LV systolic dysfunction. In this subset, beyond haemodynamic stress, the production and secretion of natriuretic peptides could be elicited by a higher sympathetic drive and chronic periodical occurrence of hypoxia during CSR [24,25], two recognized additional stimuli for the release of natriuretic peptides [26]. The plasma concentration of natriuretic peptides could be confirmed not merely as a marker of myocardial dysfunction, but rather as an index of overall neurohormonal activation, contributing to its diagnostic and prognostic value in HF.

The observation made by Cheyne [27] and Stokes [28] two centuries ago on periodic breathing during the daytime in awake patients with HF has been confirmed and associated with sleep-time respiratory abnormalities [29,30]. Daytime [31,32] and night-time [33,34] CSR have been associated with poor prognosis in patients with HF. Increased chemosensitivity, in particular to hypercapnia, has been suggested to play a central pathophysiological role in the onset of CSR [10,11,35,36], together with altered haemodynamics (or delayed circulatory time) [37]. Our present results support this hypothesis, extending this concept to the daytime period. Indeed, all patients with normal chemosensitivity had a normal breathing pattern, whereas there was a progressive increase in the occurrence of CSR from patients with isolated to those with combined enhancement of chemosensitivity.

Using multivariate analysis, increased chemosensitivity to hypercapnia was an independent predictor of both daytime and night-time CSR, suggesting that CO2 changes could be more relevant than oxygen variations in the genesis of CSR, as proposed previously [38,39]. This provides further support for the concept that the two entities share a common pathophysiological basis. The BNP level was an independent predictor only of daytime CSR, suggesting a stronger influence of haemodynamic or neurohormonal factors during wakefulness.

Combined enhancement of the hypoxic and hypercapnic ventilatory responses was associated with an increased ventilatory response to exercise, as expressed by the V̇E/V̇CO2 slope, a recognized prognostic marker in HF [40,41]. The level of chemosensitivity activation was inversely related to ventilatory efficiency, thus confirming the powerful influence of chemoreceptors on ventilation during exercise [9,41,42]. Indeed, patients with an altered chemoreflex had a worse clinical status, as expressed by a higher NYHA class, independently of the degree of LV dysfunction. The lack of a significant association between the chemoreflex and peak V̇O2 in the present study suggests a secondary role of the chemoreflex on functional capacity on effort, mainly dependent upon altered haemodynamics [43] and loss of skeletal muscle mass and function [44].

Finally, the combined enhancement of HVR and HCVR was associated with a higher incidence of both chronic atrial fibrillation and NSVTs during 24-h ECG recording, probably related to the increased sympathetic tone caused by the reflex alteration, supporting previous observations [45].

In conclusion, despite optimal pharmacological treatment, increased chemosensitivity to hypoxia and hypercapnia influences several pathophysiological pathways associated with disease progression in HF, such as neurohormonal activation, control of respiration at rest and during effort, and ventricular arrhythmias. Hence we suggest that the assessment of chemoreflex sensitivity may contribute to the diagnostic definition of high-risk patients with HF [12] and that treating an abnormal chemoreflex, either by optimization of conventional treatment or by specific intervention [42,46], could be considered a novel therapeutic target in HF.

Abbreviations

     
  • ACE

    angiotensin-converting enzyme

  •  
  • AHI

    apnoea/hypopnoea index

  •  
  • AUC

    area under the curve

  •  
  • BMI

    body mass index

  •  
  • BNP

    B-type natriuretic peptide

  •  
  • CPT

    cardiopulmonary exercise testing

  •  
  • CSR

    Cheyne–Stokes respiration

  •  
  • FEV1

    forced expiratory volume in 1 s

  •  
  • FVC

    forced vital capacity

  •  
  • HCVR

    hypercapnic ventilatory response

  •  
  • HF

    heart failure

  •  
  • HVR

    hypoxic ventilatory response

  •  
  • LV

    left ventricular

  •  
  • LVEF

    LV ejection fraction

  •  
  • NSVT

    non-sustained ventricular tachycardia

  •  
  • NT-proBNP

    N-terminal pro-BNP

  •  
  • NYHA

    New York Heart Association

  •  
  • PaO2

    arterial partial pressure of oxygen

  •  
  • pNN50

    number of adjacent RR intervals differing by more than 50 ms, as a percentage of the total number of RR intervals

  •  
  • PRA

    plasma renin activity

  •  
  • ROC

    receiver operating characteristic

  •  
  • SaO2

    arterial oxygen saturation

  •  
  • SDANN

    S.D. of 5 min mean values of RR intervals

  •  
  • SDNN

    S.D. of all RR intervals

  •  
  • V̇CO2

    carbon dioxide uptake

  •  
  • V̇E

    minute ventilation

  •  
  • V̇O2

    oxygen uptake

  •  
  • V̇T

    tidal volume

We are grateful to Fabio Micheletti and Mauro Micalizzi for their technical support.

References

References
1
The Task Force for the Diagnosis and Treatment of Chronic Heart Failure of the European Society of Cardiology
Guidelines for the diagnosis and treatment of chronic heart failure
Eur. Heart J.
2005
, vol. 
26
 (pg. 
1115
-
1140
)
2
Packer
 
M.
 
The neurohormonal hypothesis: a theory to explain the mechanism of disease progression in heart failure
J. Am. Coll. Cardiol.
1992
, vol. 
20
 (pg. 
248
-
254
)
3
Floras
 
J. S.
 
Clinical aspects of sympathetic activation and parasympathetic withdrawal in heart failure
J. Am. Coll. Cardiol.
1993
, vol. 
22
 
Suppl.
(pg. 
72A
-
84A
)
4
Francis
 
G. S.
Benedict
 
C.
Johnstone
 
D. E.
, et al 
Comparison of neuroendocrine activation in patients with left ventricular dysfunction with and without congestive heart failure. A substudy of the Studies of Left Ventricular Dysfunction (SOLVD)
Circulation
1990
, vol. 
82
 (pg. 
1724
-
1729
)
5
Clerico
 
A.
Emdin
 
M.
 
Diagnostic accuracy and prognostic relevance of the measurement of the cardiac natriuretic peptides: a review
Clin. Chem.
2004
, vol. 
50
 (pg. 
33
-
50
)
6
La Rovere
 
M. T.
Pinna
 
G. D.
Hohnloser
 
S. H.
, et al 
Autonomic tone and reflexes after myocardial infarction. Baroreflex sensitivity and heart rate variability in the identification of patients at risk for life-threatening arrhythmias: implications for clinical trials
Circulation
2001
, vol. 
103
 (pg. 
2072
-
2077
)
7
Piepoli
 
M.
Clark
 
A. L.
Volterrani
 
M.
Adamopoulos
 
S.
Sleight
 
P.
Coats
 
A. J.
 
Contribution of muscle afferents to the hemodynamic, autonomic, and ventilatory responses to exercise in patients with chronic heart failure: effects of physical training
Circulation
1996
, vol. 
93
 (pg. 
940
-
952
)
8
Narkiewicz
 
K.
Pesek
 
C. A.
van de Borne
 
P. J.
Kato
 
M.
Somers
 
V. K.
 
Enhanced sympathetic and ventilatory responses to central chemoreflex activation in heart failure
Circulation
1999
, vol. 
100
 (pg. 
262
-
267
)
9
Chua
 
T. P.
Clark
 
A. L.
Amadi
 
A. A.
Coats
 
A. J.
 
The relationship between chemosensitivity and the ventilatory response to exercise in chronic heart failure
J. Am. Coll. Cardiol.
1996
, vol. 
27
 (pg. 
650
-
657
)
10
Javaheri
 
S.
 
A mechanism of central sleep apnea in patients with heart failure
N. Engl. J. Med.
1999
, vol. 
341
 (pg. 
949
-
954
)
11
Solin
 
P.
Roebuck
 
T.
Johns
 
D. P
Walters
 
E. H.
Naughton
 
M. T.
 
Peripheral and central ventilatory responses in central sleep apnea with and without congestive heart failure
Am. J. Respir. Crit. Care Med.
2000
, vol. 
162
 (pg. 
2194
-
2200
)
12
Ponikowski
 
P.
Chua
 
T. P.
Anker
 
S. D.
, et al 
Peripheral chemoreceptor hypersensitivity: an ominous sign in patients with chronic heart failure
Circulation
2001
, vol. 
104
 (pg. 
544
-
549
)
13
Read
 
D. J.
 
A clinical method for assessing the ventilatory response to carbon dioxide
Australas. Ann. Med.
1967
, vol. 
16
 (pg. 
20
-
32
)
14
Rebuck
 
A. S.
Campbell
 
E. J.
 
A clinical method for assessing the ventilatory response to hypoxia
Am. Rev. Respir. Dis.
1974
, vol. 
109
 (pg. 
345
-
350
)
15
Emdin
 
M.
Passino
 
C.
Prontera
 
C.
, et al 
Cardiac natriuretic hormones, neuro-hormones, thyroid hormones and cytokines in normal subjects and patients with heart failure
Clin. Chem. Lab. Med.
2004
, vol. 
42
 (pg. 
627
-
636
)
16
Mortara
 
A.
Sleight
 
P.
Pinna
 
G. D.
, et al 
Abnormal awake respiratory patterns are common in chronic heart failure and may prevent evaluation of autonomic tone by measures of heart rate variability
Circulation
1997
, vol. 
96
 (pg. 
246
-
252
)
17
Yamashiro
 
Y.
Kryger
 
M. H.
 
Sleep in heart failure
Sleep
1993
, vol. 
16
 (pg. 
513
-
523
)
18
Yamada
 
K.
Asanoi
 
H.
Ueno
 
H.
, et al 
Role of central sympathoexcitation in enhanced hypercapnic chemosensitivity in patients with heart failure
Am. Heart J.
2004
, vol. 
148
 (pg. 
964
-
970
)
19
Morgan
 
B. J.
Crabtree
 
D. C.
Palta
 
M.
Skatrud
 
J. B.
 
Combined hypoxia and hypercapnia evokes long-lasting sympathetic activation in humans
J. Appl. Physiol.
1995
, vol. 
79
 (pg. 
205
-
213
)
20
Naughton
 
M. T.
Benard
 
D. C.
Liu
 
P. P.
Rutherford
 
R.
Rankin
 
F.
Bradley
 
T. D.
 
Effects of nasal CPAP on sympathetic activity in patients with heart failure and central sleep apnea
Am. J. Respir. Crit. Care Med.
1995
, vol. 
152
 (pg. 
473
-
479
)
21
van de Borne
 
P.
Oren
 
R.
Abouassaly
 
C.
Anderson
 
E.
Somers
 
V. K.
 
Effect of Cheyne-Stokes respiration on muscle sympathetic nerve activity in severe congestive heart failure secondary to ischemic or idiopathic dilated cardiomyopathy
Am. J. Cardiol.
1998
, vol. 
81
 (pg. 
432
-
436
)
22
Mansfield
 
D.
Kaye
 
D. M.
Brunner La Rocca
 
H.
Solin
 
P.
Esler
 
M. D.
Naughton
 
M. T.
 
Raised sympathetic nerve activity in heart failure and central sleep apnea is due to heart failure severity
Circulation
2003
, vol. 
107
 (pg. 
1396
-
1400
)
23
Ciarka
 
A.
Cuylits
 
N.
Vachiery
 
J. L.
, et al 
Increased peripheral chemoreceptors sensitivity and exercise ventilation in heart transplant recipients
Circulation
2006
, vol. 
113
 (pg. 
252
-
257
)
24
Carmona-Bernal
 
C.
Quintana-Gallego
 
E.
Villa-Gil
 
M.
Sanchez-Armengol
 
A.
Martinez-Martinez
 
A.
Capote
 
F.
 
Brain natriuretic peptide in patients with congestive heart failure and central sleep apnea
Chest
2005
, vol. 
127
 (pg. 
1667
-
1673
)
25
Pepperell
 
J. C.
Maskell
 
N. A.
Jones
 
D. R.
, et al 
A randomized controlled trial of adaptive ventilation for Cheyne-Stokes breathing in heart failure
Am. J. Respir. Crit. Care Med.
2003
, vol. 
168
 (pg. 
1109
-
1114
)
26
Clerico
 
A.
Recchia
 
F. A.
Passino
 
C.
Emdin
 
M.
 
Cardiac endocrine function is an essential component of the homeostatic regulation network: physiological and clinical implications
Am. J. Physiol. Heart Circ. Physiol.
2006
, vol. 
290
 (pg. 
H17
-
H29
)
27
Cheyne
 
J.
 
A case of apoplexy, in which the fleshy part of the heart was converted into fat
Dublin Hosp. Rep.
1818
, vol. 
2
 (pg. 
216
-
223
)
28
Stokes
 
W.
 
Observations on some cases of permanently slow pulse
Dublin Q. J. Med. Sci.
1846
, vol. 
2
 (pg. 
73
-
83
)
29
Sin
 
D. D.
Fitzgerald
 
F.
Parker
 
J. D.
Newton
 
G.
Floras
 
J. S.
Bradley
 
T. D.
 
Risk factors for central and obstructive sleep apnea in 450 men and women with congestive heart failure
Am. J. Respir. Crit. Care Med.
1999
, vol. 
160
 (pg. 
1101
-
1106
)
30
Javaheri
 
S.
 
Sleep disorders in systolic heart failure: a prospective study of 100 male patients
The final report. Int. J. Cardiol.
2006
, vol. 
106
 (pg. 
21
-
28
)
31
Ponikowski
 
P.
Anker
 
S. D.
Chua
 
T. P.
, et al 
Oscillatory breathing patterns during wakefulness in patients with chronic heart failure: clinical implications and role of augmented peripheral chemosensitivity
Circulation
1999
, vol. 
100
 (pg. 
2418
-
2424
)
32
La Rovere
 
M. T.
Pinna
 
G. D.
Maestri
 
R.
, et al 
Clinical relevance of short-term day-time breathing disorders in chronic heart failure patients
Eur. J. Heart Failure
2007
, vol. 
9
 (pg. 
949
-
954
)
33
Lanfranchi
 
P. A.
Braghiroli
 
A.
Bosimini
 
E.
, et al 
Prognostic value of nocturnal Cheyne-Stokes respiration in chronic heart failure
Circulation
1999
, vol. 
99
 (pg. 
1435
-
1440
)
34
Hanly
 
P. J.
Zuberi-Khokhar
 
N. S.
 
Increased mortality associated with Cheyne-Stokes respiration in patients with congestive heart failure
Am. J. Respir. Crit. Care Med.
1996
, vol. 
153
 (pg. 
272
-
276
)
35
Cherniack
 
N. S.
Longobardo
 
G. S.
 
Cherniack
 
N. S.
Widdicombe
 
J. G.
 
Abnormalities in respiratory rhythm. Handbook of Physiology, vol. 2: The Respiratory System
Control of Breathing
1986
Bethesda, MD
American Physiological Society
(pg. 
729
-
749
)
36
Francis
 
D. P.
Willson
 
K.
Davies
 
L. C.
Coats
 
A. J.
Piepoli
 
M.
 
Quantitative general theory for periodic breathing in chronic heart failure and its clinical implications
Circulation
2000
, vol. 
102
 (pg. 
2214
-
2221
)
37
Hall
 
M. J.
Xie
 
A.
Rutherford
 
R.
Ando
 
S.
Floras
 
J. S.
Bradley
 
T. D.
 
Cycle length of periodic breathing in patients with and without heart failure
Am. J. Respir. Crit. Care Med.
1996
, vol. 
154
 (pg. 
376
-
381
)
38
Duffin
 
J.
 
The chemoreflex control of breathing and its measurement
Can. J. Anaesth.
1990
, vol. 
37
 (pg. 
933
-
942
)
39
Mohan
 
R.
Duffin
 
J.
 
The effect of hypoxia on the ventilatory response to carbon dioxide in man
Respir. Physiol.
1997
, vol. 
108
 (pg. 
101
-
115
)
40
Francis
 
D. P.
Shamim
 
W.
Davies
 
L. C.
, et al 
Cardiopulmonary exercise testing for prognosis in chronic heart failure: continuous and independent prognostic value from V̇E/V̇CO2 slope and peak V̇O2
Eur. Heart J.
2000
, vol. 
21
 (pg. 
154
-
161
)
41
Ponikowski
 
P.
Francis
 
D. P.
Piepoli
 
M. F.
, et al 
Enhanced ventilatory response to exercise in patients with chronic heart failure and preserved exercise tolerance: marker of abnormal cardiorespiratory reflex control and predictor of poor prognosis
Circulation
2001
, vol. 
103
 (pg. 
967
-
972
)
42
Chua
 
T. P.
Harrington
 
D.
Ponikowski
 
P.
Webb-Peploe
 
K.
Poole-Wilson
 
P. A.
Coats
 
A. J.
 
Effects of dihydrocodeine on chemosensitivity and exercise tolerance in patients with chronic heart failure
J. Am. Coll. Cardiol.
1997
, vol. 
29
 (pg. 
147
-
152
)
43
Wasserman
 
K.
 
Measures of functional capacity in patients with heart failure
Circulation
1990
, vol. 
81
 
Suppl. II
(pg. 
1
-
4
)
44
Piepoli
 
M. F.
Kaczmarek
 
A.
Francis
 
D. P.
, et al 
Reduced peripheral skeletal muscle mass and abnormal reflex physiology in chronic heart failure
Circulation
2006
, vol. 
114
 (pg. 
126
-
134
)
45
Chua
 
T. P.
Ponikowski
 
P.
Webb-Peploe
 
K.
, et al 
Clinical characteristics of chronic heart failure patients with an augmented peripheral chemoreflex
Eur. Heart J.
1997
, vol. 
18
 (pg. 
480
-
486
)
46
Spicuzza
 
L.
Gabutti
 
A.
Porta
 
C.
Montano
 
N.
Bernardi
 
L.
 
Yoga and chemoreflex response to hypoxia and hypercapnia
Lancet
2000
, vol. 
356
 (pg. 
1495
-
1496
)