International Journal of Cardiology 189 (2015) 134–140
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Impact of chronic obstructive pulmonary disease on exercise ventilatory efficiency in heart failure☆ Anna Apostolo a, PierAntonio Laveneziana b,c,d,e, Paolo Palange f, Cecilia Agalbato a, Roberta Molle a, Dejana Popovic g, Maurizio Bussotti h, Mattia Internullo f, Susanna Sciomer i, Matteo Bonini f, Maria Clara Alencar j, Laurent Godinas b,k,l, Flavio Arbex j, Gilles Garcia b,k,l, J. Alberto Neder j,m, Piergiuseppe Agostoni a,n,⁎ a
Centro Cardiologico Monzino, IRCCS, Milano, Italy AP-HP, Hôpital Universitaire de Bicêtre, Service d'Explorations Fonctionnelles Respiratoires, Centre de Référence de l'Hypertension Pulmonaire Sévère, DHU TORINO “Thorax Innovation”, Le Kremlin-Bicêtre, France c Sorbonne Universités, UPMC Univ. Paris 06, UMR_S 1158, Neurophysiologie Respiratoire Expérimentale et Clinique, Paris, France d INSERM, UMR_S 1158, Neurophysiologie Respiratoire Expérimentale et Clinique, Paris, France e AP-HP, Groupe Hospitalier Pitié-Salpêtrière Charles Foix, Service des Explorations Fonctionnelles de la Respirationde l'Exercice et de la Dyspnée, Paris, France f Dipartimento di Medicina Clinica, University La Sapienza, Viale Università 37, Rome, Italy g Division of Cardiology, Faculty of Medicine, University of Belgrade, Visegradska 26, 11000 Belgrade, Serbia h Cardiac Rehabilitation Unit, Fondazione Salvatore Maugeri, IRCCS, Scientific Institute of Milan, Italy i Dipartimento di Scienze Cardiovascolari, Respiratorie, Anestesiologiche, Nefrologiche e Geriatriche, “La Sapienza”, Rome,Italy j Respiratory Division, Dept. of Medicine, Federal University of São Paulo — Paulista School of Medicine (UNIFESP-EPM), Brazil k Univ. Paris-Sud 11, Faculté de médecine, Le Kremlin-Bicêtre, France l INSERM U999, LabEx LERMIT, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France m Division of Respiratory and Critical Care Medicine, Faculty of Health Sciences, Queen's University, Kingston, ON, Canada n Department of Clinical Sciences and Community Health, Cardiovascular Section, University of Milan, Italy b
a r t i c l e
i n f o
a b s t r a c t
Article history: Received 2 December 2014 Received in revised form 13 March 2015 Accepted 30 March 2015 Available online 31 March 2015
Background: Heart failure (HF) and chronic obstructive pulmonary disease (COPD) coexistence increases morbidity and mortality. The intercept of ventilation (VEint) on the VE vs. carbon dioxide production (VCO2) relationship during exercise has been found to vary in proportion with dead space (VD) in HF. Considering that increased VD is the key pathophysiological abnormality in COPD but a secondary finding in HF we hypothesized that a high VEint would be useful in suggesting COPD as HF co-morbidity. Our aim was to assess whether an elevated VEint suggests the presence of COPD in HF. Methods: In a multicenter retrospective study, the VE–VCO2 relationship was analyzed both as slope and intercept in HF (n = 108), HF–COPD (n = 106) and COPD (n = 95). Patients with pulmonary arterial hypertension (PAH) (n = 85) and healthy subjects (HF) (n = 56) served as positive and negative controls relative to VE–VCO2 abnormalities, respectively. Results: Slope and VEint varied in opposite directions in all groups (p b 0.05) being VE–VCO2 slope highest and lowest in PAH and healthy subjects, respectively. No slope differences were observed among HF, HF–COPD and COPD (32 ± 7, 31 ± 7, and 31 ± 6, respectively). VEint was higher in HF–COPD and COPD compared to HF, PAH and controls (4.8 ± 2.4 L/min, 5.9 ± 3.0 L/min, 3.0 ± 2.6 L/min, 2.3 ± 3.3 L/min and 3.9 ± 2.5 L/min, respectively; p b 0.01). A VEint ≥4.07 L/min identified patients with high probability of having COPD or HF–COPD (sensitivity of 71.6% and specificity of 72.0%). Conclusion: These data provide novel evidence that a high VEint (≥ 4.07 L/min) should be valued to suggest coexistent COPD in HF patients. © 2015 Published by Elsevier Ireland Ltd.
Keywords: Ventilatory efficiency Exercise Dead space Heart failure Lung diseases
Abbreviations:VE, linear ventilation;VCO2, carbon dioxide output; COPD, chronic obstructive pulmonary disease; HF, heart failure;VEint, intercept of the linear ventilation; VD, dead space ventilation; HF–COPD, heart failure and chronic obstructive pulmonary disease; PAH, pulmonary arterial hypertension; NYHA, New York Heart Association; CPET, cardiopulmonary exercise test; VC, lung vital capacity; FVC, forced vital capacity; FEV1, forced expiratory volume in 1 s; VO2, oxygen uptake. ☆ All authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation. ⁎ Corresponding author at: Centro Cardiologico Monzino, IRCCS, Department of Clinical Sciences and Community Health, Cardiovascular Section, University of Milan, Via Parea, 4, 20138 Milano, Italy. E-mail address: [email protected] (P. Agostoni).
http://dx.doi.org/10.1016/j.ijcard.2015.03.422 0167-5273/© 2015 Published by Elsevier Ireland Ltd.
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
1. Introduction Heart failure (HF) with reduced left ventricular ejection fraction and chronic obstructive pulmonary disease (COPD) are chronic-degenerative diseases frequently found in the general population. Both provide a poor prognosis, both have a negative influence on patients' quality of life, and both are the causes of relevant costs for the health care system [1]. Unfortunately, HF and COPD often coexist, leading to further increases in patients' disability and mortality [2,3]. Identification of COPD in patients with established HF, however, might be troublesome in individual patients as the diseases share the same symptoms of dyspnea and fatigability on exertion and resting lung function abnormalities may occur in HF on isolation [4–6]. This state of affairs explains the clinical relevance of identifying novel functional indexes able to indicate the presence of COPD as co-morbidity of HF [1,2]. It is noteworthy that during exercise
in both HF and COPD more ventilation (VE) is needed to meet the peripheral metabolic demands (as expressed, for example, by changes in carbon
dioxide output (VCO2)). Interestingly, while in HF this mainly reflects the
functional adaptation of VE to a chronically-increased sympathetic tonus
[7,8], the VE–VCO2 relationship in COPD is strongly modulated by the
extent at whichVE is “wasted” in the dead space (VD) [9]. From a practical
perspective, this excessive exerciseVE has been traditionally quantified by
the slope of the linear VE–VCO2 relationship [10,11]. Notably, however, whereas milder ventilatory abnormalities do not preclude HF patients in meeting the increased ventilatory demands (i.e., the slope does increase as disease progresses) [12,13], mechanical constraints preclude or restrict COPD patients to attend those requirements (i.e., the slope diminishes as disease worsens) [14,15]. The opposite effects of HF and COPD on the slope, therefore, make it unlikely that this parameter would be of value in separating patients with HF from those with HF–COPD overlap. There is, however, another VE–VCO2 parameter that may carry important information relative to the magnitude of VD increase whereas
not being constrained by lung mechanical abnormalities — the VE inter
cept (V Eint) [16,17]. V Eint represents the ventilatory requirements
when pulmonary gas exchange is nil (VCO2 = 0), i.e., the very definition
of VD. In fact, as normal lungs have a small VD, V Eint has a positive value in more than 95% of healthy subjects [11,18,19]. Of note, we
showed that VEint increased in tandem with VD when the latter was artificially increased during exercise in both HF and healthy subjects [20].
Teopompi and colleagues [21] described significantly higher VEint in patients with COPD compared to their counterparts with HF despite similar maximal exercise capacity. Moreover, Neder and colleagues recently
showed that increases in VEint better reflected the progression of functional impairment from mild to end-stage COPD [9]. Altogether, this
135
specifically, Centro Cardiologico Monzino, IRCCS, Milan, Italy; Cardiologia Riabilitativa, Fondazione S Maugeri, IRCCS, Milan, Italy; Department of Public Health and Infectious Diseases, Division of Pulmonary Research, “La Sapienza” University, Rome, Italy; Université de Paris 06, Equipe de Recherche ER 10 UPMC, Laboratoire de Physio-Pathologie Respiratoire, Faculté de Médecine Pierre et Marie Curie (site Pitié-Salpêtrière), Paris, France; Division of Respiratory Diseases, Department of Medicine, Federal University of Sao Paulo (UNIFESP), Sao Paulo Brazil; and Clinical Center Serbia, Cardiology Department, Medical Faculty University of Belgrade. All studied patients were well known at the referring study center, and all were in stable clinical conditions under optimal medical treatment. Patients were recruited consecutively in each center, but only data of patients with an arbitrarily judged high quality and maximal effort test were considered. Specifically, the following criteria for definition of maximal tests were used: exercise was interrupted by the patients when they stated they had performed a maximal effort, test surveillance medical personnel confirming the maximal effort of the patients, peak exercise RQ N1.05. Moreover, we considered only tests interrupted because of dyspnea or fatigue or both, in the absence of cardiac arrhythmias limiting exercise performance and ECG signs of severe ischemia. Finally, we excluded tests with an irregular breathing pattern which makes exercise evaluation unclear. For this analysis, subjects were divided into five groups according to clinical and laboratory diagnoses (n = 450). Patients with HF, HF–COPD and COPD were recruited in all participating centers. Vice versa, PAH patients were only recruited at Centro Cardiologico Monzino, Milan, at “La Sapienza” University Roma and at Université de Paris, while normal subjects were all recruited at Centro Cardiologico Monzino. a) HF group (n = 108): New York Heart Association (NYHA) classes I to III, echocardiographic evidence of reduced left ventricular ejection fraction: ejection fraction ≤40% according to Simpson rule, with no evidence of COPD according to clinical history and spirometry. b) HF–COPD group (n = 106): HF patients according to the abovementioned characteristics and evidence of COPD (clinical history and spirometric findings). c) COPD group (N = 95): evidence of COPD (as indicated above) without clinical/echocardiographic evidence of HF. d) PAH (n = 85): class 1 and 4 patients [24] showing mean pulmonary pressure ≥25 mm Hg and normal wedge pressure as documented by right heart catheterization with normal left ventricular function and dimensions (left ventricular ejection fraction N55%, left ventricular end-diastolic volume b75 mL/m2, left ventricular end systolic volume b30 mL/m2). e) Healthy subjects (n = 56): subjects with no history and/or clinical evidence of any cardiovascular or pulmonary or systemic disease or any condition requiring daily medications.
preliminary evidence led us to hypothesize whether increased VEint would be particularly helpful in suggesting COPD as co-morbidity of HF. This large scale, multicenter study was therefore undertaken to con
trast the parameters of the VE–VCO2 relationship (VEint and slope) in patients with HF, COPD and HF–COPD overlap. Normal subjects and patients with pulmonary artery hypertension (PAH) served as negative and positive controls relative to V E–V CO2 abnormalities [11,22,23]. We specifically hypothesized that increases in VEint – but not changes in V E– V CO2 slope – would be useful in indicating the presence of COPD in patients with HF.
2. Methods 2.1. Study design and population This is a multicenter retrospective study involving 450 patients followed in 6 Centers where CPETs are performed by experts —
2.2. Measurements Anthropometric parameters were recorded for all included patients. All subjects underwent standard spirometry and cardiopulmonary exercise test (CPET) (Centro Cardiologico Monzino and Cardiologia Riabilitativa, Fondazione S Maugeri, Milan: Vmax 29C, SensorMedics, Yorba Linda, CA, US; Department of Public Health and Infectious Diseases, Division of Pulmonary Research, “La Sapienza” University, Rome, Italy: COSMED Quark CPET, Rome, Italy; Université de Paris 06, Equipe de Recherche ER 10 UPMC, Laboratoire de PhysioPathologie Respiratoire, Faculté de Médecine Pierre et Marie Curie (site Pitié-Salpêtrière), Paris, France: Vmax 229d with Autobox 6,200 DL, SensorMedics, Yorba Linda, CA; Division of Respiratory Diseases, Department of Medicine, Federal University of Sao Paulo (UNIFESP), Sao Paulo Brazil and Clinical Center Serbia, Cardiology Department, Medical Faculty University of Belgrade: Cardiovit CS200 Ergospiro Schiller Baar, Switzerland). Pulmonary function tests
136
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
were performed according to the recommended technique [25]. Measurements are reported as a percentage of predicted normal values. At spirometry, we measured lung vital capacity (VC), forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1). According to guidelines, we considered as normal FEV1 ≥80% of predicted, and FEV1/FVC N0.70. All CPETs were performed on a cycle-ergometer. Patients were instructed to pedal at the speed of 60 rpm. In all subjects, we applied a progressively increasing workload (ramp protocol). Physicians personalized the workload for every individual according to the information about the exercise capacity self-reported by patients or, whenever possible, considering the results of a previous CPET. The aim was to achieve an active loaded exercise duration of around 10 min; CPETs were selfterminated by the subjects when they claimed that they had reached maximal effort, regardless of the respiratory quotient reached. During the exercise test, twelve-lead ECG, blood pressure, and arterial oxygen saturation were monitored. Oxygen uptake (VO2), V E and V CO2 were measured breath by breath. V O2, V E and V CO2 data are reported as 20-second average. Linear regression was applied to V E– V CO2 relationship from 1 min after the beginning of loaded pedaling to the end of the isocapnic buffering period [17,18]. Slope and VEint (L/min) were recorded. Centro Cardiologico Monzino was responsible for data collection, while individual investigators were responsible for their own records.
Inverse correlation between VE–VCO2 slope and VEint.
Entire population HF HF–COPD COPD PAH HEALTHY
Spearman correlation
p value
−0.41 −0.32 −0.46 −0.37 −0.63 −0.50
b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001
HF = heart failure patients; HF–COPD = heart failure patients with coexisting chronic obstructive pulmonary disease; COPD = patients with chronic obstructive pulmonary disease; PAH = pulmonary arterial hypertension patients; HEALTHY = healthy subjects.
analysis was performed by using IBM SPSS Statistics 20 and SAS v. 9.2 (SAS Institute Inc., Cary, NC, USA).
2.3. Statistical analysis Continuous variables are expressed as mean ± standard deviation (SD) and compared using ANOVA, and unpaired t-test was used for group comparison. Categorical variables are expressed as frequencies and percentages and compared using chi-square test. The normality of data distribution was assessed using the Shapiro–Wilkins test. Skewed distributed variables were reported as median and interquartile
range and compared by Kruskal Wallis test. Association between VEint
Table 2
3. Results 3.1. Resting variables Descriptive statistics are reported in Table 1. HF, COPD and HF–COPD had similar age while PAH patients and healthy controls were younger. PAH patients were more frequently female. Left ventricular ejection fraction was 34 ± 7% and 32 ± 8% in HF–COPD and HF, respectively. As expected, COPD and HF–COPD patients showed evidence of an obstructive ventilatory defect, which, based on both FEV1 and FEV1/FVC, was more severe in COPD patients. Indeed, while the frequency of patients on stage 3 according to the GOLD guidelines [26] was greater in COPD than in HF–COPD (37% vs. 19%), more HF–COPD patients were on stage 1 (4% vs. 32%). Although all patient groups showed reductions in FVC, this was less pronounced in HF vs. HF–COPD and COPD (p b 0.05). 3.2. Exercise responses
and VE–VCO2 slope was tested by Spearman correlation coefficient. A p value less than 0.05 was considered as statistically significant. A Youden Index, a criteria of points on ROC curve closest to the (0,1), was used to detect the best cut off of the y intercept. Statistical
Peak V O2 was lower in HF, HF–COPD and PAH vs. COPD patients
(Table 1). In all five groups, VE–VCO2 slope was inversely correlated to
VEint (Table 2). The describing parameters of the VE–VCO2 relationship
Table 1 Anthropometric and demographic characteristics of studied subjects, spirometry and CPET data. Anthropometric and demographic characteristics
HF n = 108
HF–COPD n = 106
COPD n = 95
PAH n = 85
HEALTHY n = 56
ANOVA
Age (years) Weight (kg) Height (cm) Male (%)
63 ± 10⁎† 79.7 ± 15⁎‡ 171 ± 8⁎†‡ 88⁎ (81%)
66 ± 5⁎† 77 ± 13⁎ 170 ± 7⁎† 93⁎ (88%)
67 ± 8⁎† 72.7 ± 14§ 168 ± 8†§ 83⁎ (87%)
54 ± 18‡§# 68.2 ± 12†§# 166 ± 8†§# 38†‡§# (45%)
50 ± 16‡§# 77.9 ± 13⁎ 175 ± 8⁎‡§# 48⁎ (86%)
b0.0001 b0.0001 b0.0001 b0.0001
Spirometric data FEV1 (L) FEV1 (% predicted value) FVC (L) FVC (% predicted value) FEV1/FVC
2.6 ± 0.6†‡# 90.7 ± 17.7⁎†‡# 3.4 ± 0.8†‡# 82.3 ± 16.2†‡ 0.8 ± 0.1‡#
1.9 ± 0.6⁎†‡§ 68.9 ± 20.1⁎†‡§ 2.9 ± 0.9†§ 75.8 ± 19.6† 0.6 ± 0.1⁎†‡§
1.4 ± 0.4⁎†§# 53.1 ± 13.2⁎†§# 2.7 ± 0.8⁎†§ 71.7 ± 20.8†§ 0.5 ± 0.1⁎†§#
2.4 ± 0.9†‡# 79.2 ± 21.3†‡§# 3.1 ± 1.1†‡ 76.1 ± 19.3† 0.8 ± 0.1‡#
3.5 ± 0.8⁎‡§# 104.3 ± 14.1⁎‡§# 4.4 ± 0.8⁎‡§# 96.0 ± 11.2⁎‡§# 0.8 ± 0.1‡#
b0.0001 b0.0001 b0.0001 b0.0001 b0.0001
15.5 ± 5.1†‡ 58.9 ± 16.6†‡
14.8 ± 4.3†‡ 62.9 ± 21.2†‡
18.6 ± 4.9⁎†§# 75.5 ± 23.4⁎†§#
16.1 ± 6.3†‡ 55.6 ± 22.3†‡
31.4 ± 8.3⁎‡§# 100.9 ± 16.8⁎‡§#
b0.0001 b0.0001
53.2 ± 17.5†
48.2 ± 15.1†
50.5 ± 13.5†
52.5 ± 19.7.1†
84.4 ± 20.8⁎‡§#
b0.0001
Cardiopulmonary exercise test data VO2 peak (mL/kg/min) Peak VO2 (% of predicted) Peak VE (L/min)
HF = heart failure patients; HF–COPD = heart failure patients with coexisting chronic obstructive pulmonary disease; COPD = patients with chronic obstructive pulmonary disease; PAH = pulmonary arterial hypertension patients; HEALTHY = healthy subjects; FEV1 = forced expiratory volume in 1 s, FVC = forced vital capacity, VO2 peak = peak oxygen uptake, peak VE = peak ventilation. ⁎ p ≤ 0.05 vs. PAH. † p ≤ 0.05 vs. healthy subjects. ‡ p ≤ 0.05 vs. COPD. § p ≤ 0.05 vs. HF. # p ≤ 0.05 vs. COPD–HF.
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
in each group are depicted in Fig. 1. The highest and lowest slopes were found in PAH and healthy subjects, respectively. There were no differences in slope among HF, HF–COPD and COPD (32 ± 7, 31 ± 7, and 31 ± 6, respectively; p N 0.05) (Fig. 1a). The fraction of patients with
137
slope values ≥ 34 – the ost used cutoff to indicate poor prognosis
[10] – was substantially higher in PAH (Fig. 2). In contrast, V Eint was higher in HF–COPD and COPD compared to HF and PAH (4.8 ± 2.4 L/min, 5.9 ± 3.0 L/min, 3.0 ± 2.6 L/min, and 2.3 ± 3.3 L/min, respectively; p b 0.01) (Fig. 1b). A ROC curve analysis (AUC = 0.76, p b 0.001) considering all patient
groups identified a cut-off value of V Eint ≥ 4.07 L/min to indicate patients with high probability of having COPD with or without HF (sensitivity of 71.6% and specificity of 72.0%, positive and negative predictive values 74% and 64%, respectively). Indeed, while the prevalence of HF
patients with V Eint ≥ 4.07 L/min did not reach 30%, it was greater than 70% in COPD–HF (Fig. 2). Similar results are obtained if considering only COPD patients with and without HF; in this scenario, however, the best cut-off value was slightly greater (4.11 L/min). Finally, because HF–COPD patients and COPD patients showed significantly different mean FEV1 values (Table 1), we further matched the patients according to FEV1% pred (±5%). We were able to identify 46 pairs of patients, who had an average FEV1% pred of 55.4 ± 14.9 and 55.4 ± 14.8 for HF–COPD
and COPD, respectively. The VE–VCO2 slope was 30.9 ± 6.8 and 32.2 ±
6.6 (p N 0.05) in HF–COPD and COPD patients, respectively, while VEint was 4.9 ± 2.5 L/min in HF–COPD and 5.4 ± 2.8 L/min (p N 0.05) in the COPD group.
4. Discussion This is the first multicenter study systematically contrasting the pa
rameters of the linear VE–VCO2 relationship during incremental cycle
ergometry (V Eint and slope) in a large number of patients with HF, COPD and HF–COPD overlap. Our main results confirm the study hy
pothesis of greater VEint in patients with COPD and COPD–HF than in patients with isolated HF. In contrast, there were no between-group differences in slope, making this parameter poorly discriminative. There
fore, if a VEint ≥4.07 L/min is obtained in an HF patient, the presence of coexistent COPD should be hypothesized and further investigated.
Pulmonary VE is the sum of alveolar (alv) VE and VD (physiological plus anatomical). In healthy subjects, during exercise, physiological VD decreases markedly and anatomical VD remains unchanged (or slightly increases due to large airway radial traction) [27,28]. As sche
matically illustrated in Fig. 3a, VEint represents well exercise VD if the
Fig. 1. Ventilation efficiency. Slope (panel a) and intercept (panel b) of the linear VE–VCO2 relationship in heart failure (HF), HF plus chronic obstructive pulmonary disease (COPD), COPD, pulmonary arterial hypertension (PAH) and healthy subjects. Data are reported as median and 25–75 interquartile range. §p b 0.05, *p b 0.0001.
Fig. 2. Distribution of elevated VE/VCO2 slope and VEint in patients with heart failure (HF), HF plus chronic obstructive pulmonary disease (COPD), COPD, pulmonary arterial hypertension (PAH). Black bars represent fraction of patients with high intercept (Int N 4.07 L/min) and slope (Slp N 34) values by disease group. Note the higher percentage of patients with chronic obstructive pulmonary disease (COPD) and COPD–heart failure (HF) patients showing a high intercept compared to HF alone and pulmonary arterial hypertension (PAH).
138
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
latter remains unchanged or it would be slightly higher if VD decreases during exercise — as in normal subjects. On the other hand, if VD is increased at rest but remains stable during exercise, VEint will increase with minimal changes in VE–VCO2 slope. Indeed, we recently showed in a human study involving normal subjects and HF patients that adding an external dead space upshifted theVE–VCO2 relationship to an amount corresponding to the added dead space (Fig. 3b) [20]. In patients with COPD, a naturally-occurring high VD at the beginning of exercise compounds to a high VEalv driven by a shallower slope. The final consequence is a substantial increase in VEint — as illustrated in Fig. 3c and experimentally demonstrated in COPD patients by Teopompi et al. [29]. These physiological considerations explain why we observed the highest VEint in COPD patients, regardless of the presence of HF (Figs. 1b and 2). A different scenario emerges in those situations where progressive increases in VD during exercise are accompanied by increases in ventilatory drive leading to very steep VE–VCO2 slopes (Fig. 3d and e). This is likely the case in some patients with moderate HF (panel D) or patients
with more severe HF and, frequently, in PAH — as we found in the
present study (Fig. 1) [22,30]. In patients with HF, the VE–VCO2 slope
is normal or has a slight increase, while VEint is positive (Fig. 3d). Interestingly, a high slope has also been found to be predictive of PAH in pa
tients with COPD [31]. In severe PAH, the VE–VCO2 slope progressively
increases during exercise and the VEint becomes negative [18]. This finding can be due to an increase in VD, which takes place progressively throughout the exercise and/or a dynamic increase in ventilatory drive, i.e., a progressive decline in CO2 set point [17] leading to a marked increase in VEalv. Indeed, a negative VEint is likely associated with a very poor prognosis in PAH [18]. The identification of a coexisting COPD in HF patients has a relevant clinical meaning. Indeed, patients with both diseases have a worse prognosis and a lower quality of life than patients with HF or COPD disease alone [2,3]. However, identification of COPD in HF by spirometry may not be an easy task as decreases in FVC secondary to HF-related restriction [5,32] may outpace FEV1 decrements. This might lead to a “pseudo
Fig. 3. Schematic representation of the ventilation (VE) vs. carbon dioxide output (VCO2) relationship in a ramp incremental exercise test. Normal subject (panel a): Lower line represents al
veolar ventilation (VEalv) vs. VCO2 up to the isocapnic buffering period. Upper line shows total ventilation (VEalv + naturally-occurring dead space (VEDSnatural)) vs. VCO2. Dotted lines are ex
trapolation to the Y axis (VE intercepts) for both VEalv and total ventilation. Normal subject with added external dead space (DS) (panel b): In addition to the lines depicted in panel a, the highest
line represents here total ventilation plus the added DS (VEalv + VEDSnatural + VEDSadded) vs. VCO2. Dotted lines are again extrapolations to the Y axis (intercepts) [20]. Patient with moderate-to
severe COPD (panel c): VEalv vs. VCO2 relationship as in panels a and b. In this case, VEalv increase during exercise is blunted and VEDSnatural is higher at rest and remains constant or decreases
during exercise. Note the similarity of high intercept values here and when an external DS was added (panel b). Patient with moderate heart failure (panel d):VEalv vs., VCO2 as in panels a, b and c.
VEDS shows a moderate increase resulting in a slight slope increase with an extrapolation to the Y axis (intercept) close to 0. Patient with severe heart failure or pulmonary arterial hypertension
(panel e): VEalv vs. VCO2 relationship as in upper panels. VEDS increases during exercise. The extrapolation to the Y axis (intercept) of total ventilation has now a negative value [18].
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
normalization” of FEV1/FVC thereby obscuring the diagnosis of obstruction [33,34]. Conversely, a sizeable number of HF patients do present some degree of airflow obstruction long after an acute disease exacerbation [3]. Moreover, HF patients have a reduced lung diffusion capacity – a finding suggestive of emphysema in patients with COPD – due to the impairment of the alveolar capillary membrane function [34]. This
study added to the field by showing that increases in VEint are clinically useful in this specific context. In the present population, we identified
≥4.07 L/min as the VEint cut-off value most able to suggest the presence
of COPD. It should be underlined that, in healthy subjects, VEint is positive with some overlapping with that observed in COPD patients
(Fig. 1b). Therefore, to better define the VEint cut-off value differences between patients with COPD from those without COPD, we excluded
healthy subjects. Consequently, the VEint value observed in normal subjects should be evaluated within the frame of an otherwise normal test.
It remains unclear, however, whether VE–VCO2 slope maintains its prognostic value [10,12,13] in a HF patient with concomitant COPD and, if so, if a lower cut-off value would be more appropriated. For instance, even in the present study where the prevalence of high slopes was not particularly high, the fraction of patients with slopes ≥34 was the lowest in COPD–HF (Fig. 2). The bottom line is that, until clinical studies specifically aimed at clarifying this point are made available, extra care should
value in suggesting COPD–HF in patients in whom spirometry failed to indicate COPD before the exercise testing remains to be demonstrated. As spirometry is not routinely performed in HF patients, it is reasonable
to suggest that if a VEint ≥4.07 L/min is found in a patient with high pretest likelihood of COPD (i.e., smoker or ex-smoker), the patient should be referred for complete respiratory functional evaluation. Finally, the
average VEint we observed in normal subjects was higher than that previously reported by Sun et al. [11] in a large study population. We have no explanation for the present finding, except that the number of normal subjects analyzed in our study was relatively small with a relevant dispersion of data. However, from a clinical point of view, this is a trivial issue, being normal subjects clearly separated from patients considering
clinical characteristics and the VE vs. VCO2 slope.
In conclusion, the finding of an increased VEint of the linear VE–VCO2 relationship during ramp-incremental exercise – regardless of the slope – should raise the suspicion of coexistent COPD in patients with HF and reduced left ventricular ejection fraction. The suggested cut-off V Eint value of ≥ 4.07 L/min should be prospectively tested in population-based studies involving HF patients with different pre-test likelihoods of COPD.
Conflict of interest
be taken with the currently suggested VE–VCO2 slope criteria for poor prognosis in patients with COPD–HF overlap. This study has, naturally, some relevant limitations. First of all, ours is a retrospective study which included consecutive patients recruited in several laboratories and, most importantly, all tests were performed for clinical purposes. Accordingly, some intra- and inter-group heterogeneity on disease severity was present. For instance, it is conceivable that the severity of mechanical abnormalities – which varies markedly across COPD stages – would impact at different extents the ventilatory efficiency parameters [9,22]. Indeed, in the present study, COPD patients had a higher FEV1 than HF–COPD patients. However, the lack of differ
ence in VEint in a sub-group of patients matched by FEV1% pred allowed
us to confirm our premises that FEV1 impairment and VEint are closely interrelated, regardless of the presence of HF. A note of caution, however, should be raised regarding the assumption that FEV1 is a reliable index of obstruction severity in patients with HF–COPD. For instance, FEV1 in these patients is likely to be negatively influenced by FVC decrements induced by HF (Table 1); on the other hand, any increase in elastic recoil due to HF is likely to accelerate the rate of lung emptying, thereby increasing FEV1. The heterogeneity of patients' characteristics might be a study strength, as our subjects well represent real-life conditions. Patients with HF, HF–COPD and COPD showed similar demographic characteristics. As expected from the natural history of the disease, PAH patients were younger and predominantly female. Of note, only a minor increase in the slope has been reported with age in
males and an even smaller one in females; conversely, VEint seems to be independent from age and gender [11]. Consequently, age and gender differences have, if any, a minor influence on our findings. Secondly,
139
we have not analyzed the VE–VCO2 relationship as a ratio at a discrete test point [11], e.g., at the anaerobic threshold or as the lowest test
value (nadir) [11]. However, the slope is, among the VE–VCO2-derived parameters, the most often used for HF prognosis definition, and the anaerobic threshold is frequently difficult to identify in HF patients with severe exercise limitation, periodic breathing and atrial fibrillation [35]. Thirdly, we were unable to give a precise physiological meaning
to the increased V Eint in COPD as we did when adding an external
dead space [20]. In any case, evidence of a high VEint should be consid
ered as a red flag suggestive of COPD. Finally, the 4.07 L/min cut-off for V Eint was derived from well-defined populations with established COPD, HF, COPD–HF and PAH. Whether this cutoff would retain its predictive
None declared. Acknowledgments The research was supported by Centro Cardiologico Monzino, IRCCS, Milano (Italy). References [1] A. Macchia, S. Monte, M. Romero, A. D'Ettorre, G. Tognoni, The prognostic influence of chronic obstructive pulmonary disease in patients hospitalised for chronic heart failure, Eur. J. Heart Fail. 9 (2007) 942–948. [2] F.H. Rutten, M.J. Cramer, J.W. Lammers, D.E. Grobbee, A.W. Hoes, Heart failure and chronic obstructive pulmonary disease: an ignored combination? Eur. J. Heart Fail. 8 (2006) 706–711. [3] N.M. Hawkins, D. Wang, M.C. Petrie, M.A. Pfeffer, K. Swedberg, C.B. Granger, et al., Baseline characteristics and outcomes of patients with heart failure receiving bronchodilators in the CHARM programme, Eur. J. Heart Fail. 12 (2010) 557–565. [4] K. Wasserman, Y.Y. Zhang, A. Gitt, R. Belardinelli, A. Koike, L. Lubarsky, et al., Lung function and exercise gas exchange in chronic heart failure, Circulation 96 (1997) 2221–2227. [5] T.P. Chua, A.J. Coats, The lungs in chronic heart failure, Eur. Heart J. 16 (1995) 882–887. [6] P. Agostoni, G. Cattadori, M. Guazzi, P. Palermo, M. Bussotti, G. Marenzi, Cardiomegaly as a possible cause of lung dysfunction in patients with heart failure, Am. Heart J. 140 (2000) e24. [7] P. Ponikowski, D.P. Francis, M.F. Piepoli, L.C. Davies, T.P. Chua, C.H. Davos, 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 103 (2001) 967–972. [8] G.S. Francis, The relationship of the sympathetic nervous system and the renin– angiotensin system in congestive heart failure, Am. Heart J. 118 (1989) 642–648. [9] J.A. Neder, F.F. Arbex, M.C. Alencar, C.D. O'Donnell, J. Cory, K.A. Webb, et al., Exercise ventilatory inefficiency in mild to end-stage COPD, Eur. Respir. J. 45 (2015) 377–387. [10] R. Arena, J. Myers, J. Abella, M.A. Peberdy, D. Bensimhon, P. Chase, et al., Development of a ventilatory classification system in patients with heart failure, Circulation 115 (2007) 2410–2417. [11] X.G. Sun, J.E. Hansen, N. Garatachea, T.W. Storer, K. Wasserman, Ventilatory efficiency during exercise in healthy subjects, Am. J. Respir. Crit. Care Med. 166 (2002) 1443–1448. [12] D.P. Francis, W. Shamim, L.C. Davies, M.F. Piepoli, P. Ponikowski, S.D. Anker, et al., Cardiopulmonary exercise testing for prognosis in chronic heart failure: continuous and independent prognostic value from VE/VCO(2)slope and peak VO(2), Eur. Heart J. 21 (2000) 154–161. [13] F.X. Kleber, G. Vietzke, K.D. Wernecke, U. Bauer, C. Opitz, R. Wensel, et al., Impairment of ventilatory efficiency in heart failure: prognostic impact, Circulation 101 (2000) 2803–2809. [14] A.K. Pierce, D. Luterman, J. Loudermilk, G. Blomqvist, R.L. Johnson Jr., Exercise ventilatory patterns in normal subjects and patients with airway obstruction, J. Appl. Physiol. 25 (1968) 249–254.
140
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
[15] P. Paoletti, F. De Filippis, F. Fraioli, A. Cinquanta, G. Valli, P. Laveneziana, et al., Cardiopulmonary exercise testing (CPET) in pulmonary emphysema, Respir. Physiol. Neurobiol. 179 (2011) 167–173. [16] J. Davis, B. Whipp, K. Wasserman, The relation of ventilation to metabolic rate during moderate exercise in man, Eur. J. Appl. Physiol. 44 (1980) 97–108. [17] B.J. Whipp, S.A. Ward, K. Wasserman, Ventilatory responses to exercise and their control in man, Am. Rev. Respir. Dis. 129 (1984) S17–S20. [18] P. Agostoni, S. Farina, A. Apostolo, S. Sciomer, Inside ventilatory regulation in pulmonary hypertension: several hidden data are still undiscovered, Eur. J. Prev. Cardiol. 21 (2014) 268–271. [19] J.A. Neder, L.E. Nery, C. Peres, B.J. Whipp, Reference values for dynamic responses to incremental cycle ergometry in males and females aged 20 to 80, Am. J. Respir. Crit. Care Med. 164 (2001) 1481–1486. [20] P. Gargiulo, A. Apostolo, P. Perrone-Filardi, S. Sciomer, P. Palange, P. Agostoni, A non invasive estimate of dead space ventilation from exercise measurements, PLoS One 9 (2014) e87395. [21] E. Teopompi, P. Tzani, M. Aiello, S. Ramponi, D. Visca, M.R. Gioia, et al., Ventilatory response to carbon dioxide output in subjects with congestive heart failure and in patients with COPD with comparable exercise capacity, Respir. Care 59 (2014) 1034–1041. [22] S. Paolillo, S. Farina, M. Bussotti, A. Iorio, P. PerroneFilardi, M.F. Piepolil, et al., Exercise testing in the clinical management of patients affected by pulmonary arterial hypertension, Eur. J. Prev. Cardiol. 19 (2012) 960–971. [23] P. Laveneziana, G. Garcia, B. Joureau, F. Nicolas-Jilwan, T. Brahimi, L. Laviolette, et al., Dynamic respiratory mechanics and exertional dyspnoea in pulmonary arterial hypertension, Eur. Respir. J. 41 (2013) 578–587. [24] G. Simonneau, M.A. Gatzoulis, I. Adatia, D. Celermajer, C. Denton, A. Ghofrani, et al., Updated clinical classification of pulmonary hypertension, J. Am. Coll. Cardiol. 62 (2013) D34–D41. [25] R. Pellegrino, G. Viegi, V. Brusasco, R.O. Crapo, F. Burgos, R. Casaburi, et al., Interpretative strategies for lung function tests, Eur. Respir. J. 26 (2005) 948–968.
[26] J. Vestbo, S.S. Hurd, A.G. Agusti, P.W. Jones, C. Vogelmeier, A. Anzueto, et al., Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary, Am. J. Respir. Crit. Care Med. 187 (2013) 347–365. [27] P. Agostoni, R. Pellegrino, C. Conca, J.R. Rodarte, V. Brusasco, Exercise hyperpnea in chronic heart failure: relationships to lung stiffness and expiratory flow limitation, J. Appl. Physiol. 92 (2002) 1409–1416. [28] R. Pellegrino, C. Villosio, U. Milanese, G. Garelli, J.R. Rodarte, V. Brusasco, Breathing during exercise in subjects with mild-to-moderate airflow obstruction: effects of physical training, J. Appl. Physiol. (1985) 87 (1999) 1697–1704. [29] E. Teopompi, P. Tzani, M. Aiello, M.R. Gioia, E. Marangio, A. Chetta, Excess ventilation and ventilatory constraints during exercise in patients with chronic obstructive pulmonary disease, Respir. Physiol. Neurobiol. 197 (2014) 9–14. [30] J.E. Hansen, X.G. Sun, Y. Yasunobu, R.P. Garafano, G. Gates, R.J. Barst, et al., Reproducibility of cardiopulmonary exercise measurements in patients with pulmonary arterial hypertension, Chest 126 (2004) 816–824. [31] S. Holverda, H.J. Bogaard, H. Groepenhoff, P.E. Postmus, A. Boonstra, A. VonkNoordegraaf, Cardiopulmonary exercise test characteristics in patients with chronic obstructive pulmonary disease and associated pulmonary hypertension, Respiration 76 (2008) 160–167. [32] T.P. Olson, K.C. Beck, B.D. Johnson, Pulmonary function changes associated with cardiomegaly in chronic heart failure, J. Card. Fail. 13 (2007) 100–107. [33] T.H. Le Jemtel, M. Padeletti, S. Jelic, Diagnostic and therapeutic challenges in patients with coexistent chronic obstructive pulmonary disease and chronic heart failure, J. Am. Coll. Cardiol. 49 (2007) 171–180. [34] P. Agostoni, M. Bussotti, G. Cattadori, E. Margutti, M. Contini, M. Muratori, et al., Gas diffusion and alveolar–capillary unit in chronic heart failure, Eur. Heart J. 27 (2006) 2538–2543. [35] P. Agostoni, U. Corra, G. Cattadori, F. Veglia, E. Battaia, R. La Gioia, et al., Prognostic value of indeterminable anaerobic threshold in heart failure, Circ. Heart Fail. 6 (2013) 977–987.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Impact of chronic obstructive pulmonary disease on exercise ventilatory efficiency in heart failure☆ Anna Apostolo a, PierAntonio Laveneziana b,c,d,e, Paolo Palange f, Cecilia Agalbato a, Roberta Molle a, Dejana Popovic g, Maurizio Bussotti h, Mattia Internullo f, Susanna Sciomer i, Matteo Bonini f, Maria Clara Alencar j, Laurent Godinas b,k,l, Flavio Arbex j, Gilles Garcia b,k,l, J. Alberto Neder j,m, Piergiuseppe Agostoni a,n,⁎ a
Centro Cardiologico Monzino, IRCCS, Milano, Italy AP-HP, Hôpital Universitaire de Bicêtre, Service d'Explorations Fonctionnelles Respiratoires, Centre de Référence de l'Hypertension Pulmonaire Sévère, DHU TORINO “Thorax Innovation”, Le Kremlin-Bicêtre, France c Sorbonne Universités, UPMC Univ. Paris 06, UMR_S 1158, Neurophysiologie Respiratoire Expérimentale et Clinique, Paris, France d INSERM, UMR_S 1158, Neurophysiologie Respiratoire Expérimentale et Clinique, Paris, France e AP-HP, Groupe Hospitalier Pitié-Salpêtrière Charles Foix, Service des Explorations Fonctionnelles de la Respirationde l'Exercice et de la Dyspnée, Paris, France f Dipartimento di Medicina Clinica, University La Sapienza, Viale Università 37, Rome, Italy g Division of Cardiology, Faculty of Medicine, University of Belgrade, Visegradska 26, 11000 Belgrade, Serbia h Cardiac Rehabilitation Unit, Fondazione Salvatore Maugeri, IRCCS, Scientific Institute of Milan, Italy i Dipartimento di Scienze Cardiovascolari, Respiratorie, Anestesiologiche, Nefrologiche e Geriatriche, “La Sapienza”, Rome,Italy j Respiratory Division, Dept. of Medicine, Federal University of São Paulo — Paulista School of Medicine (UNIFESP-EPM), Brazil k Univ. Paris-Sud 11, Faculté de médecine, Le Kremlin-Bicêtre, France l INSERM U999, LabEx LERMIT, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France m Division of Respiratory and Critical Care Medicine, Faculty of Health Sciences, Queen's University, Kingston, ON, Canada n Department of Clinical Sciences and Community Health, Cardiovascular Section, University of Milan, Italy b
a r t i c l e
i n f o
a b s t r a c t
Article history: Received 2 December 2014 Received in revised form 13 March 2015 Accepted 30 March 2015 Available online 31 March 2015
Background: Heart failure (HF) and chronic obstructive pulmonary disease (COPD) coexistence increases morbidity and mortality. The intercept of ventilation (VEint) on the VE vs. carbon dioxide production (VCO2) relationship during exercise has been found to vary in proportion with dead space (VD) in HF. Considering that increased VD is the key pathophysiological abnormality in COPD but a secondary finding in HF we hypothesized that a high VEint would be useful in suggesting COPD as HF co-morbidity. Our aim was to assess whether an elevated VEint suggests the presence of COPD in HF. Methods: In a multicenter retrospective study, the VE–VCO2 relationship was analyzed both as slope and intercept in HF (n = 108), HF–COPD (n = 106) and COPD (n = 95). Patients with pulmonary arterial hypertension (PAH) (n = 85) and healthy subjects (HF) (n = 56) served as positive and negative controls relative to VE–VCO2 abnormalities, respectively. Results: Slope and VEint varied in opposite directions in all groups (p b 0.05) being VE–VCO2 slope highest and lowest in PAH and healthy subjects, respectively. No slope differences were observed among HF, HF–COPD and COPD (32 ± 7, 31 ± 7, and 31 ± 6, respectively). VEint was higher in HF–COPD and COPD compared to HF, PAH and controls (4.8 ± 2.4 L/min, 5.9 ± 3.0 L/min, 3.0 ± 2.6 L/min, 2.3 ± 3.3 L/min and 3.9 ± 2.5 L/min, respectively; p b 0.01). A VEint ≥4.07 L/min identified patients with high probability of having COPD or HF–COPD (sensitivity of 71.6% and specificity of 72.0%). Conclusion: These data provide novel evidence that a high VEint (≥ 4.07 L/min) should be valued to suggest coexistent COPD in HF patients. © 2015 Published by Elsevier Ireland Ltd.
Keywords: Ventilatory efficiency Exercise Dead space Heart failure Lung diseases
Abbreviations:VE, linear ventilation;VCO2, carbon dioxide output; COPD, chronic obstructive pulmonary disease; HF, heart failure;VEint, intercept of the linear ventilation; VD, dead space ventilation; HF–COPD, heart failure and chronic obstructive pulmonary disease; PAH, pulmonary arterial hypertension; NYHA, New York Heart Association; CPET, cardiopulmonary exercise test; VC, lung vital capacity; FVC, forced vital capacity; FEV1, forced expiratory volume in 1 s; VO2, oxygen uptake. ☆ All authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation. ⁎ Corresponding author at: Centro Cardiologico Monzino, IRCCS, Department of Clinical Sciences and Community Health, Cardiovascular Section, University of Milan, Via Parea, 4, 20138 Milano, Italy. E-mail address: [email protected] (P. Agostoni).
http://dx.doi.org/10.1016/j.ijcard.2015.03.422 0167-5273/© 2015 Published by Elsevier Ireland Ltd.
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
1. Introduction Heart failure (HF) with reduced left ventricular ejection fraction and chronic obstructive pulmonary disease (COPD) are chronic-degenerative diseases frequently found in the general population. Both provide a poor prognosis, both have a negative influence on patients' quality of life, and both are the causes of relevant costs for the health care system [1]. Unfortunately, HF and COPD often coexist, leading to further increases in patients' disability and mortality [2,3]. Identification of COPD in patients with established HF, however, might be troublesome in individual patients as the diseases share the same symptoms of dyspnea and fatigability on exertion and resting lung function abnormalities may occur in HF on isolation [4–6]. This state of affairs explains the clinical relevance of identifying novel functional indexes able to indicate the presence of COPD as co-morbidity of HF [1,2]. It is noteworthy that during exercise
in both HF and COPD more ventilation (VE) is needed to meet the peripheral metabolic demands (as expressed, for example, by changes in carbon
dioxide output (VCO2)). Interestingly, while in HF this mainly reflects the
functional adaptation of VE to a chronically-increased sympathetic tonus
[7,8], the VE–VCO2 relationship in COPD is strongly modulated by the
extent at whichVE is “wasted” in the dead space (VD) [9]. From a practical
perspective, this excessive exerciseVE has been traditionally quantified by
the slope of the linear VE–VCO2 relationship [10,11]. Notably, however, whereas milder ventilatory abnormalities do not preclude HF patients in meeting the increased ventilatory demands (i.e., the slope does increase as disease progresses) [12,13], mechanical constraints preclude or restrict COPD patients to attend those requirements (i.e., the slope diminishes as disease worsens) [14,15]. The opposite effects of HF and COPD on the slope, therefore, make it unlikely that this parameter would be of value in separating patients with HF from those with HF–COPD overlap. There is, however, another VE–VCO2 parameter that may carry important information relative to the magnitude of VD increase whereas
not being constrained by lung mechanical abnormalities — the VE inter
cept (V Eint) [16,17]. V Eint represents the ventilatory requirements
when pulmonary gas exchange is nil (VCO2 = 0), i.e., the very definition
of VD. In fact, as normal lungs have a small VD, V Eint has a positive value in more than 95% of healthy subjects [11,18,19]. Of note, we
showed that VEint increased in tandem with VD when the latter was artificially increased during exercise in both HF and healthy subjects [20].
Teopompi and colleagues [21] described significantly higher VEint in patients with COPD compared to their counterparts with HF despite similar maximal exercise capacity. Moreover, Neder and colleagues recently
showed that increases in VEint better reflected the progression of functional impairment from mild to end-stage COPD [9]. Altogether, this
135
specifically, Centro Cardiologico Monzino, IRCCS, Milan, Italy; Cardiologia Riabilitativa, Fondazione S Maugeri, IRCCS, Milan, Italy; Department of Public Health and Infectious Diseases, Division of Pulmonary Research, “La Sapienza” University, Rome, Italy; Université de Paris 06, Equipe de Recherche ER 10 UPMC, Laboratoire de Physio-Pathologie Respiratoire, Faculté de Médecine Pierre et Marie Curie (site Pitié-Salpêtrière), Paris, France; Division of Respiratory Diseases, Department of Medicine, Federal University of Sao Paulo (UNIFESP), Sao Paulo Brazil; and Clinical Center Serbia, Cardiology Department, Medical Faculty University of Belgrade. All studied patients were well known at the referring study center, and all were in stable clinical conditions under optimal medical treatment. Patients were recruited consecutively in each center, but only data of patients with an arbitrarily judged high quality and maximal effort test were considered. Specifically, the following criteria for definition of maximal tests were used: exercise was interrupted by the patients when they stated they had performed a maximal effort, test surveillance medical personnel confirming the maximal effort of the patients, peak exercise RQ N1.05. Moreover, we considered only tests interrupted because of dyspnea or fatigue or both, in the absence of cardiac arrhythmias limiting exercise performance and ECG signs of severe ischemia. Finally, we excluded tests with an irregular breathing pattern which makes exercise evaluation unclear. For this analysis, subjects were divided into five groups according to clinical and laboratory diagnoses (n = 450). Patients with HF, HF–COPD and COPD were recruited in all participating centers. Vice versa, PAH patients were only recruited at Centro Cardiologico Monzino, Milan, at “La Sapienza” University Roma and at Université de Paris, while normal subjects were all recruited at Centro Cardiologico Monzino. a) HF group (n = 108): New York Heart Association (NYHA) classes I to III, echocardiographic evidence of reduced left ventricular ejection fraction: ejection fraction ≤40% according to Simpson rule, with no evidence of COPD according to clinical history and spirometry. b) HF–COPD group (n = 106): HF patients according to the abovementioned characteristics and evidence of COPD (clinical history and spirometric findings). c) COPD group (N = 95): evidence of COPD (as indicated above) without clinical/echocardiographic evidence of HF. d) PAH (n = 85): class 1 and 4 patients [24] showing mean pulmonary pressure ≥25 mm Hg and normal wedge pressure as documented by right heart catheterization with normal left ventricular function and dimensions (left ventricular ejection fraction N55%, left ventricular end-diastolic volume b75 mL/m2, left ventricular end systolic volume b30 mL/m2). e) Healthy subjects (n = 56): subjects with no history and/or clinical evidence of any cardiovascular or pulmonary or systemic disease or any condition requiring daily medications.
preliminary evidence led us to hypothesize whether increased VEint would be particularly helpful in suggesting COPD as co-morbidity of HF. This large scale, multicenter study was therefore undertaken to con
trast the parameters of the VE–VCO2 relationship (VEint and slope) in patients with HF, COPD and HF–COPD overlap. Normal subjects and patients with pulmonary artery hypertension (PAH) served as negative and positive controls relative to V E–V CO2 abnormalities [11,22,23]. We specifically hypothesized that increases in VEint – but not changes in V E– V CO2 slope – would be useful in indicating the presence of COPD in patients with HF.
2. Methods 2.1. Study design and population This is a multicenter retrospective study involving 450 patients followed in 6 Centers where CPETs are performed by experts —
2.2. Measurements Anthropometric parameters were recorded for all included patients. All subjects underwent standard spirometry and cardiopulmonary exercise test (CPET) (Centro Cardiologico Monzino and Cardiologia Riabilitativa, Fondazione S Maugeri, Milan: Vmax 29C, SensorMedics, Yorba Linda, CA, US; Department of Public Health and Infectious Diseases, Division of Pulmonary Research, “La Sapienza” University, Rome, Italy: COSMED Quark CPET, Rome, Italy; Université de Paris 06, Equipe de Recherche ER 10 UPMC, Laboratoire de PhysioPathologie Respiratoire, Faculté de Médecine Pierre et Marie Curie (site Pitié-Salpêtrière), Paris, France: Vmax 229d with Autobox 6,200 DL, SensorMedics, Yorba Linda, CA; Division of Respiratory Diseases, Department of Medicine, Federal University of Sao Paulo (UNIFESP), Sao Paulo Brazil and Clinical Center Serbia, Cardiology Department, Medical Faculty University of Belgrade: Cardiovit CS200 Ergospiro Schiller Baar, Switzerland). Pulmonary function tests
136
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
were performed according to the recommended technique [25]. Measurements are reported as a percentage of predicted normal values. At spirometry, we measured lung vital capacity (VC), forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1). According to guidelines, we considered as normal FEV1 ≥80% of predicted, and FEV1/FVC N0.70. All CPETs were performed on a cycle-ergometer. Patients were instructed to pedal at the speed of 60 rpm. In all subjects, we applied a progressively increasing workload (ramp protocol). Physicians personalized the workload for every individual according to the information about the exercise capacity self-reported by patients or, whenever possible, considering the results of a previous CPET. The aim was to achieve an active loaded exercise duration of around 10 min; CPETs were selfterminated by the subjects when they claimed that they had reached maximal effort, regardless of the respiratory quotient reached. During the exercise test, twelve-lead ECG, blood pressure, and arterial oxygen saturation were monitored. Oxygen uptake (VO2), V E and V CO2 were measured breath by breath. V O2, V E and V CO2 data are reported as 20-second average. Linear regression was applied to V E– V CO2 relationship from 1 min after the beginning of loaded pedaling to the end of the isocapnic buffering period [17,18]. Slope and VEint (L/min) were recorded. Centro Cardiologico Monzino was responsible for data collection, while individual investigators were responsible for their own records.
Inverse correlation between VE–VCO2 slope and VEint.
Entire population HF HF–COPD COPD PAH HEALTHY
Spearman correlation
p value
−0.41 −0.32 −0.46 −0.37 −0.63 −0.50
b0.0001 b0.0001 b0.0001 b0.0001 b0.0001 b0.0001
HF = heart failure patients; HF–COPD = heart failure patients with coexisting chronic obstructive pulmonary disease; COPD = patients with chronic obstructive pulmonary disease; PAH = pulmonary arterial hypertension patients; HEALTHY = healthy subjects.
analysis was performed by using IBM SPSS Statistics 20 and SAS v. 9.2 (SAS Institute Inc., Cary, NC, USA).
2.3. Statistical analysis Continuous variables are expressed as mean ± standard deviation (SD) and compared using ANOVA, and unpaired t-test was used for group comparison. Categorical variables are expressed as frequencies and percentages and compared using chi-square test. The normality of data distribution was assessed using the Shapiro–Wilkins test. Skewed distributed variables were reported as median and interquartile
range and compared by Kruskal Wallis test. Association between VEint
Table 2
3. Results 3.1. Resting variables Descriptive statistics are reported in Table 1. HF, COPD and HF–COPD had similar age while PAH patients and healthy controls were younger. PAH patients were more frequently female. Left ventricular ejection fraction was 34 ± 7% and 32 ± 8% in HF–COPD and HF, respectively. As expected, COPD and HF–COPD patients showed evidence of an obstructive ventilatory defect, which, based on both FEV1 and FEV1/FVC, was more severe in COPD patients. Indeed, while the frequency of patients on stage 3 according to the GOLD guidelines [26] was greater in COPD than in HF–COPD (37% vs. 19%), more HF–COPD patients were on stage 1 (4% vs. 32%). Although all patient groups showed reductions in FVC, this was less pronounced in HF vs. HF–COPD and COPD (p b 0.05). 3.2. Exercise responses
and VE–VCO2 slope was tested by Spearman correlation coefficient. A p value less than 0.05 was considered as statistically significant. A Youden Index, a criteria of points on ROC curve closest to the (0,1), was used to detect the best cut off of the y intercept. Statistical
Peak V O2 was lower in HF, HF–COPD and PAH vs. COPD patients
(Table 1). In all five groups, VE–VCO2 slope was inversely correlated to
VEint (Table 2). The describing parameters of the VE–VCO2 relationship
Table 1 Anthropometric and demographic characteristics of studied subjects, spirometry and CPET data. Anthropometric and demographic characteristics
HF n = 108
HF–COPD n = 106
COPD n = 95
PAH n = 85
HEALTHY n = 56
ANOVA
Age (years) Weight (kg) Height (cm) Male (%)
63 ± 10⁎† 79.7 ± 15⁎‡ 171 ± 8⁎†‡ 88⁎ (81%)
66 ± 5⁎† 77 ± 13⁎ 170 ± 7⁎† 93⁎ (88%)
67 ± 8⁎† 72.7 ± 14§ 168 ± 8†§ 83⁎ (87%)
54 ± 18‡§# 68.2 ± 12†§# 166 ± 8†§# 38†‡§# (45%)
50 ± 16‡§# 77.9 ± 13⁎ 175 ± 8⁎‡§# 48⁎ (86%)
b0.0001 b0.0001 b0.0001 b0.0001
Spirometric data FEV1 (L) FEV1 (% predicted value) FVC (L) FVC (% predicted value) FEV1/FVC
2.6 ± 0.6†‡# 90.7 ± 17.7⁎†‡# 3.4 ± 0.8†‡# 82.3 ± 16.2†‡ 0.8 ± 0.1‡#
1.9 ± 0.6⁎†‡§ 68.9 ± 20.1⁎†‡§ 2.9 ± 0.9†§ 75.8 ± 19.6† 0.6 ± 0.1⁎†‡§
1.4 ± 0.4⁎†§# 53.1 ± 13.2⁎†§# 2.7 ± 0.8⁎†§ 71.7 ± 20.8†§ 0.5 ± 0.1⁎†§#
2.4 ± 0.9†‡# 79.2 ± 21.3†‡§# 3.1 ± 1.1†‡ 76.1 ± 19.3† 0.8 ± 0.1‡#
3.5 ± 0.8⁎‡§# 104.3 ± 14.1⁎‡§# 4.4 ± 0.8⁎‡§# 96.0 ± 11.2⁎‡§# 0.8 ± 0.1‡#
b0.0001 b0.0001 b0.0001 b0.0001 b0.0001
15.5 ± 5.1†‡ 58.9 ± 16.6†‡
14.8 ± 4.3†‡ 62.9 ± 21.2†‡
18.6 ± 4.9⁎†§# 75.5 ± 23.4⁎†§#
16.1 ± 6.3†‡ 55.6 ± 22.3†‡
31.4 ± 8.3⁎‡§# 100.9 ± 16.8⁎‡§#
b0.0001 b0.0001
53.2 ± 17.5†
48.2 ± 15.1†
50.5 ± 13.5†
52.5 ± 19.7.1†
84.4 ± 20.8⁎‡§#
b0.0001
Cardiopulmonary exercise test data VO2 peak (mL/kg/min) Peak VO2 (% of predicted) Peak VE (L/min)
HF = heart failure patients; HF–COPD = heart failure patients with coexisting chronic obstructive pulmonary disease; COPD = patients with chronic obstructive pulmonary disease; PAH = pulmonary arterial hypertension patients; HEALTHY = healthy subjects; FEV1 = forced expiratory volume in 1 s, FVC = forced vital capacity, VO2 peak = peak oxygen uptake, peak VE = peak ventilation. ⁎ p ≤ 0.05 vs. PAH. † p ≤ 0.05 vs. healthy subjects. ‡ p ≤ 0.05 vs. COPD. § p ≤ 0.05 vs. HF. # p ≤ 0.05 vs. COPD–HF.
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
in each group are depicted in Fig. 1. The highest and lowest slopes were found in PAH and healthy subjects, respectively. There were no differences in slope among HF, HF–COPD and COPD (32 ± 7, 31 ± 7, and 31 ± 6, respectively; p N 0.05) (Fig. 1a). The fraction of patients with
137
slope values ≥ 34 – the ost used cutoff to indicate poor prognosis
[10] – was substantially higher in PAH (Fig. 2). In contrast, V Eint was higher in HF–COPD and COPD compared to HF and PAH (4.8 ± 2.4 L/min, 5.9 ± 3.0 L/min, 3.0 ± 2.6 L/min, and 2.3 ± 3.3 L/min, respectively; p b 0.01) (Fig. 1b). A ROC curve analysis (AUC = 0.76, p b 0.001) considering all patient
groups identified a cut-off value of V Eint ≥ 4.07 L/min to indicate patients with high probability of having COPD with or without HF (sensitivity of 71.6% and specificity of 72.0%, positive and negative predictive values 74% and 64%, respectively). Indeed, while the prevalence of HF
patients with V Eint ≥ 4.07 L/min did not reach 30%, it was greater than 70% in COPD–HF (Fig. 2). Similar results are obtained if considering only COPD patients with and without HF; in this scenario, however, the best cut-off value was slightly greater (4.11 L/min). Finally, because HF–COPD patients and COPD patients showed significantly different mean FEV1 values (Table 1), we further matched the patients according to FEV1% pred (±5%). We were able to identify 46 pairs of patients, who had an average FEV1% pred of 55.4 ± 14.9 and 55.4 ± 14.8 for HF–COPD
and COPD, respectively. The VE–VCO2 slope was 30.9 ± 6.8 and 32.2 ±
6.6 (p N 0.05) in HF–COPD and COPD patients, respectively, while VEint was 4.9 ± 2.5 L/min in HF–COPD and 5.4 ± 2.8 L/min (p N 0.05) in the COPD group.
4. Discussion This is the first multicenter study systematically contrasting the pa
rameters of the linear VE–VCO2 relationship during incremental cycle
ergometry (V Eint and slope) in a large number of patients with HF, COPD and HF–COPD overlap. Our main results confirm the study hy
pothesis of greater VEint in patients with COPD and COPD–HF than in patients with isolated HF. In contrast, there were no between-group differences in slope, making this parameter poorly discriminative. There
fore, if a VEint ≥4.07 L/min is obtained in an HF patient, the presence of coexistent COPD should be hypothesized and further investigated.
Pulmonary VE is the sum of alveolar (alv) VE and VD (physiological plus anatomical). In healthy subjects, during exercise, physiological VD decreases markedly and anatomical VD remains unchanged (or slightly increases due to large airway radial traction) [27,28]. As sche
matically illustrated in Fig. 3a, VEint represents well exercise VD if the
Fig. 1. Ventilation efficiency. Slope (panel a) and intercept (panel b) of the linear VE–VCO2 relationship in heart failure (HF), HF plus chronic obstructive pulmonary disease (COPD), COPD, pulmonary arterial hypertension (PAH) and healthy subjects. Data are reported as median and 25–75 interquartile range. §p b 0.05, *p b 0.0001.
Fig. 2. Distribution of elevated VE/VCO2 slope and VEint in patients with heart failure (HF), HF plus chronic obstructive pulmonary disease (COPD), COPD, pulmonary arterial hypertension (PAH). Black bars represent fraction of patients with high intercept (Int N 4.07 L/min) and slope (Slp N 34) values by disease group. Note the higher percentage of patients with chronic obstructive pulmonary disease (COPD) and COPD–heart failure (HF) patients showing a high intercept compared to HF alone and pulmonary arterial hypertension (PAH).
138
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
latter remains unchanged or it would be slightly higher if VD decreases during exercise — as in normal subjects. On the other hand, if VD is increased at rest but remains stable during exercise, VEint will increase with minimal changes in VE–VCO2 slope. Indeed, we recently showed in a human study involving normal subjects and HF patients that adding an external dead space upshifted theVE–VCO2 relationship to an amount corresponding to the added dead space (Fig. 3b) [20]. In patients with COPD, a naturally-occurring high VD at the beginning of exercise compounds to a high VEalv driven by a shallower slope. The final consequence is a substantial increase in VEint — as illustrated in Fig. 3c and experimentally demonstrated in COPD patients by Teopompi et al. [29]. These physiological considerations explain why we observed the highest VEint in COPD patients, regardless of the presence of HF (Figs. 1b and 2). A different scenario emerges in those situations where progressive increases in VD during exercise are accompanied by increases in ventilatory drive leading to very steep VE–VCO2 slopes (Fig. 3d and e). This is likely the case in some patients with moderate HF (panel D) or patients
with more severe HF and, frequently, in PAH — as we found in the
present study (Fig. 1) [22,30]. In patients with HF, the VE–VCO2 slope
is normal or has a slight increase, while VEint is positive (Fig. 3d). Interestingly, a high slope has also been found to be predictive of PAH in pa
tients with COPD [31]. In severe PAH, the VE–VCO2 slope progressively
increases during exercise and the VEint becomes negative [18]. This finding can be due to an increase in VD, which takes place progressively throughout the exercise and/or a dynamic increase in ventilatory drive, i.e., a progressive decline in CO2 set point [17] leading to a marked increase in VEalv. Indeed, a negative VEint is likely associated with a very poor prognosis in PAH [18]. The identification of a coexisting COPD in HF patients has a relevant clinical meaning. Indeed, patients with both diseases have a worse prognosis and a lower quality of life than patients with HF or COPD disease alone [2,3]. However, identification of COPD in HF by spirometry may not be an easy task as decreases in FVC secondary to HF-related restriction [5,32] may outpace FEV1 decrements. This might lead to a “pseudo
Fig. 3. Schematic representation of the ventilation (VE) vs. carbon dioxide output (VCO2) relationship in a ramp incremental exercise test. Normal subject (panel a): Lower line represents al
veolar ventilation (VEalv) vs. VCO2 up to the isocapnic buffering period. Upper line shows total ventilation (VEalv + naturally-occurring dead space (VEDSnatural)) vs. VCO2. Dotted lines are ex
trapolation to the Y axis (VE intercepts) for both VEalv and total ventilation. Normal subject with added external dead space (DS) (panel b): In addition to the lines depicted in panel a, the highest
line represents here total ventilation plus the added DS (VEalv + VEDSnatural + VEDSadded) vs. VCO2. Dotted lines are again extrapolations to the Y axis (intercepts) [20]. Patient with moderate-to
severe COPD (panel c): VEalv vs. VCO2 relationship as in panels a and b. In this case, VEalv increase during exercise is blunted and VEDSnatural is higher at rest and remains constant or decreases
during exercise. Note the similarity of high intercept values here and when an external DS was added (panel b). Patient with moderate heart failure (panel d):VEalv vs., VCO2 as in panels a, b and c.
VEDS shows a moderate increase resulting in a slight slope increase with an extrapolation to the Y axis (intercept) close to 0. Patient with severe heart failure or pulmonary arterial hypertension
(panel e): VEalv vs. VCO2 relationship as in upper panels. VEDS increases during exercise. The extrapolation to the Y axis (intercept) of total ventilation has now a negative value [18].
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
normalization” of FEV1/FVC thereby obscuring the diagnosis of obstruction [33,34]. Conversely, a sizeable number of HF patients do present some degree of airflow obstruction long after an acute disease exacerbation [3]. Moreover, HF patients have a reduced lung diffusion capacity – a finding suggestive of emphysema in patients with COPD – due to the impairment of the alveolar capillary membrane function [34]. This
study added to the field by showing that increases in VEint are clinically useful in this specific context. In the present population, we identified
≥4.07 L/min as the VEint cut-off value most able to suggest the presence
of COPD. It should be underlined that, in healthy subjects, VEint is positive with some overlapping with that observed in COPD patients
(Fig. 1b). Therefore, to better define the VEint cut-off value differences between patients with COPD from those without COPD, we excluded
healthy subjects. Consequently, the VEint value observed in normal subjects should be evaluated within the frame of an otherwise normal test.
It remains unclear, however, whether VE–VCO2 slope maintains its prognostic value [10,12,13] in a HF patient with concomitant COPD and, if so, if a lower cut-off value would be more appropriated. For instance, even in the present study where the prevalence of high slopes was not particularly high, the fraction of patients with slopes ≥34 was the lowest in COPD–HF (Fig. 2). The bottom line is that, until clinical studies specifically aimed at clarifying this point are made available, extra care should
value in suggesting COPD–HF in patients in whom spirometry failed to indicate COPD before the exercise testing remains to be demonstrated. As spirometry is not routinely performed in HF patients, it is reasonable
to suggest that if a VEint ≥4.07 L/min is found in a patient with high pretest likelihood of COPD (i.e., smoker or ex-smoker), the patient should be referred for complete respiratory functional evaluation. Finally, the
average VEint we observed in normal subjects was higher than that previously reported by Sun et al. [11] in a large study population. We have no explanation for the present finding, except that the number of normal subjects analyzed in our study was relatively small with a relevant dispersion of data. However, from a clinical point of view, this is a trivial issue, being normal subjects clearly separated from patients considering
clinical characteristics and the VE vs. VCO2 slope.
In conclusion, the finding of an increased VEint of the linear VE–VCO2 relationship during ramp-incremental exercise – regardless of the slope – should raise the suspicion of coexistent COPD in patients with HF and reduced left ventricular ejection fraction. The suggested cut-off V Eint value of ≥ 4.07 L/min should be prospectively tested in population-based studies involving HF patients with different pre-test likelihoods of COPD.
Conflict of interest
be taken with the currently suggested VE–VCO2 slope criteria for poor prognosis in patients with COPD–HF overlap. This study has, naturally, some relevant limitations. First of all, ours is a retrospective study which included consecutive patients recruited in several laboratories and, most importantly, all tests were performed for clinical purposes. Accordingly, some intra- and inter-group heterogeneity on disease severity was present. For instance, it is conceivable that the severity of mechanical abnormalities – which varies markedly across COPD stages – would impact at different extents the ventilatory efficiency parameters [9,22]. Indeed, in the present study, COPD patients had a higher FEV1 than HF–COPD patients. However, the lack of differ
ence in VEint in a sub-group of patients matched by FEV1% pred allowed
us to confirm our premises that FEV1 impairment and VEint are closely interrelated, regardless of the presence of HF. A note of caution, however, should be raised regarding the assumption that FEV1 is a reliable index of obstruction severity in patients with HF–COPD. For instance, FEV1 in these patients is likely to be negatively influenced by FVC decrements induced by HF (Table 1); on the other hand, any increase in elastic recoil due to HF is likely to accelerate the rate of lung emptying, thereby increasing FEV1. The heterogeneity of patients' characteristics might be a study strength, as our subjects well represent real-life conditions. Patients with HF, HF–COPD and COPD showed similar demographic characteristics. As expected from the natural history of the disease, PAH patients were younger and predominantly female. Of note, only a minor increase in the slope has been reported with age in
males and an even smaller one in females; conversely, VEint seems to be independent from age and gender [11]. Consequently, age and gender differences have, if any, a minor influence on our findings. Secondly,
139
we have not analyzed the VE–VCO2 relationship as a ratio at a discrete test point [11], e.g., at the anaerobic threshold or as the lowest test
value (nadir) [11]. However, the slope is, among the VE–VCO2-derived parameters, the most often used for HF prognosis definition, and the anaerobic threshold is frequently difficult to identify in HF patients with severe exercise limitation, periodic breathing and atrial fibrillation [35]. Thirdly, we were unable to give a precise physiological meaning
to the increased V Eint in COPD as we did when adding an external
dead space [20]. In any case, evidence of a high VEint should be consid
ered as a red flag suggestive of COPD. Finally, the 4.07 L/min cut-off for V Eint was derived from well-defined populations with established COPD, HF, COPD–HF and PAH. Whether this cutoff would retain its predictive
None declared. Acknowledgments The research was supported by Centro Cardiologico Monzino, IRCCS, Milano (Italy). References [1] A. Macchia, S. Monte, M. Romero, A. D'Ettorre, G. Tognoni, The prognostic influence of chronic obstructive pulmonary disease in patients hospitalised for chronic heart failure, Eur. J. Heart Fail. 9 (2007) 942–948. [2] F.H. Rutten, M.J. Cramer, J.W. Lammers, D.E. Grobbee, A.W. Hoes, Heart failure and chronic obstructive pulmonary disease: an ignored combination? Eur. J. Heart Fail. 8 (2006) 706–711. [3] N.M. Hawkins, D. Wang, M.C. Petrie, M.A. Pfeffer, K. Swedberg, C.B. Granger, et al., Baseline characteristics and outcomes of patients with heart failure receiving bronchodilators in the CHARM programme, Eur. J. Heart Fail. 12 (2010) 557–565. [4] K. Wasserman, Y.Y. Zhang, A. Gitt, R. Belardinelli, A. Koike, L. Lubarsky, et al., Lung function and exercise gas exchange in chronic heart failure, Circulation 96 (1997) 2221–2227. [5] T.P. Chua, A.J. Coats, The lungs in chronic heart failure, Eur. Heart J. 16 (1995) 882–887. [6] P. Agostoni, G. Cattadori, M. Guazzi, P. Palermo, M. Bussotti, G. Marenzi, Cardiomegaly as a possible cause of lung dysfunction in patients with heart failure, Am. Heart J. 140 (2000) e24. [7] P. Ponikowski, D.P. Francis, M.F. Piepoli, L.C. Davies, T.P. Chua, C.H. Davos, 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 103 (2001) 967–972. [8] G.S. Francis, The relationship of the sympathetic nervous system and the renin– angiotensin system in congestive heart failure, Am. Heart J. 118 (1989) 642–648. [9] J.A. Neder, F.F. Arbex, M.C. Alencar, C.D. O'Donnell, J. Cory, K.A. Webb, et al., Exercise ventilatory inefficiency in mild to end-stage COPD, Eur. Respir. J. 45 (2015) 377–387. [10] R. Arena, J. Myers, J. Abella, M.A. Peberdy, D. Bensimhon, P. Chase, et al., Development of a ventilatory classification system in patients with heart failure, Circulation 115 (2007) 2410–2417. [11] X.G. Sun, J.E. Hansen, N. Garatachea, T.W. Storer, K. Wasserman, Ventilatory efficiency during exercise in healthy subjects, Am. J. Respir. Crit. Care Med. 166 (2002) 1443–1448. [12] D.P. Francis, W. Shamim, L.C. Davies, M.F. Piepoli, P. Ponikowski, S.D. Anker, et al., Cardiopulmonary exercise testing for prognosis in chronic heart failure: continuous and independent prognostic value from VE/VCO(2)slope and peak VO(2), Eur. Heart J. 21 (2000) 154–161. [13] F.X. Kleber, G. Vietzke, K.D. Wernecke, U. Bauer, C. Opitz, R. Wensel, et al., Impairment of ventilatory efficiency in heart failure: prognostic impact, Circulation 101 (2000) 2803–2809. [14] A.K. Pierce, D. Luterman, J. Loudermilk, G. Blomqvist, R.L. Johnson Jr., Exercise ventilatory patterns in normal subjects and patients with airway obstruction, J. Appl. Physiol. 25 (1968) 249–254.
140
A. Apostolo et al. / International Journal of Cardiology 189 (2015) 134–140
[15] P. Paoletti, F. De Filippis, F. Fraioli, A. Cinquanta, G. Valli, P. Laveneziana, et al., Cardiopulmonary exercise testing (CPET) in pulmonary emphysema, Respir. Physiol. Neurobiol. 179 (2011) 167–173. [16] J. Davis, B. Whipp, K. Wasserman, The relation of ventilation to metabolic rate during moderate exercise in man, Eur. J. Appl. Physiol. 44 (1980) 97–108. [17] B.J. Whipp, S.A. Ward, K. Wasserman, Ventilatory responses to exercise and their control in man, Am. Rev. Respir. Dis. 129 (1984) S17–S20. [18] P. Agostoni, S. Farina, A. Apostolo, S. Sciomer, Inside ventilatory regulation in pulmonary hypertension: several hidden data are still undiscovered, Eur. J. Prev. Cardiol. 21 (2014) 268–271. [19] J.A. Neder, L.E. Nery, C. Peres, B.J. Whipp, Reference values for dynamic responses to incremental cycle ergometry in males and females aged 20 to 80, Am. J. Respir. Crit. Care Med. 164 (2001) 1481–1486. [20] P. Gargiulo, A. Apostolo, P. Perrone-Filardi, S. Sciomer, P. Palange, P. Agostoni, A non invasive estimate of dead space ventilation from exercise measurements, PLoS One 9 (2014) e87395. [21] E. Teopompi, P. Tzani, M. Aiello, S. Ramponi, D. Visca, M.R. Gioia, et al., Ventilatory response to carbon dioxide output in subjects with congestive heart failure and in patients with COPD with comparable exercise capacity, Respir. Care 59 (2014) 1034–1041. [22] S. Paolillo, S. Farina, M. Bussotti, A. Iorio, P. PerroneFilardi, M.F. Piepolil, et al., Exercise testing in the clinical management of patients affected by pulmonary arterial hypertension, Eur. J. Prev. Cardiol. 19 (2012) 960–971. [23] P. Laveneziana, G. Garcia, B. Joureau, F. Nicolas-Jilwan, T. Brahimi, L. Laviolette, et al., Dynamic respiratory mechanics and exertional dyspnoea in pulmonary arterial hypertension, Eur. Respir. J. 41 (2013) 578–587. [24] G. Simonneau, M.A. Gatzoulis, I. Adatia, D. Celermajer, C. Denton, A. Ghofrani, et al., Updated clinical classification of pulmonary hypertension, J. Am. Coll. Cardiol. 62 (2013) D34–D41. [25] R. Pellegrino, G. Viegi, V. Brusasco, R.O. Crapo, F. Burgos, R. Casaburi, et al., Interpretative strategies for lung function tests, Eur. Respir. J. 26 (2005) 948–968.
[26] J. Vestbo, S.S. Hurd, A.G. Agusti, P.W. Jones, C. Vogelmeier, A. Anzueto, et al., Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary, Am. J. Respir. Crit. Care Med. 187 (2013) 347–365. [27] P. Agostoni, R. Pellegrino, C. Conca, J.R. Rodarte, V. Brusasco, Exercise hyperpnea in chronic heart failure: relationships to lung stiffness and expiratory flow limitation, J. Appl. Physiol. 92 (2002) 1409–1416. [28] R. Pellegrino, C. Villosio, U. Milanese, G. Garelli, J.R. Rodarte, V. Brusasco, Breathing during exercise in subjects with mild-to-moderate airflow obstruction: effects of physical training, J. Appl. Physiol. (1985) 87 (1999) 1697–1704. [29] E. Teopompi, P. Tzani, M. Aiello, M.R. Gioia, E. Marangio, A. Chetta, Excess ventilation and ventilatory constraints during exercise in patients with chronic obstructive pulmonary disease, Respir. Physiol. Neurobiol. 197 (2014) 9–14. [30] J.E. Hansen, X.G. Sun, Y. Yasunobu, R.P. Garafano, G. Gates, R.J. Barst, et al., Reproducibility of cardiopulmonary exercise measurements in patients with pulmonary arterial hypertension, Chest 126 (2004) 816–824. [31] S. Holverda, H.J. Bogaard, H. Groepenhoff, P.E. Postmus, A. Boonstra, A. VonkNoordegraaf, Cardiopulmonary exercise test characteristics in patients with chronic obstructive pulmonary disease and associated pulmonary hypertension, Respiration 76 (2008) 160–167. [32] T.P. Olson, K.C. Beck, B.D. Johnson, Pulmonary function changes associated with cardiomegaly in chronic heart failure, J. Card. Fail. 13 (2007) 100–107. [33] T.H. Le Jemtel, M. Padeletti, S. Jelic, Diagnostic and therapeutic challenges in patients with coexistent chronic obstructive pulmonary disease and chronic heart failure, J. Am. Coll. Cardiol. 49 (2007) 171–180. [34] P. Agostoni, M. Bussotti, G. Cattadori, E. Margutti, M. Contini, M. Muratori, et al., Gas diffusion and alveolar–capillary unit in chronic heart failure, Eur. Heart J. 27 (2006) 2538–2543. [35] P. Agostoni, U. Corra, G. Cattadori, F. Veglia, E. Battaia, R. La Gioia, et al., Prognostic value of indeterminable anaerobic threshold in heart failure, Circ. Heart Fail. 6 (2013) 977–987.
