by Excerpta Medica, Inc. All rights reserved /04/$ see front matter

Transcription

1 Differential Contribution of Dead Space Ventilation and Low Arterial pco 2 to Exercise Hyperpnea in Patients With Chronic Heart Failure Secondary to Ischemic or Idiopathic Dilated Cardiomyopathy Roland Wensel, MD, Panagiota Georgiadou, MD, Darrel P. Francis, MA, MRCP, Stephanie Bayne, Adam C. Scott, PhD, Sabine Genth-Zotz, MD, Stefan D. Anker, MD, PhD, Andrew J.S. Coats, DM, and Massimo F. Piepoli, MD, PhD From the Department of Clinical Cardiology, National Heart & Lung Institute, Imperial College of Science, Technology and Medicine, London, United Kingdom; Klinik und Poliklinik für Innere Medizin II, Klinikum der Universität Regensburg, Regensburg, Germany; Medizinische Klinik II, Johannes Gutenberg Universität Mainz, Mainz, Germany; and Universitätsklinik (Charité, Campus Virchow) der Humboldt-Universität zu Berlin, Berlin, Germany. Dr. Wensel was supported by the Ernst Schering Research Foundation, Berlin, Germany. Dr. Piepoli was supported by the Wellcome Trust Advanced Research Fellowship, London, United Kingdom. Dr. Georgiadou was supported by the Wellcome Trust, London, United Kingdom. Dr. Coats was supported by the Viscount Royston Trust, London, United Kingdom. Manuscript received May 16, 2003; revised manuscript received and accepted September 29, Address for reprints: Roland Wensel, MD, Klinik und Poliklinik für Innere Medizin II, Klinikum der Universität Regensburg, Franz-Josef- Strau -Allee11,93053Regensburg,Germany. roland.wensel@ klinik.uni-regensburg.de. In chronic heart failure (CHF), the abnormally large ventilatory response to exercise (VE/VCO 2 slope) has 2 conceptual elements: the requirement of restraining arterial partial pressure of carbon dioxide (pco 2) from increasing (because of an increased ratio between increased physiologic dead space and tidal volume [VD/ VT]) and the depression of arterial pco 2 by further increased ventilation, which necessarily implies an important non-carbon dioxide stimulus to ventilation. We aimed to assess the contribution of these 2 factors in determining the elevated VE/VCO 2 slope in CHF. Thirty patients with CHF underwent cardiopulmonary exercise testing (age years, left ventricular ejection fraction 34 15%, peak oxygen uptake ml/kg/min, VE/VCO 2 slope 36.4). At rest and during exercise, arterial pco 2 was measured and VD was calculated and separated into serial and alveolar compo- bnormally enhanced ventilatory response to ex- is a characteristic feature of chronic heart Aercise failure (CHF), but its mechanism has not been universally agreed upon. The conventional view is that an underlying defect is inefficiency of gas exchange, necessitating an increase in ventilation to maintain arterial partial pressure of carbon dioxide (pco 2 ). 1,2 nents. VD/VT decreased from 0.57 at rest to 0.44 at peak exercise (p <0.01). VE/VCO 2 slope was correlated with peak exercise VD/VT (r 0.67), the serial VD/VT ratio (r 0.64), and alveolar VD/VT ratio (r 0.51) at peak exercise (all p <0.01). VE/VCO 2 slope was also correlated with arterial pco 2 (r 0.75, p <0.001). Despite this, arterial pco 2 was not related to peak oxygen uptake (r 0.2) or to arterial lactate (r 0.25) and only weakly to New York Heart Association functional class (F 3.7). First, the increased VE/VCO 2 slope was caused by both the high VD/VT ratio and by other mechanisms, as shown by low arterial pco 2 during exercise. Second, this latter component (depression of arterial pco 2 ) was not related to conventional measures of heart failure severity by Excerpta Medica, Inc. (Am J Cardiol 2004;93: ) The alternative view is that there is an abnormal stimulus for ventilation ( primary hyperventilation ) that overpowers the mechanisms that normally control arterial pco 2, thus causing arterial pco 2 to decrease. 3 The key finding supporting the inefficiency of gas exchange view is that physiologic dead space (VD) increases as a proportion of tidal volume (VT) as heart failure worsens. 1,2 Moreover, arterial pco 2 is not lower in the sickest patients, 1 suggesting that severe disease per se does not generate hyperventilation. The alternative primary hyperventilation view recognizes that, during exercise in heart failure, arterial pco 2 is no longer tightly regulated itself but becomes determined by other processes, for example, increased sensitivities of chemo- and ergoreflexes. 3 6 The key prediction of this latter view, namely, that arterial pco 2 will be lowest in patients with the greatest exercise ventilatory response, has never been truly tested. 1,2 We therefore set out to test these competing hypotheses by looking specifically at the relation between arterial pco 2 and the exercise ventilatory response. We then quantified systemic acidosis, which is a plausible stimulus for ventilation. We also assessed the serial and alveolar components of VD by Excerpta Medica, Inc. All rights reserved /04/$ see front matter The American Journal of Cardiology Vol. 93 February 1, 2004 doi: /j.amjcard

2 TABLE 1 Clinical Characteristics and Exercise Test Results of the Study Population Heart Failure Patients (n 30) METHODS Patients: We studied 30 patients (Table 1) (ages years; left ventricular ejection fraction 34 15%; 29 men) with CHF due to ischemic heart disease (n 22) or dilated cardiomyopathy (n 8). New York Heart Association (NYHA) functional classification showed 5 patients in class I, 14 in class II, and 11 in class III. Forced expiratory 1-second volume and forced vital capacity were 79 17% and 86 16% of predicted values, respectively. Patients medications included angiotensin-converting enzyme inhibitors (n 25), angiotensin II receptor antagonists (n 5), blockers (n 18), amiodarone (n 5), loop diuretics (n 25), spironolactone (n 10), aspirin (n 16), digitalis (n 5), hydroxymethylglutaryl coenzyme A-reductase inhibitors (statins; n 17), nitrates (n 7), and warfarin (n 11). Medication had been constant for 6 weeks before the study. An age-matched group of 39 healthy volunteers (Table 1) was studied to show how the VE/VCO 2 relation during exercise behaves in subjects without heart failure. The local ethics committee has approved the study protocol, which conformed to the protocols of the Declaration of Helsinki. All patients gave written informed consent. Exercise testing: After 5 minutes of standing at rest, a symptom-limited treadmill cardiopulmonary exercise test was performed in all patients. We used the modified Naughton protocol, which is an incremental test with stages of 2 minutes and increments in both slope and velocity of the treadmill that simulate an increment of 1 MET ( 3.5 ml oxygen kg 1 min 1 ) per stage. Pulmonary gas exchange was analyzed using a metabolic cart consisting of a flowmeter and a mass spectrometer (AMIS 2000, Innovision, Odense, Denmark). Patients breathed through a mouthpiece (volume 35 ml). Raw data (flow, oxygen and carbon dioxide concentrations) were stored for later offline breath-by-breath analysis of gas exchange and calculation of serial dead space. Peak oxygen uptake was defined as the highest 30-second average of oxygen uptake in the last minute of exercise. The anaerobic threshold was calculated using the V-slope method. 7 When no clear kink in the VCO 2 /VO 2 relation was observed, the start of the increase in end tidal po 2 and the increase of the ventilatory equivalent for oxygen were used to determine the anaerobic threshold. VE/VCO 2 ratios were calculated at rest, at the Healthy Controls (n 39) Age (yrs) Men 29 (97%) 36 (92%) Left ventricular ejection fraction, % Peak oxygen uptake (ml kg 1 min 1 ) * Oxygen uptake at anaerobic threshold * (ml kg 1 min 1 ) Respiratory exchange ratio at peak exercise * VE/VCO 2 slope 36.4* 25 *p 0.05 versus healthy controls. anaerobic threshold, and at peak exercise. The VE/VCO 2 slope was measured as the slope of the linear regression relating ventilation to VCO 2. 8 Measurement of dead space ventilation: A 20-gauge arterial catheter (Leader; Vygon, Ecoven, France) was inserted into the radial artery. Arterial blood samples were obtained at rest and during the second minute of each stage of exercise. Each sample was drawn over a period of approximately 10 seconds, and when the respiratory rate was 30/min, special care was taken to draw the sample over an integral number of respiratory cycles. Arterial blood gases, lactate, and hydrogen ion concentrations were measured with electrodes by standard methods, and physiologic VD was calculated from gas exchange and arterial pco 2. Serial dead space was measured from the expiratory capnogram using the modified Fowler technique. 9,10 In brief, this method consists of construction of a sharp boundary between serial dead space and alveolar gas by graphic integration. Alveolar dead space was then defined as the difference between physiologic and serial dead space. The ratio of VD/VT, between VD and VT, was then calculated. Statistical analysis: Data are shown as mean SD. The changes in the parameters during exercise have been analyzed by repeated-measures analysis of variance. The patient dead space to tidal volume ratios, arterial pco 2, lactate and hydrogen ion concentrations were correlated with VE/VCO 2 ratio, VE/VCO 2 slope and peak oxygen uptake using the Pearson correlation. When data did not follow normal distribution the median value is shown and Friedman repeated measures analysis of variance on ranks and the Spearman correlation was used, respectively. Differences in VD/VT and arterial pco 2 with respect to NYHA class were analyzed with an analysis of variance. A p value 0.05 was considered significant. RESULTS The results of the exercise tests are given in Tables 1and 2. Compared with normal controls, patients with heart failure had a higher VE/VCO 2 slope and higher VE/VCO 2 ratios throughout the exercise test. In both groups the VE/VCO 2 ratios significantly decreased from at rest to the anaerobic threshold. From the anaerobic threshold to peak exercise, no further reduction occurred in both groups. Contribution of VD/VT and arterial pco 2 to the ventilatory response in heart failure: We examined whether the exercise ventilatory response was related to VD (which would support the inefficiency of gas exchange hypothesis) or to arterial pco 2 (which would support the primary hyperventilation hypothesis). At rest, the VE/VCO 2 ratio correlated significantly only with the VD/VT ratio (r 0.75, p 0.001) but not with arterial pco 2 (r 0.36, p 0.053). On multiple linear regression analysis, VD/VT and arte- HEART FAILURE/EXERCISE HYPERPNEA IN HEART FAILURE 319

3 FIGURE 2. Relation between peak oxygen uptake and VE/VCO 2 slope. FIGURE 1. Relation between (A) arterial pco 2 and VE/VCO 2 slope; (B) VD/VT and VE/VCO 2 slope; (C) arterial pco 2 and peak oxygen uptake; and (D) VD/VT and peak oxygen uptake. rial pco 2 (all p 0.001) predicted the VE/VCO 2 ratio. Neither the VD/VT ratio at rest nor the arterial pco 2 at rest predicted the exercise VE/VCO 2 slope (r 0.1, p 0.59; and r 0.01, p 0.96, respectively). At the anaerobic threshold, the VE/VCO 2 ratio correlated clearly with the VD/VT ratio (r 0.9, p 0.001) but not with arterial pco 2 (r 0.33, p 0.077), but on multiple linear regression analysis both variables predicted VE/VCO 2 (all p 0.001). Both VD/VT ratio and arterial pco 2 at the anaerobic threshold predicted the exercise VE/VCO 2 slope (r 0.51, p 0.01 and r 0.65, p 0.005, respectively). At peak exercise, the VE/VCO 2 ratio correlated with both the VD/VT ratio and arterial pco 2 (r 0.88, p and r 0.60, p 0.001). VD/VT and arterial pco 2 at peak exercise were also clearly correlated with VE/VCO 2 slope (Figure 1). VD/VT correlated not only with VE/VCO 2 slope but also with peak oxygen uptake (Figure 1) and NYHA class (F 13.7, p 0.001). Although VE/ VCO 2 slope and peak oxygen uptake did correlate (Figure 2), arterial pco 2, however, showed a significant correlation only with VE/VCO 2 slope and not with peak oxygen uptake (Figure 1), and only a weak relation to NYHA class (F 3.7, p 0.04). Contribution of systemic acidosis to the ventilatory response in heart failure: If systemic lactic acidosis was the cause of the exercise ventilatory response, we would expect plasma lactate and hydrogen ion levels to correlate positively with VE/VCO 2 slope and negatively with arterial pco 2. Our results, however, were that plasma lactate levels at peak exercise were not correlated with a high VE/VCO 2 slope (r 0.04), to low arterial pco 2 (r 0.25, p 0.2), or to the increase in the VE/VCO 2 ratio from the anaerobic threshold to peak exercise (r 0.004). In contrast, plasma hydrogen ion concentrations showed a significant tendency to be lower in patients with higher VE/VCO 2 slopes (r 0.46, p 0.05) or lower arterial pco 2 at peak exercise (r 0.43, p 0.05). The same relation was seen at the anaerobic threshold (lactate vs VE/VCO 2 slope: r 0.19, p 0.3; lactate vs arterial pco 2 :r 0.14, p 0.45; hydrogen vs VE/VCO 2 slope: r 0.42, p 0.05; hydrogen vs arterial pco 2 :r 0.41, p 0.05). The increase in arterial lactate concentrations from the anaerobic threshold to peak exercise was correlated with the concomitant decrease in arterial pco 2 (r 0.43, p 0.05) but not with the concomitant increase in VE/ VCO 2 ratio (r 0.09). Relative contribution of the serial and alveolar components of physiologic dead space: At rest, the VD/VT was (range 0.28 to 0.8), which consisted of a serial dead space to VT ratio of and an alveolar dead space to VT ratio of During exercise, VD/VT decreased to at the anaerobic threshold (p 0.05 vs rest) with no further reduction at peak exercise ( , p 0.05 vs rest, p 0.9 vs anaerobic threshold). The change at the anaerobic threshold resulted from a reduction of both serial (to , p 0.05 vs rest) and alveolar (to , p 0.05 vs rest) components of the VD/VT ratio. There was no significant change in these components between the anaerobic threshold and peak exercise, although they remained significantly different from at rest (serial dead space to VT ratio ; alveolar dead space to VT ratio , both p 0.05 vs rest). VE/VCO 2 slope was correlated not only with VD/VT at peak exercise (r 0.67, p 0.001) and at the anaerobic threshold (r 0.51, p 0.01) but also with serial dead space to VT ratio (r 0.64, p and r 0.45, p 0.05, respectively) and 320 THE AMERICAN JOURNAL OF CARDIOLOGY VOL. 93 FEBRUARY 1, 2004

4 TABLE 2 VE/VCO 2, Arterial pco 2, Lactate, and ph at Different Stages of Exercise At Rest Anaerobic Threshold Peak Exercise VE/VCO 2 Heart failure * 40.7* Controls 34 27* 29* Arterial pco 2 (mm/hg) Lactate (mmol/l) * * ph *p 0.05 versus at rest; p 0.05 versus anaerobic threshold; p 0.05 versus at rest and anaerobic threshold; p 0.05 versus healthy controls. No blood gas data were obtained from the control group. with the alveolar dead space to VT ratio (r 0.51, p 0.01 and r 0.42, p 0.05, respectively). A high serial dead space to VT ratio at peak exercise had 2 contributing factors; VE/VCO 2 slope was correlated low VT (r 0.44, p 0.05) and showed a trend toward higher serial dead space (r 0.33, p 0.076). FIGURE 3. Time course of the VE/VCO 2 ratio during the incremental exercise test in 1 patient. The VE/VCO 2 ratio is relatively stable at rest (phase I), starts to decrease at the beginning of exercise and reaches a plateau (phase II), and increases again toward the end of exercise (phase III). DISCUSSION In our study, patients with a higher VE/VCO 2 slope had both an increased VD/VT ratio and lower arterial carbon dioxide during exercise. This cannot be explained by isolated inefficiency of gas exchange, which would cause the former, but certainly not the latter. Nor can it be explained by an isolated primary hyperventilation during exercise, (which would cause the latter, but not directly cause the former. Thus, the abnormally high ventilatory response in patients with steep VE/VCO 2 slopes may most plausibly result from a combination of gas exchange inefficiency (synonymous with an increase in VD/VT) and a primary increase in ventilatory drive on exercise (synonymous with a decrease in arterial pco 2 ). Ventilatory response to exercise is often measured as the slope of the VE/VCO 2 relation, because this relation is almost linear, and, therefore, the VE/VCO 2 slope provides a stable measure of the required increase in ventilation for an increase in carbon dioxide output during exercise. However, as evident from our data and that of others, 11 the VE/VCO 2 ratio changes considerably during exercise (Table 2 and Figure 3), and the VE/VCO 2 relation is not linear (Figure 4). The change of the VE/VCO 2 ratio during exercise results from corresponding changes of the VD/VT ratio and arterial pco 2. The VE/VCO 2 slope gives a single value for the potentially complex VE/VCO 2 relation during exercise. Because it is a simplification, it is of limited value in delineating the mechanisms that determine the VE/VCO 2 ratio at different stages of exercise. At rest and during aerobic exercise, the VE/ VCO 2 ratio is predominantly related to the VD/VT ratio. As anaerobic metabolism ensues, the hyperventilatory response affects the VE/VCO 2 ratio, and the 2 variables are clearly related at peak exercise. Despite being present only during anaerobic exercise, hyperventilation leads to an increase in the overall VE/ VCO 2 slope, and, consequently, there is a significant relation between VE/VCO 2 slope and the degree of hyperventilation. Our data (Figure 1) can be reconciled with previous studies 1,2 that seemed to reject the possibility of a primary hyperventilatory mechanism; the answer may lie in Figure 1. If a primary hyperventilatory mechanism were present, the arterial pco 2 would be expected to be low in the affected patients. Previous studies have examined this low arterial carbon dioxide level, but sought a correlation with peak oxygen uptake 1 or NYHA class; 2 these studies found no significant relation and thus rejected the hypothesis. However, the primary hyperventilation hypothesis predicts only that low arterial carbon dioxide would correlate with high ventilatory responses (namely, VE/VCO 2 slope). It does not directly predict whether there should be a relation between arterial carbon dioxide and either peak oxygen uptake or NYHA class. Our finding that arterial pco 2 is lower in patients with increased VE/VCO 2 slope (Figure 1) confirms the existence of a primary hyperventilatory mechanism. It may at first seem surprising that arterial pco 2 was correlated with VE/VCO 2 slope but not with peak oxygen uptake when the latter 2 are generally highly correlated, as seen in this study (Figure 2). However the physiopathologic mechanisms responsible for VE/ VCO 2 slope and the peak oxygen uptake are not identical. 8,12,13 The fact that arterial pco 2 correlates with VE/VCO 2 slope and not with peak oxygen uptake, indicates that even although arterial pco 2 is not significantly linked to conventional severity, it still varies from patient to patient, and this variation cova- HEART FAILURE/EXERCISE HYPERPNEA IN HEART FAILURE 321

5 FIGURE 4. Nonlinearity of the VE/VCO 2 relation during exercise. (A) Despite the high r value, the different phases of the VE/VCO 2 ratio, as shown in Figure 3, are apparent as segments of the VE/VCO 2 plot that clearly are not identical in terms of their regression slope. Omission of phase I (B), and of phases I and III (C) result in a change in the regression slope. ries with VE/VCO 2 slope. In other words, arterial pco 2 has a specific link to VE/VCO 2 slope that is not the result of confounding by conventional markers of disease severity. What is the origin of the primary hyperventilatory stimulus that decreases the arterial pco 2 in some patients? Hyperventilation becomes a relevant factor only once anaerobic metabolism increases, suggesting that it does not result from the volitional aspects of exercise but from the systemic or local (muscular) accumulation of anaerobic metabolites. Decreased aerobic exercise capacity and early development of systemic lactic acidosis is characteristic of patients with heart failure However, we observed no significant relation between arterial pco 2 and plasma lactate levels at peak exercise. Patients with a lower arterial pco 2 had a lower (rather than higher) hydrogen ion concentration. These findings are evidence against systemic lactic acidosis as the key drive to excess ventilation. An alternative stimulus to excess ventilation is activation of the muscle metaboreflex (ergoreflex), which has been shown to stimulate ventilation. 5 Recently, we found that accumulation of hydrogen in the skeletal muscle during exercise is a major trigger of the ergoreflex afferents (located in the muscular interstitium), and prevention of this accumulation completely abolished the ergoreflex-mediated ventilatory response to exercise. 17 Interstitial muscular concentrations of lactate and hydrogen differ considerably from concentrations in arterial blood. 18,19 Hence, arterial lactate accumulation might coincide with the start of ergoreflex activation during exercise without predicting the magnitude of the ergoreflex response. Although the mechanism remains uncertain, it can be concluded that hyperventilation contributes to the enhanced ventilatory response to exercise in heart failure. Our data confirm that it is the alveolar component that is the major contributor to the high VD/VT ratio in patients with heart failure. It has been proposed that this may result from suboptimal distribution of blood flow to alveoli (ventilation/perfusion mismatch). 20 In our study, we observed that only the peak exercise alveolar VD/VT ratio was correlated with VE/ VCO 2 slope; the values at rest did not. This implies that the pattern of ventilation/perfusion mismatch at rest is not a useful predictor of exercise response. It is noteworthy that high VE/VCO 2 slope is related to a lower cardiac output during exercise. 2,21 It could be speculated that ventilation/perfusion matching can only improve with exercise if pulmonary blood flow increases adequately. More specifically, high VE/ VCO 2 slopes have been related to pulmonary vasoconstriction and pulmonary hypertension in CHF. 22 The high VE/VCO 2 slope in patients with CHF results from the combination of increased VD ventilation, which is related to impaired pulmonary hemodynamics and augmented hyperventilation, which is related to the peripheral changes that occur in are characteristic of the syndrom of CHF. This underlying pathophysiology explains the outstanding value of the VE/VCO 2 slope in the assessment of disease severity and prognosis of patients with CHF. 1. Wasserman K, Zhang YY, Gitt A, Belardinelli R, Koike A, Lubarsky L, Agostoni PG. Lung function and exercise gas exchange in chronic heart failure. Circulation 1997;96: Tanabe Y, Hosaka Y, Ito M, Ito E, Suzuki K. Significance of end-tidal P(CO(2)) response to exercise and its relation to functional capacity in patients with chronic heart failure. Chest 2001;119: Clark AL, Poole-Wilson PA, Coats AJ. Exercise limitation in chronic heart failure: central role of the periphery. J Am Coll Cardiol 1996;28: Chua TP, Clark AL, Amadi AA, Coats AJ. Relation between chemosensitivity 322 THE AMERICAN JOURNAL OF CARDIOLOGY VOL. 93 FEBRUARY 1, 2004

6 and the ventilatory response to exercise in chronic heart failure. J Am Coll Cardiol 1996;27: Piepoli M, Clark AL, Volterrani M, Adamopoulos S, Sleight P, Coats AJ. 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;93: Piepoli M, Clark AL, Coats AJ. Muscle metaboreceptors in hemodynamic, autonomic, and ventilatory responses to exercise in men. Am J Physiol 1995;269: H1428 H Beaver WL, Wasserman K, Whipp BJ. A new method for detecting anaerobic threshold by gas exchange. J Appl Physiol 1986;60: Chua TP, Ponikowski P, Harrington D, Anker SD, Webb-Peploe K, Clark AL, Poole-Wilson PA, Coats AJ. Clinical correlates and prognostic significance of the ventilatory response to exercise in chronic heart failure. J Am Coll Cardiol 1997;29: Fowler WS. Lungfunction Studies. II. The respiratory dead space. Am J Physiol 1948;154: Heller H, Könen-Bergmann M, Schuster KD. An algebraic solution to dead space determination according to Fowler s graphical method. Comput Biomed Res 1999;32: Clark AL, Poole-Wilson PA, Coats AJ. Relation between ventilation and carbon dioxide production in patients with chronic heart failure. J Am Coll Cardiol 1992;20: Francis DP, Shamim W, Davies LC, Piepoli MF, Ponikowski P, Anker SD, Coats AJ. 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 2000;21: Ponikowski P, Francis DP, Piepoli MF, Davies LC, Chua TP, Davos CH, Florea V, Banasiak W, Poole-Wilson PA, Coats AJ, Anker SD. 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;103: Massie BM, Conway M, Rajagopalan B, Yonge R, Frostick S, Ledingham J, Sleight P, Radda G. Skeletal muscle metabolism during exercise under ischemic conditions in congestive heart failure. Evidence for abnormalities unrelated to blood flow. Circulation 1988;78: Sullivan MJ, Green HJ, Cobb FR. Altered skeletal muscle metabolic response to exercise in chronic heart failure. Relation to skeletal muscle aerobic enzyme activity. Circulation 1991;84: Wilson JR, Mancini DM. Factors contributing to the exercise limitation of heart failure. J Am Coll Cardiol 1993;22(4 suppl A):93A 98A. 17. Scott AC, Wensel R, Davos CH, Georgiadou P, Kemp M, Hooper J, Coats AJ, Piepoli MF. Skeletal muscle reflex in heart failure patients: role of hydrogen. Circulation 2003;107: MacLean DA, Bangsbo J, Saltin B. Muscle interstitial glucose and lactate levels during dynamic exercise in humans determined by microdialysis. J Appl Physiol 1999;87: Street D, Bangsbo J, Juel C. Interstitial ph in human skeletal muscle during and after dynamic graded exercise. J Physiol 2001;537: Wada O, Asanoi H, Miyagi K, Ishizaka S, Kameyama T, Seto H, Sasayama S. Importance of abnormal lung perfusion in excessive exercise ventilation in chronic heart failure. Am Heart J 1993;125: Metra M, Dei Cas L, Panina G, Visioli O. Exercise hyperventilation chronic congestive heart failure, and its relation to functional capacity and hemodynamics. Am J Cardiol 1992;70: Reindl I, Wernecke KD, Opitz C, Wensel R, Konig D, Dengler T, Schimke I, Kleber FX. Impaired ventilatory efficiency in chronic heart failure: possible role of pulmonary vasoconstriction. Am Heart J 1998;136: HEART FAILURE/EXERCISE HYPERPNEA IN HEART FAILURE 323