Intercostal Muscle Blood Flow Limitation during Exercise in Chronic Obstructive Pulmonary Disease

Transcription

1 Intercostal Muscle Blood Flow Limitation during Exercise in Chronic Obstructive Pulmonary Disease Ioannis Vogiatzis 1,2, Dimitris Athanasopoulos 1,2, Helmut Habazettl 3,4, Andrea Aliverti 5, Zafiris Louvaris 1,2, Evgenia Cherouveim 2, Harrieth Wagner 6, Charis Roussos 1, Peter David Wagner 6, and Spyros Zakynthinos 1 1 Department of Critical Care Medicine and Pulmonary Services, Evangelismos Hospital, M. Simou, and G.P. Livanos Laboratories, National and Kapodistrian University of Athens; 2 Department of Physical Education and Sport Sciences, National and Kapodistrian University of Athens, Greece; 3 Institute of Physiology, Charité Campus Benjamin Franklin, Berlin; 4 Institute of Anesthesiology, German Heart Institute, Berlin, Germany; 5 Dipartimento di Bioingegneria, Politecnico di Milano, Milano, Italy, and 6 Department of Medicine, University of California San Diego, La Jolla, California Rationale: It has been hypothesized that, because of the high work of breathing sustained by patients with chronic obstructive pulmonary disease (COPD) during exercise, blood flow may increase in favor of the respiratory muscles, thereby compromising locomotor muscle blood flow. Objectives: To test this hypothesis by investigating whether, at the same work of breathing, intercostal muscle blood flow during exercise is as high as during resting isocapnic hyperpnea when respiratory and locomotor muscles do not compete for the available blood flow. Methods: Intercostal and vastus lateralis muscle perfusion was measured simultaneously in 10 patients with COPD (FEV % predicted) by near-infrared spectroscopy using indocyanine green dye. Measurements and Main Results: Measurements were made at several exercise intensities up to peak work rate (WRpeak) and subsequently during resting hyperpnea at minute ventilation levels up to those at WRpeak. During resting hyperpnea, intercostal muscle blood flow increased with the power of breathing to ml/min per 100 g at the same ventilation recorded at WRpeak. Conversely, during graded exercise, intercostal muscle blood flow remained unchanged from rest up to 50% WRpeak ( ml/min per 100 g) and then fell to ml/min per 100 g at WRpeak (P ). Cardiac output plateaued above 50% WRpeak ( l/min), whereas vastus lateralis muscle blood flow increased progressively, reaching ml/min per 100 g at WRpeak. Conclusions: During intense exercise in COPD, restriction of intercostal muscle perfusion but preservation of quadriceps muscle blood flow along with attainment of a plateau in cardiac output represents the inability of the circulatory system to satisfy the energy demands of locomotor and respiratory muscles. Keywords: COPD; exercise; intercostal muscle blood flow; quadriceps muscle blood flow; NIRS In patients with chronic obstructive pulmonary disease (COPD), exercise intolerance is multifactorial, involving ventilatory, gas exchange, cardiovascular, and peripheral muscle abnormalities that prevent the increased oxygen and carbon dioxide transport demands of the peripheral muscles to be met (1 3). In the majority of patients with COPD, abnormal lung mechanics and (Received in original form February 4, 2010; accepted in final form July 8, 2010) Supported by Thorax Foundation and by grants from the A. Perotti visiting Professorship fund of the Thorax Foundation. Correspondence and requests for reprints should be addressed to Ioannis Vogiatzis, Ph.D., Thorax Foundation, 3 Ploutarhou Str Athens, Greece. gianvog@phed.uoa.gr This article has an online supplement, which is accessible from this issue s table of contents at Am J Respir Crit Care Med Vol 182. pp , 2010 Originally Published in Press as DOI: /rccm OC on July 9, 2010 Internet address: AT A GLANCE COMMENTARY Scientific Knowledge on the Subject In healthy subjects, restriction of blood flow to the intercostal muscles during heavy exercise reflects the inability of the circulatory system to meet the increasing energy demands of respiratory and locomotor muscles. What This Study Adds to the Field The study demonstrates that intercostal muscle blood flow is restricted during high-intensity exercise while quadriceps muscle perfusion is preserved. This happens under conditions of restricted cardiac output, possibly occurring secondary to the increase in expiratory abdominal muscle recruitment, and challenges the concept of blood flow redistribution in favor of the respiratory muscles during exercise in patients with chronic obstructive pulmonary disease. expiratory flow limitation during exercise increase the work of breathing several-fold, thereby exacerbating the competition between the respiratory and the locomotor muscles for the available energy and oxygen supplies (4). In this context, unloading the respiratory muscles during exercise (by means of heliox breathing or by proportional assisted ventilation) is associated with improved peripheral muscle oxygenation and exercise performance in COPD (5, 6). These findings have been interpreted to indicate that a significant fraction of the cardiac output, otherwise devoted to the respiratory muscles during exercise, can be redirected to the peripheral muscles. However, actual respiratory or locomotor muscle blood flow measurements have not been performed (5, 6). In healthy, fit subjects (7), measuring intercostal muscle perfusion (8) revealed that, at essentially the same work of breathing, blood flow to the intercostal muscles during maximal exercise was lower compared with that during voluntary resting hyperpnea (i.e., when the locomotor muscles were inactive). This finding was taken as evidence against the concept of blood flow redistribution in favor of the respiratory muscles during exercise in healthy subjects (9, 10). Whether the same finding would apply to patients with COPD, who normally experience greater work of breathing than healthy subjects at rest and during exercise, is unknown. The purpose of the present study was to investigate, in patients with COPD, whether during exercise intercostal muscle blood flow is different from that recorded at the same respiratory muscle load during resting hyperpnea (i.e., when respiratory and locomotor muscles do not compete for the

2 1106 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL available blood flow and thus respiratory muscle blood flow is expected to be the maximal attainable). Simultaneous measurements of intercostal and vastus lateralis muscle blood flow were performed during graded exercise up to maximal levels and during resting hyperpnea at the same levels of minute ventilation as during exercise. It was reasoned that if intercostal muscle blood flow during exercise was similar to that recorded during resting hyperpnea and locomotor muscle blood flow was decreased at maximal levels of exercise, this would support the notion that the work of breathing normally incurred during exercise causes preferential redistribution of blood flow from the locomotor to the respiratory muscles (9, 10). If intercostal muscle blood flow during exercise was lower than during resting hyperpnea and locomotor muscle blood flow was preserved, this would challenge the concept of blood flow redistribution in favor of the respiratory muscles during exercise in patients with COPD. METHODS Subjects Ten patients with clinically stable COPD participated in the study. Inclusion criteria are detailed in the online supplement. The study was approved by the University Hospital Ethics Committee. Experimental Design Experiments were conducted in two visits. In visit 1, patients underwent a preliminary incremental exercise test to the limit of tolerance (peak work rate [WRpeak]). In visit 2, patients performed a graded exercise test to WRpeak (protocol 1), which was followed 2 hours later by resting isocapnic hyperpnea trials (protocol 2) sustained at the same tidal volume, breathing frequency, and minute ventilation recorded at rest and during exercise. During both protocols, intercostal and vastus lateralis muscle blood flow (assessed by near-infrared spectroscopy [NIRS] and the light-absorbing tracer indocyanine green dye [ICG]) and cardiac output were measured during the final minute of each of the exercise or hyperpnea trials. NIRS was used to continuously record muscle oxygen saturation, which is commonly adopted as an index of tissue oxygen availability, reflecting the balance between muscle oxygen supply and demand (11). End-inspiratory and end-expiratory compartmental (rib cage and abdominal) chest wall volumes (measured by optoelectronic plethysmography) as well as esophageal and gastric pressures (to measure tidal excursions in pleural and gastric pressures and to calculate transdiaphragmatic pressure, power of breathing, and the pressure time product for the diaphragm and expiratory abdominal muscles) were continuously recorded during the exercise and the hyperpnea trails. Details are provided in the online supplement. Subject Preparation Subjects were prepared with arterial and venous catheters and then with esophageal and gastric balloons. The catheters were used to inject ICG and sample blood after each injection for cardiac output and muscle blood flow measurements (see online supplement). Protocol 1: Graded Exercise During graded exercise testing, patients completed four bouts of constant-load exercise corresponding to the following targeted intensities: (1) 25% WRpeak for 5 minutes, (2) 50% WRpeak for 5 minutes, (3) 75% WRpeak for 3 to 4 minutes, and (4) 100% WRpeak for 2 to 3 minutes. Pulmonary gas exchange and ventilatory variables were recorded breath-by-breath, whereas arterial blood pressure was measured every minute (see online supplement). Protocol 2: Resting Isocapnic Hyperpnea Trials Patients were asked to maintain targeted ventilation equal to their own mean ventilation recorded at rest and during exercise at 25, 50, 75, and 100% WRpeak. Experimenters provided verbal guidance to instruct the subjects to adjust the rate and depth of their breathing such that the target ventilation was obtained and held constant to within 5%. Isocapnia was maintained by having subjects inspire from a Douglas bag containing 5% CO 2, 21% O 2, and 74% N 2. Statistical Analysis Data are reported as means 6 SEM unless otherwise stated. SEM was chosen rather than SD because we were interested in the variance of the mean values rather than the intersubject variance. The minimum sample size was calculated based on 80% power, 5% a error, and a two-sided 0.05 significance level. A sample size capable of detecting a between-condition (i.e., exercise, hyperpnea) difference of 30% was estimated for the change in intercostal muscle blood flow at WRpeak using the standard deviations from our previous study (7). The critical sample size was estimated to be nine patients. Two-way ANOVA with repeated measures was used to identify statistically significant differences across the exercise and the hyperpnea tests. When ANOVA detected statistical significance, pairwise differences were identified using Tukey s honestly significant difference post hoc procedure. Pearson s product moment correlation was used to assess the level of association between continuous variables. The level of significance for all analyses was set at P, RESULTS Subject Characteristics and Exercise Capacity As anticipated from the inclusion criteria, patients had moderate to severe airflow obstruction with increased static lung volumes, moderate reductions in diffusion capacity, and mildly altered resting blood gases (Table 1). Four patients were GOLD stage II, three patients were GOLD stage III, and three patients were GOLD stage IV. Subjects exhibited reduced maximal exercise capacity with moderate oxyhemoglobin desaturation (Table 2). Power of Breathing and Respiratory Kinematics The power of breathing was similar during exercise and hyperpnea trials at equivalent rates of minute ventilation (Figure 1A). End-inspiratory and end-expiratory rib cage volumes significantly increased compared with resting values throughout the exercise TABLE 1. PULMONARY FUNCTION DATA Parameter Value* Age, yr Height, cm Weight, kg BMI, kg/m FFMI, kg/m FEV 1,L FEV 1, % predicted FVC, L FVC, % predicted TLC, L TLC, % predicted RV, L RV, % predicted FRC, L FRC, % predicted DL CO, % predicted Pa O2,mmHg Pa CO2,mmHg Sa O2,% ph Definition of abbreviations: BMI 5 body mass index; DL CO 5 diffusing factor of the lung for carbon monoxide; FFMI 5 fat free mass index; RV 5 residual volume. * Values are means 6 SD for 10 subjects.

3 Vogiatzis, Athanasopoulos, Habazettl, et al.: Respiratory Muscle Blood Flow in COPD 1107 TABLE 2. PEAK EXERCISE DATA* Parameter Value WR peak,w VO 2peak, ml/kg/min HR peak, beats/min VE peak, L/min VT peak, L/min f peak, breaths/min Sp O2,% Borg dyspnea scores Borg leg effort scores Definition of abbreviations: f peak 5 peak breathing frequency; HR peak 5 peak heart rate; Sp O2 5 arterial oxygen saturation measured by pulse oximetry; V Epeak 5 peak minute ventilation; VO 2peak 5 peak oxygen uptake; V Tpeak 5 peak tidal volume; WR peak 5 peak work rate. * Exercise data depict the results of the incremental exercise test. Values are means 6 SD for 10 subjects. and hyperpnea trials (Figure 1C). Rib cage volume changes at the end of inspiration and expiration (reflecting, in part, the contribution of other respiratory muscles and the activity of the external and internal intercostal muscles, respectively) were not different between the exercise and the resting hyperpnea trials (Figure 1C). Similarly, volume changes of the abdominal compartment at the end of inspiration (reflecting the activity of the diaphragm) were not different between exercise and hyperpnea trials (Figure 1D). However, end-expiratory abdominal volume changes (reflecting the activity of the expiratory abdominal muscles) were greater (P ) during exercise than during hyperpnea (Figure 1D). Consequently, end-expiratory total chest wall volume changes were greater (P ) during exercise compared with hyperpnea (Figure 1B). Peak expiratory gastric pressure (Figure 2A) and the pressure time product for expiratory abdominal muscles (Figure 2D) were also significantly greater (P and P , respectively) during exercise compared with hyperpnea. Tidal Pdi excursion and the pressure time product for the diaphragm and tidal excursions of gastric over pleural pressure expressing the relative contribution of the diaphragm to the pressure generated by all inspiratory muscles were not different between exercise and hyperpnea (Figures 2B, 2D, and 2C, respectively). Breathing pattern (tidal volume, frequency, inspiratory and expiratory times, and duty cycle) was not different between exercise and hyperpnea sustained by definition at identical levels of minute ventilation (Figure 3). Cardiorespiratory and Hemodynamic Responses Whole-body oxygen uptake and systemic hemodynamics are shown in Figure 4. During exercise, whole-body oxygen uptake increased linearly (r ) with work rate up to WRpeak (Figure 4A), whereas cardiac output increased up to a load corresponding to 50% WRpeak and leveled off thereafter (Figure 4B). The plateau in cardiac output above 50% WRpeak was due to a fall in stroke volume (Figure 4C) because heart rate continued to increase linearly (r ) with work rate up to WRpeak (Figure 4D). The pattern of change in intercostal muscle blood flow was very different (P ) between the exercise and the hyperpnea tests (Figure 5A). During exercise, intercostal muscle blood flow remained unchanged from rest with increasing rates of work up to 50% WRpeak but significantly decreased thereafter (Figure 5A). In contrast, during isocapnic hyperpnea, intercostal muscle blood flow significantly increased up to 75% WRpeak (Figure 5A); the change from rest in intercostal muscle blood flow was linear with respect to the power of Figure 1. Power of breathing and chest wall volume regulation. (A) Power of breathing, (B) chest wall, (C ) rib cage, and (D) abdominal volume regulation at the end of inspiration and expiration recorded at different fractions of peak work rate (WRpeak) during exercise (open triangles) and during isocapnic hyperpnea trials (filled triangles) that were sustained at levels of minute ventilation similar to those recorded during exercise. Values are means 6 SEM for 10 subjects. Asterisks denote significant differences between the exercise and hyperpnea tests at different fractions of peak work rate. Crosses denote significant differences across the two protocols.

4 1108 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL A C B D Figure 2. Mechanics of breathing. (A) Peak expiratory gastric pressure, (B) tidal excursion in transdiaphragmatic pressure (DPdi), and (C ) tidal excursion in gastric pressure (DPga) to tidal excursion in pleural pressure (DPpl) ratio (expressing the relative contribution of the diaphragm to the pressure generated by the whole inspiratory muscles) measured at different fractions of peak work rate (WRpeak) during exercise (open triangles) and during resting isocapnic hyperpnea trials (closed triangles) that were sustained at levels of minute ventilation similar to those recorded during exercise. (D) Pressure time product (PTP) of respiratory muscles. Closed squares correspond to PTP for expiratory abdominal muscles (PTPab) during the isocapnic hyperpnea trials; open squares refer to PTPab during exercise. Closed triangles correspond to PTP for the diaphragm (PTPdi) during isocapnic resting hyperpnea; open triangles refer to PTPdi during exercise. Values are means 6 SEM for 10 subjects. Asterisks denote significant differences between the exercise and hyperpnea tests at different fractions of peak work rate. Crosses denote significant differences across the two protocols. breathing (r ). During exercise, quadriceps muscle blood flow increased with increasing work rate up to 100% WRpeak (r ), whereas during hyperpnea, quadriceps muscle blood flow did not change with increasing minute ventilation (Figure 5C). During the exercise and hyperpnea trials, intercostal muscle oxygen saturation significantly decreased from baseline (Figure 5B), albeit the magnitude of decrease was greater during exercise. Similarly, during exercise, quadriceps oxygen saturation significantly decreased from baseline but remained unchanged during resting hyperpnea (Figure 5D). During exercise, mean arterial pressure and systemic vascular conductance increased with increasing work rate up to a load corresponding to 50% WR max. During isocapnic hyperpnea, neither mean arterial pressure nor systemic vascular conductance changed from baseline with increasing minute ventilation (Figures 6A and 6C). The pattern of change in intercostal muscle vascular conductance was different between the exercise and the hyperpnea tests (P ) (Figure 6B). During exercise, intercostal muscle vascular conductance remained unchanged with increasing work rate up to 50% WRpeak and significantly decreased thereafter. In contrast, during isocapnic hyperpnea, intercostal muscle vascular conductance significantly increased up to 75% WRpeak and leveled off thereafter. During exercise, quadriceps muscle vascular conductance increased linearly with increasing work rate, peaking at up to 100% WRpeak (Figure 6D). During hyperpnea, quadriceps muscle vascular conductance did not change with increasing ventilation (Figure 6D). DISCUSSION The present study investigated, in patients with COPD, whether there is redistribution of blood flow to the intercostal muscles during graded exercise. Because we were unable to directly measure diaphragmatic blood flow and because with disease progression the act of breathing becomes more dependent on the inspiratory intercostal muscles (12), respiratory muscle blood flow was assessed by measuring intercostal muscle perfusion using indocyanine green dye and a NIRS probe positioned over the seventh intercostal space (8). Intercostal muscle blood flow during resting hyperpnea (when respiratory muscle perfusion was presumably the maximal attainable and not limited by competition from the locomotor muscles) was compared with that recorded during graded exercise where limb and respiratory muscles had to compete for the available blood flow. There are two novel findings of this study. First, intercostal muscle blood flow did not rise above resting levels but fell progressively with increasing effort such that during exercise above WRpeak perfusion to the intercostal muscles was significantly lower than during resting hyperpnea (Figure 5). Second, although cardiac output, blood pressure, and systemic vascular conductance reached a plateau at exercise intensities above 50% WRpeak (Figures 4 and 6, respectively), quadriceps muscle vascular conductance and perfusion continued to increase, reaching maximal values at peak exercise (Figures 5 and 6). Collectively, these observations suggest that during highintensity exercise in COPD, restriction of intercostal muscle blood flow, but preservation of quadriceps muscle perfusion along with the attainment of a plateau in cardiac output, represent the inability of the circulatory system to satisfy the simultaneous energy demands of locomotor and respiratory muscles. Blood Flow Distribution during Exercise The concept of blood flow redistribution from the locomotor to the respiratory muscles during exercise was originally evolved

5 Vogiatzis, Athanasopoulos, Habazettl, et al.: Respiratory Muscle Blood Flow in COPD 1109 Figure 3. Breathing pattern. (A) time of inspiration (T i ), (B) time of expiration (T e ), (C ) duty cycle ( T i /T tot ), (D) minute ventilation (V E ), (E ) tidal volume (V T ), and (F ) breathing frequency (f) recorded at different fractions of peak work rate ( WRpeak) during exercise (open triangles) and during isocapnic hyperpnea trials (closed triangles) that were sustained at levels of minute ventilation similar to those recorded during exercise. Values are means 6 SEM for 10 subjects. from studies in healthy subjects showing reductions in limb blood flow with respiratory muscle loading and an increase in locomotor muscle blood flow with respiratory muscle unloading (13, 14). In COPD, unloading the respiratory muscles by means of heliox breathing or proportional assisted ventilation has been shown to improve peripheral muscle oxygenation (5, 6) and exercise performance (15 20). These findings were interpreted as evidence of redistribution of energy supplies from the respiratory to locomotor muscles. However, none of these studies (5, 6, 15 20) directly assessed the interaction between locomotor and respiratory muscle blood flow during exercise in COPD. The novelty of the present investigation is that simultaneous measurements of intercostal and quadriceps muscle blood flow and vascular conductance were performed under conditions where the amount of respiratory and leg muscle work concurrently increased during graded discontinuous exercise to the limit of tolerance and was compared with conditions under which the respiratory muscles produced similar work while the legs remained inactive. Because intercostal muscle blood flow was lower during exercise than hyperpnea, although the increase in quadriceps muscle blood flow was preserved throughout incremental exercise (Figure 5), our findings in patients with COPD do not support the notion of blood flow redistribution in favor of the respiratory muscles during exercise that was originally proposed from studies in healthy subjects (13, 14). In this context, our results differ from those of previous studies in healthy individuals using inspiratory resistances (13, 14) or expiratory loading techniques (21, 22) to artificially increase the work of breathing. The most likely reason for this discrepancy is that artificially increasing the work of the inspiratory or expiratory muscles in healthy subjects is an acute procedure that causes abrupt adjustments in the regulation of systemic blood flow (14, 21, 22), possibly forcing redistribution of blood flow from the locomotor to the respiratory muscles (13). In contrast, in COPD the phenomenon of increased breathing work is chronic, and therefore certain central or peripheral muscle adjustments in blood flow regulation may take place to accommodate the energy supply according to demand. Our findings are in agreement with those in healthy subjects in that, at essentially the same work of breathing, intercostal muscle blood flow during high-intensity exercise was lower compared with that during voluntary resting hyperpnea (7). In that study (7), after an initial significant increase from rest, intercostal muscle blood flow and vascular conductance fell at intensities exceeding 80% of maximal capacity, whereas cardiac output and quadriceps muscle blood flow reached a plateau. This finding was interpreted as a reflection of the inability of the circulatory system to meet the increasing energy demands of locomotor and intercostal muscles at exercise intensities exceeding 80% of maximal (7). In the present study, however, intercostal muscle blood flow and vascular conductance during

6 1110 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL Figure 4. Metabolic and central hemodynamic responses. (A) Oxygen uptake, (B) cardiac output, (C ) stroke volume, and (D) heart rate recorded at different fractions of peak work rate (WRpeak) during exercise (open triangles) and during isocapnic hyperpnea trials (closed triangles) that were sustained at levels of minute ventilation similar to those recorded during exercise. Values are means 6 SEM for 10 subjects. Asterisks denote significant differences between the exercise and hyperpnea tests at different fractions of WRpeak. Crosses denote significant differences across the two trials. submaximal exercise did not rise above baseline value, whereas at higher exercise intensities (> 50% of peak work capacity), intercostal muscle blood flow and vascular conductance significantly decreased below baseline levels (Figures 5 and 6), most likely reflecting increased sympathetic vasoconstrictor activity to the intercostal muscles (7, 23 25). Figure 5. Intercostal and quadriceps muscle blood flow and tissue oxygen saturation responses. (A) Intercostal muscle blood flow, (B) intercostal muscle oxygen saturation, (C ) quadriceps muscle blood flow, and (D) quadriceps muscle oxygen saturation recorded at different fractions of peak work rate (WRpeak) during exercise (open triangles) and during isocapnic hyperpnea trials (closed triangles) that were sustained at levels of minute ventilation similar to those recorded during exercise. Values are means 6 SEM for 10 subjects. Asterisks denote significant differences between the two conditions at different fractions of WRpeak. Crosses denote significant differences across the two protocols.

7 Vogiatzis, Athanasopoulos, Habazettl, et al.: Respiratory Muscle Blood Flow in COPD 1111 Figure 6. Central and peripheral vascular responses. (A) Mean arterial blood pressure, (B) intercostal muscle vascular conductance, (C) systemic vascular conductance, and (D) quadriceps muscle vascular conductance recorded at different fractions of peak work rate ( WRpeak) during exercise (open triangles) and during isocapnic hyperpnea trials (closed triangles) that were sustained at levels of minute ventilation similar to those recorded during exercise. Values are means 6 SEM for 10 subjects. Asterisks denote significant differences between exercise and hyperpnea tests at different fractions of WRpeak. Crosses denote significant differences across the two trials. There are several lines of evidence suggesting that, in patients with COPD exhibiting significant hyperinflation and expiratory flow limitation, there may be a secondary cardiovascular defect that limits the increase in cardiac output during exercise (15, 26 29). Our results are in agreement with those of earlier studies in COPD (15, 26 29) showing a blunted increase in cardiac output during incremental exercise. In fact, in the present study the reduction in stroke volume and the compensatory increase in heart rate at intensities greater than WRpeak (Figure 4) are in tandem with the results on healthy individuals mimicking important pathopysiologic features of COPD during exercise by increasing expiratory loading (21) or by limiting expiratory flow through a Starling resistor (22). Both strategies produced excessive expiratory muscle recruitment and adverse effects on the increase in stroke volume and cardiac output (21, 22). Our patients showed evidence of significant expiratory abdominal muscle recruitment at intensities greater than WRpeak compared with resting hyperpnea (i.e., decrease in end-expiratory abdominal volume and increase in peak expiratory gastric pressure and pressure time product for the abdominal muscles) (Figures 1 and 2), which most likely hindered the increase in cardiac output beyond that intensity. Indeed, elevations in gastric pressure during expiration caused by increased abdominal muscle recruitment have been shown to act as a Valsava maneuver, decreasing cardiac output and producing blood shifts from trunk to extremities (1, 4, 22). In the present study, it is possible that in the face of a blunted cardiac output, blood flow to the intercostal muscles was reduced in favor of blood flow to the locomotor muscles, which exhibited a linear increase up to peak work capacity, thereby confirming earlier studies (30, 31) that the peripheral muscle hemodynamic and metabolic capacity is preserved during maximal exercise in COPD. Nevertheless, the progressive reduction in quadriceps muscle oxygen saturation (Figure 5) against the background of a steadily increased muscle blood flow during exercise suggests that the increase in quadriceps muscle blood flow was insufficient to prevent the mismatch between oxygen delivery to, and demand of, the quadriceps muscles, making them vulnerable to muscle fatigue (31). Regulation of Blood Flow within the Respiratory Muscles The peripheral muscles were possibly not the only ones having a preferential share of the cardiac output during high-intensity exercise. Volume variations of the rib cage and abdominal chest wall compartments and pressure measurements revealed that, although the activation of the external and internal intercostal muscles (primarily represented by volume changes of the rib cage compartment) and activation of the diaphragm (reflected by volume changes of the abdominal compartment at end of inspiration and evidenced by tidal Pdi excursions and pressure time product for the diaphragm) were similar between exercise and hyperpnea, volume variations of the abdominal compartment, peak expiratory gastric pressures, and pressure time product of the expiratory abdominal muscles (all indicating abdominal muscle recruitment) were significantly greater during exercise compared with hyperpnea (Figures 1 and 2). It is thus plausible that blood flow to the intercostal muscles and the diaphragm may have been redirected to support the increased energy demands of the abdominal muscles. Measurement of Respiratory Muscle Blood Flow in COPD In healthy subjects, blood flow measured over the seventh intercostal space using NIRS and ICG has been considered as perfusing mostly the external and internal intercostal muscles and to a lesser extent the costal segment of the diaphragm (8). This is due to the substantial distance between the sampling

8 1112 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL point of NIRS on the skin and the diaphragmatic appositional area (8) compared with the shorter distance to the intercostals. In COPD, monitoring intercostal muscle blood flow during exercise has additional important implications because there is evidence to indicate that chronic hyperinflation makes these patients use their intercostal muscles more vigorously than normal subjects (12) during exercise, whereas the diaphragm makes a relatively limited contribution to the generation of maximal levels of ventilation (32, 33), likely due to its flattened shape. Our patients exhibited a delayed pattern of exerciseinduced dynamic hyperinflation (34) secondary to a significant reduction in end-expiratory abdominal volume and increased abdominal muscle activity. Nevertheless, the progressive reduction in intercostal muscle oxygen saturation (Figure 5) against the background of initially stable and later decreasing intercostal muscle blood flow during graded exercise is consistent with a mismatch between oxygen delivery to, and demand of, the intercostal muscles, thereby suggesting that intercostal muscle blood flow limitation may have rendered them vulnerable to muscle fatigue (35). In contrast, the decrease from baseline in intercostal muscle oxygen saturation was less pronounced across increasing levels of respiratory muscle work during resting hyperpnea (Figure 5), suggesting that increasing oxygen demand of the intercostal muscles was more adequately met by proportionally increased blood flow and oxygen delivery. On the other hand, abdominal muscle blood flow could not be measured to provide evidence of redistribution in respiratory muscle perfusion because placing the NIRS optodes on the abdomen generally provides a poor signal that underestimates blood flow due to the contribution of often substantial subcutaneous adipose tissue to the light-absorption signal. Study Limitations The invasive, intensive, and exhausting nature of the present study precluded repeated testing across the different work rates so as to establish reproducibility of cardiac output and regional muscle blood flow measurements in patients with COPD. Furthermore, inclusion of an age-matched, healthy control group would have allowed comparisons between healthy subjects and patients with COPD with respect to the regulation of intercostal muscle blood flow during the exercise and hyperpnea protocols. In addition, performing the exercise protocol first and the hyperpnea protocol afterward on the same day may have influenced respiratory muscle oxygen uptake and delivery kinetic responses during the hyperpnea tests owing to muscle warm-up (36). However, the finding that baseline intercostal muscle blood flow was not different between the two protocols suggests that the time elapsed (2 h) between the two protocols eliminated any effect of prior exercise on intercostal muscle blood flow regulation during hyperpnea. In conclusion, during exercise in patients with moderate COPD, the circulatory system appears unable to meet the increasing demands of locomotor and intercostal muscles. Intercostal muscle blood flow does not rise above resting levels and falls progressively with increasing locomotor muscle effort, whereas quadriceps muscle blood flow is preserved even in the presence of a plateau in cardiac output. We suggest that the reduction in vascular conductance and blood flow of the intercostal, but not the quadriceps muscles, appears to be the result of a blunted cardiac output response to exercise in an attempt to maintain blood flow to the locomotor and possibly also to the expiratory abdominal muscles. The findings of the present study reinforce the impact of interactions between respiratory and cardiovascular systems and the need for an integrative approach of their function during exercise, which would potentially reveal the factors limiting exercise performance in patients with COPD. Author Disclosure: I.V. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. D.A. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. H.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. A.A. is a coinventor of Opto-Electronic Plethysmography (OEP), whose patents rights are held by his Institution, the Politecnico di Milano, which licensed OEP to BTS Bio-engineering, Italy. Z.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. E.C. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. H.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. C.R. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. P.D.W. received grant support from CSL Behring ($50,001 $100,000). S.Z. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. References 1. Aliverti A, Macklem P. The major limitation to exercise performance in COPD is inadequate energy supply to the respiratory and locomotor muscles. J Appl Physiol 2008;105: Debigare R, Maltais F. The major limitation to exercise performance in COPD is lower limb muscle dysfunction. J Appl Physiol 2008;105: O Donnell D, Webb K. The major limitation to exercise performance in COPD is dynamic hyperinflation. J Appl Physiol 2008;105: Macklem PT. Exercise in COPD: damned if you do and damned if you don t. Thorax 2005;60: Borghi-Silva A, Carneiro Oliveira C, Carrascosa C, Maia J, Berton DC, Queiroga F Jr, Ferreira EMV, Ribeiro D, Nery LE, Neder JA. Respiratory muscle unloading improves leg muscle oxygenation during exercise in patients with COPD. Thorax 2008;63: Chiappa CR, Queiroga F Jr, Meda E, Ferreira LF, Diefenthaeler F, Nunes M, Vaz M, Machado MCL, Nery LE, Neder JA. Heliox improves oxygen delivery and utilization during dynamic exercise in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2009;179: Vogiatzis I, Athanasopoulos D, Habazettl H, Kuebler W, Wagner H, Roussos C, Wagner PD, Zakynthinos S. Intercostal muscle blood flow limitation during exercise in trained athletes. J Physiol 2009;587: Guenette JA, Vogiatzis I, Zakynthinos S, Athanasopoulos D, Koskolou M, Golemati S, Vasilopoulou M, Wagner HE, Roussos C, Wagner PD, et al. Human respiratory muscle blood flow measured by nearinfrared spectroscopy and indocyanine green. J Appl Physiol 2008; 104: Dempsey JA, Sheel AW, St Croix CM, Morgan BJ. Respiratory influences on sympathetic vasomotor outflow in humans. Respir Physiol Neurobiol 2002;130: Dempsey JA, Romer L, Rodman J, Miller J, Smith C. Consequences of exercise-induced respiratory muscle work. Respir Physiol Neurobiol 2006;151: DeLorey DS, Kowalchuk JM, Paterson DH. Relationship between pulmonary O 2 uptake kinetics and muscle deoxygenation during moderate-intensity exercise. J Appl Physiol 2003;95: De Troyer A, Pride NB. The chest wall and respiratory muscles in chronic obstructive pulmonary disease. In: Roussos C, editor. The thorax. New York: Dekker; pp Harms CA, Babcock MA, McClaran SR, Pegelow DF, Nickele GA, Nelson WB, Dempsey JA. Respiratory muscle work compromises leg blood flow during maximal exercise. JApplPhysiol1997;82: Harms CA, Thomas A, Wetter J, McClaran SR, Pegelow DF, Nickele GA, Nelson WB, Hanson P, Dempsey JA. Effects of respiratory muscle work on cardiac output and its distribution during maximal exercise. J Appl Physiol 1998;85: Oelberg DA, Kacmarek M, Pappagianopoulos PP, Ginns LC, Systrom DM. Ventilatory and cardiovascular responses to inspired He-O2 during exercise in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;179: Bianchi L, Foglio K, Pagani M. Effect of proportional assist vetilation on exercsie tolerance in COPD patients with chronic hyperpnoea. Eur Respir J 1998;11: Richardson RS, Sheldon J, Poole DS, Hopkins SR, Ries AL, Wagner PD. Evidence of skeletal muscle metabolic reserve during whole body

9 Vogiatzis, Athanasopoulos, Habazettl, et al.: Respiratory Muscle Blood Flow in COPD 1113 exercise in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999;159: Hawkins P, Johnson LC, Nikoletou D, Hamnegard CH, Sherwod R, Polkey MI, Moxham J. Proportional assist ventilation as an aid to exercise training in severe chronic obstructive pulmonary disease. Thorax 2002;57: Palange P, Valli G, Onorati P, Antonucci R, Paoletti P, Rosato A, Manfredi F, Serra P. Effect of heliox on lung dynamic hyperinflation, dyspnea, and exercise endurance capacity in COPD patients. J Appl Physiol 2004;85: Eves ND, Petersen SR, Haykowsky MJ, Wong EY, Jones RL. Heliumhyperoxia, exercsie, and respiratory mechanics in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2006;174: Stark-Leyva KN, Beck KC, Johnson BD. Influence of expiratory loading and hyperinflation on cardiac output during exercise. J Appl Physiol 2004;96: Aliverti A, Dellaca RL, Lotti P, Bertini S, Duranti R, Scano G, Heyman J, Lo Mauro A, Pedotti A, Macklem PT. Influence of expiratory flowlimitation during exercise on systemic oxygen delivery in humans. Eur J Appl Physiol 2005;95: Mortensen SP, Dawson EA, Yoshiga CC, Dalsgaard MK, Damsgaard R, Secher NH, Gonzalez-Alonso J. Limitations to systemic and locomotor limb muscle oxygen delivery and uptake during maximal exercise in humans. J Physiol 2005;566: Calbet JA, Gonz alez-alonso J, Helge JW, Sondergaard H, Munch- Andersen T, Boushel R, Saltin B. Cardiac output and leg and arm blood flow during incremental exercise to exhaustion on the cycle ergometer. J Appl Physiol 2007;103: Saito M, Tsukanaka A, Yanagihara D, Mano T. Muscle sympathetic nerve responses to graded leg cycling. JApplPhysiol1993;75: Stewart RI, Lewis CM. Cardiac output during exercise in patients with COPD. Chest 1986;89: Chiappa GR, Borghi-Silva A, Ferreira LF, Carrascosa C, Oliveira CC, Maia J, Gimenes AC, Queiroga F, Berton D, Ferreira EM, et al. Kinetics of muscle deoxygenation are accelerated at the onset of heavy intensity exercise in patients with COPD: relationship to central cardiovascular dynamics. J Appl Physiol 2008;104: Wehr KL, Johnson RL. Maximal oxygen consumption in patients with lung disease. J Clin Invest 1976;58: Light RW, Mintz HM, Linden GS, Brown SE. Hemodynamics of patients with severe chronic obstructive pulmonary disease during progressive upright exercise. Am Rev Respir Dis 1984;130: Richardson RS, Leek BT, Gavin TP, Haseler LJ, Mudaliar SR, Henry R, Mathieu-Costello O, Wagner PD. Reduced mechanical efficiency in COPD, but normal peak VO2 with small muscle exercise. Am J Respir Crit Care Med 2004;169: Maltais F, Simon M, Jobin J, Desmeules M, Sullivan MJ, Belanger M, LeBlanc P. Effects of oxygen on lower limb blood flow and O2 uptake during exercise in COPD. Med Sci Sports Exerc 2001;33: Polkey M, Kyrrousis D, Hamnegard CH, Mills GH, Hughes PD, Gren M, Moxham J. Diaphragm performance during maximal voluntary ventilation in chronic obstructive pulmonary disease. Am Rev Respir Dis 1997;155: Decramer M, De Troyer A. Respiratory changes in parasternal intercostal length. J Appl Physiol 1984;57: Vogiatzis I, Georgiadou O, Golemati S, Aliverti A, Kosmas E, Kastanakis E, Nanas S, Zakynthinos S, Ch Roussos. Respiratory kinematics during exercise and recovery in patients with severe chronic obstructive pulmonary disease. Thorax 2005;60: Babcock MA, Johnson BD, Pegelow DF, Suman OE, Griffin D, Dempsey JA. Hypoxic effects on exercise-induced diaphragmatic fatigue in normal healthy humans. J Appl Physiol 1995;78: Buchheit M, Laursen PB, Ahmaidi S. Effect of prior exercise on pulmonary O2 uptake and estimated muscle capillary blood flow kinetics during moderate-intensity field running in men. J Appl Physiol 2009; 107: