© 2002 American Thoracic Society
Cardiopulmonary Responses to Exercise in Women with Sickle Cell AnemiaPulmonary and Critical Care Unit, Department of Medicine, University of Rochester Medical Center, Rochester, New York; Departments of Medicine, Physiology, and Endocrinology; and The Georgia Prevention Institute, Medical College of Georgia, Augusta, Georgia Correspondence and requests for reprints should be addressed to Leigh A. Callahan, M.D., University of Rochester Medical Center, Department of Medicine, Pulmonary and Critical Care Unit, Box 692, 601 Elmwood Avenue, Rochester, NY 14642-8692. E-mail: leighann_callahan{at}urmc.rochester.edu
Multiple factors contribute to exercise intolerance in patients with sickle cell anemia, but little information exists regarding the safety of maximal cardiopulmonary exercise testing (CPET) or the mechanisms of exercise limitation in these patients. The purpose of the present study was to examine these issues. Seventeen adult women with sickle cell anemia underwent symptom-limited maximal CPET using cycle ergometry and ramp protocols; blood gases and lactate concentrations were measured every 2 minutes. All patients completed CPET without complications. No patient demonstrated a mechanical ventilatory limitation to exercise or had evidence of myocardial ischemia. However, we observed three pathophysiologic patterns of response to exercise in these patients. Eleven patients had low peak V·O2, low anaerobic threshold (AT), gas exchange abnormalities, and high ventilatory reserve; this pattern is consistent with exercise limitation due to pulmonary vascular disease in this patient subgroup. Three patients had low peak V·O2, low AT, no gas exchange abnormalities, and a high heart rate reserve, a pattern consistent with peripheral vascular disease and/or a myopathy. The remaining three patients had low peak V·O2, low AT, no gas exchange abnormalities, and a low heart rate reserve; this pattern of exercise limitation is best explained by anemia.
Key Words: sickle cell anemia • cardiopulmonary exercise testing • pulse oximetry • lactate concentrations
Sickle cell anemia (SCA) is a genetic disease characterized by the production of abnormal hemoglobin (Hb), chronic hemolytic anemia, and acute as well as chronic tissue damage due to repeated episodes of vasoocclusion. As therapies for the treatment of SCA have evolved, median life expectancy has increased (1), and although mortality has improved, morbidity remains significant, and many adults with SCA complain of significant exercise intolerance. In theory, multiple factors could contribute to exercise limitation in these patients including: (1) reduced O2 carrying capacity secondary to low Hb; (2) structural and functional cardiac adaptations resulting from chronic anemia (2–4); (3) pulmonary changes from repeated episodes of acute chest syndrome (5–8); or (4) peripheral vascular impairments related to recurrent microvascular occlusion. Whether one or more of these factors reduces exercise capacity in patients with SCA, however, is not known. The primary purpose of the present study was to determine the mechanism(s) of exercise limitation in patients with SCA. A secondary goal was to determine the safety of performing invasive, maximal cardiopulmonary exercise testing (CPET) in this population. Eligible patients underwent symptom-limited CPET, with invasive arterial blood pressure monitoring and serial measurements of arterial blood gases (ABGs) and lactic acid concentrations.
Seventeen women with HbSS were recruited to participate in a study to evaluate the safety and effectiveness of a physical training program for patients with SCA. Approval was obtained through the Medical College of Georgia Institutional Review Board; all patients gave written informed consent. Eligibility criteria were: females age 18–50 years (males excluded due to stipulations and restrictions of funding source), electrophoretically diagnosed SCA, and ability to pedal a bicycle. Patients were excluded for: avascular necrosis, stroke, pregnancy, leg ulcers, cardiac arrhythmias, myocardial ischemia, or echocardiographic evidence of significant heart disease. Patients underwent a history and physical exam, baseline laboratory testing (complete blood cell count, Hb electrophoresis, chemistry panel), echocardiography, and symptom-limited CPET. Echocardiograms were performed to exclude valvular heart disease (defined as stenosis or regurgitation of greater than trace or mild degree) ventricular dysfunction, pericardial disease, or cardiomyopathy. One patient was excluded because of moderate mitral regurgitation. Spirometry was measured using a 9-L water-sealed spirometer; lung volumes were determined by plethysmography (MedGraphics 1070; Medical Graphics Corporation, St. Paul, MN). Single-breath carbon monoxide diffusing capacity (DLCO) and maximal voluntary ventilation were also measured. Predicted values were from Crapo and coworkers for FEV1, FVC, TLC, and DLCO (9), with race corrections using American Thoracic Society recommendations (10). DLCO was corrected for Hb and carboxyhemoglobin (COHb)(11). Brachial artery catheters (3.0-French, 20-gauge, 8-cm polyethylene; Cook Instruments, Indianapolis, IN) were inserted for blood pressure monitoring and arterial blood sampling. Symptom-limited CPET was performed on an electronically braked cycle ergometer (Sensormedics 2900 metabolic cart; Sensormedics Corporation, Yorba Linda, CA). Resting measurements were obtained, followed by unloaded cycling to steady state; work rate was increased using an incremental ramp protocol (7, 10, or 15 watts [W]) until the patient could not continue; work rate increment was determined using prediction equations of Wasserman and coworkers for peak V·O2 (12). Ventilatory volumes were measured using a low dead-space valve connected to a pneumotach; heart rate (HR), blood pressure, and pulse oximetry saturations were monitored continuously, and electrocardiograms (ECGs) recorded every 2 minutes. Oxygen uptake (V·O2, STPD), carbon dioxide output (V·CO2, STPD), minute ventilation (V·E, BTPS), end-tidal oxygen and carbon dioxide tensions (PETO2 and PETCO2, respectively), and related variables were calculated breath-by-breath and displayed as 30-second time averages. ABGs and lactate samples were obtained every 2 minutes from baseline into the recovery period. ABGs were collected over a 30-second time period corresponding to respired gas analysis, maintained on ice, and measurements of pH, PCO2, PO2, total Hb (tHb), oxyhemoglobin (O2Hb), carboxyHb (COHb), methemoglobin (metHb), and deoxyhemoglobin (HHb) percentages were performed with a blood gas analyzer and co-oximeter (CIBA Corning Models 288 and 270; Bayer Diagnostics, Medfield, MA) within 5 minutes of collection. Lactate concentrations were measured using a modification of the Marbach and Weil method (13). Measured CPET parameters were compared with predicted normal values from Wasserman and colleagues (12), and are reported as percent predicted; these specific parameters classify responses in women based on the best available data with respect to age and sex (12). Data represent the average ± SEM. Paired t tests were performed for comparisons in the same individual, and t tests for group comparisons. A value of p < 0.05 was considered statistically significant.
The characteristics of the subjects who participated in the study are shown in Table 1. Average age was 29.0 ± 1.9 years (range 20–45 years). The Hb concentration averaged 8.6 ± 0.3 g/dl for all subjects (range 7.0–10.0 g/dl). Several patients were receiving treatment with hydroxyurea.
Pulmonary Function Studies Results of pulmonary function testing are shown in Table 2. One patient had evidence for restrictive ventilatory impairment (Patient 6), and one patient had evidence for an obstructive impairment (Patient 7). DLCO/ VA, corrected for Hb and COHb was reduced in four patients.
Safety All patients completed the CPET without serious complications. Fifteen of 17 stopped exercise due to leg discomfort; 6 of 17 patients complained of mild shortness of breath. Patient 1 experienced lightheadedness at peak exercise and underwent a repeat CPET with cerebral Doppler studies that showed no abnormalities in cerebral blood flow; we were unable to detect any physiologic variable that could explain her symptoms. Two patients complained of mild back pain at the end of the study, but on follow up, these symptoms had resolved without intervention. No patient developed an acute crisis as a result of CPET. In addition, no patient developed ischemic changes on ECG and no significant dysrhythmias were noted. Sixteen of 17 patients underwent brachial artery cannulation. Patient 16 refused catheter insertion after two unsuccessful attempts; a baseline ABG was obtained. No patient experienced complications related to brachial artery catheterization other than bruising and soreness at the insertion site, which resolved within 24 to 48 hours.
Cardiopulmonary Responses to Exercise
Heart Rate Response to Exercise In all patients, the heart rate response to exercise was steeper than predicted for the V·O2. Peak exercise heart rate averaged 171 ± 3 beats per minute compared with the average predicted peak heart rate of 191 ± 3 beats per minute. Six of 17 patients reached their predicted peak heart rate (220-age), with a heart rate reserve (HRR) (predicted heart rate-peak heart rate) of less than 15 beats. In five of the remaining patients, a flattened O2 pulse was observed at the end of exercise, suggesting that despite an increase in the HRR, the study was maximal or near maximal. Patient 10 did not meet the criteria for a maximal study.
· VO2 Responses to Exercise
Results of peak V·O2/HR (oxygen pulse) measurements are shown in Figure 3 . In all patients, the peak V·O2/HR was less than 80% of the predicted values. In 12 patients (see Table 3), the pattern of the V·O2/HR response was flat and unchanging at the maximum work rates achieved, suggestive of abnormalities in stroke volume or inability to further extract oxygen (unchanging c[a-v]O2 difference).
Abnormalities in the  V·O2/WR were also observed and averaged 8.0 ± 0.2 ml/minute/W. If the criterion of 10.3 ml/minute/W is used as normal (12), no patient demonstrated a "normal" V·O2/WR relationship. However, 6/16 patients demonstrated V·O2/WR above the 95% confidence interval (> 8.6 ml/minute/W).
Ventilatory Responses to Exercise
Gas Exchange Responses
Pulse Oximetry Measurements Comparisons of pulse oximetry saturations (SpO2) and the measurements obtained from ABGs (SaO2) using co-oximetry, performed according to the recommendations of Ortiz and colleagues (14), are shown in Figure 7 . As demonstrated, pulse oximetry measurements during exercise in women with SCA were frequently inaccurately low when compared with measurements obtained by ABGs.
Lactate Concentrations Baseline lactate concentrations averaged 0.8 ± 0.1 mEq/L (range 0.3–2.2 mEq/L) and 2 minutes recovery lactate concentrations averaged 6.9 ± 0.5 mEq/L (range 3.5–10.7 mEq/L). The pattern of lactate production as a function of V·O2 in the women with SCA is shown in Figure 8 . This data indicates a pathologic onset of lactic production (15). Lactate concentrations were significantly elevated during recovery, but arterial pH was well preserved (average 7.38 ± 0.01). The lowest pH measured was 7.33 and occurred at 2 minutes into recovery in Patient 8 whose lactate concentration was 8.5 mEq/L.
Safety and Technical Issues with CPET in Patients with SCA The results of the current study show that volitionally limited invasive CPET can be safely performed in women with SCA. No significant complications occurred during or following CPET and importantly, no patient developed an acute crisis as a result of exercise. Sudden death during exercise has been reported with sickle cell trait (16), but a recent review suggests that the majority of such deaths were related to exertional heat illness, rather than to sickle trait (17). Although our patients performed short bouts of exercise under precisely controlled conditions and were relatively "healthy," our findings are compatible with previous reports that SCA is not clearly associated with an increased risk of exercise-related complications (17). However, it is possible that increased complications may occur during CPET in more debilitated patients with SCA. We also found that noninvasive pulse oximetry saturation measurements (SpO2) were not accurate or reliable at rest or during exercise in patients with sickle cell when compared with arterial oxygen saturation measured by co-oximetry (SaO2). In fact, SpO2 significantly underestimated SaO2, and had we relied solely on SpO2, we would have prematurely stopped the CPET in 14 patients (see Figure 7). We are not aware of any other studies that have compared the accuracy and reliability of pulse oximetry with ABGs in patients with SCA during exercise. Although Ortiz and coworkers (14) have suggested that SpO2 is accurate and reliable in estimating SaO2 in adult patients with SCA during acute vasoocclusive crises, other studies indicate conflicting results with regard to the accuracy of SpO2 in patients with SCA, arguing that estimates are correct, too high, or too low (18, 19). Anemia (20), dark skin pigmentation (21, 22), and irregular photoplethysmographic waves (14) have been implicated as potential factors contributing to the inaccuracy of pulse oximetry recordings in SCA. We believe that the current findings represent a clinically important observation that should be considered in further studies evaluating patients with SCA during exercise. Our data also demonstrate that arterial catheterization can be safely performed in this patient population. Furthermore, our findings argue that ABGs are crucial during diagnostic CPET, both to prevent premature discontinuation of exercise due to unreliable and inaccurate information obtained from noninvasive measurements on pulse oximetry and to reveal subtle changes in gas exchange that cannot be detected with noninvasive testing or be predicted to occur based on resting pulmonary function tests.
Factors Determining Exercise Limitation in SCA
Anemia, per se, can cause significant exercise impairment and can produce a low peak V·O2, a low AT, a low O2 pulse, an increased heart rate response for the level of work, and increased V·E/VCO2 during exercise (23). All of our patients had significant anemia and, it is likely that low Hb concentrations contributed to exercise intolerance in all of these individuals. Moreover, in three of our patients (Patients 9, 13, and 16), detailed analysis suggests that anemia alone accounted for exercise limitation; none of these patients had evidence of pulmonary gas exchange abnormalities (a sign of possible pulmonary vascular disease) or an inappropriately high HRR at maximum exercise (a sign of possible peripheral vascular disease or myopathy). In keeping with the role of anemia in producing exercise intolerance in SCA, one study indicates that patients with SCA who undergo exchange transfusion show improved work performance, either as a direct result of the Hb or due to improvements in blood flow related to decreases in viscosity with transfusion of normal erythrocytes (24). Although anemia is sufficient to account for the pattern of physiologic abnormalities observed in 3 of our 17 patients during exercise, the remaining 14 patients had physiologic derangements that, arguably, cannot be accounted for on the basis of anemia alone. Most of these patients (11 out of 17) developed significant gas exchange abnormalities during exercise. These gas exchange abnormalities included: increases in peak exercise P(A-a)O2 greater than 30 mm Hg, VD/VT abnormalities at rest and with exercise, peak exercise P(a-ET)CO2 greater than zero, and excessively high V·E/VCO2 levels. In support of the argument that anemia alone does not produce significant gas exchange abnormalities during exercise, Lewis and coworkers (25) performed exercise testing in patients with renal disease before and after treatment with erythropoietin and found that anemia per se did not produce abnormalities in VD/VT, reductions in arterial PaO2, or widening of the P(A-a)O2 (i.e., the PaO2 at maximal exercise was 109 ± 14 mm Hg before and 105 ± 8 mm Hg after correction of anemia) (25). Of note, the mean Hgb concentrations in the study by Lewis and coworkers are similar to those in the present study (i.e., 7.0 ± 1.2 versus 8.6 ± 0.3 g/dl) (25). Although some studies report that anemia alone can produce modest increases in V·E/VCO2 with exercise, this effect is generally seen at more severe degrees of anemia and/or at higher exercise workloads than achieved by the patients in our study. Specifically, increases in V·E/VCO2 during exercise in patients with established anemia are seen with very high workloads (levels of V·O2 > 25 ml/minute/kg) (26), but our patients achieved peak V·O2 values that averaged only 15.6 ml/minute/kg. Therefore, we believe that the pulmonary gas exchange abnormalities seen in 11 of our 17 patients with SCA (Patients 1, 3, 5–8, 11, 12, 14, 15, 17) cannot be explained by anemia alone. The response to exercise in these 11 patients revealed a pattern that included gas exchange abnormalities, low peak V·O2, low AT, and a high ventilatory reserve at maximum exercise. This pattern is consistent with that seen in patients with known pulmonary vascular disease (12), arguing that pulmonary vascular disease plays a significant role in limiting exercise in the majority of our patients with SCA. The remaining three patients (Patients 2, 4, and 10) demonstrated a pattern of physiologic alterations during exercise that could not be accounted for by anemia alone and that was also not consistent with the presence of pulmonary vascular disease (i.e., no gas exchange abnormalities were present). The pattern in this group of patients revealed low peak V·O2, low AT, no gas exchange abnormalities, a high ventilatory reserve, and an inappropriately high HRR at maximum exercise. This last finding is not consistent with the exercise response seen in anemia alone, which is usually associated with a low HRR at maximum exercise. Instead, this combination of findings is characteristically seen in patients with peripheral vascular diseases and myopathies (which limit oxygen extraction by tissues in the face of adequate central circulatory function). Therefore, it seems possible that peripheral vascular disease or myopathy may have been present in these three patients with SCA.
Potential Mechanisms of Pulmonary Vascular and Peripheral Vascular Disease in Patients with SCA Another possible explanation for the pathophysiologic abnormalities seen during exercise in our patients may be related to the effects of SCA on pulmonary vasoregulation. Stamler and coworkers have demonstrated that nitric oxide binding to Hb is important for vasoregulation (28–30). It has further been suggested that individuals with SCA may have alterations in both pulmonary nitric oxide scavenging and loss of hypoxic pulmonary vasoregulation (31). It is possible that alterations in pulmonary artery nitric oxide metabolism may limit vasodilatation during exercise in SCA, increasing ventilation–perfusion mismatch and producing, in turn, the impaired gas exchange that was present in the majority of the patients we exercised. In three of our patients, we observed a pattern exercise limitation that is characteristically seen in patients with peripheral vascular disease or myopathic syndromes (i.e., high HRR, low AT, and low peak V·O2 at maximum exercise). Although there are a limited number of reports of myositis, myonecrosis, and myofibrosis in SCA (32–35), it seems reasonable to speculate that in the setting of vasoocclusion, the skeletal muscles of these patients are exposed to repetitive episodes of ischemia reperfusion, which could lead to significant long-term derangements in muscle metabolism and function. It is also possible that subtle peripheral microvascular disease may develop as a consequence of vascular damage resulting from recurrent sickling crises, albeit little work has been performed to examine this latter issue. Finally, impaired peripheral vasoregulation due to derangements in nitric oxide metabolism and/or altered nitric oxide–hemoglobin interactions may also occur, creating mismatched peripheral blood flow during exercise and resultant tissue dysoxia.
Implications The findings from the present study suggest that a substantial number of patients with SCA who have no overt clinical evidence of pulmonary vascular disease develop significant pulmonary complications, leading to gas exchange abnormalities that are sufficiently severe to impair their exercise performance. We propose that cardiopulmonary exercise testing might facilitate early detection of significant pulmonary involvement in patients with SCA; this information could lead to stratification and identification of those who are at risk for increased morbidity and mortality, and who may require more aggressive clinical interventions such as those that are currently being used to treat other forms of secondary pulmonary hypertension.
The authors express their appreciation to Teresa Moss, Lisa Benton, and Janet Cofer for their expert technical assistance in performing the cardiopulmonary exercise tests and blood gas analyses. They also thank Dr. Gerald Supinski for his suggestions and careful review of the manuscript.
This research was supported by NIH grant HD 35063. Received in original form February 2, 2000; accepted in final form December 10, 2001
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
