(Chest. 2005;128:3434-3439.)
© 2005
American College of Chest Physicians
Ventilation and Exercise Performance After Phrenic Nerve and Multiple Intercostal Nerve Transfers for Avulsed Brachial Plexus Injury*
Ming-Lung Chuang, MD;
David C.C. Chuang, MD;
I-Feng Lin, Dr.PH;
Janine R.E. Vintch, MD;
Jean J.W. Ker, BS and
Thomas C.Y. Tsao, MD, PhD
* From the Division of Pulmonary and Critical Care Medicine (Drs. M.-L. Chuang, Ker, and Tsao), Buddhist Taipei Tzu Chi General Hospital, Xindian City, Taipei, Taiwan; the Department of Plastic Surgery (Dr. D.C.C. Chuang), Chang Gung Memorial Hospital, Taipei, Taiwan; the Department of Social Medicine (Dr. Lin), National Yang Ming University, Taipei, Taiwan; and the Division of Respiratory and Critical Care Physiology and Medicine (Dr. Vintch), Department of Medicine, Harbor-UCLA Medical Center, Torrance, CA.
Correspondence to: Ming-Lung Chuang, MD, Division of Pulmonary and Critical Care Medicine, Buddhist Taipei Tzu Chi General Hospital, No. 289 Jianguo Rd, Xindian City, Taipei 23142, Taiwan, Republic of China; e-mail: yuan1007{at}ms36.hinet.net
 |
Abstract
|
|---|
Background: Diaphragmatic excursion, lung function, exercise performance, and clinical symptoms have not been previously described in patients after phrenic nerve transfer (PNT) and/or multiple intercostal nerve transfer (MIT) for the repair of avulsed brachial plexus injury (ABPI) to prevent functional musculoskeletal impairment in the shoulder.
Setting: A university-based hospital.
Methods: Dyspnea scores, chest ultrasonography to assess diaphragmatic excursion, and pulmonary function testing were performed to assess ventilation in patients sustaining trauma to their brachial plexus. In addition, cardiopulmonary exercise testing was also performed. These studies were obtained prior to surgical intervention, and were repeated postoperatively at 6, 12, 18, 24, and 36 months. The results obtained preoperatively were compared to those obtained throughout the postoperative monitoring period.
Results: This study demonstrates that the PNT-MIT procedure results in permanent ipsilateral diaphragmatic paralysis accompanied by an approximately 8% decrease in inspiratory capacity, FVC, and total lung capacity. There was also an 11% increase in diffusing capacity noted during the period between 6 months and 3 years after PNT-MIT procedure. Despite these measurable changes in lung function, the patients reported amelioration of their dyspnea complaint within 6 months of undergoing this procedure, which was due mainly to an improvement in their cardiovascular exercise performance related to increased daily activity.
Conclusions: This study demonstrates that the PNT-MIT procedure is a safe method for the restoration of drop shoulder incurred by ABPI. This surgery has an impact on measurable diaphragmatic and lung function but with minimal impact in terms of postoperative clinical symptoms and exercise performance.
Key Words: cardiopulmonary exercise testing • diaphragmatic paralysis • postoperative • preoperative • spirometry
|
Introduction
|
|---|
Avulsed brachial plexus injury (ABPI) has been reported to be one of the major injuries experienced by victims of motorcycle accidents. This injury can lead to significant shoulder function impairment. Shoulder abduction can be restored by performing a phrenic nerve transfer (PNT) combined with multiple intercostal nerve transfer (MIT).1 However, the phrenic nerve controls the movement of the diaphragm, and the intercostal nerves assist in ventilation. Therefore, impaired phrenic nerve and/or intercostal nerve function after this type of injury or during surgical repair can affect ventilation. It has been previously reported23 that resting ventilatory function is temporarily impaired and recovers within 2 years after PNT. On the other hand, the mean time to recovery from accompanying dyspnea on exertion and orthopnea has been reported in the literature to range from 4 months to 4 years.4 It has also been reported4 that seven patients with unilateral diaphragmatic paralysis had hypoxemia (PaO2 mean [± SD], 73 ± 11 mm Hg), despite little decrement in overall ventilatory capacity in patients with either unilateral or bilateral phrenic nerve interruption.56 Therefore, it seems worthy to reevaluate the long-term effects of this surgical intervention with regard to clinical symptoms, resting lung function, and exercise performance.
Hypoxemia and/or hypercapnia and a reduced breathing reserve are very unusual in healthy subjects.78 An increase in end-tidal PCO2, and/or a decrease in oxyhemoglobin saturation, or a limited breathing reserve (ie, < 15 L/min of the measured maximal voluntary ventilation [MVV]) at peak exercise indicate inadequate ventilation in response to exercise.7 We hypothesized the following: (1) lung volume measurements might reflect the presence of impaired ventilation due to unilateral diaphragmatic paralysis; (2) the impaired lung function may be permanent after a period of time following the PNT-MIT because the resected phrenic nerve and/or intercostal nerves do not regenerate; and (3) the unilateral "paralytic" diaphragm may unmask limited ventilation in response to exercise, thereby leading to exertional dyspnea.
|
Materials and Methods
|
|---|
For the 1-year period of enrollment from May 2000 to April 2001 with a follow-up of up to 3 years, a total of 19 patients having limited range of motion of the injured arm and shoulder drop after sustaining brachial plexus injury were enrolled in the study (Table 1
). Nine of the 19 patients were enrolled within 1 week of undergoing the PNT-MIT procedure, while 10 patients were enrolled after the surgical procedure had been completed. These 10 subjects were included because they had no major medical conditions noted before or after the ABPI, and the measured lung function and exercise capacity could be normalized as a percentage of normal predicted values. Last, to avoid contaminating the preoperative data obtained in patients with normal diaphragmatic function, we retained the data of these 10 patients postoperatively. Their mean (± SD) age was 25 ± 6 years (range, 17 to 38 years). Their mean weight was 65 ± 11 kg, and their mean height was 169 ± 8 cm. The ABPI was a result of a motorcycle accident in 17 patients, and a chemical explosion accident in the other 2 patients. Despite this injury, all patients were able to perform their daily activities without any obvious limitation. Patients were instructed in and encouraged to perform general exercise training at home but were not in a hospital-based training or rehabilitation program. Patients were encouraged to abstain from tobacco use at the time of their enrollment. All patients were in stable condition after the trauma for a mean time of 3.7 ± 1.2 months before the PNT-MIT procedure was performed, which is consistent with the results of our previous report.1 Fifteen patients underwent a complete PNT-MIT with or without spinal accessory nerve, ipsilateral C7, or cervical motor nerve roots, while 2 patients underwent only the MIT portion of the procedure without the PNT, as phrenic nerve paralysis had been sustained. The data from these two patients were retained for analysis as postoperative data only. Two other patients underwent only the PNT procedure. All patients took part in a shoulder rehabilitation program postoperatively including physical therapy and electrical stimulation, which was initiated 3 weeks postoperatively and was performed 3 days per week for up to 2 years.1 The shoulder abduction improved from a drop shoulder to a mean angle of 95 ± 49°. For details of the actual surgical operation and a shoulder rehabilitation program, please refer to the study by Chuang et al.1 The medical research committee approved the study protocol. Each patient gave informed consent to participate in the study.
Study Design
The variables of the following tests were measured: (1) self-assessed dyspnea score; (2) diaphragmatic excursion evaluated with chest ultrasonography; (3) pulmonary function testing; and (4) symptom-limited (eg, leg fatigue or dyspnea) increasing ramp-pattern cardiopulmonary exercise testing (CPET) that was performed on a stationary cycle ergometer. The variables were obtained preoperatively and postoperatively for a period of up to 3 years.
Protocols and Measurements
Dyspnea evaluation:
A total of 19 patients reported the severity of the effect of dyspnea on performing daily activities before and after undergoing the PNT-MIT. The severity of dyspnea was graded subjectively on a scale including nil, mild, moderate, or severe.
Chest Ultrasonography:
Chest ultrasonography was performed in each subject in the upright position to evaluate the diaphragmatic excursion before and at approximately 6, 12, 18, 24, and 36 months after the PNT-MIT procedure was performed. A 3.5-MHz convex probe (model SSD2000; Aloka; Tokyo, Japan) was used. Diaphragmatic excursion was measured semiquantitatively, and was graded as 0 for no movement, 0.5 for movement of less than one intercostal space (ICS), 1 for one ICS, 1.5 for more than one and less than two ICSs, 2 for two ICSs, and 3 for more than two ICSs.
Pulmonary Function Testing:
Tobacco, other forms of nicotine, caffeinated products, alcohol, and medications were prohibited for 24 h before any testing was performed. For convenience, all subjects performed pulmonary function testing before the exercise testing was performed. Maximally forced expiratory flow curves were performed on all patients using a pneumotachographic spirometer (Pre-Vent CardiO2; Medical Graphics; St. Paul, MN). FEV1 and FVC values were obtained in all patients, with the best of three technically satisfactory readings being used. A pressure-sensitive plethysmograph (6200 Autobox DL; SensorMedics; Yorba Linda, CA) measured total lung capacity (TLC), residual volume (RV), and vital capacity. The results were reported at body temperature, ambient atmospheric pressure, and fully saturated. Single-breath diffusing capacity of the lung for carbon monoxide (DLCO) was also measured. Simple volume calibration was conducted with a 3-L syringe before each test. For a detailed description of the lung function testing performed, please refer to the studies by Chuang et al.910
Maximal CPET:
After obtaining stable exercise gas exchange while the subject sat on the cycle ergometer (CardiO2; Medical Graphics), data was then collected during a 2-min period of rest followed by a 2-min period of unloaded cycling. The exercise test continued with an increasing ramp-pattern cycle at a work rate between 5 and 20 W/min until the subject felt exhausted and could not go further. The rate of the increment of exercise was based on the patient’s preassessment level of fitness and with a goal to complete the test in 10 min. During each test, a pedaling frequency of 60 revolutions per minute was maintained with the aid of a visual pedal rate indicator. Oxygen uptake (
O2), carbon dioxide output (CO2) [ml/min], and minute ventilation (E) were computed breath by breath, and the data were displayed every 15 s using an online computer. We designated the peak O2 as the symptom-limited O2 that the patients achieved. Twelve-lead ECG, heart rate, and oxygen saturation using a pulse oximeter (Ohmeda 3740; BOC Health Care Company; Louisville, CO) were measured continuously. The lactic acidosis threshold (LAT) was measured by the V-slope method.11 The details have been described in our previous studies.910
Statistical Analysis
The patients’ characteristics were described as the mean ± SD for numerical measurements or as the frequency/percentage for categoric variables. The generalized estimating equation was used to compare the means of each variable between the preoperative and 6-month postoperative groups, and to test the linear trend from 6 months through 3 years. We did not abandon the patients who were enrolled after the operation or were missing data after the operation because most causes of incompleteness were not medically related. The generalized estimating equation takes into account the repeated measures at different time points for the same individual with consideration of the incomplete values.12 Pearson correlation coefficients were used to quantify the relationship between two concerned variables. All tests were two-sided, and a p value of < 0.05 was considered to be statistically significant. These procedures were carried out using statistical software packages (SAS, version 8.2; SAS Institute Inc; Cary, NC; and Microcal Origin, version 4.0; Microcal Software Inc; Northampton, MA).
|
Results
|
|---|
Dyspnea
Dyspnea developed prior to the surgical intervention in four patients. In eight patients, mild dyspnea was reported within the first 6 months after they had undergone the PNT-MIT procedure (p < 0.05) and resolved thereafter (Fig 1
, top).
View larger version (12K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1.. Top: median dyspnea score before and after PNT and/or MIT. Horizontal bars = SE of time; dashed line = the time of PNT-MIT; * = p < 0.05. Bottom: diaphragmatic excursion evaluated with chest ultrasonography as a function of time before and after PNT-MIT. Vertical bars = SE of diaphragmatic excursion; horizontal bars = SE of time; ** = p < 0.01. A.U. = absolute unit (for details, please see "Materials and Methods" section).
|
|
Chest Ultrasonography
Seven patients had normal diaphragmatic excursion before undergoing the PNT-MIT procedure (Fig 1, bottom). Two patients were excluded from the preoperative data because their ipsilateral diaphragm was injured by trauma of the lower cervical spine. The data for these two patients were retained and integrated as postoperative data. The ipsilateral diaphragm was paralyzed for up to 36 months following PNT-MIT compared to the preoperative state (p < 0.01).
Pulmonary Function Testing
Before undergoing the PNT-MIT operation, seven patients had normal lung function test results, as follows: FVC, 3.4 ± 0.7 L; FEV1, 2.9 ± 0.7 L; MVV, 107 ± 16 L/min; TLC, 5 ± 0.8 L; and inspiratory capacity (IC), 2.3 ± 0.4 L (Fig 2
). The FVC, TLC, and IC values decreased by 8% within 6 months after undergoing the PNT-MIT procedure (all p < 0.05). The decrease in FVC, TLC, and IC persisted for up to 3 years following the PNT-MIT procedure. Expiratory reserve volume and RV were preserved both before and after the PNT-MIT procedure. These findings of decreased FVC, TLC, and IC, and preserved expiratory reserve volume and RV reflect the presence of a paralytic diaphragm, which is the major muscle of inspiration. DLCO was decreased after the trauma and had not changed 6 months after the PNT-MIT procedure (difference not significant). However, a noted increase in DLCO by 11% (p < 0.01) occurred during the period between 6 months and 3 years after PNT-MIT procedure. The increase in DLCO appears to be related to the improvement in exercise performance and ventilatory efficiency postoperatively (Table 2
). The direct MVV is influenced by the status of the ventilatory muscles, the effort and cooperation of the subjects, the compliance of the lungs and chest wall, the condition of the ventilatory control mechanisms, and the resistance offered by the airway and tissues. The decreased direct MVV before the PNT-MIT procedure (69 ± 12%) can be due to neuromuscular injury by trauma, uncooperativeness of the patients, or obstructive ventilatory disease.1314 However, the averaged values for lung volume and FEV1/FVC ratio were normal before the PNT-MIT was performed, and only two patients having phrenic nerve injury before the PNT-MIT procedure and their data were retained for analysis postoperatively. Thus, a lower direct MVV than an indirect MVV (= FEV1 x 40)7 suggested that patient cooperativeness played an important role while performing the maneuver after trauma and shortly after the PNT-MIT procedure. However, the fact that no patients developed further significant decreases in MVV after the PNT-MIT procedure inferred a cancel-out effect due to a sustained paralytic diaphragm and an improved learning effect on performing the MVV maneuver.
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2.. Top: averaged FVC percent predicted, FEV1 percent predicted, and FEV1/FVC ration before and after the PNT and/or MIT procedure. Horizontal bars = SE of time; vertical bars = SE of dependent variables (n = 7, 12, 12, 11, 11, and 10 at each time point, respectively); dashed line = the time of PNT-MIT; * = p < 0.05 (before vs 6 months after PNT-MIT procedure). Middle: TLC percent predicted, IC percent predicted, and RV/TLC ratio as a function of time before and after the PNT and/or MIT procedure. Bottom: DLCO percent predicted as a function of time before and after the PNT and/or MIT procedure. ** = p < 0.01 (testing the linear trend from 6 months through 3 years).
|
|
Exercise Performance
Before the PNT-MIT operation, the peak O and peak E were 1.59 ± 0.36 and 62 ± 16 L/min, respectively. Most patients experience deconditioning rather than ventilatory limitation, which was most likely related to restricted daily activity and exercise after the accident and before the PNT-MIT procedure (Fig 3
). Within the first 6 months following the PNT-MIT procedure, no ventilatory limitation was noted despite a 9% increment in E at similar peak O2 values as in the baseline study (p < 0.05). This increase in ventilation was primarily attributed to a proportional increase in breathing frequency by 7% (p < 0.05). Strikingly, the peak O2, LAT, O2 pulse at peak exercise, and the ratio of the increase in O2 and the increase in work rate (WR) continued to improve significantly 6 months later compared to the data obtained within the first 6 months following the PNT-MIT procedure (all p < 0.0.5, respectively) [Fig 3]. Ventilation in response to peak exercise increased by 6% per half year after the PNT-MIT procedure due in part to the increase in exercise performance, revealing an overall 8% improvement in peak O2 (r = 0.54; p < 0.001).
|
Discussion
|
|---|
According to previous reports,1 shoulder abduction function after sustaining an ABPI can be restored by performing a combined PNT-MIT procedure. Lung function has been evaluated in patients undergoing PNT or intercostal nerve transfer alone but has not been fully explored with the combined procedure.23456 Daily clinical symptoms, lung function, and the demanded ventilatory response to exercise related to a paralyzed hemidiaphragm incurred during a PNT-MIT procedure have not been described. This study demonstrates that the decrements in FVC, TLC, and IC reflect the impaired ventilation due to unilateral diaphragmatic paralysis as confirmed by chest ultrasonography. These defects persisted for up to 3 years following the PNT-MIT procedure (Fig 1, 2) and were not regained. However, the unilateral "paralytic" diaphragm did not appear to limit ventilation in response to exercise in the short-term (6 months) or long-term (up to 3 years) follow-up period after the PNT-MIT procedure (Fig 3). Dyspnea on exertion temporarily worsened within the first 6 months following the procedure and gradually improved thereafter (Fig 1). Cardiovascular function improved gradually during the first 6 months and continued to improve up to 3 years after the PNT-MIT procedure as demonstrated by CPET findings. Based on the persistent impaired lung volumes and improving cardiovascular performance, we considered deconditioning shortly after the injury and surgical procedure, and reconditioning after recovery, 6 months after the trauma and PNT-MIT procedure, play an important role in the mechanism of dyspnea on exertion that was experienced by these patients (Fig 123, Table 2).
It has been reported3 that transient dyspnea occurred in only 1 of 65 patients undergoing the PNT procedure for an ABPI. The lung function in the previous report including FVC, TLC, functional residual capacity, and MVV were only temporarily impaired for 12 months after the PNT procedure.3 In contrast, four of our patients reported dyspnea before the PNT-MIT procedure, and eight patients reported dyspnea after the PNT-MIT procedure. The peak time of this shortness-of-breath complaint occurred during the first 6 months with recovery by 1 year after the PNT procedure. Only two patients had a limited diaphragmatic excursion, as demonstrated by chest ultrasonography before undergoing the PNT-MIT procedure. Restricted activity with associated deconditioning after the accident or trauma may be one of the other possible explanations for the dyspnea that these patients experienced preoperatively. In this study, FVC, TLC, and IC but not functional residual capacity remained persistently decreased by approximately 8% after the PNT-MIT procedure. This is consistent with the results of previous reports in the literature.15 Gu and Ma3 described 5 of 12 patients with persistent limited diaphragmatic excursion by chest radiography at 5 years after the PNT procedure. In our study, the diaphragmatic paralysis appears to be permanent, as evaluated by multiple chest ultrasonography sessions over the course of follow-up in this patient population. The discrepancy between our study and previous reports might be attributed to the different operative procedures and techniques, as well as the different evaluation methods used. Seventeen of 19 patients in this study underwent PNT, and/or spinal accessory nerve transfer, and/or MIT, while most patients in the study of Gu and Ma3 underwent PNT only. Chalidapong et al16 and Giddins et al17 both reported separately that the MIT procedure transiently and mildly impaired ventilatory function, if at all, compared to the PNT procedure. The study revealed that the PNT-MIT procedure might impair lung function more seriously compared to either the PNT or MIT procedure alone.
The improvement in DLCO between 6 and 36 months postoperatively correlated with cardiovascular exercise performance, as demonstrated positively by the 
O2/WR ratio, heart rate, WR, and systolic BP at peak exercise, and negatively by the E/CO2 ratio18 (Table 2). The significant correlation of DLCO with E and breathing frequency at peak exercise might be due to an improvement in exercise capacity. The improvement of pulmonary blood flow along with the improvement of cardiovascular exercise performance might be anticipated and therefore may result in the improvement of DLCO.19
In summary, despite the permanent impaired lung volumes caused by unilateral diaphragmatic paralysis after the PNT-MIT procedure, dyspnea on exertion was transient and could be improved along with the improvement in cardiovascular performance with reconditioning in the postoperative period. Ventilatory limitation in response to exercise rarely occurred in these patients. We conclude that PNT-MIT remains a safe procedure for the treatment of ABPI to regain shoulder function with minimal impact on respiratory symptoms and cardiopulmonary performance. However, because of the relatively young age of our enrolled patients, one should be conservative in extrapolating the results to senile patients and/or infants with brachial plexus injury.
|
Footnotes
|
|---|
Abbreviations: ABPI = avulsed brachial plexus injury; CPET = cardiopulmonary exercise testing; DLCO = diffusing capacity of the lung for carbon monoxide; IC = inspiratory capacity; ICS = intercostal space; LAT = lactic acidosis threshold; MIT = multiple intercostal nerve transfer; MVV = maximum voluntary ventilation; PNT = phrenic nerve transfer; RV = residual volume; TLC = total lung capacity; CO2 = carbon dioxide output; E = minute ventilation; O2 = oxygen uptake; WR = work rate
This research was supported by the Chang Gung Medical Research Committee under CMRP contracts No. 443 and No. 989.
Received for publication February 3, 2005.
Accepted for publication July 5, 2005.
|
References
|
|---|
- Chuang, DCC, Lee, GW, Hashem, F, et al (1995) Restoration of shoulder abduction by nerve transfer in avulsed brachial plexus injury: evaluation of 99 patients with various nerve transfers. Plast Reconstr Surg 96,122-128[ISI][Medline]
- Gu, Y-D, Wu, M-M, Zhen, Y-L, et al Phrenic nerve transfer for brachial plexus motor neurotization. Microsurgery 1989;10,287-289[ISI][Medline]
- Gu, Y-D, Ma, M-K Use of the phrenic nerve for brachial plexus reconstruction. Clin Orthop 1996;323,119-121
- Graham, DR, Kaplan, D, Evans, CC, et al Diaphragmatic plication for unilateral diaphragmatic paralysis: a 10-year experience. Ann Thorac Surg 1990;49,248-252[Abstract]
- Dowman, CE Relief of diaphragmatic tic, following encephalitis, by section of phrenic nerves. JAMA 1927;88,95
- Eisele, JH, Noble, MIM, Katz, J, et al Bilateral phrenic-nerve block in man. Anesthesiology 1972;37,64-69[CrossRef][Medline]
- Wasserman, K, Hansen, JE, Sue, DY, et al Physiology of exercise. Wasserman, K eds. Principles of exercise testing and interpretation 1999,10-61 Lea & Febiger. Philadelphia, PA:
- Dillard, TA Ventilatory limitation of exercise: prediction in COPD. Chest 1987;92,195-196[Free Full Text]
- Chuang, ML, Lin, IF, Wasserman, K The body weight-walking distance product as related to lung function, anaerobic threshold and peak O2 in COPD patients. Respir Med 2001;95,618-626[CrossRef][ISI][Medline]
- Chuang, ML, Lin, I-F, Vintch, JRE Comparison of estimated VO2 and measured VO2 during exercise in patients with COPD. Int Med J 2004;34,469-474[CrossRef]
- Beaver, WL, Wasserman, K, Whipp, BJ A new method for detecting anaerobic threshold by gas exchange. J Appl Physiol 1986;60,2020-2027[Abstract/Free Full Text]
- Diggle, PJ, Heagerty, P, Liang, KY, et al Analysis of longitudinal data 2nd ed. 1994,282-316 Oxford University Press. New York, NY:
- Campbell, S A comparison of the maximum voluntary ventilation with the forced expiratory volume in one second: an assessment of subject cooperation. J Occup Med 1982;24,531-533[ISI][Medline]
- Fairshter, RD, Carilli, A, Soriano, A, et al The MVV/FEV1 ratio in normal and asthmatic subjects. Chest 1989;95,76-81[Abstract/Free Full Text]
- Luedemann, W, Hamm, M, Blomer, U, et al Brachial plexus neurotization with donor phrenic nerves and its effect on pulmonary function. J Neurosurg 2002;96,523-526[Medline]
- Chalidapong, P, Sananpanich, K, Kraisarin, J, et al Pulmonary and biceps function after intercostal and phrenic nerve transfer for brachial plexus injury. J Hand Surg [Br] 2004;29,8-11[Medline]
- Giddins, G, Kakkar, N, Alltree, J, et al The effect of unilateral intercostal nerve transfer upon lung function. J Hand Surg [Br] 1995;20,675-676[Medline]
- Guazzi, M, Reina, G, Tumminello, G, et al Exercise ventilation insufficiency and cardiovascular mortality in heart failure: the critical independent prognostic value of the arterial CO2 partial pressure. Eur Heart J 2005;26,472-480[Abstract/Free Full Text]
- Guazzi, M, Reina, G, Tumminello, G, et al Improvement of alveolar-capillary membrane diffusing capacity with exercise training in chronic heart failure. J Appl Physiol 2004;97,1866-1873[Abstract/Free Full Text]
TOP
Abstract
Introduction
