INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung transplantation has become a viable option for the management of end-stage cardiopulmonary illness. In the absence of complications, lung transplant recipients have near-normal cardiopulmonary function at rest, and effectively sustain the levels of physical activity required for daily life. However, lung allograft recipients are abnormal when evaluated with cardiopulmonary exercise testing (1); maximal work rates and maximal oxygen uptake (VO_2max) are depressed by 40 to 60% as compared with predicted values (1, 6, 7). VO_2max is determined by three major factors: ventilatory capacity, maximal cardiac output, and oxygen extraction of the working muscles (8). Lung allograft recipients typically exhibit normal cardiac and ventilatory responses to exercise, and skeletal muscle dysfunction is therefore thought to set these subjects' exercise limit (1).

This belief is supported by reports that muscle strength is lower than predicted in lung allograft recipients (1, 9). However, few studies have included a control population, and little is known about the relative susceptibility to weakness of respiratory versus limb skeletal muscles. We therefore evaluated respiratory muscle strength and ankle dorsiflexor function in stable lung allograft recipients at an average of 37 mo after transplantation (range: 5 to 102 mo). Data from lung recipients were compared with measurements in healthy age- and sex-matched control subjects. Two hypotheses were tested as follows:

Hypothesis 1: Respiratory muscle strength is diminished in lung allograft recipients. Indices of respiratory muscle strength (maximal inspiratory pressure [MIP] and maximal expiratory pressure [MEP]) are below predicted values in heart-lung and lung allograft recipients (1, 10, 11). However, the MIP of allograft recipients does not appear to be abnormal when compared with that of appropriately matched control subjects (1, 9), and the MEP of transplant recipients has not been assessed in a controlled trial. We therefore assessed both the MIP and MEP of transplant recipients, evaluating the resulting data against measurements made in control subjects and against published control values.

Hypothesis 2: Limb muscles are weaker and more susceptible to fatigue in transplant recipients. Earlier studies have found that power output by leg flexor and extensor muscle groups is less than predicted in heart-lung allograft recipients (11), and that time to exhaustion during knee extensor exercise is diminished in lung transplant recipients (12). However, the limb muscle strength of allograft recipients has not been assessed in a controlled trial, nor have contractile properties or fatigue characteristics been evaluated using muscle stimulation. We therefore measured the strength of ankle dorsiflexor muscle groups during maximal volitional contractions. We also measured the force-frequency characteristics and fatigue properties of the tibialis anterior muscle, using transcutaneous electrical stimulation. Data from transplant recipients were compared with values obtained from control subjects.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Human Subjects

Our experimental protocol was approved by the Baylor Affiliates Review Board for Human Subject Research before subject recruitment for our study. Criteria for participation in the study were: (1) age  65 yr; (2) status > 3 mo posttransplantation; (3) no evidence of myopathy, cardiovascular disease, anemia, hepatic dysfunction, renal disease, or cystic fibrosis; (4) no infection or allograft rejection requiring admission, invasive testing, or intravenous therapy within the preceding 30 d; (5) no currently administered muscle relaxants; and (6) ability to perform experimental maneuvers and give informed consent. At the time of the study, 41 lung transplant recipients at Baylor College of Medicine were screened as participants; 21 met our enrollment criteria; nine were willing to participate in the study. Nine healthy sex- and age-matched volunteers were recruited as control subjects; sedentary individuals were preferentially included. Anthropometric data from transplant recipients and control subjects are shown in Table 1.

Clinical characteristics of the lung recipients are summarized in Table 2. The recipients' ages ranged from 30 to 65 yr. Preoperative diagnoses included interstitial pulmonary fibrosis, chronic obstructive pulmonary disease, and primary pulmonary hypertension. Eight individuals received single-lung transplants and one received two lungs; transplants were performed from 5 to 102 mo before the study. After transplantation, pulmonary function was measured according to American Thoracic Society Guidelines (13), and was compared with the reference population of Knudson and coworkers (14); severity of disease was assessed according to the criteria of Enright and Hyatt (15) from the most recent spirometric examination of each individual. All transplant recipients except Subject 6 reported a history of exercise intolerance, defined as tiredness, or shortness of breath during physical activity. None of the recipients exhibited a decline in oxygen saturation with exercise, as measured either in a cardiopulmonary stress test or in ambulation during a clinic visit. Each of six subjects underwent a symptom-limited cardiopulmonary exercise test within 12 mo of the study; VO_2max during bicycle ergometry was below the predicted value in every individual (range: 34 to 58%); the symptoms reported to limit exercise were leg tiredness or overall body fatigue. Cardiac function was normal in all transplant recipients as measured with two-dimensional echocardiography within 12-mo of enrollment in the study.

Medications that might have affected muscle function in our transplant recipients were limited to corticosteroids and cyclosporine. All subjects were receiving prednisone, at an average dose over the preceding 6 mo of 10.3 mg/d (range: 7.5 to 15 mg/d). Subjects also took cyclosporine daily to achieve a serum level of 200 to 300 ng/ml.

Respiratory Muscle Testing

Respiratory muscle function was tested with a method adapted from Black and Hyatt (16). Each individual sat upright, wore noseclips, and breathed through either of two mouthpieces. MEP maneuvers were performed with a plastic cylinder whose circular surface, which was in contact with the subject's face, had dimensions comparable to those of the rubber device used by Black and Hyatt. Subjects held the device firmly against the perioral region during expiratory efforts; soft-tissue deformation prevented detectable leaks in any of the maneuvers accepted for analysis. MIP maneuvers were performed with a standard commercial mouthpiece that was held between the teeth and sealed with the lips. The mouthpiece assembly incorporated a small leak (~ 2-mm diameter) to prevent pressure development by cheek-muscle activity. The mouthpiece was attached to a differential pressure transducer (range ± 500 cm H₂O; Validyne Corp., Northridge, CA) for measurement of mouth pressure and to a Fleisch pneumotachograph (Fleisch, Lausanne, Switzerland) by which airflow was measured. The flow signal was integrated to reflect changes in lung volume; pressure and volume tracings were displayed on a strip-chart recorder. Maneuvers were performed against a closed airway following a maximal inspiration (MIP) or a maximal expiration (MEP). The pressures we recorded were maintained for  1 s. Each subject performed multiple maneuvers with coaching and instruction until three technically satisfactory measurements were recorded. MEP and MIP values reported herein reflect the largest expiratory and inspiratory pressures developed by each individual.

Limb Muscle Testing

We measured the functional properties of ankle dorsiflexor muscle groups with methods developed previously (17, 18). Each subject lay in the supine position with either leg positioned in an orthopedic leg frame as illustrated in Figure 1. The hip and knee were semiflexed; the foot was positioned against a vertical footplate and was supported by a heel pad. A canvas strap was placed across the metatarsal heads and was connected to a strain gauge (Model 3397-100; Eaton Corp., Troy, MI); the strap was tightened to apply a resting force of 30 Newtons (N). The leg and thigh were strapped to the frame so as to minimize the effect of agonist muscle activity without occluding blood flow. The strain gauge measured forces developed by ankle dorsiflexor muscle groups, including the tibialis anterior muscle. Forces were recorded on a strip-chart recorder.

View larger version (41K): [in this window] [in a new window]   Figure 1.   Measurement of ankle dorsiflexor muscle contractile function. Diagram depicts subject with leg positioned for contractile study; straps secure the leg within frame; electrodes positioned over motor point of tibialis anterior are connected to stimulator used to activate the muscle; footstrap transmits forces developed by ankle dorsiflexor muscles to load cell attached to underside of footplate; force signal is displayed on strip-chart recorder.

After positioning the subject's leg, we located the motor point of the tibialis anterior by administering repetitive transcutaneous stimuli (0.2 ms square-wave pulses of 10 to 15 mA and 1 Hz) with a stimulator (Model S48; Grass Medical Instruments, Quincy, MA), constant-current stimulus isolator (Model A385; World Precision Instruments, Sarasota, FL), and bipolar probe (interelectrode distance = 7 cm). The strength of evoked contractions was assessed by palpation of the tendon at the ankle and by recordings of twitch force. The motor point was defined as the stimulation site that elicited the most forceful twitch contraction. Carbon-coated rubber electrodes (5 cm²) were positioned over the motor point (cathode) and 7 cm distally (anode). Maximum tolerable current was determined with repetitive tetanic stimuli; the current was increased until an 850-ms stimulus train at 40 Hz was reported by the subject to be painful (i.e., pain threshold); the current was then decreased (typically by 2 to 5 mA) to a level just below the pain threshold. In the absence of pain, current was limited to 100 mA.

A standardized protocol was used to obtain data from each subject. The strength of the ankle dorsiflexor muscles was determined by having the subject pull maximally against the foot strap, with this action defined as a maximal voluntary contraction (MVC). This maneuver was performed in triplicate. Peak forces developed during three MVC maneuvers were averaged to obtain a single MVC value for each subject. Transcutaneous electrical stimulation was used to determine the force-frequency relationship of the tibialis anterior, using stimulus frequencies of 1 (twitch), 10, 20, 40, 80, and 120 Hz. Muscle contractions were stimulated at 30-s intervals with an 850-ms train duration. The entire force-frequency relationship was measured in duplicate in each subject. Fatigue was induced by stimulating the tibialis anterior to contract repetitively for 15 min through the application of 1-s trains of 10-Hz stimuli delivered every 2 s (duty cycle = 0.5).

Statistical Analyses

Data sets were evaluated for normality using Kolmogorov-Smirnov criteria (19); one-way analysis of variance (ANOVA) (19) was used to assess differences between normally distributed data (MEP, maximal tetanic force, fatigue indices); the Mann-Whitney rank sum test (19) was used to evaluate nonnormal data sets (MIP, MVC). Differences between force-frequency relationships were evaluated through ANOVA for repeated measures (20). Data are reported as means ± SEM. Differences were deemed significant at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Respiratory Muscle Function

The maximal respiratory pressures developed by individual subjects are shown in Figure 2. Separate plots are shown for data from females and males relative to published values for healthy individuals (16). Pressures developed by control subjects fell within the normal range during both inspiratory and expiratory maneuvers. The inspiratory pressures developed by transplant recipients also were within normal limits. However, during maximal expiratory maneuvers, the pressures developed by transplant recipients tended to fall below the data recorded for healthy subjects. Figure 3 shows a direct comparison of average respiratory pressures developed by transplant recipients and control subjects. Transplant recipients developed MEPs that averaged 30% less than contemporaneously measured control values (p < 0.04) and 35% less than values predicted (196 ± 13; p < 0.015) from regression equations fitted to age- and gender-specific data provided by Black and Hyatt (Figure 2). In contrast, inspiratory pressures developed by transplant recipients were not significantly different from pressures of control subjects or from predicted normal values (98 ± 7%).

View larger version (17K): [in this window] [in a new window]   Figure 2.   Maximal static respiratory pressures developed by female subjects (left panel ) and male subjects (right panel ). Closed triangles depict data from transplant recipients; dotted triangles represent data from control subjects; open circles represent measurements of healthy individuals replotted from Black and Hyatt (16); the latter data were used to compute regression equations for normal relationships between age and MEP (females: y = 0.5644x + 174.9, r 2 = 0.053; males: y = 0.8112x + 264.1, r 2 = 0.069), and between age and MIP (females: y = 0.6736x  110.4, r 2 = 0.166; males: y = 0.7197x  147.5, r 2 = 0.153); these relationships were used to compute age- and sex-specific predicted values (see text).

View larger version (27K): [in this window] [in a new window]   Figure 3.   Average values for maximal static pressures developed during expiratory efforts (upper panel ) and inspiratory efforts (lower panel ). Closed bars depict values from transplant recipients (mean ± SEM); shaded bars depict data from control subjects (*less than control, p < 0.05).

Ankle Dorsiflexor Function

Lower limb weakness also was detected in transplant recipients. Ankle dorsiflexor strength was assessed from the forces produced during maximal voluntary contractions. As shown in Figure 4, transplant recipients developed 39% less force than did control subjects during this maneuver (p < 0.05), a deficit similar to the weakness observed in expiratory efforts.

View larger version (31K): [in this window] [in a new window]   Figure 4.   Maximal isometric forces developed by ankle dorsiflexor muscle groups during volitional efforts. Closed bar depicts value from transplant recipient (mean ± SEM); shaded bar depicts value from control subjects (*less than control, p < 0.05).

Electrical stimulation of the tibialis anterior was tolerated by only six of our nine transplant recipients. The force-frequency relationships measured in unfatigued muscles are shown in Figure 5. Figure 5A depicts absolute forces. The average values measured in transplant recipients were 17 to 28% lower than control values at each level of tetanic stimulation (10 to 120 Hz). However, large variability in both data sets and a small sample size precluded statistical resolution of these differences. Figure 5B shows the normalized form of this relationship, in which relative forces are expressed as a percentage of maximal force. This plot is strikingly similar for the two study groups. Neither the position of the relative force-frequency curve nor its configuration was abnormal in transplant recipients.

View larger version (15K): [in this window] [in a new window]   Figure 5.   Force-frequency characteristics of tibialis anterior muscle plotted as absolute forces (upper panel ) or as relative forces (lower panel ). Closed circles depict data from transplant recipients (mean ± SEM); open circles represent data from control subjects. Relationships did not differ significantly in either panel.

Repetitive tetanic contractions were used to induce peripheral fatigue of the tibialis anterior. This was reflected by a monotonic decline in force over 15 min that did not differ in rate or magnitude between the two study groups. Average forces produced at the end of the fatigue protocol were not different when compared either as absolute forces (3.8 ± 1.4 N for transplant recipients versus 4.2 ± 1.2 N for controls) or as percentages of initial force (68 ± 5% versus 64 ± 5%).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study was the first controlled trial to demonstrate muscle weakness in humans following lung transplantation. Our data document significant expiratory and lower limb weakness in stable lung allograft recipients studied 3 yr after surgery. Inspiratory muscle strength was well preserved in these individuals, indicating that muscle groups differ in their response to transplantation-related insults. However, disseminated weakness of trunk and limb musculature is likely to influence the persistent exercise intolerance experienced by lung transplant recipients.

Respiratory Muscle Function in Allograft Recipients

Four previous studies have assessed inspiratory muscle strength in lung allograft recipients; all used MIP measurements as the endpoint. Sciurba and coworkers (10) studied seven heart-lung recipients 14 mo after transplantation; these individuals achieved an average MIP of only 48 cm H₂O. Ambrosino and colleagues (11) studied 11 heart-lung allograft recipients 18 mo after transplantation; their MIP averaged 65 cm H₂O or ~ 80% predicted. Williams and colleagues (1) studied 13 lung allograft recipients from 1 to 2 yr after transplantation; their MIP values averaged 90% of those predicted. None of the aforementioned studies included data from control subjects. In contrast, Sanders and colleagues (9) compared MIP values of eight female heart-lung recipients at 7 mo after transplantation, with data obtained from age-matched healthy women. MIP appeared to be low in transplant recipients, averaging only 54 cm H₂O; however, MIP was similarly depressed in healthy controls, and the difference between the two groups was not significant. Our lung allograft recipients achieved MIP values that were well within the normal range and were not different from contemporary control values. These findings are consistent with reports that diaphragm function is relatively unaffected by lung transplantation (21, 22).

Fewer data are available on expiratory muscle strength after lung allograft. Only two previous reports included MEP measurements. Sciurba and coworkers (10) measured a mean MEP of +58 cm H₂O, and Ambrosino and coworkers (11) reported an average value of +82 cm H₂O or ~ 75% predicted. Such values suggest expiratory muscle weakness, but neither study tested this postulate directly or included a control group. We found that MEP was significantly decreased in lung allograft recipients as compared either with contemporaneous measurements in healthy control subjects or with values predicted from the data of Black of Hyatt (16). This finding demonstrates loss of function in major muscle groups of the trunk, including the abdominal muscles, internal intercostal muscles, or both. The observed weakness may have been caused by atrophy of the expiratory muscles, myopathic dysfunction, or incomplete motor activation.

Interpretation of MIP and MEP results is influenced by several factors. Both maneuvers are volitional efforts that may be limited by the ability of the subject to understand and perform the required tasks. The normal magnitude and uniform distribution of MIP data from our transplant recipients indicate that learning, language, and general motor skills did not prevent normal performance of the required maneuver. Accurate measurements also require appropriate mouthpiece design. Our equipment was modeled after that of Black and Hyatt (16). A small leak prevented negative pressure generation by muscles of the mouth and throat during MIP maneuvers; a perioral mouthpiece was used for MEP maneuvers to support soft tissues of the face. Similar equipment and procedures yielded similar results; the maximal pressures measured in our control subjects are indistinguishable from the data provided by Black and Hyatt.

Limb Muscle Function in Allograft Recipients

It is commonly thought that the exercise capacity of lung allograft recipients may be limited by limb muscle function. However, only two previous studies have examined limb muscle performance in such individuals. Ambrosino and coworkers (11) measured the isokinetic power of knee extensor and flexor muscle groups in heart-lung recipients for up to 18 mo after transplantation. The power output of each muscle group remained below the predicted level throughout this postoperative period. No attempt was made to compare transplant recipients with a healthy control population. More recently, Evans, and associates (12) studied nine individuals who had received lung allografts an average of 24 mo earlier. As compared with age- and sex-matched control subjects, transplant recipients had a shorter time to exhaustion during knee-extension exercise. This was associated with a lower VO_2max and exaggerated acidosis in the quadriceps muscle as measured with ³¹P-magnetic resonance spectroscopy. These data demonstrate decreased endurance of knee extensor muscles in transplant recipients, and indicate that metabolism is altered in these muscles.

In the present study we measured several aspects of limb muscle function. Ankle dorsiflexor strength was assessed by measuring MVC, a volitional, quasiisometric maneuver analogous to those used for MIP and MEP measurements. MVC values of the control population were lower and more variable than in our previous studies (17, 18, 23) because older, less fit individuals were studied in the current protocol. As compared with control subjects, the ankle dorsiflexor muscle strength of transplant recipients was significantly depressed. The magnitude of this depression was similar to that observed in expiratory maneuvers, suggesting that the two respective muscle groups have comparable susceptibilities to transplant-associated losses. The muscles that contribute to ankle dorsiflexion include the tibialis anterior, extensor digitorum longus, extensor hallucis longus, and peroneus tertius. The decrement in maximal force that we observed may reflect atrophy or myopathy of these muscles or may reflect a deficit in motor activation.

We used electrical stimulation to selectively assess the contractile function of an individual ankle dorsiflexor muscle, the tibialis anterior. In the six transplant recipients who participated in this assessment, absolute forces were below control values at all tetanic stimulation frequencies, ranging from 10 to 120 Hz. The magnitude of this force depression was similar to that seen in MVC data; however, the small sample size and large variability precluded statistical resolution. Plots of relative force versus stimulus frequency were more informative. The averaged curve obtained from muscles of transplant recipients was superimposable on the relationship observed in control muscle. The configuration of these two curves and their positions in the frequency domain were virtually identical. These properties indicate that the fiber composition of activated motor units was similar in the transplant and control groups (24), and provide no functional evidence of gross myopathy.

We also used electrical stimulation to assess the fatigue characteristics of the tibialis anterior. This strategy avoided central factors that can contribute to fatigue (e.g., decreased effort, impaired neural transmission, and altered motor-unit recruitment) (25). As was observed in previous studies (17, 18, 23), repetitive tetanic contractions caused a progressive decline in contractile force that reflected peripheral fatigue. The time course and magnitude of this decline were indistinguishable in transplant recipients and controls, as were the relative forces developed at the beginning of the protocol. Thus, fatigue characteristics of the distal motor nerve and muscle fibers of the tibialis anterior were not abnormal in our transplant recipients. This finding is inconsistent with a metabolic myopathy, which Tirdel and associates (7) have suggested may occur in lung transplant recipients.

Our results also contrast with the report by Evans and coworkers (12) that the endurance time of transplant recipients was diminished during volitional knee-extension exercise. The most likely explanation for this discrepancy is that different methods were used to assess muscle function in their study and ours. Electrically stimulated fatigue protocols are relatively insensitive to the influence of muscle weakness (26). In our two study groups, force-frequency characteristics were equivalent before fatigue; 10-Hz stimulation therefore evoked contractions of comparable intensities (25 to 30% of maximal force) at the beginning of the fatigue protocol. Since relative forces were similar at the outset, fatigue developed at similar rates in the two groups. In contrast, the subjects studied by Evans and coworkers (12) worked against a standardized mechanical load. If knee extensor muscle groups were weakened in their transplant recipients, as we observed for ankle dorsiflexor muscles in our subjects, then knee extensors would be expected to maintain a given work or power output for less time (see the subsequent discussion of muscle weakness and exercise capacity). Two other factors also may have contributed to the different findings in our study and that of Evans and coworkers: (1) ankle dorsiflexors may be less susceptible to transplant-associated dysfunction than knee extensors; or (2) the central component of fatigue could be exaggerated by the neuropathic changes associated with solid-tissue transplantation (27) and chronic cyclosporine administration (28); exaggeration of central fatigue would decrease endurance during volitional exercise without altering the peripheral fatigue of electrically-stimulated muscle.

Transcutaneous stimulation produces data that are reproducible over periods of days to weeks (17), and which are highly sensitive to functional changes within individual muscles. Using subjects as their own controls, we have used transcutaneous stimulation to detect contractile differences between dominant and nondominant limbs (17), to document the rightward shift of the force-frequency curve produced by fatigue (23), and to establish that antioxidant pretreatment inhibits fatigue (23). However, in the current study, subjects could not be used as their own controls. This limited our ability to resolve differences in absolute force production, which varies importantly among individuals. In part, this variability reflects differences in muscle mass and contractile function. But the effectiveness of electrical stimulation also influences absolute force. Skin resistance, subcutaneous fat deposits, motor-point anatomy, and the subjective response to electrical stimulation are independent variables that affect stimulation efficacy and vary markedly among subjects.

Possible Causes of Muscle Weakness

Several factors predispose lung allograft recipients to muscle dysfunction. First, repeated episodes of infection and rejection favor chronic increases in cytokines and other inflammatory mediators that are deleterious to muscle (29). We excluded lung recipients who had evidence of rejection or active infection within 30 d of beginning of our study. However, we could not exclude the possible long-term effects of earlier inflammatory events.

Second, chronic use of antiinflammatory drugs can compromise limb and respiratory muscle function. Several reports implicate glucocorticoid myopathy as a cause of respiratory muscle weakness (32, 33). The average daily dose of glucocorticoids used by our transplant recipients over the 6 mo preceding our study was 10.3 mg/d, exceeding the dose reported by Decramer and coworkers (32) to cause respiratory muscle weakness. Also, cyclosporine A induces mitochondrial myopathy and decreases exercise endurance time in rats (34), and has been implicated as a myopathic agent in humans (35, 36). All of the lung recipients in our study were receiving chronic cyclosporine therapy.

Third, underlying pulmonary disease can affect the behavior of lung allograft recipients, causing them to lead a sedentary life. Inactivity causes muscle deconditioning, with decrements in strength and aerobic adaptation (37). All but one of our transplant recipients reported a history of exercise intolerance, suggesting that these individuals were predisposed to physical inactivity.

Fourth, skeletal muscle can be indirectly compromised by nutritional deficits (38) and systemic organ dysfunction (39). Our study population was selected to minimize these effects. Nutritional problems were not detected in any subject, and major organ dysfunction was an explicit exclusion criterion.

Muscle Weakness and Exercise Capacity

The present study demonstrates expiratory and ankle dorsiflexor muscle weakness in mature lung allograft recipients. Muscle groups were not affected uniformly by this pathologyinspiratory muscle strength remained normalbut the observed weakness has important implications for exercise capacity. Maximal muscle force production is directly related to the endurance of both limb muscles (8) and respiratory muscles (42). Weakness lessens the absolute force, work, or power that can be sustained during volitional exercise (8). Conversely, for a given load, muscle weakness tends to reduce the time during which an exercise task can be maintained. Limb muscle weakness predisposes transplant recipients to abbreviated walking times and submaximal performance during cycle ergometry. Expiratory muscle weakness tends to limit volitional tasks that have an expulsive component (e.g., lifting). Our findings are consistent with the general thesis that muscle dysfunction contributes to exercise intolerance in transplant recipients. Further studies are needed to define the muscle groups that exhibit transplant-associated dysfunction and to determine the mechanism(s) underlying such dysfunction.