INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: COPD; diaphragm; sarcomere; disruption; exercise

Numerous studies in both animal models and humans have shown that exertion can injure the limb muscles (1), which may involve several mechanical and metabolic processes (2, 3). Exercise-induced muscle injury is associated with morphological abnormalities such as degeneration of the cytoplasm, disruption of cell membranous structures (sarcolemma, mitochondria, sarcoplasmic reticulum, and T-tubules), and disorganization of the contractile myofibrils (including Z-band streaming, misalignment of the myofilaments, and desmin loss) (4). Although exertion-induced injury is associated with impaired muscle function (decreased strength and/or endurance) (1), muscle injury also appears to stimulate complex mechanisms that can induce adaptive repair to increased utilization and stress in skeletal muscles (i.e., training) (5, 6).

The respiratory muscles perform low-frequency, low-intensity contractions every minute throughout life (7). In patients with respiratory diseases, the diaphragm (the main respiratory muscle) can experience mechanical overloading (8). In patients with chronic obstructive pulmonary disease (COPD), diaphragm dysfunction appears to have an inverse association with survival (9) although its link to diaphragm injury is still unknown. Injury of the diaphragm has been observed in several animal models of respiratory loading (6, 10, 11). Evidence of diaphragm injury in humans is sparse, however, some evidence of diaphragm fiber injury has been described after sudden infant death syndrome (12) and fatal asthma (13). It may have important clinical implications in patients with chronic respiratory conditions and after acute episodes of an increased inspiratory workload such as exacerbation or exercise.

We postulated that acute respiratory muscle loading (e.g., due to exercise) and/or chronic increased diaphragm activity (e.g., due to COPD) could be associated with diaphragm injury. By employing a sensitive technique to detect injury in muscle biopsies, the purpose of this study was, first, to determine whether an acute inspiratory overload induces injury of the human diaphragm, and second, to determine if diaphragm injury is associated with the presence of COPD.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Population

Eighteen patients with COPD and 11 control patients with normal pulmonary function (62 ± 10 yr) undergoing thoracotomy or laparotomy for localized lung or colon neoplasm were selected for the study. Subjects were sedentary. Patients were systematically selected through the Lung Cancer Multidisciplinary Committee and the Department of Surgery at our hospital. A diagnosis of COPD was determined from a clinical history consistent with chronic bronchitis and/or emphysema, a long history of cigarette smoking, and pulmonary function tests consistent with irreversible airflow obstruction (FEV₁ < 75% predicted, and FEV₁/FVC ratio < 65% predicted) according to European Respiratory Society criteria (14). Exclusion criteria included asthma, coronary disease, undernourishment (body mass index < 20 kg/m²), chronic metabolic diseases (e.g., diabetes, hypo- or hyperthyroidism), orthopedic diseases, suspected paraneoplastic or myopathic syndromes, previous abdominal or thoracic surgery, and/or treatment with steroids, hormones, or cancer chemotherapy.

Study Design

The study was conducted in accordance with the World Medical Association guidelines for research in humans. An institutional ethics board approved the protocols. After informed consent was obtained from patients, pulmonary function and nutritional status were assessed. Respiratory muscle function was evaluated 1 to 2 d later. Biopsies from the apposition zone of the costal diaphragm (~ 0.5 cm × 0.5 cm × 0.5 cm) were obtained 24 to 48 h later, at the beginning of surgery. Diaphragm samples were taken ~ 2 cm from the costodiaphragmatic angle at the mid-axillary line. The length of time between threshold inspiratory loading and diaphragm sampling was similar for the patients with and without COPD. Each biopsy was divided into two blocks, and processed for fiber typing and electron microscopy.

Measurements

Nutritional assessment. Nutritional assessment consisted of determining body mass index (BMI = weight/height²); the percentage of ideal body weight (%IBW) calculated using predicted values (15) provided by the World Health Organization (%IBW = [actual weight/ideal weight] × 100); and blood analysis of serum cholesterol, triglycerides, total protein, albumin, globulins, albumin/globulin index, and prothrombin consumption time.

Pulmonary function tests. Forced spirometry and inspiratory capacity (Datospir 92; Sibel, Barcelona, Spain), thoracic gas volume (Masterlab; Jaegger, Würzburg, Germany), and the transfer factor for carbon monoxide (TL_CO, single-breath method) (Masterlab; Jaegger) were measured in each patient and compared with reference values from a mediterranean population (16, 17). Arterial blood samples were analyzed for Pa_O2, Pa_CO2, and pH (ABL 330; Radiometer, Copenhagen, Denmark).

Respiratory muscle strength and breathing pattern. While patients breathed through a two-way valve (Hans Rudolph, Kansas City, MO), transdiaphragmatic pressure during both quiet breathing (Pdi) and a maximal sniff maneuver at functional residual capacity (Pdi_max) was measured (18). The following were also determined: tidal volume (VT), respiratory rate (RR), inspiratory time (TI), and total respiratory time (Ttot) obtained from a pneumotachometer (Screenmate; Jaegger) inserted in the inspiratory circuit. Pdi and Pdi_max were quantified using esophageal and gastric balloons (19) connected to pressure transducers (BP1050, Biopac Systems, Goleta, CA). Maximal inspiratory mouth pressure (PI_max) was measured using a manometer (BP1050, Biopac Systems) with an occludable mouthpiece (Sibelmed, Sibel) according to the technique described by Black and Hyatt (20) and referenced to values from Wilson and coworkers (21). Calibrations were performed at the beginning and end of each study. The tension-time index (22) was calculated from the following formula: TTdi = (Pdi/ Pdi_max) · (TI/Ttot).

Threshold inspiratory loading tests. These tests were performed using a threshold device (23) (after respiratory muscle strength measurements) as a part of the presurgical evaluation in a subset of seven patients with COPD and five control subjects. The assignment to each of both groups (loaded or nonloaded) was performed at random according to the study period. In the first inspiratory loading test, the volunteers breathed against incremental loads (50 g every 2 min) until maximal inspiratory threshold pressure was reached, which was defined as the maximal threshold pressure that a volunteer tolerated for 60 or more s (24). In the second inspiratory loading test, subjects breathed against a submaximal constant load equivalent to 80% maximal inspiratory threshold pressure until task failure. The elapsed time was defined as the inspiratory threshold endurance time. Inspiratory loading tests I and II were terminated when subjects could not produce airflow in five sequential inspiratory efforts and/or had limiting symptoms. Loading tests I and II were carried out at least 30 min apart in each patient. The following variables were continuously recorded: mouth pressure (BP1050, Biopac Systems), airflow (Screenmate, Jaegger), chest and abdominal movements (RSP100A, Biopac Systems), oxyhemoglobin saturation (Artema MM205, Medical AB, Stockholm, Sweden), and electrocardiographic signal (ECG100A, Biopac Systems).

Analysis of Diaphragm

Fiber type sizes and proportions. One block from each biopsy was quick-frozen in isopentane cooled in N₂(l) and stored at 70° C. Ten-micrometer-thick sections were cut varying the inclination of the holder by 5° increments until the minimum cross-sectional area was obtained, which was defined as truly transverse (25, 26). Cross sections were processed for ATPase (27). Cross-sectional area, mean least diameter, and proportions of types I and II fibers were assessed using light microscopy (OLYMPUS, Series BX50F3, Olympus Optical Co., Shinjukuku, Tokyo, Japan) coupled to an image-digitizing camera (Pixera Studio, Version 1.2, Pixera Corporation, CA) and a morphometry program (NIH Image, Version 1.60). At least 100 fibers were measured from each diaphragm biopsy. Fiber diameters between 40 and 80 µm were considered normal (27, 28). Atrophy and hypertrophy factors were calculated for each sample (27).

Evaluation of muscle injury. Samples were processed according to standard methods (Tissue Processor E9200, Biorad, Watford, UK) described elsewhere (29). Micrographs of the muscle samples were taken from 16 randomly selected fields at constant calibrated magnifications (×1900) using a transmission electron microscope (Philips CM100, Amsterdam, The Netherlands). An accelerating voltage of 60 kV was used. Disrupted sarcomeres were evaluated as a sign of muscle injury and defined as a zone with distinct distortion of the usual sarcomeric architecture, defined by the following six criteria: discontinuity of a group of myofibrils, A- and I-band disruption, Z-band streaming, embedded subcellular components (mitochondria or collagen), preserved adjacent sarcomere, and absence of regional sectioning artifacts (scratches, holes, or chatters). To quantify the injury, both density (i.e., number of areas containing disrupted sarcomere, expressed as n/100 µm²) and proportion (i.e., abnormal area fraction, expressed as percentage) of disrupted sarcomere were normalized to the micrographed area. To control a potential contraction artifact, the contraction index of individual sarcomere was calculated (29). This analysis of sarcomere disruption was performed in duplicate using a double-blind approach. The mean value obtained by the two observers was used for the statistical analysis.

Statistical Analysis

Data are presented as mean ± standard deviation (SD) and ranges. Comparisons between the non-COPD versus COPD, and nonloaded versus inspiratory loaded groups were performed using the Mann- Whitney U tests. Spearman's coefficient was calculated to assess correlation of the variables. A linear regression analysis was used when appropriate. p Value  0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Nutrition, Pulmonary Function, and Respiratory Muscle Function

Nutritional status, assessed by both anthropometrical and biochemical variables, was found to be normal in all patients without COPD (n = 11) and with COPD (n = 18) (Table 1). In patients with COPD, the FEV₁ showed a wide range of values (22 to 74% predicted). Ten patients showed signs of air trapping (residual volume [RV] greater than 120% predicted) and five showed pulmonary hyperinflation (total lung capacity [TLC] greater than 120% predicted). In all patients without COPD and most patients with COPD, inspiratory muscle strength (Table 2) was within the normal predicted range as defined by Wilson and coworkers (21). Six of 18 patients with COPD showed a moderate to severe decrease in PI_max (range, 25 to 67% predicted) compared with normal predicted values (21). PI_max correlated with airflow obstruction (with FEV₁ %predicted: r = 0.405, p < 0.05) and pulmonary hyperinflation (with TLC %predicted: r = 0.679; with RV %predicted: r = 0.555; p < 0.01, both). Pdi (9 ± 2 cm H₂O for both groups) and Pdi_max (86 ± 26 versus 99 ± 30 cm H₂O; range: 54 to 134 cm H₂O) showed no differences between patients without COPD and patients with COPD (Table 2). In contrast, values obtained from inspiratory loading tests were significantly lower in the patients with COPD compared with the control subjects: maximal inspiratory threshold pressure and endurance time were 27% and 40% lower, respectively.

Light Microscopy Analysis

Diaphragm fibers showed normal cross-sectional area and mean least diameter in both patients without COPD and patients with COPD (Table 3). The proportion of type I and II fibers was similar to each other, and between patients without COPD and patients with COPD. The atrophy and hypertrophy indices tended to be slightly greater (not significant) in COPD. Signs of necrosis, inflammatory cells, or specific fiber type grouping were not observed in the diaphragm from either group of patients.

Electron Microscopy Analysis

Signs of injury usually observed were A- and I-band disruption and Z-band streaming (Figure 1). The variables indicating the amount of sarcomere disruptions (both density and abnormal area fraction) were found to be closely related (p < 0.01, r = 0.786). No relationships were found between sarcomere disruption and the sarcomere contraction index.

View larger version (216K): [in this window] [in a new window]   Figure 1.   Ultramicrophotography (electron microscopy) showing areas of both normal (A) and disrupted (B) sarcomere in the diaphragm sample from a patient with COPD. For definition criteria, see text.

Injury of the diaphragm in nonloaded control subjects and nonloaded patients with COPD. All diaphragm samples showed signs of sarcomere disruptions in both non-COPD and COPD groups, even in those who had not performed the inspiratory loading tests (n = 17, Table 3). However, COPD was associated with greater injury. FEV₁ was inversely correlated with both the sarcomere disruption density (p < 0.01, Figure 2A) and area fraction (r = 0.604, p < 0.01). We found that pulmonary hyperinflation explained up to 40% of the degree of diaphragm injury. Specifically, both the %RV and %RV/TLC ratio significantly correlated with the sarcomere disruption density (r = 0.723 for %RV and r = 0.795 for %RV/TLC; p < 0.01, both). This association is shown in Figure 2B. No relationships were found between the amount of sarcomere disruption and either time of patient recruitment (month or year of patient recruitment), age, or anthropometrical or biochemical nutritional parameters.

View larger version (28K): [in this window] [in a new window]   View larger version (24K): [in this window] [in a new window]   Figure 2.   (A) Relationship between airflow obstruction (% predicted FEV1) and density of sarcomere disruptions (expressed as n/100 µm2). The shaded area represents the normal predicted range for FEV1. Open triangles and circles: control volunteers and patients with COPD, respectively, who did not perform the inspiratory loading tests; closed triangles and circles: control volunteers and patients with COPD, respectively, who performed the inspiratory loading tests. (B) Relationship between pulmonary air trapping (% predicted RV) and density of sarcomere disruptions (expressed as n/100 µm2). Shaded area represents the normal predicted range for RV. Open circles: patients with COPD who did not perform inspiratory loading tests; closed circles: patients with COPD who performed the inspiratory loading tests.

Injury of the diaphragm in loaded control subjects and loaded patients with COPD. Inspiratory-loaded control subjects tended to show an 89% greater density (19.5 ± 7.5 versus 10.3 ± 8.7 sarcomere disruptions/100 µm², respectively; p = 0.07; Figure 3) and an 83% greater area fraction (7.5 ± 2.1 versus 4.1 ± 2.7%, respectively; not significant) of sarcomere disruption compared with nonloaded control subjects. The inspiratory-loaded patients with COPD showed a 38% greater density (39.0 ± 2.7 versus 28.2 ± 10.1 sarcomere disruptions/100 µm², p < 0.05) and a 32% greater area fraction (14.4 ± 1.7 versus 10.9 ± 3.3; p < 0.05) of damage compared with nonloaded COPD (Figure 3). Neither the inspiratory muscle strength (as assessed by PI_max and Pdi_max) nor endurance (maximal threshold pressure or endurance time) correlated with the amount of muscle injury.

View larger version (31K): [in this window] [in a new window]   Figure 3.   Density of sarcomere disruptions (expressed as n/100 µm2) in different patient groups according to presence of COPD and the performance of the threshold inspiratory loading tests. Means ± SD are shown.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study is the first to show sarcomere disruption in the human diaphragm. The presence of COPD was associated with greater sarcomere disruption, and the severity of airflow obstruction directly correlated with this injury. Further, exertion-induced diaphragm injury occurred during or after acute, high-intensity inspiratory loading. This provides the first evidence that diaphragm injury may be precipitated during or after performance of an inspiratory loading test in either control subjects or patients with COPD. Finally, susceptibility to sarcomere injury in response to inspiratory loading appears to be higher in the diaphragm of patients with COPD who are clinically stable and show no significant comorbidity, despite breathing against a lower inspiratory load than control subjects. In fact, patients with COPD showed a lower tolerance to breathe against an inspiratory load as shown by the lower maximal threshold pressure and endurance time sustained.

The morphological signs of injury are well defined and have been used in previous analyses of injury in respiratory and peripheral muscles (1). Transmission electron microscopy (EM) is considered to be a very sensitive method for assessing diverse signs of muscle injury (1). Previous studies have demonstrated that when animals perform strenuous exercise, less than 5% of the fibers in the contracting muscle show overt signs of injury when examined using light microscopy techniques (1), whereas EM studies can show disruption in myofilaments and other organelles in as many as 50% of the fibers in muscle biopsy specimens (1).

Because all diaphragm samples showed signs of injury (and artifacts were excluded), we postulate that sarcomere disruption is a normal phenomenon of damage/repair dynamics in muscle fibers. We found that diaphragm injury is increased in patients with COPD (with the greatest injury after inspiratory loading) even when clinically stable, and even when malnutrition, hypoxemia, hypercapnia, steroid treatment, deconditioning, and suspicion of underlying paraneoplastic syndrome are absent. Diaphragms from patients with COPD were found to be more susceptible (up to three times) to additional injury when breathing against inspiratory loads. This is supported by the finding that the diaphragm showed a similar increase of sarcomere disruption in patients with COPD submitted to inspiratory loading despite the fact that both the external load tolerated during the tests and the time spent breathing against inspiratory loads were significantly lower (40 and 70%, respectively) when compared with control subjects.

The ultimate cause of the increased diaphragm injury present in COPD is not clear. However, chronic increased loads may increase regional stress and strain on the diaphragm (8). Although Pdi was similar in patients with COPD and control subjects, this measure does not account for differences in regional stresses and strains in the diaphragm of patients with COPD due to altered diaphragmatic geometry and increased workload during exertion associated with activities of daily living. In addition to mechanical stress, metabolic factors (i.e., increased protease activity, free radicals, oxidation) might induce or accentuate diaphragm injury (30). Calpain, a cytosolic protease capable of degrading cytoskeletal and myofibrillar proteins, is increased in the diaphragm of rabbits after 1.5 h of inspiratory resistive loading (31). The excised diaphragms of hamsters after 6 d of tracheal banding were more susceptible to degradation by calpain (31). In a similar manner, the diaphragm in humans with stable COPD may show increased calpain activity, increased susceptibility to calpain degradation, and/or may undergo mechanical stresses that cause injury at a greater rate than those found in control subjects and/or noninspiratory loaded patients.

Diaphragm injury in our study appears to be similar to that observed in the diaphragm of an animal model of tracheal banding (10). We postulate that inspiratory loading likely increased mechanical and metabolic stress to the diaphragm. Injury may be worse after inspiratory loading in patients with COPD versus control subjects because greater fatigue may result in incoordination and cocontraction of agonists and antagonists (8). Increased loading could increase eccentric loads in the diaphragm (32) and increase the heterogeneity of sarcomere lengthening and, thus, their susceptibility to injury (33). Further, inspiratory loading may also increase protease activity (i.e., calpain) associated with exertion and might increase oxidation of proteins susceptible to degradation in the diaphragm of patients with COPD. Belcastro and coworkers postulated that increased oxidation of proteins may be important in marking proteins for degradation by decreasing their conformational stability (30).

The limitations of the study include its cross-sectional design and sample size. Obtaining muscle samples during thoracotomy restricts the population to patients healthy enough for surgery such that most had mild or moderate lung disease. Elective laparotomy for abdominal disease was used as a means to obtain diaphragm samples from patients with a wider range of pulmonary function (healthy and severe COPD). Future studies need to be performed to provide a broader perspective of frequency and severity of diaphragm injury in a larger group of individuals. Differences in diaphragm biopsy location do not explain the differences shown in the data as all biopsies were sampled in a similar location at the apposition zone of the costal diaphragm. This small sample may not represent the entire diaphragm. Animal studies, however, have shown that more injury is found in the costal diaphragm than other regions (10, 34).

Clinical Implications

Although structural evidence of diaphragm injury is mounting, the clinical implications are less well described. The functional losses and time course of repair of injury observed at the ultrastructural level in the present study need to be determined. However, diaphragm injury could be associated with two phenomena.

Diaphragm dysfunction. Muscle injury can precipitate decreased strength and/or endurance of the diaphragm (35). The functional impact of injury observed in both limb muscles (36) and the diaphragm (34) (e.g., force loss) is proportionately much higher than the morphological signs of injury. Further, the recovery of functional losses has been described as lasting up to 6 wk (35). The injury we observed in this study, if not accompanied by cell necrosis and inflammation, might be more quickly reversible than injury previously described in animals. These findings could be clinically important because many respiratory disorders, such as COPD, acute bronchospasm, obstructive breathing events occurring in snoring or sleep apnea syndromes, and weaning trials from mechanical ventilation, are characterized by an increase in respiratory resistance, diaphragm activity, and eventual respiratory muscle dysfunction (8). Dysfunction of respiratory muscles may result in inadequate ventilation and has been implicated as a cause of respiratory failure (8), although its specific etiology is still unknown. It is possible that respiratory muscle fatigue, weakness, or injury (alone or in concert) may contribute to respiratory failure in some cases (10). Previous reports in animals (4, 10), a few postmortem observations in humans (12, 13), and a recent study in subjects going for thoracotomy surgery (37) together with our data support the postulate that ventilatory loading (e.g., COPD) can precipitate diaphragm injury.

Diaphragm remodeling. Muscle injury associated with exertion might be an essential stage in adaptation and improved function (6). Several signs of muscle injury such as sublethal plasma membrane injury occur frequently in vivo and appears to provide a route for molecular traffic directly into and out of the cell (6). In particular, the basic fibroblast growth factor (bFGF, a potent mitogen for satellite cells and fibroblasts) has been associated with muscle remodeling but it lacks the classical peptide for export by exocytosis. As a result, membrane disruptions would permit the releasing of bFGF from the cell (6). Similarly, sarcomere injury may facilitate protein turnover such that the subsequent regenerated myofibrils have improved diaphragm strength and/or endurance.

Conclusions

This study provides the first evidence that diaphragm muscle injury can be precipitated in humans during or after acute, high-intensity inspiratory loading. Of interest, the diaphragm of patients with COPD not only showed greater injury but also greater susceptibility to additional injury during and/or after inspiratory loading. The elucidation of potential factors leading to diaphragm injury in humans may help determine the etiology of respiratory muscle dysfunction observed during clinical stability and acute respiratory failure. The potential effects of such injury on muscle function, remodeling mechanisms, and clinical outcomes warrant further studies. Defining the relationships between muscle dysfunction, muscle injury, and adaptive processes will facilitate the development of more effective training regimens in the context of pulmonary rehabilitation and may be important in the critical care setting when weaning patients from mechanical ventilation.