Abstract
Although circulating tumor necrosis factor (TNF)- α levels have been found to be increased in weight-losing patients with chronic obstructive pulmonary disease (COPD), the main causes for this phenomenon remain to be elucidated. Since hypoxia itself can enhance the production of the TNF- α in vitro, we studied the relationship between hypoxemia and activities of the TNF- α system, including circulating TNF- α and soluble TNF-receptors (sTNF-R; sTNF-R55 and -R75) levels, in 27 COPD patients and 15 age-matched healthy controls. The COPD patients showed a significant weight loss (body mass index = 18.1 ± 2.8 versus 22.8 ± 2.2 [mean ± SD] kg/m2; p < 0.0001. % fat = 16.3 ± 5.9 versus 24.3 ± 4.9 %; p < 0.001), and hypoxemia (PaO2 = 62.2 ± 9.5 versus 88.6 ± 5.9 mm Hg; p < 0.0001) as compared with the healthy controls. Serum TNF- α (6.15 ± 1.08 versus 5.41 ± 1.60 pg/ml; p < 0.05) and plasma sTNF-R55 (1.15 ± 0.49 versus 0.67 ± 0.13 ng/ml; p < 0.0001) and sTNF-R75 (3.54 ± 1.16 versus 2.25 ± 0.43; p < 0.0001) levels were significantly higher in the COPD patients than in the healthy controls. There were inverse correlations between PaO2 and circulating TNF- α and sTNF-R levels in patients with COPD (TNF- α ; r = − 0.426, p = 0.0297; sTNF-R55: r = − 0.587, p = 0.0027; sTNF-R75: r = − 0.573, p = 0.0035). In addition, we found inverse correlations between sTNF-R levels and % fat in COPD patients (sTNF-R55: r = − 0.442, p = 0.0272; sTNF-R75: r = − 0.484, p = 0.0155). TNF- α levels correlated well with sTNF-R levels (sTNF-R55: r = 0.488, p = 0.0127; sTNF-R75: r = 0.609, p = 0.0019). These relationships were not observed in the healthy controls. These data suggest that systemic hypoxemia noted in patients with COPD is associated with activation of the TNF- α system in vivo, which may be a factor contributing to the weight loss in patients with the disease.
Unexplained weight loss is common in patients with chronic obstructive pulmonary disease (COPD) (1, 2). Several studies have reported that tumor necrosis factor (TNF)-α, a multifunctional cytokine, may play a potential role in the weight loss noted in COPD patients as well as in other cachexic patients with chronic wasting diseases (3-10). Although circulating TNF-α levels were increased in COPD patients, the main causes of this phenomenon remain to be clarified (9, 10). Since no direct correlations between circulating TNF-α levels and weight loss were found in patients with COPD, unidentified factors that could interfere with both the TNF-α system and cachexia in COPD patients have been considered (9, 10).
Recently, it has been suggested that tissue hypoxia (less than 10% O2 exposure) markedly induces the local expression of inflammatory cytokines, including TNF-α and interleukin (IL)-1, and that overproduction of these cytokines plays pathophysiologic roles in clinical disorders, such as preeclampsia (11) or ischemic brain injury (12). Exposure to hypoxia (10 to 12% O2) potentiates the deleterious effects of the adult respiratory distress syndrome (ARDS) (13, 14). Experimentally, hypoxia (1% O2 or less) accelerates TNF-α production in lipopolysaccharide (LPS)-stimulated human alveolar macrophages (AM) (15), peripheral blood mononuclear cells (16), and in the human macrophage cell line THP-1 (17). Although hypoxic mechanisms of regulation of these cytokines have been delineated in in vitro studies, the question of whether the systemic hypoxemia observed in COPD patients could have influences on the production of these cytokines in vivo has not been extensively studied.
Against this background, we tested the hypothesis that systemic hypoxemia observed in patients with COPD might contribute to activation of the TNF-α system and therefore cause weight loss. We also evaluated the circulating levels of two kinds of soluble TNF-receptor (sTNF-R) in patients with COPD, since both of these kinds of sTNF-R are known to serve as sensitive, reliable markers for monitoring the activity of the TNF-α system (18, 19). Investigating this hypothesis may provide profound insight into the pathophysiology of the cachexia observed in patients with COPD.
The study population consisted of 27 male patients with COPD diagnosed according to the criteria of the American Thoracic Society (20). Their irreversible chronic airflow obstruction was confirmed by spirography. The patients had been clinically stable for at least 3 mo, and lacked clinical signs of exacerbation. Patients who had conditions known to affect serum TNF-α levels, such as infection, heart failure, cancer, or collagen vascular diseases were strictly excluded (3-8). C-reactive protein levels in the COPD patients were not increased, and the patients were not receiving any nutritional support therapy. Eight patients in the study were receiving supplemental oxygen, and the oxygen was discontinued approximately 1 h before the study began.
Fifteen age-matched health male volunteers were studied as control subjects. These control subjects had no medical illnesses, had normal results of physical examination and normal peripheral blood counts and blood chemistry findings, and showed no symptoms or signs of infection at the time of study.
After an overnight fast, all subjects had anthropometric measurements made and were tested for body composition with bioelectric impedance analysis, using an instrument and software made by the Omron Corporation (HBF-301; Tokyo, Japan) (21). Percent of normal body weight (% of normal BW) was calculated on the basis of the 1995 National Nutritional Assessment made by the Japanese Ministry of Health and Welfare (21). Informed consent was obtained from all subjects in the study.
FVC and FEV1 were measured with standard spirometric techniques (CHESTAC-25 part II EX; Chest Corp., Tokyo, Japan). The highest value from at least three spirometric maneuvers was used. Reference values were those proposed by Quanjer and colleagues (22). Arterial blood gas tensions were analyzed with the subject breathing room air while in the sitting position (280 Blood Gas System; Ciba Corning Diagnostics Corp., Medfield, MA).
Blood samples were collected from the subjects' cubital vein after an overnight fast. Both serum and plasma were separated from blood cells by centrifugation at 1,000 × g for 5 min. All samples were stored at −70° C until analyzed.
Serum TNF-α and plasma sTNF-R55 and sTNF-R75 concentrations were measured in duplicate with enzyme-linked immunosorbent assay (ELISA) kits (Quantikine; R&D Systems, Minneapolis, MN) (23, 24). Briefly, a microtiter plate was coated with a murine monoclonal antibody specific for TNF-α, sTNF-R55, or sTNF-R75. Standards and samples were added to individual wells. After several washes to remove unbound proteins, an enzyme-linked (conjugated to alkaline phosphatase for TNF-α and to horseradish peroxidase for sTNF-R) polyclonal antibody specific for TNF-α, sTNF-R55, or sTNF-R75 was added to the wells. After several additional washes to remove unbound antibody–enzyme reagent, a substrate solution (lyophilized nicotinamide adenine dinucleotide phosphate for TNF-α, and hydrogen peroxide and tetramethylbenzidine for sTNF-R) was added. For the measurement of TNF-α, amplifier solution (lyophilized amplifier enzymes) was added (high-sensitivity kit). The reaction was stopped with 2 N sulfuric acid. The color generated by the reaction was quantitated with a spectrophotometric microtiter plate reader (Model 450; Bio-Rad, Richmond, CA) by measuring the optical density at 490 nm for TNF-α and 450 nm for sTNF-R. The three measurements were specific for the particular analytes, and there was no cross-reactivity.
Because the normality hypothesis was not always fulfilled for most of the variables investigated, except for height, statistical analysis of differences between the two study groups was done with the Mann– Whitney U test for nonparametric data. The relations between continuous variables were evaluated with Spearman's rank correlation technique. Results are expressed as mean ± SD. Significance was determined at the 5% level. Statistical analysis was done with the Statview Statistical Package (Statview, Inc., Berkeley, CA).
Clinical characteristics and serum TNF-α and plasma sTNF-R levels of both the COPD patients and the healthy controls are summarized in Table 1. The COPD patients had significantly lower values of BW, body mass index (BMI), and percent body fat (% fat) than did the control subjects. The COPD patients also had airflow limitation, decreased values of PaO2 , and increased values of PaCO2 . The control subjects had normal arterial blood gas values, and normal values for % FVC and % FEV1 on spirography. We found that serum TNF-α and plasma sTNF-R55 and sTNF-R75 concentrations in the COPD patients were significantly higher than those of the healthy controls.
| COPD Patients (n = 27) | Healthy Subjects (n = 15) | p Value* | ||||
|---|---|---|---|---|---|---|
| Age, yr | 73.2 ± 6.9 | 69.8 ± 6.0 | NS | |||
| Height, m | 1.59 ± 0.05 | 1.63 ± 0.07 | < 0.05 | |||
| BW, kg | 46.0 ± 7.9 | 61.1 ± 8.5 | < 0.0001 | |||
| % of normal BW, % | 82.3 ± 12.9 | 103.7 ± 9.8 | < 0.0001 | |||
| BMI, kg/m2 | 18.1 ± 2.8 | 22.8 ± 2.2 | < 0.0001 | |||
| % fat, % | 16.3 ± 5.9 | 24.3 ± 4.9 | < 0.001 | |||
| % FVC, % | 62.7 ± 20.2 | 109.7 ± 17.6 | < 0.0001 | |||
| % FEV1, % | 52.4 ± 20.3 | 74.3 ± 6.9 | < 0.001 | |||
| Arterial Po 2, mm Hg | 62.2 ± 9.5 (42.7–78.0)† | 88.6 ± 5.9 (80.4–99.0)† | < 0.0001 | |||
| Arterial Pco 2, mm Hg | 47.0 ± 8.0 | 40.2 ± 3.0 | < 0.01 | |||
| TNF-α, pg/ml | 6.15 ± 1.08 | 5.41 ± 1.60 | < 0.05 | |||
| sTNF-R55, ng/ml | 1.15 ± 0.49 | 0.67 ± 0.13 | < 0.0001 | |||
| STNF-R75, ng/ml | 3.54 ± 1.16 | 2.25 ± 0.43 | < 0.0001 |
We investigated the relationships between PaO2 and levels of TNF-α and sTNF-R55 and −75 as continuous variables (Table 2). Inverse relationships between PaO2 and serum TNF-α or plasma sTNF-R levels were found in the COPD patients (TNF-α: r = −0.426, p = 0.0297; sTNF-R55: r = −0.587, p = 0.0027 [Figure 1A]; sTNF-R75: r = −0.573, p = 0.0035 [Figure 1B]). In addition the levels of TNF-α and sTNF-R55 and -R75 in COPD patients whose PaO2 levels were below 60 mm Hg (n = 10) were significantly higher than those of COPD patients whose PaO2 levels were above 60 mm Hg (n = 17) (Table 3). Further, we found inverse correlations between levels of both sTNF-R55 and -75 and % fat in patients with COPD (sTNF-R55: r = −0.442, p = 0.0272 [Figure 2A]; sTNF-R75: r = −0.484, p = 0.0155 [Figure 2B]). Serum TNF-α levels correlated well with plasma sTNF-R levels (sTNF-R55: r = 0.488, p = 0.0127; sTNF-R75: r = 0.609, p = 0.0019). No significant correlations between any of these variables and any other were observed in healthy controls.
| TNF-α | sTNF-R55 | sTNF-R75 | PaO2 | |||||
|---|---|---|---|---|---|---|---|---|
| % Fat | NS | r = −0.442 | r = −0.484 | NS | ||||
| p = 0.0272 | p = 0.0155 | |||||||
| TNF-α | r = 0.488 | r = 0.609 | r = −0.426 | |||||
| p = 0.0127 | p = 0.0019 | p = 0.0297 | ||||||
| sTNF-R55 | r = 0.912 | r = −0.587 | ||||||
| p < 0.0001 | p = 0.0027 | |||||||
| sTNF-R75 | r = −0.573 | |||||||
| p = 0.0035 |
Fig. 1. PaO2 and plasma sTNF-R55 (A) and sTNF-R75 (B) levels in patients with COPD. PaO2 and plasma sTNF-R55 and -R75 levels were determined as described in Methods. The negative correlations found in COPD patients were not observed in healthy controls.
[More] [Minimize]| COPD (PaO2 > 60 mm Hg) | COPD (PaO2 < 60 mm Hg) | p Value† | ||||
|---|---|---|---|---|---|---|
| n | 17 | 10 | ||||
| TNF-α, pg/ml | 5.79 ± 0.80 | 6.76 ± 1.25 | 0.031 | |||
| sTNF-R55, ng/ml | 0.96 ± 0.25 | 1.48 ± 0.63 | 0.010 | |||
| sTNF-R75, ng/ml | 3.03 ± 0.67 | 4.42 ± 1.33 | 0.005 |
Fig. 2. % Fat and plasma sTNF-R55 (A) and sTNF-R75 (B) levels in patients with COPD. % Fat and plasma sTNF-R55 and -R75 levels were determined as described in Methods. The negative correlations found in COPD patients were not observed in healthy controls.
[More] [Minimize]In order to examine the hypothesis that activation of the TNF-α system by systemic hypoxemia may relate to the general BW loss and/or cachexia in patients with COPD, we divided our COPD patients into two groups according to their nutritional status (Table 4). One group consisted of relatively malnourished patients with COPD (n = 13, % fat < 16%), the other consisted of relatively normally nourished patients with COPD (n = 13, % fat > 16%). We took % fat rather than BMI, as a standard with which to assess nutritional status because BMI is a less accurate and reliable measurement for clinical assessment of body composition than is % fat (25, 26). We used 16% fat as a cutoff value because of the balance that was required for the statistical analysis. Levels of both sTNF-R55 and -R75 were significantly higher in malnourished COPD patients than in normally nourished COPD patients, although the differences in the two groups' TNF-α and PaO2 levels did not reach statistical significance. However, trends toward higher TNF-α and lower PaO2 levels were observed in malnourished COPD patients.
| COPD (% Fat > 16%) | COPD (% Fat < 16%) | p Value† | ||||
|---|---|---|---|---|---|---|
| n | 13 | 13 | ||||
| TNF-α, pg/ml | 5.81 ± 0.93 | 6.45 ± 1.19 | 0.191 | |||
| sTNF-R55, ng/ml | 0.93 ± 0.27 | 1.30 ± 0.57 | 0.043 | |||
| sTNF-R75, ng/ml | 2.93 ± 0.68 | 4.01 ± 1.23 | 0.013 | |||
| PaO2 , mm Hg | 64.5 ± 10.5 | 60.3 ± 8.6 | 0.272 |
The malnourished COPD patients (n = 13, % fat < 16%) were further divided into two groups according to their PaO2 levels (Table 5). Group A consisted of patients whose PaO2 levels were over 60 mm Hg (n = 7, % fat < 16%). Group B had PaO2 levels below 60 mm Hg (n = 6, % fat < 16%). TNF-α and sTNF-R55 and -R75 levels were significantly higher in Group B than in Group A. The normally nourished COPD patients (n = 13, % fat > 16%) were similarly divided into two groups (Table 6). Group C (n = 6, % fat > 16%, PaO2 > 70 mm Hg) and Group D (n = 7, % fat > 16%, PaO2 < 70 mm Hg). We used 70 mm Hg as a cutoff value for Pao 2 because a balance was required for the statistical analysis. The mean levels of TNF-α, sTNF-R55, and sTNF-R75 were higher in Group D than in Group C, although there was no statistically significant difference between the two groups in these variables.
| Group A (PaO2 > 60 mm Hg) | Group B (PaO2 < 60 mm Hg) | p Value† | ||||
|---|---|---|---|---|---|---|
| n | 7 | 6 | ||||
| TNF-α, pg/ml | 5.73 ± 0.48 | 7.30 ± 1.24 | 0.022 | |||
| sTNF-R55, ng/ml | 0.96 ± 0.16 | 1.69 ± 0.63 | 0.003 | |||
| sTNF-R75, ng/ml | 3.15 ± 0.60 | 5.02 ± 0.98 | 0.003 |
| Group C (PaO2 > 70 mm Hg) | Group D (PaO2 < 70 mm Hg) | p Value† | ||||
|---|---|---|---|---|---|---|
| n | 6 | 7 | ||||
| TNF-α, pg/ml | 5.60 ± 1.09 | 6.00 ± 0.81 | 0.317 | |||
| sTNF-R55, ng/ml | 0.84 ± 0.20 | 1.03 ± 0.31 | 0.253 | |||
| sTNF-R75, ng/ml | 2.78 ± 0.71 | 3.06 ± 0.67 | 0.567 |
In general, the results of subgrouping of the COPD patients were consistent with our hypothesis that systemic hypoxemia might contribute to activation of the TNF-α system, and might therefore relate to malnutrition in COPD patients.
The purpose of this study was to investigate the effect of chronic hypoxemia on activation of the TNF-α system in patients with COPD. We found inverse correlations between PaO2 levels and the TNF-α system in terms of circulating TNF-α and sTNF-R levels in patients with COPD. These results suggest that systemic hypoxemia per se is associated with activation of the TNF-α system in patients with COPD. Moreover, the COPD patients in our study were clinically stable and showed no clinical signs of exacerbation of COPD from infection or heart failure.
In this study, the inverse correlation between PaO2 and circulating TNF-α levels was less profound than that between PaO2 and circulating sTNF-R levels in patients with COPD. Biologically active TNF-α is difficult to detect because of its short half-life (approximately 6 to 7 min), the formation of complexes with sTNF-R, and its renal clearance (18, 19). As compared with TNF-α, circulating sTNF-R levels are more sensitive markers for monitoring the activity of the TNF-α system, because the clearance of circulating sTNF-R is relatively slow (18, 19). In addition, Scannell and coworkers (17) showed that exposure of THP-1 cells to nonlethal levels of hypoxia (1% O2) causes rapid release of TNF-α and sTNF-R in vitro. The pattern of release of TNF-α is biphasic, and its level starts to decline about 18 h after hypoxic exposure, whereas sTNF-R is released at a constant rate in a time-dependent manner. Taken together, our findings suggest that systemic hypoxemia itself is associated with activation of the TNF-α system in vivo in patients with COPD.
The hypoxic induction of the TNF-α system observed in several experiments done with cell culture systems (15-17) may not be directly applicable to COPD patients because of the intense hypoxic environment in cell culture studies (1% O2 exposure or less), and because of a multitude of interacting and uncontrollable factors in the “intact organism” (i.e., humans with COPD). In addition, tissue hypoxia can be prevented by increasing cardiac output and shifting the oxygen– hemoglobin dissociation curve (27-29), with the result that the levels of hypoxemia seen in our COPD patients may not have been associated with tissue hypoxia to the point at which activation of the TNF-α system was documented in several experimental studies (15-17).
Nonetheless, several considerations warrant further discussion. First, when systemic arterial blood reaches the tissue capillaries, a substantial decrease in Po 2 occurs as oxygen passes to cells (and ultimately to the mitochondria, where oxygen is utilized), essentially by passive diffusion (the process known as the oxygen cascade) (27-29). Although the reduction in Po 2 for tissues or cells seems to be far less in COPD patients than observed in the several experiments done with cell culture systems (15-17), any decrease in PaO2 levels must be reflected in a reduced tissue Po 2 in the oxygen cascade (27-29). Second, since the cytologic environment is considerably different in vivo and in vitro, the hypoxic response in the two settings is not necessarily equivalent. For example, although the tissue Po 2 differs throughout the body, the Po 2, at least in some cells, is as low as 1 to 5 mm Hg, even under the physiologic conditions present in healthy subjects (27-29). This level is similar to those found with intense hypoxic exposure in experimental cell studies (15-17). This may suggest that the baseline level of oxygen tension is already different under the two sets of circumstances, and that oxygen sensitivity is considerably different between the two. Third, hypoxia may be enhanced rather than compensated in tissues, since the oxygen cascade can be influenced by numerous factors occurring in pathologic states such as COPD (27-29). For instance, impairment of microcirculation, altered levels of metabolism in tissues, and imbalance between the volumes of intra- and extravascular fluid can all enhance tissue hypoxia (27-29). Additionally, hypoxemia may activate other systems, which may or may not interact with the TNF-α system (30, 31). All of these effects may combine to enhance the effect of chronic hypoxemia on activation of the TNF-α system in patients with COPD.
We found inverse correlations between both sTNF-R levels and % fat in our COPD patients, although the relationship between TNF-α levels and % fat in these patients did not reach statistical significance. TNF-α is known to be involved in adipose tissue waste and loss of skeletal muscle protein, leading to cachexia (4). That serum TNF-α levels correlated well with plasma sTNF-R levels in our COPD patients may suggest that the endogenous activation of the TNF-α system caused by chronic hypoxemia plays at least a partial role in forcing the metabolic balance of adipose tissue and skeletal muscle toward the catabolic side in patients with COPD (4).
The findings in our subgroups of COPD patients are consistent with our hypothesis that activation of the TNF-α system by systemic hypoxemia may relate to the general BW loss and/or cachexia in patients with COPD, although the numbers of COPD patients in each subgroup were relatively small. However, more significant differences in circulating TNF-α and sTNF-R levels were observed in malnourished COPD patients (Table 5) than in normally nourished COPD patients (Table 6). These results may suggest that the effect of chronic hypoxemia on activation of the TNF-α system is more prominent in malnourished COPD patients.
The genes for the inflammatory cytokines TNF-α and IL-1 can be included in a growing number of genes regulated by low O2 tension (32). A potential explanation for this regulation is that hypoxia alters the transcriptional levels of these genes by unknown mechanisms (11, 12, 15, 32). Human AM (15), peripheral blood mononuclear cells (16), and monocytic cells lines (17) incubated under hypoxic conditions exhibit increased TNF-α and IL-1β synthesis. Hypoxic endothelial cells are also known to produce IL-1α (33), IL-6 (34), and IL-8 (35). This hypoxia-enhanced production of proinflammatory cytokines acts collectively to produce leukocyte-mediated tissue damage in sites of ischemic injury (33-35). The production site of TNF-α in COPD patients was not investigated in our study. Mononuclear cells may be the source of the TNF-α, since hypoxemia induces marked alterations in the immune system, and because mononuclear cells are especially sensitive to hypoxemia in healthy subjects (36).
In conclusion, we found inverse correlations between PaO2 levels and the TNF-α system in terms of circulating TNF-α and sTNF-R levels in patients with COPD. In addition, we found inverse correlations between sTNF-R levels and % fat in the COPD patients in our study. The results of this study may suggest that systemic hypoxemia activates the TNF-α system, and that the activated TNF-α system contributes to the cachexic status noted in patients with COPD, although causal relationships between these events were not demonstrated in the present study. We believe that these relationships are associated with a variety of pathophysiologic features of COPD through the multifunctional effects of the activated TNF-α system.
In Figure 2A, three data points above 1,500 pg/ml for sTNF-R55 appear to be outliers, without which very little correlation would seem to remain. This is also true for two data points in Figure 2B, for sTNF-R75 above 5,000 pg/ml. To investigate these observations we evaluated the correlation without the three data points above 1,500 pg/ml in Figure 2A. The correlation remained unchanged (r = −0.500, p = 0.0191). A similar result was obtained without the two data points above 5,000 pg/ml in Figure 2B (r = −0.492, p = 0.0182).
The values for TNF-α, sTNF-R55, and sTNF-R75 in the present study are consistent with those in several other studies of these substances (7, 9, 37, 38).
The physiologic significance of the slightly increased TNF-α levels in our COPD patients is difficult to explain, but there are several possible explanations. First, TNF-α exerts its effect at least partly in a paracrine/autocrine fashion in each tissue, and there is no significant correlation between TNF-α levels at local sites and in the systemic circulation (39). It may be that circulating TNF-α is spilled into the systemic circulation from the TNF-α–producing tissues or cells of COPD patients (40). Second, weight loss in COPD patients tends to develop very slowly. Although TNF-α levels are slightly increased, persistent elevation of circulating TNF-α levels may have unknown chronic effects on metabolic abnormalities in COPD patients.
The authors wish to thank Drs. T. Sayama and K. Shida, of the Department of Pulmonary Medicine, Tohoku Central Hospital, Yamagata, Japan, for their helpful discussion; Mr. Arjuna J. Celaya for his English help; and the Cosmic Corporation (Tokyo, Japan) for technical assistance.
Supported in part by grant 09670597 from the Ministry of Education, Science, Sports and Culture, Japan.
| 1. | Schols A. M. W. J., Soeters P. B., Mostert R., Saris W. H. M., Wouters E. F. M.Energy balance in chronic obstructive pulmonary disease. Am. Rev. Respir. Dis.143199112481252 |
| 2. | Schols A. M. W. J., Soeters P. B., Dingemans A. M. C., Mostert R., Frantzen P. J., Wouters E. F. M.Prevalence and characteristics of nutritional depletion in patients with stable COPD eligible for pulmonary rehabilitation. Am. Rev. Respir. Dis.147199311511156 |
| 3. | Tracey K. J., Wei H., Manogue K. R., Fong Y., Hesse D. G., Nguyen H. T., Kuo G. C., Beutler B., Cotran R. S., Cerami A.Cachectin/tumor necrosis factor induced cachexia, anemia, and inflammation. J. Exp. Med.167198812111227 |
| 4. | Argiles J. M., Lopez-Soriano J., Busquets S., Lopez-Soriano F. J.Journey from cachexia to obesity by TNF. FASEB J.111997743751 |
| 5. | Balkwill F., Osborne R., Burke F., Naylor S., Talbot D., Durbin H., Tavernier J., Fiers W.Evidence for tumor necrosis factor/ cachectin production in cancer. Lancet2198712291232 |
| 6. | Norman D., Elborn J. S., Cordon S. M., Rayner R. J., Wiseman M. S., Hiller E. J., Shale D. J.Plasma tumour necrosis factor alpha in cystic fibrosis. Thorax4619919195 |
| 7. | Nozaki N., Yamaguchi S., Shirakabe M., Nakamura H., Tomoike H.Soluble tumor necrosis factor receptors are elevated in relation to severity of congestive heart failure. Jpn. Circ. J.611997657664 |
| 8. | Maury C. P. J., Teppo A. M.Tumor necrosis factor in the serum of patients with systemic lupus erythematosus. Arthritis Rheum.321989146150 |
| 9. | Di Francia M., Barbier D., Mege J. L., Orehek J.Tumor necrosis factor alpha levels and weight loss in chronic obstructive pulmonary disease. Am. J. Respir. Crit. Care Med.150199414531455 |
| 10. | de Godoy I., Donahoe M., Calhoun W. J., Mancino J., Rogers R. M.Elevated TNF-α production by peripheral blood monocytes of weight-losing COPD patients. Am. J. Respir. Crit. Care Med.1531996633637 |
| 11. | Benyo D. F., Miles T. M., Conrad K. P.Hypoxia stimulates cytokine production by villous explants from the human placenta. J. Clin. Endocrinol. Metab.82199715821588 |
| 12. | Szaflarski J., Burtrum D., Silverstein F. S.Cerebral hypoxia-ischemia stimulates cytokine gene expression in perinatal rats. Stroke26199510931100 |
| 13. | Larsen G. L., Webster R. O., Worthen G. S., Gumbay R. S., Henson P. M.Additive effect of intravascular complement activation and brief episodes of hypoxia in producing increased permeability in the rabbit lung. J. Clin. Invest.751985902910 |
| 14. | Nuytink J. K. S., Goris R. J. A., Weerts J. G. E., Schillings P. H. M., Stekhoven J. H. S.Acute generalized microvascular injury by activated complement and hypoxia: the basis of the adult respiratory distress syndrome and multiple organ failure. J. Exp. Pathol.671986537548 |
| 15. | Hempel S. L., Monick M. M., Hunninghake G. W.Effect of hypoxia on release of IL-1 and TNF by human alveolar macrophages. Am. J. Respir. Cell Mol. Biol.141996170176 |
| 16. | Ghezzi P., Dinarello C. A., Bianchi M., Rosandich M. E., Repine J. E., White C. W.Hypoxia increases production of interleukin-1 and tumor necrosis factor by human mononuclear cells. Cytokine31991189194 |
| 17. | Scannell G., Waxman K., Kaml G. J., Ioli G., Gatanaga T., Yamamoto R., Granger G. A.Hypoxia induces a human macrophage cell line to release tumor necrosis factor-alpha and its soluble receptors in vitro. J. Surg. Res.541993281285 |
| 18. | Bemelmans M. H. A., van Tits L. J. H., Buurman W. A.Tumor necrosis factor: function, release and clearance. Crit. Rev. Immunol.161996111 |
| 19. | Diez-Ruiz A., Tilz G. P., Zangerle R., Baier-Bitterlich G., Wachter H., Fuchs D.Soluble receptors for tumor necrosis factor in clinical laboratory diagnosis. Eur. J. Haematol.54199518 |
| 20. | American Thoracic SocietyStandards for the diagnosis and care of patients with chronic obstructive pulmonary disease (COPD) and asthma. Am. Rev. Respir. Dis.1361987225243 |
| 21. | Takabatake N., Nakamura H., Abe S., Hino T., Saito H., Yuki H., Kato S., Tomoike H.Circulating leptin in patients with chronic obstructive pulmonary disease. Am. J. Respir. Crit. Care Med.159199912151219 |
| 22. | Quanjer, P. H., editor. Standardized lung function testing. 1983. Bull. Eur. Physiopathol. Respir. 19:7–44. |
| 23. | Hino T., Nakamura H., Shibata Y., Abe S., Kato S., Tomoike H.Elevated levels of type II soluble tumor necrosis factor receptors in the bronchoalveolar lavage fluids of patients with sarcoidosis. Lung1751997187193 |
| 24. | Nakamura H., Hino T., Kato S., Shibata Y., Takahashi H., Tomoike H.Tumor necrosis factor receptor gene expression and shedding in human whole lung tissue and pulmonary epithelium. Eur. Respir. J.9199616431647 |
| 25. | Curtin F., Morabia A., Pichard C., Slosman D. O.Body mass index compared to dual-energy X-ray absorptiometry: evidence for a spectrum bias. J. Clin. Epidemiol.501997837843 |
| 26. | Smalley K. J., Knerr A. N., Kendrick Z. V., Colliver J. A., Owen O. E.Reassessment of body mass indices. Am. J. Clin. Nutr.521990405408 |
| 27. | Murray, J. F., and J. A. Nadel. 1994. Textbook of Respiratory Medicine, 2nd ed. W. B. Saunders Company, Philadelphia. Section B, Ch. 3. |
| 28. | West, J. B. 1995. Respiratory Physiology—The Essentials, 5th ed. Williams & Wilkins, Baltimore. Ch. 5, 6. |
| 29. | Shapiro, B. A., W. T. Peruzzi, and R. Kozelowski-Templin. 1994. Clinical Application of Blood Gases, 5th ed. Mosby-Year Book, St. Louis. Ch. 2, 4, 8. |
| 30. | Tanaka M., Ito H., Adachi S., Akimoto H., Nishikawa T., Kasajima T., Marumo F., Hiroe M.Hypoxia induces apoptosis with enhanced expression of Fas antigen messenger RNA in cultured neonatal rat cardiomyocytes. Circ. Res.751994426433 |
| 31. | Stefanelli C., Stanic I., Bonavita F., Muscari C., Pignatti C., Rossoni C., Caldarera C. M.Oxygen tension influences DNA fragmentation and cell death in glucocorticoid-treated thymocytes. Biophys. Res. Commun.2121995300306 |
| 32. | Fandrey J.Hypoxia-inducible gene expression. Respir. Physiol.1011995110 |
| 33. | Shreeniwas R., Koga S., Karakurum M., Pinski D., Kaiser E., Brett J., Wolitzky B. A., Norton C., Plocinski J., Benjamin W., Burns D. K., Goldstein A., Stern D.Hypoxia-mediated induction of endothelial cell interleukin-1α. J. Clin. Invest.90199223332339 |
| 34. | Yan S. F., Tritto I., Pinski D., Liao H., Huang J., Fuller G., Brett J., May L., Stern D.Induction of interleukin 6 (IL-6) by hypoxia in vascular cells: central role of the binding site for nuclear factor-IL-6. J. Biol. Chem.27019951146311471 |
| 35. | Karakurum M., Shreeniwas R., Chen J., Pinsky D., Yan S.-D., Anderson M., Sunouchi K., Major J., Hamilton T., Kuwabara K., Rot A., Nowygrod R., Stern D.Hypoxic induction of interleukin-8 gene expression in human endothelial cells. J. Clin. Invest.93199415641570 |
| 36. | Pedersen B. K., Kappel M., Klokker M., Nielsen H. B., Secher N. H.The immune system during exposure to extreme physiologic conditions. Int. J. Sports Med.151994S116S121 |
| 37. | Okuyama M., Yamaguchi S., Nozaki N., Yamaoka M., Shirakabe M., Tomoike H.Serum levels of soluble form of Fas molecule in patients with congestive heart failure. Am. J. Cardiol.79199716981701 |
| 38. | Schols A. M. W. J., Buurman W. A., Staal-van den Brekel A. J., Dentener M. A., Wouters E. F. M.Evidence for a relation between metabolic derangements and increased levels of inflammatory mediators in a subgroup of patients with chronic obstructive pulmonary disease. Thorax511996819824 |
| 39. | Torre-Amione G., Kapadia S., Lee J., Durand J. B., Bies R. D., Young J. B., Mann D. L.Tumor necrosis factor-α and tumor necrosis factor receptors in the failing human heart. Circulation931995704711 |
| 40. | Rich E. A., Panuska J. R., Wallis R. S., Wolf C. B., Leonard M. L., Ellner J. J.Dyscoordinate expression of tumor necrosis factor-alpha by human blood monocytes and alveolar macrophages. Am. Rev. Respir. Dis.139198910101016 |
