Proinflammatory Activation Is Linked to Apoptotic Mediator, Soluble Fas Level in Patients With Chronic Heart Failure Toru Kinugawa,1 MD, Masahiko Kato,2 MD, Kazuhiro Yamamoto,2 MD, Ichiro Hisatome,3 MD, and Ryuji Nohara,4 MD

Summary The Fas/Fas system is a major signaling pathway that is up-regulated in patients with chronic heart failure (CHF). Serum soluble Fas (sFas) levels increase in proportion to the CHF severity and may have prognostic value, therefore, sFas is a promising biomarker of heart failure. In this study, we attempted to identify the determinants of sFas levels in patients with CHF. Serum levels of (TNF)-α and its soluble receptors (sTNF-R1 & sTNF-R2), (IL)-6, soluble IL-6 (sIL-6R), glycoprotein (gp)130, and sFas were measured in 106 pa- tients with CHF and 39 controls. All subjects performed a symptom-limited cycle ergometer exercise test with expired gas analysis. CHF patients had higher levels of TNF-α, sTNF-R1, sTNF-R2, IL-6, and gp130. Serum levels of sFas (con- trols versus CHF; 2.60 ± 0.88 versus 3.38 ± 1.23 ng/mL, P = 0.0004) were higher in CHF. On univariate analysis, age (P

= 0.0003), NYHA functional class (P = 0.0012), peak VO2 (P < 0.0001), plasma norepinephrine (P = 0.0013), log IL-6 (P < 0.0001), log TNF-α (P = 0.0002), log sTNF-R1 (P < 0.0001), and log TNF-R2 (P < 0.0001) were significantly re- lated to log sFas levels. Multivariate analysis showed that age and log IL-6 and log sTNF-R1 levels were independently associated with log sFas levels (overall R = 0.603, P < 0.0001). Serum levels of sFas were increased in patients with CHF, and age and serum IL-6 and sTNF-R1 levels were independent determinants of sFas levels. These data suggest that proinflammatory cytokine activation is linked to the Fas/ system in patients with CHF. (Int Heart J 2012; 53: 182-186)

Key words: Heart failure, Soluble Fas, Proinflammatory cytokine, Cardiopulmonary exercise test

hronic heart failure (CHF) is a complex clinical syn- Methods drome characterized by exercise intolerance, hemo- C dynamic alterations, and neurohormonal and cytokine Subjects: We studied 106 patients with CHF (84 men and 22 activation. Studies have shown that proinflammatory women, 59.2 ± 11.9 years). Every patient had a left ventricular such as interleukin (IL)-6 and tumor necrosis factor (TNF)-α, ejection fraction < 45%. We also studied 39 age-matched as well as soluble TNF receptors (sTNF-R1 and sTNF-R2) are healthy individuals as controls (31 men and 8 women). The re- increased in circulating blood in patients with CHF.1-5) Apop- sults for the medical history, physical examination, electrocar- totic mediators have also been investigated and Fas/Fas Ligand diograms, chest X-rays, and echocardiograms were negative system activity was reported to be up-regulated in patients with for cardiovascular disease in these 39 control subjects. Control CHF.6) subjects did not take medications. We excluded patients with Fas is a type I cell-surface belonging to the TNF inducible ischemia, significant pulmonary disease, intermittent family.7) Fas can regulate the apoptotic process by acting as a claudication, or other disorders limiting exercise performance receptor for Fas Ligand.8,9) Fas is considered to be an inhibitor other than cardiac disease. Patients with chronic inflammatory of apoptosis, while Fas Ligand is an inducer of apoptosis. Al- disease, collagen disease, acute infection, and/or neoplasm though Fas acts as an inhibitory regulator, serum levels of the were also excluded. The protocol was approved by the Ethics soluble form of Fas (sFas) reflect the Fas/Fas Ligand system Committee of Tottori University and all subjects gave their activity.10) Previous studies have confirmed that serum sFas written informed consent to participate in the study. levels are increased in proportion to the severity of CHF,11-13) Measurement of blood samples: Blood specimens were drawn and may have prognostic value.14,15) Therefore, serum sFas can through a plastic cannula placed in the forearm vein after 30 be a promising biomarker of heart failure. The aim of this minutes of supine rest. Serum IL-6 levels were measured using study was to investigate the significant determinants of sFas chemiluminescent enzyme immunoassay (CLEIA, Fujirebio, levels among cardiopulmonary exercise variables, neurohor- Tokyo). Concentrations of serum TNF-α were determined by monal parameters, and proinflammatory cytokines and their the high-sensitivity human TNF-α test (Quantikine HS, R&D receptors in patients with CHF. Systems, Minneapolis, MN, USA). Serum soluble TNF recep-

From 1 Kinugawa Cardiology Clinic, Osaka, Departments of 2 Cardiovascular Medicine, Faculty of Medicine and 3 Genetic Medicine and Regenerative Therapeutics, Graduate School of Medical Science, Tottori University, Tottori, and 4 Heart Center, Kitano Hospital, The Tazuke Kofukai Medical Research Institute, Osaka, Japan. Address for correspondence: Toru Kinugawa, MD, Kinugawa Cardiology Clinic, 2-13-28 Tsukamoto, Yodogawa-ku, Osaka, Osaka 532-0026, Japan. Received for publication January 17, 2012. Revised and accepted March 16, 2012. 182 Vol 53 No 3 SFAS IN HEART FAILURE 183 tor type 1 (sTNF-R1), type 2 (sTNF-R2), Table I. Clinical, Echocardiographic, Neurohormonal, and Cardiopulmo- (gp130), soluble IL-6 receptor (sIL-6R), and sFas levels were nary Exercise Parameters in Control Subjects and in Patients With Chronic assessed according to the manufacturer’s specifications using Heart Failure an ELISA (Quantikine). Plasma catecholamines were de- Controls CHF termined by high performance liquid chromatography using (n = 39) (n = 106) the diphenylethylene diamine method. ANP and BNP levels Age (years) 55.1 ± 10.4 59.2 ± 11.9 were determined using a Shionoria RIA Kit and an S-1215 Male/Female 31/8 84/22 RIA kit, respectively. The intra-assay variability values in our Body height (cm) 166 ± 9 163 ± 8 laboratory for TNF-α, IL-6, sTNF-R1, sTNF-R2, gp130, sIL- Body weight (kg) 62.3 ± 10.6 61.7 ± 12.3 6R, sFas, norepinephrine, epinephrine, ANP, and BNP are Body mass index (kg/m2) 22.6 ± 2.8 23.0 ± 3.5 5.3%, 2.2%, 2.9%, 2.5%, 4.3%, 2.6%, 1.7%, 1.8%, 6.0%, New York Heart Association (I/II/III) - 43/46/17 6.1%, and 3.3%, respectively. Etiology Dilated cardiomyopathy - 55 (52%) Cardiopulmonary exercise testing: Cardiopulmonary exercise Old myocardial infarction - 21 (20%) testing was performed using an upright bicycle ergometer with Valvular heart disease - 22 (20%) 16) ramp protocol as described previously. Briefly, after 3 min- Hypertensive heart disease - 4 (4%) utes of unloaded cycling, the exercise load was increased in 10 Others - 4 (4%) or 20 watt/minute increments to symptom-limited maximal Echocardiographic data * work. Patients stopped exercise when they had severe leg fa- Left ventricular end-diastolic dimen- 49 ± 5 63 ± 11 tigue and/or dyspnea. Oxygen uptake (VO ), carbon dioxide sion (mm) 2 Left ventricular end-systolic dimension 32 ± 4 52 ± 11* output (VCO2), and minute ventilation (VE) were measured at (mm) rest and throughout the exercise period using a 280E Aero- Left ventricular ejection fraction (%) 68.3 ± 6.6 34.8 ± 4.8* monitor (Minato Medical Science, Osaka, Japan). Anaerobic Cardiothoracic ratio (%) 46 ± 4 57 ± 7* threshold was determined by the V-slope method. Peak VO2 Neurohormones Plasma norepinephrine (pg/mL) 206 ± 67 311 ± 224* was defined as the maximal VO2 attained during exercise. The slope of the VE-VCO relationship was calculated by linear re- Plasma epinephrine (pg/mL) 35 ± 25 34 ± 27 2 71.3 ± 64.2* gression analysis using the values of VE and VCO . Plasma atrial natriuretic peptide 21.2 ± 18.2 2 (pg/mL) Statistical analysis: Comparisons of the continuous variables Plasma brain natriuretic peptide 14.9 ± 14.1 187.9 ± 217.0* between the two groups were performed using the Mann- (pg/mL) Whitney U-test. Linear regression analysis was used to com- Cardiopulmonary exercise variables pare the relationship between the two variables. Univariate and Peak work rate (watts) 183 ± 85 106 ± 48* Anaerobic threshold (mL/minute/kg) 16.5 ± 3.8 11.1 ± 2.5* stepwise multivariate linear regression analyses were used to * Peak VO2 (mL/minute/kg) 29.2 ± 8.7 18.3 ± 4.9 detect independent factors associated with serum sFas levels * VE/VCO2 slope 26.7 ± 3.1 33.6 ± 7.4 among 14 variables. All analyses were performed with a Peak RER 1.20 ± 0.10 1.17 ± 0.11 StatView statistical program (Version 5.0, SAS Institute Inc. Cary, NC, USA). The differences were considered significant Values are mean ± SD. *P < 0.05 versus Controls. CHF indicates chronic when P values were less than 0.05. Data are expressed as the heart failure; VO2, oxygen uptake; VE, minute ventilation; VCO2, carbon mean ± SD. dioxide output; and RER, respiratory exchange ratio.

Results Table II. Cytokines and Their Receptors in Control Subjects and in Pa- tients With Chronic Heart Failure Baseline characteristics in controls and in patients with CHF: Controls CHF The baseline characteristics of the study subjects are summa- (n = 39) (n = 106) rized in Table I. Two groups were matched for age, male to fe- male ratio, and body mass index (BMI). There were 43 pa- TNF-α (pg/mL) 2.66 ± 0.94 4.26 ± 4.36* * tients with New York Heart Association (NYHA) functional Soluble TNF-R1 (pg/mL) 626 ± 161 916 ± 438 Soluble TNF-R2 (pg/mL) 1231 ± 342 1841 ± 816* class I, 46 patients with class II, and 17 patients with class III. IL-6 (pg/mL) 1.17 ± 0.54 2.38 ± 2.38* The etiology of CHF was idiopathic dilated cardiomyopathy in Soluble IL-6R (pg/mL) 31.0 ± 13.5 33.0 ± 15.2 55 patients, old myocardial infarction in 21 patients, valvular gp130 (pg/mL) 252 ± 56 355 ± 174* heart disease in 22 patients, hypertensive heart disease in 4 pa- * tients, and others in 4 patients. At the time of the study, patients Values are mean ± SD. P < 0.05 versus Controls. CHF indicates chronic heart failure; TNF, tumor necrosis factor; R1, receptor type 1; R2, receptor were treated with loop-diuretics (70%), digitalis (65%), angi- type 2; IL, interleukin; and gp, glycoprotein. otensin-converting enzyme inhibitors/angiotensin II type 1 re- ceptor antagonists (83%), β-blockers (41%), calcium channel blockers (13%), antiplatelet agents (25%), warfarin (27%), and antiarrhythmic agents (10%). Echocardiographic examination significantly higher in patients with CHF. Peak work rate, revealed that patients with CHF had significantly larger left anaerobic threshold, and peak VO2 were lower and VE/VCO2 ventricular end-diastolic and end-systolic dimensions with a slope was higher in patients with CHF. The peak respiratory significantly reduced left ventricular ejection fraction of 34.8 ± exchange ratio was not significantly different between the two 4.8%. groups, indicating that subjects in both groups performed max- Plasma levels of norepinephrine, ANP and BNP were imal exercise. Int Heart J 184 KINUGAWA, ET AL May 2012

Table III. Univariate and Multivariate Linear Model of sFas Levels in Pa- tients With Chronic Heart Failure

Multivariate Univariate Variable P Beta-Coeff P Corr Coeff (SE)

Age (years) 0.345 0.0003 0.002 (0.001) 0.0315 Body mass index (kg/m2) -0.168 0.0843 NS NYHA functional class 0.31 0.0012 NS LV ejection fraction (%) 0.022 0.8194 NS

Peak VO2 (mL/minute/kg) -0.287 < 0.0001 NS Plasma norepinephrine (pg/mL) 0.309 0.0013 NS Plasma ANP (pg/mL) 0.1026 0.7943 NS Plasma BNP (pg/mL) 0.153 0.1166 NS log IL-6 (pg/mL) 0.444 < 0.0001 0.125 (0.042) 0.0037 log sIL-6R (pg/mL) 0.159 0.1103 NS log gp130 (pg/mL) 0.027 0.7823 NS log TNF-α (pg/mL) 0.359 0.0002 NS log sTNF-R1 (pg/mL) 0.467 < 0.0001 0.295 (0.078) 0.0003 log sTNF-R2 (pg/mL) 0.462 < 0.0001 NS

Corr indicates correlation; Coeff, coefficient; NYHA, New York Heart Asso-

ciation; LV, left ventricular; VO2, oxygen uptake; ANP, atrial natriuretic pep- tide; BNP, brain natriuretic peptide; IL-6, interleukin-6; sIL-6R, soluble IL-6 receptor; TNF-α, tumor necrosis factor-α; sTNF-R1, soluble TNF receptor Figure 1. Serum sFas levels in control subjects (hatched bar) and in pa- type 1; and sTNR-R2, soluble TNF receptor type 2. tients with chronic heart failure (closed bar). Values are mean ± SD.

Figure 2. Correlation between log IL-6 levels and log sFas levels (left), and correlation between log TNF-R1 levels and log sFas levels (right) in patients with chronic heart failure. Both log IL-6 and log TNF-R1 levels were significantly correlated with log sFas levels.

Serum levels of proinflammatory cytokines and their re- hormonal factors, proinflammatory cytokines/their receptors, ceptors are summarized in Table II. Patients with CHF had ele- cardiopulmonary exercise variables, as well as important clini- vated TNF-α, sTNF-R1, sTNF-R2, IL-6, and gp130 levels cal factors were investigated. On univariate analysis, age (r = compared to those of normal control subjects. 0.345, P = 0.0003), NYHA functional class (r = 0.310, P = sFas levels in controls and in patients with CHF: Circulating 0.0012), peak VO2 (r = -0.287, P < 0.0001), plasma norepine- levels of sFas are shown in Figure 1 for control subjects and phrine (r = 0.309, P = 0.0013), log IL-6 (r = 0.444, P < for patients with CHF. Serum levels of sFas were significantly 0.0001), log TNF-α (r = 0.359, P = 0.0002), log sTNF-R1 elevated in patients with CHF compared to those of normal (r = 0.467, P < 0.0001), and log sTNF-R2 (r = 0.462, P < control subjects (Controls versus CHF; 2.60 ± 0.88 versus 3.38 0.0001) were significantly related to log sFas levels. Multivari- ± 1.23 ng/mL, P = 0.0004). ate analysis showed that age and log IL-6 and log sTNF-R1 Relation between sFas levels and other variables in patients levels were independently associated with log sFas levels with CHF: The relationships between sFas levels and neuro- (overall R = 0.603, P < 0.0001) (Table III). Vol 53 No 3 SFAS IN HEART FAILURE 185

The correlations between log IL-6 and log sFas, and be- sistent with their findings that age is a significant factor for tween log sTNF-R1 and log sFas are shown in Figure 2. Log sFas levels. sFas levels were positively correlated with log IL-6 levels (r = In accordance with another study,5) our patients with CHF 0.44, P < 0.0001), and log sFas levels were also positively cor- had elevated IL-6 and gp130 levels, but not IL-6R concentra- related with log sTNF-R1 levels (r = 0.47, P < 0.0001) in pa- tions. Gp130 is a common signal-transducing receptor subunit tients with CHF. in the IL-6 cytokine family,27) and cardiac expression of IL-6 mRNA expression is increased in the myocardium in patients with advanced heart failure.28) It is possible that the enhanced Discussion IL-6 activity in the failing myocardium may contribute to the increased gp130 and IL-6 levels in our patients with CHF. The main findings in this study were that; in patients with We confirmed previously reported findings4,5) of elevated CHF, (1) serum levels of proinflammatory cytokines and their sTNF-R1 and sTNF-R2 levels in CHF. Decreased excretion29) receptors, such as TNF-α IL-6, sTNF-R1, sTNF-R2, and and increased TNF system activity4) may account for the ele- gp130 were elevated; (2) serum levels of sFas were significant- vated blood levels of these receptors. The biologic significance ly increased; and (3) age and serum IL-6 and sTNF-R1 levels of these soluble receptors is not fully understood, but it seems were independent determinants of sFas levels. that soluble TNF receptors may inhibit TNF activity in the Fas is a transmembrane protein of 45 kDa which belongs presence of a low TNF/soluble TNF receptor ratio.4,30) Our pa- to the TNF superfamily, which can initiate apoptosis.7) The tients exhibited higher TNF/sTNF-R1 (or TNF/sTNF-R2) clinical importance of apoptosis lies in the fact that apoptosis compared with those of control subjects (data not shown), sug- is evident in the myocardium of patients with end-stage heart gesting that the TNF system was activated in our patients with failure,17) and the resultant loss of cardiac myocytes contributes CHF. to heart failure progression.18) Previous studies have shown that Limitations of this study include; 1) the cross-sectional heart failure treatment, either by drug19) or by exercise train- study design (with few longitudinal data for cytokines and ap- ing,20) decreased the serum sFas concentration in patients with optotic mediators), 2) exclusion of NYHA class IV CHF pa- CHF. In our CHF patients, one 60-year-old male patient exhib- tients (whose immune activation may be the highest among ited amelioration of sFas levels (from 3.46 ng/mL to 2.63 ng/ CHF patients), and 3) lack of soluble form of Fas Ligand mL) with increased peak VO2 (from 11.0 mL/minute/kg to measurements (a specific agonistic , and an inducer of 13.3 mL/minute/kg) during 3 months of medical therapy. apoptosis). Determinations of both sFas and sFas Ligand con- In addition, other investigators reported that the Fas/Fas centrations would give us more detailed information on the Ligand system activity (as reflected by sFas levels) is related to Fas/Fas Ligand system in patients with CHF. the severity of heart failure11-13) and may have prognostic infor- In conclusion, the present study showed that serum levels mation.14,15) Thus, the sFas level can be a useful biomarker for of sFas were increased in patients with CHF, and that age and heart failure. In this study, we found that higher sFas levels serum IL-6 and sTNF-R1 levels were independent determi- were related to NYHA functional classification, exercise intol- nants of sFas levels. These data suggest that proinflammatory erance (with lower peak VO2), sympathetic activation (with cytokine activation is linked to the Fas/Fas Ligand system in higher norepinephrine concentration), and cytokine activation, patients with CHF. suggesting a close relation between sFas levels and CHF sever- ity. It is of interest to speculate that sFas-guided heart failure therapy results in a better prognosis, just like BNP-guided References heart failure therapy.21) The mechanisms of increased sFas levels in heart failure 1. Levine B, Kalman J, Mayer L, Fillit HM, Packer M. Elevated cir- have not been elucidated, but a low BMI with cachexia,15) hy- culating levels of tumor necrosis factor in severe chronic heart poxic stimuli,22) and immune activation12) may have contributed failure. N Engl J Med 1990; 323: 236-41. 2. Torre-Amione G, Kapadia S, Benedict C, Oral H, Young JB, to the sFas elevation in CHF. In the current study, we investi- Mann DL. Proinflammatory cytokine levels in patients with de- gated the independent determinants of sFas levels, and found pressed left ventricular ejection fraction: a report from the Studies that age and serum IL-6 and sTNF-R1 levels were significant of Left Ventricular Dysfunction (SOLVD). J Am Coll Cardiol determinants of sFas levels. The relationships between sFas 1996; 27: 1201-6. levels and these selected variables (IL-6, sTNR-R1 and age) 3. Mann DL. Recent insights into the role of tumor necrosis factor in deserve some comment. IL-6 enhances the cytokine cascade the failing heart. Heart Fail Rev 2001; 6: 71-80. (Review) and IL-6 is a potent inducer of apoptosis.23) In an experimental 4. Ferrari R, Bachetti T, Confortini R, et al. Tumor necrosis factor 24) soluble receptors in patients with various degrees of congestive study using human astrocytes, Choi, et al demonstrated that heart failure. Circulation 1995; 92: 1479-86. Fas expression is up-regulated by the application of IL-6. Thus, 5. Aukrust P, Ueland T, Lien E, et al. Cytokine network in congestive high IL-6 may causally relate to the activated Fas/Fas Ligand heart failure secondary to ischemic or idiopathic dilated cardiomy- system. The TNF system can also lead to the higher sFas lev- opathy. Am J Cardiol 1999; 83: 376-82. els. Kubota, et al25) examined the expression of apoptosis-relat- 6. Kang PM, Izumo S. Apoptosis and heart failure: A critical review ed genes in mice with cardiac-specific overexpression of of the literature. Circ Res 2000; 86: 1107-13. (Review) 7. Itoh N, Yonehara S, Ishii A, et al. The polypeptide encoded by the TNF-α. They found that the transcript for Fas was up-regulated cDNA for human cell surface antigen Fas can mediate apoptosis. in this transgenic mice myocardium. Therefore, an activated Cell 1991; 66: 233-43. TNF system can lead to the Fas/Fas Ligand system activation. 8. Yonehara S, Ishii A, Yonehara M. A cell-killing monoclonal anti- Concerning age, a previous study26) demonstrated that older body (anti-Fas) to a cell surface antigen co-downregulated with age was associated with higher sFas levels. Our data are con- the receptor of tumor necrosis factor. J Exp Med 1989; 169: 1747- Int Heart J 186 KINUGAWA, ET AL May 2012

56. 20. Adamopoulos S, Parissis J, Karatzas D, et al. Physical training 9. Suda T, Takahashi T, Goldstein P, Nagata S. Molecular cloning modulates proinflammatory cytokines and the soluble Fas/soluble and expression of the Fas ligand, a novel member of the tumor Fas ligand system in patients with chronic heart failure. J Am Coll necrosis factor family. Cell 1993; 75: 1169-78. Cardiol 2002; 39: 653-63. 10. Iio S, Hayashi N, Mita E, et al. Serum levels of soluble Fas anti- 21. Porapakkham P, Porapakkham P, Zimmet H, Billah B, Krum H. gen in chronic hepatitis C patients. J Hepatol 1998; 29: 517-23. B-type natriutetic peptide-guided heart failure therapy: A meta- 11. Okuyama M, Yamaguchi S, Nozaki N, Yamaoka M, Shirakabe M, analysis. Arch Intern Med 2010; 170: 507-14. Tomoike H. Serum levels of soluble form of Fas molecule in pa- 22. Tanaka M, Ito H, Adachi S, et al. Hypoxia induces apoptosis with tients with congestive heart failure. Am J Cardiol 1997; 79: 1698- enhanced expression of Fas antigen messenger RNA in cultured 701. neonatal rat cardiomyocytes. Circ Res 1994; 75: 426-33. 12. Nishigaki K, Minatoguchi S, Seishima M, et al. Plasma Fas lig- 23. Wollert KC, Drexler H. The role of interleukin-6 in the failing and, an inducer of apoptosis, and plasma soluble Fas, an inhibitor heart. Heart Fail Rev 2001; 6: 95-103. (Review) of apoptosis, in patients with chronic congestive heart failure. J 24. Choi C, Park JY, Lee J, et al. Fas ligand and Fas are expressed Am Coll Cardiol 1997; 29: 1214-20. constitutively in human astrocytes and the expression increases 13. Kawakami H, Shigematsu Y, Ohtsuka T, et al. Increased circulat- with IL-1, IL-6, TNF-alpha, or IFN-gamma. J Immunol 1999; ing soluble form of Fas in patients with dilated cardiomyopathy. 162: 1889-95. Jpn Circ J 1998; 62: 873-6. 25. Kubota T, Miyagishima M, Frye CS, et al. Overexpression of tu- 14. Tsutamoto T, Wada A, Maeda K, et al. Relationship between plas- mor necrosis factor-alpha activates both anti- and pro-apoptotic ma levels of cardiac natriuretic peptides and soluble Fas: plasma pathways in the myocardium. J Mol Cell Cardiol 2001; 33: 1331- soluble Fas as a prognostic predictor in patients with congestive 44. heart failure. J Card Fail 2001; 7: 322-8. 26. Seishima M, Takemura M, Saito K, et al. Highly sensitive ELISA 15. Niessner A, Hohensinner PJ, Rychli K, et al. Prognostic value of for soluble Fas in serum: increased soluble Fas in the elderly. Clin apoptosis markers in advanced heart failure. Eur Heart J 2009; 30: Chem 1996; 42: 1911-4. 789-96. 27. Kishimoto T, Akira S, Narazaki M, Toga T. Interleukin-6 family of 16. Kinugawa T, Ogino K, Noguchi N, et al. Sympathetic nervous re- cytokines and gp130. Blood 1995; 86: 1243-54. (Review) sponse relative to the adenosine triphosphate supply-demand im- 28. Plenz G, Song ZF, Tjan TD, et al. Activation of the cardiac inter- balance during exercise is augmented in patients with heart failure. leukin-6 system in advanced heart failure. Eur J Heart Fail 2001; 3: Chest 1998; 114: 1295-300. 415-21. 17. Narula J, Haider N, Virmani R, et al. Apoptosis in myocytes in 29. Anker SD, Clark AL, Kemp M, et al. Tumor necrosis factor and end-stage heart failure. N Engl J Med 1996; 335: 1182-9. steroid metabolism in chronic heart failure: Possible relation to 18. Bennett MR. Apoptosis in the cardiovascular system. Heart 2002; muscle wasting. J Am Coll Cardiol 1997; 30: 997-1001. 87: 480-7. (Review) 30. Aukrust P, Lien E, Kristoffersen AK, et al. Persistent activation of 19. Lanfear DE, Hasan R, Gupta RC, et al. Short term effects of milri- the tumor necrosis factor system in a subgroup of patients with none on biomarkers of necrosis, apoptosis, and inflammation in common variable immunodeficiency--possible immunological and patients with severe heart failure. J Transl Med 2009; 7: 67. clinical consequences. Blood 1996; 87: 674-81.