- Email: [email protected]
HIGH TUMOUR NECROSIS FACTOR-α LEVELS ARE ASSOCIATED WITH EXERCISE INTOLERANCE AND NEUROHORMONAL ACTIVATION IN CHRONIC HEART FAILURE PATIENTS
doi:10.1006/cyto.2001.0918, available online at http://www.idealibrary.com on
HIGH TUMOUR NECROSIS FACTOR- LEVELS ARE ASSOCIATED WITH EXERCISE INTOLERANCE AND NEUROHORMONAL ACTIVATION IN CHRONIC HEART FAILURE PATIENTS Mariantonietta Cicoira,1,2 Aidan P. Bolger,1 Wolfram Doehner,1 Mathias Rauchhaus,1,3 Constantinos Davos,1 Rakesh Sharma,1 Faisal O. Al-Nasser,1 Andrew J. S. Coats,1 Stefan D. Anker1,4 Immune activation plays an important role in the progression of chronic heart failure (CHF). We sought to investigate whether different degrees of tumor necrosis factor- (TNF-) activation are associated with exercise intolerance, neurohormonal activation and alterations in muscle mass and function in patients with CHF without cardiac cachexia. Patients were divided into quartiles according to their TNF levels (first quartile: 0.98–4.90 pg/ml, second quartile: 5.00–6.60 pg/ml; third quartile 6.80–9.00 pg/ml; fourth quartile 9.80–32.00 pg/ml). Patients underwent cardiopulmonary exercise testing, quadriceps muscle strength test, quadriceps fatigue test, and assessment of thigh muscle and fat cross-sectional area (CSA) by computerized tomography scanning. Patients in the highest TNF quartile had the lowest peak oxygen consumption [13.1 (4.1) ml/kg/min vs 18.1 (5.3), 18.8 (4.8) and 18.7 (5.6) ml/kg/min, P<0.01] the greatest relation of ventilation and dioxide production (VE/VCO2) slope (P<0.05) and the most elevated catecholamine levels (P<0.05) compared to patients in the first three quartiles. Patients with the lowest TNF levels had preserved thigh muscle size and quadriceps strength. Strength/ muscle CSA was similar in the four groups. Muscle strength during fatigue testing was significantly lower in the fourth quartile (P=0.01) compared with the other three groups. In CHF patients only the highest levels of TNF are associated with poor functional status and neurohormonal activation. This group of patients may represent the appropriate target population for TNF antagonism. 2001 Academic Press
Chronic heart failure (CHF) is a syndrome with a dismal prognosis,1,2 which is characterised by shortness of breath and exercise intolerance. Abnormalities of skeletal muscle are associated with the genesis of fatigue and breathlessness,3–5 although the mechanisms leading to myopathy are still not fully understood. It has been suggested that activation of the inflammaFrom the 1Department of Cardiac Medicine, National Heart & Lung Institute, London, UK; 2Divisione Clinicizzata di Cardiologia, Universita Y degli Studi di Verona, Verona, Italy; 3Klinik und Poliklinik fu¨r Innere Medizin III, Halle, Germany; 4Franz Volhard Klinik (Charite´, Campus Berlin-Buch) at Max Delbru¨ck Centrum for Molecular Medicine, Berlin, Germany Correspondence to: Dr Mariantonietta Cicoira, Divisione Clinicizzata di Cardiologia, Ospedale Civile Maggiore – P.le Stefani,1 37126 Verona, Italy; E-mail: [email protected] Received 15 February 2001; accepted for publication 27 April 2001 2001 Academic Press 1043–4666/01/140080+07 $35.00/0 KEY WORDS: chronic heart failure/TNF-/exercise intolerance/ neurohormonal activation 80
tory immune system is related to skeletal muscle wasting and cardiac cachexia in CHF.6–8 Elevated levels of tumour necrosis factor (TNF) are associated with left ventricular dysfunction, remodelling9,10 and cell death.11,12 We have previously demonstrated that, particularly in cachectic CHF patients, an increase in TNF and soluble TNF receptors correlates with a low body mass index and abnormalities in steroid hormone metabolism.13 These observations suggest a possible causal relationship between increased TNF levels and progression of the CHF syndrome. The impact of increasing degrees of TNF activation on the clinical status of CHF patients without evidence of weight loss has not yet been investigated. Pharmacological antagonism or neutralization of TNF is an expensive therapeutic approach, and may not necessarily be beneficial to all patients with CHF. Initial studies on patients with NYHA class III and IV symptoms suggest that therapy with a recombinant CYTOKINE, Vol. 15, No. 2 (21 July), 2001: pp 80–86
TNF- and exercise intolerance in heart failure / 81
TABLE 1.
Clinical characteristics of study population
Variable
MeanSD
Median
Range
Age (years) NYHA class Male (%) BMI (kg/m2) % ideal body weight (kg) LVEF (%) Peak VO2 (ml/min/kg) VE/VCO2 slope SBP (mm Hg) DBP (mm Hg) Plasma sodium (mmol/l) Total protein (g/l) Albumin (g/l) Creatinine (mmol/l) Urea (mmol/l) Aldosterone (pmol/l) Renin (ng/l) Norepinephrine (nmol/l) Epinephrine (nmol/l) TNF (pg/ml)
6112 2.60.8 96 25.94.0 11118 2714 17.25.4 37.313.5 11517 7210 1373 715 433 12443 95 712717 12.113.3 3.52.5 1.21.4 8.25.9
59 3 — 25.5 110 25 16.1 33.4 115 70 137 71 44 111 7 440 6.7 2.6 0.5 6.7
30–85 1–4 — 19.3–36.9 81–163 4–48 6.5–31.9 21.0–86.4 86–160 40–98 130–142 59–84 37–50 68–254 3–25 94–3439 0.6–63.8 0.811.9 0.1–5.4 0.932.0
NYHA: New York Heart Association classification of symptoms in heart failure; BMI: body mass index; LVEF: left ventricular ejection fraction; peak VO2: peak oxygen consumption; VE/VCO2: relation of ventilation and dioxide production; SBP: systolic blood pressure; DBP: diastolic blood pressure; TNF: tumour necrosis factor.
TNF soluble receptor is safe14 and improves cardiac function and clinical status.15 Our study aimed to explore the relationship between plasma concentrations of TNF and certain clinical characteristics in a typical population with CHF, so as to provide a logical approach to defining a target population most likely to benefit from anti-TNF therapy. We aimed to assess whether various degrees of immune activation in a population of CHF patients without cachexia are associated with differing degrees of disease severity, in terms of functional status, neurohormonal activation and altered skeletal muscle mass and performance.
class and plasma sodium did not differ in the four groups (ANOVA: all P>0.1) (Table 2). The patients in group 4 were somewhat older [mean age 68(8) years vs 55(10) and 62(14) years in group 1 to 3, ANOVA P=0.0019] and had higher creatinine levels [157(42) mmol/l vs 100(28), 113(41) and 123(38) mmol/l in group 1 to 3, ANOVA P=0.0002].
Exercise capacity Patients in group 4 had a lower mean peak VO2 compared with the other three groups (ANOVA P<0.01). There were no statistically significant differences in mean peak VO2 between patients in the first three quartiles (Fig. 1). Similarly, the mean relation of ventilation and dioxide production (VE/VCO2) slope was significantly higher only in patients in group 4 compared with patients in groups 1–3 (ANOVA, P<0.05) (Fig. 1). Age correlated with peak VO2 only in patients in the first three groups (r=0.45, P=0.0008) and not in patients in group 4 (P=0.15). There was a correlation of borderline significance between peak VO2 and TNF levels only in patients in the highest TNF quartile (r=0.47, P=0.056). When adjusted for age this correlation became significant only in group 4 (P=0.044), while in groups 1–3 it remained non-significant (P>0.2).
Neurohormonal activation Patients in the highest TNF quartile had the highest levels of norepinephrine (ANOVA, P<0.05) and epinephrine (ANOVA, P=0.01) (Fig. 2). Serum aldosterone and plasma renin activity were similar in the four groups. Body mass index was significantly higher in group 1 compared with groups 2–4 (see Table 2).
Muscle size and function RESULTS Baseline characteristics of the study population are summarized in Table 1. In total 72 patients were recruited (69 male and three female). Six patients (8%) were in NYHA class I, 24 (33%) in class II, 32 (45%) in class III and 10 (14%) in class IV. The aetiology of CHF was ischaemic heart disease in 45 patients (63%), idiopathic dilated cardiomyopathy in 24 patients (33%) and valvular heart disease in three (4%). Seven patients (10%) were not taking an angiotensin-converting enzyme inhibitor due to intolerance. Thirty-five patients were receiving digoxin, 24 amiodarone, 27 warfarin, 33 aspirin, and 15 -blockers. Patients were subdivided into quartiles of 18 patients each according to serum TNF levels (group 1: 0.98–4.90 pg/ml, group 2: 5.00–6.60 pg/ml, group 3: 6.8–9.00 pg/ml, group 4: 9.80–32.00 pg/ml). Mean LVEF, NYHA functional
Muscle CSA was significantly higher in the group with lowest TNF levels (ANOVA, P<0.01) (Fig. 3). Quadriceps muscle strength in both legs was higher in patients with the lowest TNF levels compared with those in the last three quartiles (ANOVA, P<0.01). Muscle strength during fatigue at five minutes was significantly lower in group 4 compared with the other three groups (ANOVA P=0.01) (Fig. 3). See Table 3 for details.
TNF assessment using different test kits In 33 patients TNF levels were assessed twice. Using the Medgenix test kit the mean TNF level was found to be 8.2 pg/ml (age-matched healthy controls: mean TNF level 7.0 pg/ml).16 Using the R&D test kit the mean TNF level was found to be 2.0 pg/ml (agematched healthy controls: 0.8 pg/ml).17 There was a statistically significant linear correlation between TNF
82 / Cicoira et al.
CYTOKINE, Vol. 15, No. 2 (21 July, 2001: 80–86)
TABLE 2. Exercise capacity, body mass index (BMI) and neurohormonal activation of study population divided into quartile groups according to tumour necrosis factor (TNF) levels (ANOVA) Variable
First quartile (TNF 0.98–4.90 pg/ml)
Second quartile (TNF 5.00–6.60 pg/ml)
Third quartile (TNF 6.80–9.00 pg/ml)
Fourth quartile (TNF 9.80–32.00 pg/ml)
P-value
18.15.3‡ 34.814.7‡ 29.25.0*§ 2.51.4‡ 0.70.9‡ 0.830.59 2.60.2
18.84.8‡ 33.910.4‡ 26.20.7‡ 2.92.2‡ 0.61.0† 0.770.41 2.70.4
18.75.6‡ 35.714.2‡ 24.83.4 3.52.3 1.51.8 0.720.38 2.60.4
13.14.1 45.811.9 23.42.6 5.03.1 1.91.6 1.070.51 2.80.4
<0.01 <0.05 <0.001 <0.01 <0.05 ns ns
Peak VO2 (ml/min/kg) VE/VCO2 slope BMI (kg/m2) Norepinephrine (nmol/l) Epinephrine (nmol/l) Renin (ng/ml/h) Aldosterone (pmol/l)
Peak VO2: peak oxygen consumption; VE/VCO2: relation of ventilation to carbon dioxide production. †P<0.01 vs fourth quartile. ‡P<0.05 vs fourth quartile. *P<0.01 vs third quartile. §P<0.0001 vs fourth quartile.
A
35
Peak VO2 (ml/kg/min)
30 25 20
*† ‡
15 10 5
B
*†
9 Norepinephrine (nmol/l)
A
8 7 6 5 4 3 2
1st
2nd
3rd
1
4th
1st
2nd
3rd
4th
B
90
5
VE/VCO2 slope
70
*† ‡‡
60 50 40 30
Epinephrine (nmol/l)
80 4
*††
3 2 1
20 10
1st 2nd 3rd 4th Quartile groups according to TNF levels
Figure 1. The relationship between TNF- levels and (A) peak oxygen consumption (VO2) and (B) the slope of ventilation (VE) and carbon dioxide production (VO2) in 72 chronic heart failure patients. Box plot displaying the 10th, 25th, 50th, 75th and 90th percentiles. *P<0.05 vs first quartile; †P<0.05a vs second quartile; ‡P<0.05 vs third quartile.
assessment with the two methods (r2 =0.78; P<0.0001) (Fig. 4), with the equation relating the two variables being: y=0.21x. The quartile cut-off values of 4.90,
0
1st 2nd 3rd 4th Quartile groups according to TNF levels
Figure 2. The relationship between TNF- levels and (A) noradrenaline and (B) adrenaline in 72 chronic heart failure patients. Box plot displaying the 10th, 25th, 50th, 75th and 90th percentiles. *P<0.05 vs first quartile; †P<0.05 vs second quartile: ‡P<0.01 vs second quartile.
6.60, and 9.80 pg/ml (using the Medgenix kit), correspond to levels of 1.0, 1.4 and 2.1 pg/ml using the R&D kit.
TNF- and exercise intolerance in heart failure / 83
A 2
Right thigh muscle CSA (cm )
180 160 140
*
*
2nd
3rd
***
120 100 80 1st
4th
B
Fatigue 5 at minutes (% change from baseline)
100 **‡‡
90 80 70
the three lower groups. Interestingly, the mean TNF levels of patients in the highest quartile were similar to those previously reported in cachectic CHF patients [15.3(3.1) pg/ml].22 It might be hypothesised, therefore, that this group of patients exhibit a ‘‘precachectic’’ state, in which weight loss is not clinically evident, but signs of impaired muscle function and reduced muscle bulk can already be observed. Overt cachexia may develop if this state of immune activation persists, with consequent weight loss and worsening of the clinical condition. High TNF levels may contribute to the loss of muscle bulk and strength in patients with CHF through several possible mechanisms. First, TNF is able to promote apoptosis23 and the breakdown of skeletal muscle proteins;24 furthermore, it reduces leg blood flow with a consequent reduction of skeletal muscle oxygen supply.25 In addition, TNF has cytotoxic effects on endothelial cells, both directly26 and via the down-regulation of endothelial constitutive nitric oxide synthase expression.27 Such mechanisms may also help to explain the observed impairment in exercise tolerance in CHF patients.
60 1st 2nd 3rd 4th Quartile groups according to TNF levels Figure 3. The relationship between TNF- levels and (A) right thigh muscle cross sectional area (CSA) and (B) fatigue at 5 min. Box plot displaying the 10th, 25th, 50th, 75th and 90th percentiles. *P<0.05 vs first quartile; **P<0.01 vs first quartile; ***P<0.001 vs first quartile; ‡‡P<0.01 vs third quartile.
DISCUSSION TNF is a proinflammatory cytokine with pleiotropic biological effects. It can produce left ventricular dysfunction, cardiomyopathy and pulmonary oedema when overexpressed.18–20 In 1990, Levine et al.6 reported that circulating levels of TNF are particularly increased in cachectic CHF patients, and a direct relationship between TNF levels and loss of skeletal muscle mass has been reported.21 In the present study we found that only increased levels of TNF are associated with severe impairment of functional status in CHF patients. Patients with TNF levels between 9.00 pg/ml and 32.00 pg/ml, i.e. in the highest quartile, had the most severe CHF, in terms of the worst exercise capacity, the highest levels of neurohormones and evidence of skeletal muscle wasting and weakness. The patients in the lower three quartiles of TNF levels had better exercise tolerance and lower catecholamine levels than patients in the highest quartile, although no differences were found between
Implications for anti-cytokine therapy in CHF The realisation that immune activation may play a role in the deterioration of myocardial function in CHF and possibly contributes to many of the peripheral manifestations seen in this condition, has provoked great interest in therapeutic strategies directed against TNF action.28,29 Although promising results have been reported, both in experimental and clinical studies,14,15,30 the number of patients treated remains too small for any firm conclusions to be drawn. The cost of this therapeutic approach is high and the selection of a patient group most likely to benefit from it is therefore paramount. Intuitively patients with evidence of raised TNF levels might be expected to gain the most. Taking the results from this study, those patients with TNF levels in the upper quartile (9.80– 32.00 pg/ml) may be considered a reasonable target population for anti-TNF therapy. They demonstrate more severe disease, a wide range of pathophysiological abnormalities correlate to elevated TNF levels and they very likely have a poor prognosis. In these patients the prospect of regular injection therapy seems both logical and feasible. Conversely, the group of patients with the lowest TNF levels (<4.90 pg/ml) exhibit, on average, clinical features consistent with mild to moderate disease, a tendency towards obesity (BMI 29.2 kg/m2) and have evidence of preserved skeletal muscle function and bulk. The use of antiTNF approaches in this sub-group seems less convincing. The potential benefit of anti-TNF treatment
84 / Cicoira et al.
CYTOKINE, Vol. 15, No. 2 (21 July, 2001: 80–86)
TABLE 3. Skeletal muscle size and function parameters of study population divided into quartile groups according to tumour necrosis factor (TNF) levels (ANOVA) Variable Right muscle CSA (cm2) Left muscle CSA (cm2) Right strength (n) Left strength (n) Strength/CSA right (n/cm2) Strength/CSA left (n/cm2) Fatigue 5 min (%)
First quartile (TNF 0.98–4.90 pg/ml)
Second quartile (TNF 5.00–6.60 pg/ml)
Third quartile (TNF 6.80–9.00 pg/ml)
Fourth quartile (TNF 9.80–32.00 pg/ml)
P-value (ANOVA)
134.227.9*† 131.826.8*† 451.9119.1 435.9115.5*† 6.71.0 6.61.0 86.99.7†
116.118.2 109.918.4 399.6103.3 366.490.5 6.51.0 6.40.9 83.810.1
114.816.6 110.818.4 344.677.8 352.3107.7 6.31.0 6.41.3 88.56.6†
103.720.4 97.119.8 327.1150.0 305.7129.7 5.91.3 6.11.3 76.810.9
<0.01 0.001 <0.01 <0.01 ns ns 0.01
CSA: cross sectional area of the four muscles of the thigh. The fatiguability at 5 min is expressed as percentage of baseline maximal quadriceps strength. *P<0.05 vs third quartile. †P<0.01 vs fourth quartile. §P<0.01 vs third quartile.
12
TNF R&D (pg/ml)
10
n = 72 2 r = 0.78 P < 0.001
8 6 4 2
0
y = 0.21x 5
10 15 20 25 TNF Medgenics (pg/ml)
30
35
Figure 4. Linear regression analysis of TNF- levels with two different ELISA kits (Medgenix and R&D Systems).
in individuals with intermediate TNF levels (5.00– 9.00 pg/ml) is harder to gauge. Although only exhibiting symptoms of mild to moderate CHF, patients in the third quartile show somewhat elevated catecholamine levels and the exclusion of this group from anti-TNF therapy may deny treatment to a number who might benefit. Further stratification of immune abnormalities in this population to tease out higher risk patients seems desirable.
MATERIALS AND METHODS We studied 72 consecutive patients with CHF presenting to the outpatient department of the Royal Brompton Hospital, London, UK. The diagnosis of CHF was based on typical symptoms of CHF of at least six months duration, cardiomegaly on chest X-ray and evidence of reduced left ventricular ejection fraction (LVEF), as assessed by radionuclide ventriculography. At the time of investigation, no patient had signs of oedema, hepatomegaly or ascites. Patients with chronic lung disease, neuromuscular disorders, recent myocardial infarction (within six months of the study), malignancy, acquired immunodeficiency syndrome or renal
failure (creatinine >250 mol/l) were excluded from the analysis. Patients who were cachectic, i.e. those with a body mass index (BMI) <19 kg/m2, a history of weight loss of greater than 6% since the onset of CHF or weight <85% of ideal body weight were not included. The protocol of the study was in accordance with the ethical standards of the Helsinki Declaration and was approved by the local ethics committee. Every patient provided written informed consent before being enrolled in the study. Each patient underwent maximal cardiopulmonary exercise testing (treadmill, modified Bruce protocol, Amis 2000, Odense, Denmark) with on-line measurement of oxygen consumption (VO2), ventilation (VE) and carbon dioxide production (VCO2).4 Peak VO2 was taken as an index of exercise capacity. The slope of the relation between ventilation and carbon dioxide production (VE/VCO2 slope) was calculated from the exercise data and taken as an index of the ventilatory response to exercise.31 Quadriceps muscle strength was measured as previously described.4 The subject was seated in a rigid frame, with legs hanging freely. An inelastic strap was attached from the ankle to a pressure transducer. The recording (Multitrace 2, Lectromed, Jersey, UK) from the pressure transducer was used to assess strength and provide visual feedback to the subject. The loss of a superimposed 1-Hz muscle twitch during the plateau of maximum force production indicated that the contraction was maximal. The best of three voluntary contractions on each leg, with a rest period of at least one minute in between, was considered the maximal voluntary quadriceps muscle strength of the right and left leg respectively. The values for muscle strength are expressed in Newton (N). The quadriceps fatigue protocol4 was performed on the stronger leg after strength testing. Patients performed repeated voluntary contractions at 30% of the maximum, using visual feedback as a guide. Contractions of one second were followed by one second relaxations for 40 seconds, followed by 20 seconds of rest. This 60 second cycle was repeated for 20 minutes. Patients were asked to undertake a maximal voluntary contraction at 5, 10, 15 and 20 minutes. The fatigability at each time point is expressed as a percentage of baseline maximal strength. Ultrafast computerized tomography (Imatron, San Francisco, USA) was used to measure the CSA of the total
TNF- and exercise intolerance in heart failure / 85
thigh and the four major muscles of the thigh (quadriceps, hamstrings, gracilis and sartorius) in both legs.4 The scans were performed in the supine position after five minutes of rest. A single 6 mm transaxial slice was scanned at mid-femur level and marked as 12.5% of patient height above the knee joint.32 The CSA was calculated in square centimetres by semi-automatic generation of an outline of the area of interest using the console software of the computer tomography scanner. Blood samples were collected in every subject between 9 a.m. and 10 a.m. after a fasting period of at least 10 h. A catheter was inserted into an antecubital vein and after 20 min of rest in a supine position 20 ml of blood were drawn. Blood samples were centrifuged immediately and aliquots were stored at 70C until analysis. TNF was measured in each patient using an ELISA assay with a lower limit of detectability of 3.0 pg/ml (Medgenix, Fleurus, Belgium) uninfluenced by soluble TNF receptors. In order to provide a comparison of these results with other published work in this field in which different TNF test kits have been used, 33 patients had a second assessment of TNF levels using the commercially available kit from R&D Systems (Minneapolis, USA. ELISA, lower limit of detectability 0.5 pg/ml). Catecholamines, aldosterone and plasma renin activity were analysed using routine hospital analysis procedures.
Statistical analysis All results are expressed as mean value SD. Patients were divided into four groups according to quartile distribution of their serum TNF levels. The hormonal measurements were log-normally distributed, and comparisons were made between sets of log-transformed data. Comparisons between groups were made by ANOVA analysis with Fisher’s post hoc test. Univariate and multivariate regression analyses were used to establish the relationship between biochemical and clinical variables. For the regression analysis comparing the two TNF test kits raw data was used and a forced intercept model was applied. All analyses were performed using a statistical software package (StatView 4.5 Abacus Concepts Inc, Berkeley, USA). A P-value <0.05 was considered statistically significant.
REFERENCES 1. Cohn JN, Johnson G, Ziesche S, and the Vasodilator-Heart Failure Veterans Affairs Cooperative Study Group (1991) A comparison of enalapril with hydralazine-isosorbide dinitrat in the treatment of chronic congestive heart failure. N Engl J Med 325: 303–310. 2. The SOLVD Investigators (1991) Effect of enalapril on survival in patients with reduced left ventricular ejection fraction and congestive heart failure. N Engl J Med 325:296–302. 3. Mancini DM, Walter G, Reichek N et al. (1992) Contribution of skeletal muscle atrophy to exercise intolerance and altered muscle metabolism in heart failure. Circulation 85:1364– 1373. 4. Anker SD, Swan JW, Volterrani M et al. (1997) The influence of muscle mass, strength, fatiguability and blood flow on exercise capacity in cachectic and non-cachectic patients with chronic heart failure. Eur Heart J 18:259–269.
5. Volterrani M, Clark AL, Ludman PF et al. (1994) Determinants of exercise capacity in chronic heart failure. Eur Heart J 15:801–809. 6. Levine B, Kalman J, Mayer L, Fillit HM, Packer M (1990) Elevated circulating levels of tumour necrosis factor in severe chronic heart failure. N Engl J Med 323:236–241. 7. McMurray J, Abdullah I, Dargie HJ, Shapiro D (1991) Increased concentrations of tumour necrosis factor in ‘cachectic’ patients with severe chronic heart failure. Br Heart J 66:356–358. 8. Anker SD, Egerer KR, Volk H-D, Kox WJ, Poole-Wilson PA, Coats AJS (1997) Elevated soluble CD14 receptors and altered cytokines in chronic heart failure. Am J Cardiol 79:1426–1430. 9. Bozkurt B, Kribbs S, Clubb FJ Jr et al. (1998) Pathophysiologically relevant concentrations of tumour necrosis factor- promote progressive left ventricular dysfunction and remodelling in rats. Circulation 97:1382–1391. 10. Finkel MS, Oddis CV, Jacob TD, Watkins SC, Hattler BG, Simmons RL (1992) Negative inotropic effects of cytokines on the heart mediated by nitric oxide. Science 257:387–389. 11. Krown KA, Page MT, Nguyen C et al. (1996) Tumour necrosis factor alpha-induced apoptosis in cardiac myocytes; involvement of the sphingolipid signalling cascade in cardiac cell death. J Clin Invest 98:2854–2865. Agnoletti L, Curello S, Bachetti T et al. (1999) Serum from patients with severe heart failure downregulates eNOS and is proapoptotic: role of tumour necrosis factor-alpha. Circulation 100: 1983–1991. 13. Anker SD, Clark AL, Kemp M et al. (1997) Tumour necrosis factor and steroid metabolism in chronic heart failure: possible relation to muscle wasting. J Am Coll Cardiol 30:997–1001. 14. Deswal A, Bozkurt B, Yukihiro S et al. (1999) Safety and efficacy of a soluble P75 tumour necrosis factor receptor (Enbrel, etanercept) in patients with advanced heart failure. Circulation 99:3224–3226. 15. Bozkurt B, Torre-Amione G, Deswal A, Soran OZ, Whitmore J, Mann DL (1999) Regression of left ventricular remodelling in chronic heart failure after treatment with ENBREL (Etanercept, p75 TNF Receptor Fc Fusion Protein) (abstr). Circulation 100 (Suppl):1–645. 16. Anker SD, Chua TP, Ponikowski P et al. (1997) Hormonal changes and catabolic/anabolic imbalance in chronic heart failure and their importance for cardiac cachexia. Circulation 96: 526–534. 17. Torre-Amione G, Kapadia S, Benedict C, Oral H, Young JB, Mann DL (1996) Proinflammatory cytokine levels in patients with depressed left ventricular ejection fraction: a report from the Studies of Left Ventricular Dysfunction (SOLVD). J Am Coll Cardiol 27:1201–1206. 18. Suffredini AF, Fromm RE, Parker MM et al. (1989) The cardiovascular response of normal humans to the administration of endotoxin. N Engl J Med 321:280–287. 19. Hegewish S, Weh HJ, Hossfeld DK (1990) TNF-induced cardiomyopathy. Lancet 2:294–295. 20. Millar AB, Foley NM, Singer M, Meager A, Johnson NM, Rook GA (1989) Tumour necrosis factor in bronchopulmonary secretions of patients with adult respiratory distress syndrome. Lancet 2:712–714. 21. Anker SD, Ponikowski PP, Clark AL et al. (1999) Cytokines and neurohormones relating to body composition alterations in the wasting syndrome of chronic heart failure. Eur Heart J 20:683–693. 22. Anker SD, Chua TP, Ponikowski P et al. (1997) Hormonal changes and catabolic/anabolic imbalance in chronic heart failure and their importance for cardiac cachexia. Circulation 96: 526–534. 23. Adams V, Jiang H, Yu J et al. (1999) Apoptosis in skeletal myocytes of patients with chronic heart failure is associated with extensive exercise intolerance. J Am Coll Cardiol 33:959–965. 24. Warren RS, Starnes HFJr, Gabrilove JL, Oettgen HF, Brennan MF (1987) The acute metabolic effects of tumour necrosis factor administration in humans. Arch Surg 122:1396–1400. 25. Anker SD, Volterrani M, Egerer KR et al. (1998) Tumour necrosis factor alpha as a predictor of impaired leg blood flow in patients with chronic heart failure. Q J Med 91:199–203.
86 / Cicoira et al. 26. Tracey KJ, Vlassara H, Cerami A (1989) Cachectin/tumour necrosis factor. Lancet 1:1122–1126. 27. Yoshizumi M, Perrella, Burnett JC, Lee ME (1993) Tumour necrosis factor downregulates an endothelial nitric oxide synthase mRNA by shortening its half-life. Circ Res 73:205–209. 28. Feldman AM, Combes A, Wagner D et al. (2000) The role to tumour necrosis factor in the pathophysiology of heart failure. J Am Coll Cardiol 35:537–544. 29. Torre-Amione G, Bozkurt B, Deswal A, Mann DL (1999) An overview of tumour necrosis factor alpha and the failing human heart. Curr Opin Cardiol 14:206–210.
CYTOKINE, Vol. 15, No. 2 (21 July, 2001: 80–86) 30. Kubota T, Bounoutas GS, Miyagishima M et al. (2000) Soluble tumour necrosis factor receptor abrogates myocardial inflammation but not hypertrophy in cytokine-induced cardiomyopathy. Circulation 101:2518–2525. 31. Chua TP, Ponikowski P, Harrington D et al. (1997) Clinical correlates and prognostic significance of the ventilatory response to exercise in chronic heart failure. J Am Coll Cardiol 29: 1585–1590. 32. Jones DA, Round JM, Edwards RHT, Grindrod SR, Tofts PS (1983) Size and composition of the calf and quadriceps in Duchenne muscular dystrophy. J Neurol Sci 60:307–322.
HIGH TUMOUR NECROSIS FACTOR- LEVELS ARE ASSOCIATED WITH EXERCISE INTOLERANCE AND NEUROHORMONAL ACTIVATION IN CHRONIC HEART FAILURE PATIENTS Mariantonietta Cicoira,1,2 Aidan P. Bolger,1 Wolfram Doehner,1 Mathias Rauchhaus,1,3 Constantinos Davos,1 Rakesh Sharma,1 Faisal O. Al-Nasser,1 Andrew J. S. Coats,1 Stefan D. Anker1,4 Immune activation plays an important role in the progression of chronic heart failure (CHF). We sought to investigate whether different degrees of tumor necrosis factor- (TNF-) activation are associated with exercise intolerance, neurohormonal activation and alterations in muscle mass and function in patients with CHF without cardiac cachexia. Patients were divided into quartiles according to their TNF levels (first quartile: 0.98–4.90 pg/ml, second quartile: 5.00–6.60 pg/ml; third quartile 6.80–9.00 pg/ml; fourth quartile 9.80–32.00 pg/ml). Patients underwent cardiopulmonary exercise testing, quadriceps muscle strength test, quadriceps fatigue test, and assessment of thigh muscle and fat cross-sectional area (CSA) by computerized tomography scanning. Patients in the highest TNF quartile had the lowest peak oxygen consumption [13.1 (4.1) ml/kg/min vs 18.1 (5.3), 18.8 (4.8) and 18.7 (5.6) ml/kg/min, P<0.01] the greatest relation of ventilation and dioxide production (VE/VCO2) slope (P<0.05) and the most elevated catecholamine levels (P<0.05) compared to patients in the first three quartiles. Patients with the lowest TNF levels had preserved thigh muscle size and quadriceps strength. Strength/ muscle CSA was similar in the four groups. Muscle strength during fatigue testing was significantly lower in the fourth quartile (P=0.01) compared with the other three groups. In CHF patients only the highest levels of TNF are associated with poor functional status and neurohormonal activation. This group of patients may represent the appropriate target population for TNF antagonism. 2001 Academic Press
Chronic heart failure (CHF) is a syndrome with a dismal prognosis,1,2 which is characterised by shortness of breath and exercise intolerance. Abnormalities of skeletal muscle are associated with the genesis of fatigue and breathlessness,3–5 although the mechanisms leading to myopathy are still not fully understood. It has been suggested that activation of the inflammaFrom the 1Department of Cardiac Medicine, National Heart & Lung Institute, London, UK; 2Divisione Clinicizzata di Cardiologia, Universita Y degli Studi di Verona, Verona, Italy; 3Klinik und Poliklinik fu¨r Innere Medizin III, Halle, Germany; 4Franz Volhard Klinik (Charite´, Campus Berlin-Buch) at Max Delbru¨ck Centrum for Molecular Medicine, Berlin, Germany Correspondence to: Dr Mariantonietta Cicoira, Divisione Clinicizzata di Cardiologia, Ospedale Civile Maggiore – P.le Stefani,1 37126 Verona, Italy; E-mail: [email protected] Received 15 February 2001; accepted for publication 27 April 2001 2001 Academic Press 1043–4666/01/140080+07 $35.00/0 KEY WORDS: chronic heart failure/TNF-/exercise intolerance/ neurohormonal activation 80
tory immune system is related to skeletal muscle wasting and cardiac cachexia in CHF.6–8 Elevated levels of tumour necrosis factor (TNF) are associated with left ventricular dysfunction, remodelling9,10 and cell death.11,12 We have previously demonstrated that, particularly in cachectic CHF patients, an increase in TNF and soluble TNF receptors correlates with a low body mass index and abnormalities in steroid hormone metabolism.13 These observations suggest a possible causal relationship between increased TNF levels and progression of the CHF syndrome. The impact of increasing degrees of TNF activation on the clinical status of CHF patients without evidence of weight loss has not yet been investigated. Pharmacological antagonism or neutralization of TNF is an expensive therapeutic approach, and may not necessarily be beneficial to all patients with CHF. Initial studies on patients with NYHA class III and IV symptoms suggest that therapy with a recombinant CYTOKINE, Vol. 15, No. 2 (21 July), 2001: pp 80–86
TNF- and exercise intolerance in heart failure / 81
TABLE 1.
Clinical characteristics of study population
Variable
MeanSD
Median
Range
Age (years) NYHA class Male (%) BMI (kg/m2) % ideal body weight (kg) LVEF (%) Peak VO2 (ml/min/kg) VE/VCO2 slope SBP (mm Hg) DBP (mm Hg) Plasma sodium (mmol/l) Total protein (g/l) Albumin (g/l) Creatinine (mmol/l) Urea (mmol/l) Aldosterone (pmol/l) Renin (ng/l) Norepinephrine (nmol/l) Epinephrine (nmol/l) TNF (pg/ml)
6112 2.60.8 96 25.94.0 11118 2714 17.25.4 37.313.5 11517 7210 1373 715 433 12443 95 712717 12.113.3 3.52.5 1.21.4 8.25.9
59 3 — 25.5 110 25 16.1 33.4 115 70 137 71 44 111 7 440 6.7 2.6 0.5 6.7
30–85 1–4 — 19.3–36.9 81–163 4–48 6.5–31.9 21.0–86.4 86–160 40–98 130–142 59–84 37–50 68–254 3–25 94–3439 0.6–63.8 0.811.9 0.1–5.4 0.932.0
NYHA: New York Heart Association classification of symptoms in heart failure; BMI: body mass index; LVEF: left ventricular ejection fraction; peak VO2: peak oxygen consumption; VE/VCO2: relation of ventilation and dioxide production; SBP: systolic blood pressure; DBP: diastolic blood pressure; TNF: tumour necrosis factor.
TNF soluble receptor is safe14 and improves cardiac function and clinical status.15 Our study aimed to explore the relationship between plasma concentrations of TNF and certain clinical characteristics in a typical population with CHF, so as to provide a logical approach to defining a target population most likely to benefit from anti-TNF therapy. We aimed to assess whether various degrees of immune activation in a population of CHF patients without cachexia are associated with differing degrees of disease severity, in terms of functional status, neurohormonal activation and altered skeletal muscle mass and performance.
class and plasma sodium did not differ in the four groups (ANOVA: all P>0.1) (Table 2). The patients in group 4 were somewhat older [mean age 68(8) years vs 55(10) and 62(14) years in group 1 to 3, ANOVA P=0.0019] and had higher creatinine levels [157(42) mmol/l vs 100(28), 113(41) and 123(38) mmol/l in group 1 to 3, ANOVA P=0.0002].
Exercise capacity Patients in group 4 had a lower mean peak VO2 compared with the other three groups (ANOVA P<0.01). There were no statistically significant differences in mean peak VO2 between patients in the first three quartiles (Fig. 1). Similarly, the mean relation of ventilation and dioxide production (VE/VCO2) slope was significantly higher only in patients in group 4 compared with patients in groups 1–3 (ANOVA, P<0.05) (Fig. 1). Age correlated with peak VO2 only in patients in the first three groups (r=0.45, P=0.0008) and not in patients in group 4 (P=0.15). There was a correlation of borderline significance between peak VO2 and TNF levels only in patients in the highest TNF quartile (r=0.47, P=0.056). When adjusted for age this correlation became significant only in group 4 (P=0.044), while in groups 1–3 it remained non-significant (P>0.2).
Neurohormonal activation Patients in the highest TNF quartile had the highest levels of norepinephrine (ANOVA, P<0.05) and epinephrine (ANOVA, P=0.01) (Fig. 2). Serum aldosterone and plasma renin activity were similar in the four groups. Body mass index was significantly higher in group 1 compared with groups 2–4 (see Table 2).
Muscle size and function RESULTS Baseline characteristics of the study population are summarized in Table 1. In total 72 patients were recruited (69 male and three female). Six patients (8%) were in NYHA class I, 24 (33%) in class II, 32 (45%) in class III and 10 (14%) in class IV. The aetiology of CHF was ischaemic heart disease in 45 patients (63%), idiopathic dilated cardiomyopathy in 24 patients (33%) and valvular heart disease in three (4%). Seven patients (10%) were not taking an angiotensin-converting enzyme inhibitor due to intolerance. Thirty-five patients were receiving digoxin, 24 amiodarone, 27 warfarin, 33 aspirin, and 15 -blockers. Patients were subdivided into quartiles of 18 patients each according to serum TNF levels (group 1: 0.98–4.90 pg/ml, group 2: 5.00–6.60 pg/ml, group 3: 6.8–9.00 pg/ml, group 4: 9.80–32.00 pg/ml). Mean LVEF, NYHA functional
Muscle CSA was significantly higher in the group with lowest TNF levels (ANOVA, P<0.01) (Fig. 3). Quadriceps muscle strength in both legs was higher in patients with the lowest TNF levels compared with those in the last three quartiles (ANOVA, P<0.01). Muscle strength during fatigue at five minutes was significantly lower in group 4 compared with the other three groups (ANOVA P=0.01) (Fig. 3). See Table 3 for details.
TNF assessment using different test kits In 33 patients TNF levels were assessed twice. Using the Medgenix test kit the mean TNF level was found to be 8.2 pg/ml (age-matched healthy controls: mean TNF level 7.0 pg/ml).16 Using the R&D test kit the mean TNF level was found to be 2.0 pg/ml (agematched healthy controls: 0.8 pg/ml).17 There was a statistically significant linear correlation between TNF
82 / Cicoira et al.
CYTOKINE, Vol. 15, No. 2 (21 July, 2001: 80–86)
TABLE 2. Exercise capacity, body mass index (BMI) and neurohormonal activation of study population divided into quartile groups according to tumour necrosis factor (TNF) levels (ANOVA) Variable
First quartile (TNF 0.98–4.90 pg/ml)
Second quartile (TNF 5.00–6.60 pg/ml)
Third quartile (TNF 6.80–9.00 pg/ml)
Fourth quartile (TNF 9.80–32.00 pg/ml)
P-value
18.15.3‡ 34.814.7‡ 29.25.0*§ 2.51.4‡ 0.70.9‡ 0.830.59 2.60.2
18.84.8‡ 33.910.4‡ 26.20.7‡ 2.92.2‡ 0.61.0† 0.770.41 2.70.4
18.75.6‡ 35.714.2‡ 24.83.4 3.52.3 1.51.8 0.720.38 2.60.4
13.14.1 45.811.9 23.42.6 5.03.1 1.91.6 1.070.51 2.80.4
<0.01 <0.05 <0.001 <0.01 <0.05 ns ns
Peak VO2 (ml/min/kg) VE/VCO2 slope BMI (kg/m2) Norepinephrine (nmol/l) Epinephrine (nmol/l) Renin (ng/ml/h) Aldosterone (pmol/l)
Peak VO2: peak oxygen consumption; VE/VCO2: relation of ventilation to carbon dioxide production. †P<0.01 vs fourth quartile. ‡P<0.05 vs fourth quartile. *P<0.01 vs third quartile. §P<0.0001 vs fourth quartile.
A
35
Peak VO2 (ml/kg/min)
30 25 20
*† ‡
15 10 5
B
*†
9 Norepinephrine (nmol/l)
A
8 7 6 5 4 3 2
1st
2nd
3rd
1
4th
1st
2nd
3rd
4th
B
90
5
VE/VCO2 slope
70
*† ‡‡
60 50 40 30
Epinephrine (nmol/l)
80 4
*††
3 2 1
20 10
1st 2nd 3rd 4th Quartile groups according to TNF levels
Figure 1. The relationship between TNF- levels and (A) peak oxygen consumption (VO2) and (B) the slope of ventilation (VE) and carbon dioxide production (VO2) in 72 chronic heart failure patients. Box plot displaying the 10th, 25th, 50th, 75th and 90th percentiles. *P<0.05 vs first quartile; †P<0.05a vs second quartile; ‡P<0.05 vs third quartile.
assessment with the two methods (r2 =0.78; P<0.0001) (Fig. 4), with the equation relating the two variables being: y=0.21x. The quartile cut-off values of 4.90,
0
1st 2nd 3rd 4th Quartile groups according to TNF levels
Figure 2. The relationship between TNF- levels and (A) noradrenaline and (B) adrenaline in 72 chronic heart failure patients. Box plot displaying the 10th, 25th, 50th, 75th and 90th percentiles. *P<0.05 vs first quartile; †P<0.05 vs second quartile: ‡P<0.01 vs second quartile.
6.60, and 9.80 pg/ml (using the Medgenix kit), correspond to levels of 1.0, 1.4 and 2.1 pg/ml using the R&D kit.
TNF- and exercise intolerance in heart failure / 83
A 2
Right thigh muscle CSA (cm )
180 160 140
*
*
2nd
3rd
***
120 100 80 1st
4th
B
Fatigue 5 at minutes (% change from baseline)
100 **‡‡
90 80 70
the three lower groups. Interestingly, the mean TNF levels of patients in the highest quartile were similar to those previously reported in cachectic CHF patients [15.3(3.1) pg/ml].22 It might be hypothesised, therefore, that this group of patients exhibit a ‘‘precachectic’’ state, in which weight loss is not clinically evident, but signs of impaired muscle function and reduced muscle bulk can already be observed. Overt cachexia may develop if this state of immune activation persists, with consequent weight loss and worsening of the clinical condition. High TNF levels may contribute to the loss of muscle bulk and strength in patients with CHF through several possible mechanisms. First, TNF is able to promote apoptosis23 and the breakdown of skeletal muscle proteins;24 furthermore, it reduces leg blood flow with a consequent reduction of skeletal muscle oxygen supply.25 In addition, TNF has cytotoxic effects on endothelial cells, both directly26 and via the down-regulation of endothelial constitutive nitric oxide synthase expression.27 Such mechanisms may also help to explain the observed impairment in exercise tolerance in CHF patients.
60 1st 2nd 3rd 4th Quartile groups according to TNF levels Figure 3. The relationship between TNF- levels and (A) right thigh muscle cross sectional area (CSA) and (B) fatigue at 5 min. Box plot displaying the 10th, 25th, 50th, 75th and 90th percentiles. *P<0.05 vs first quartile; **P<0.01 vs first quartile; ***P<0.001 vs first quartile; ‡‡P<0.01 vs third quartile.
DISCUSSION TNF is a proinflammatory cytokine with pleiotropic biological effects. It can produce left ventricular dysfunction, cardiomyopathy and pulmonary oedema when overexpressed.18–20 In 1990, Levine et al.6 reported that circulating levels of TNF are particularly increased in cachectic CHF patients, and a direct relationship between TNF levels and loss of skeletal muscle mass has been reported.21 In the present study we found that only increased levels of TNF are associated with severe impairment of functional status in CHF patients. Patients with TNF levels between 9.00 pg/ml and 32.00 pg/ml, i.e. in the highest quartile, had the most severe CHF, in terms of the worst exercise capacity, the highest levels of neurohormones and evidence of skeletal muscle wasting and weakness. The patients in the lower three quartiles of TNF levels had better exercise tolerance and lower catecholamine levels than patients in the highest quartile, although no differences were found between
Implications for anti-cytokine therapy in CHF The realisation that immune activation may play a role in the deterioration of myocardial function in CHF and possibly contributes to many of the peripheral manifestations seen in this condition, has provoked great interest in therapeutic strategies directed against TNF action.28,29 Although promising results have been reported, both in experimental and clinical studies,14,15,30 the number of patients treated remains too small for any firm conclusions to be drawn. The cost of this therapeutic approach is high and the selection of a patient group most likely to benefit from it is therefore paramount. Intuitively patients with evidence of raised TNF levels might be expected to gain the most. Taking the results from this study, those patients with TNF levels in the upper quartile (9.80– 32.00 pg/ml) may be considered a reasonable target population for anti-TNF therapy. They demonstrate more severe disease, a wide range of pathophysiological abnormalities correlate to elevated TNF levels and they very likely have a poor prognosis. In these patients the prospect of regular injection therapy seems both logical and feasible. Conversely, the group of patients with the lowest TNF levels (<4.90 pg/ml) exhibit, on average, clinical features consistent with mild to moderate disease, a tendency towards obesity (BMI 29.2 kg/m2) and have evidence of preserved skeletal muscle function and bulk. The use of antiTNF approaches in this sub-group seems less convincing. The potential benefit of anti-TNF treatment
84 / Cicoira et al.
CYTOKINE, Vol. 15, No. 2 (21 July, 2001: 80–86)
TABLE 3. Skeletal muscle size and function parameters of study population divided into quartile groups according to tumour necrosis factor (TNF) levels (ANOVA) Variable Right muscle CSA (cm2) Left muscle CSA (cm2) Right strength (n) Left strength (n) Strength/CSA right (n/cm2) Strength/CSA left (n/cm2) Fatigue 5 min (%)
First quartile (TNF 0.98–4.90 pg/ml)
Second quartile (TNF 5.00–6.60 pg/ml)
Third quartile (TNF 6.80–9.00 pg/ml)
Fourth quartile (TNF 9.80–32.00 pg/ml)
P-value (ANOVA)
134.227.9*† 131.826.8*† 451.9119.1 435.9115.5*† 6.71.0 6.61.0 86.99.7†
116.118.2 109.918.4 399.6103.3 366.490.5 6.51.0 6.40.9 83.810.1
114.816.6 110.818.4 344.677.8 352.3107.7 6.31.0 6.41.3 88.56.6†
103.720.4 97.119.8 327.1150.0 305.7129.7 5.91.3 6.11.3 76.810.9
<0.01 0.001 <0.01 <0.01 ns ns 0.01
CSA: cross sectional area of the four muscles of the thigh. The fatiguability at 5 min is expressed as percentage of baseline maximal quadriceps strength. *P<0.05 vs third quartile. †P<0.01 vs fourth quartile. §P<0.01 vs third quartile.
12
TNF R&D (pg/ml)
10
n = 72 2 r = 0.78 P < 0.001
8 6 4 2
0
y = 0.21x 5
10 15 20 25 TNF Medgenics (pg/ml)
30
35
Figure 4. Linear regression analysis of TNF- levels with two different ELISA kits (Medgenix and R&D Systems).
in individuals with intermediate TNF levels (5.00– 9.00 pg/ml) is harder to gauge. Although only exhibiting symptoms of mild to moderate CHF, patients in the third quartile show somewhat elevated catecholamine levels and the exclusion of this group from anti-TNF therapy may deny treatment to a number who might benefit. Further stratification of immune abnormalities in this population to tease out higher risk patients seems desirable.
MATERIALS AND METHODS We studied 72 consecutive patients with CHF presenting to the outpatient department of the Royal Brompton Hospital, London, UK. The diagnosis of CHF was based on typical symptoms of CHF of at least six months duration, cardiomegaly on chest X-ray and evidence of reduced left ventricular ejection fraction (LVEF), as assessed by radionuclide ventriculography. At the time of investigation, no patient had signs of oedema, hepatomegaly or ascites. Patients with chronic lung disease, neuromuscular disorders, recent myocardial infarction (within six months of the study), malignancy, acquired immunodeficiency syndrome or renal
failure (creatinine >250 mol/l) were excluded from the analysis. Patients who were cachectic, i.e. those with a body mass index (BMI) <19 kg/m2, a history of weight loss of greater than 6% since the onset of CHF or weight <85% of ideal body weight were not included. The protocol of the study was in accordance with the ethical standards of the Helsinki Declaration and was approved by the local ethics committee. Every patient provided written informed consent before being enrolled in the study. Each patient underwent maximal cardiopulmonary exercise testing (treadmill, modified Bruce protocol, Amis 2000, Odense, Denmark) with on-line measurement of oxygen consumption (VO2), ventilation (VE) and carbon dioxide production (VCO2).4 Peak VO2 was taken as an index of exercise capacity. The slope of the relation between ventilation and carbon dioxide production (VE/VCO2 slope) was calculated from the exercise data and taken as an index of the ventilatory response to exercise.31 Quadriceps muscle strength was measured as previously described.4 The subject was seated in a rigid frame, with legs hanging freely. An inelastic strap was attached from the ankle to a pressure transducer. The recording (Multitrace 2, Lectromed, Jersey, UK) from the pressure transducer was used to assess strength and provide visual feedback to the subject. The loss of a superimposed 1-Hz muscle twitch during the plateau of maximum force production indicated that the contraction was maximal. The best of three voluntary contractions on each leg, with a rest period of at least one minute in between, was considered the maximal voluntary quadriceps muscle strength of the right and left leg respectively. The values for muscle strength are expressed in Newton (N). The quadriceps fatigue protocol4 was performed on the stronger leg after strength testing. Patients performed repeated voluntary contractions at 30% of the maximum, using visual feedback as a guide. Contractions of one second were followed by one second relaxations for 40 seconds, followed by 20 seconds of rest. This 60 second cycle was repeated for 20 minutes. Patients were asked to undertake a maximal voluntary contraction at 5, 10, 15 and 20 minutes. The fatigability at each time point is expressed as a percentage of baseline maximal strength. Ultrafast computerized tomography (Imatron, San Francisco, USA) was used to measure the CSA of the total
TNF- and exercise intolerance in heart failure / 85
thigh and the four major muscles of the thigh (quadriceps, hamstrings, gracilis and sartorius) in both legs.4 The scans were performed in the supine position after five minutes of rest. A single 6 mm transaxial slice was scanned at mid-femur level and marked as 12.5% of patient height above the knee joint.32 The CSA was calculated in square centimetres by semi-automatic generation of an outline of the area of interest using the console software of the computer tomography scanner. Blood samples were collected in every subject between 9 a.m. and 10 a.m. after a fasting period of at least 10 h. A catheter was inserted into an antecubital vein and after 20 min of rest in a supine position 20 ml of blood were drawn. Blood samples were centrifuged immediately and aliquots were stored at 70C until analysis. TNF was measured in each patient using an ELISA assay with a lower limit of detectability of 3.0 pg/ml (Medgenix, Fleurus, Belgium) uninfluenced by soluble TNF receptors. In order to provide a comparison of these results with other published work in this field in which different TNF test kits have been used, 33 patients had a second assessment of TNF levels using the commercially available kit from R&D Systems (Minneapolis, USA. ELISA, lower limit of detectability 0.5 pg/ml). Catecholamines, aldosterone and plasma renin activity were analysed using routine hospital analysis procedures.
Statistical analysis All results are expressed as mean value SD. Patients were divided into four groups according to quartile distribution of their serum TNF levels. The hormonal measurements were log-normally distributed, and comparisons were made between sets of log-transformed data. Comparisons between groups were made by ANOVA analysis with Fisher’s post hoc test. Univariate and multivariate regression analyses were used to establish the relationship between biochemical and clinical variables. For the regression analysis comparing the two TNF test kits raw data was used and a forced intercept model was applied. All analyses were performed using a statistical software package (StatView 4.5 Abacus Concepts Inc, Berkeley, USA). A P-value <0.05 was considered statistically significant.
REFERENCES 1. Cohn JN, Johnson G, Ziesche S, and the Vasodilator-Heart Failure Veterans Affairs Cooperative Study Group (1991) A comparison of enalapril with hydralazine-isosorbide dinitrat in the treatment of chronic congestive heart failure. N Engl J Med 325: 303–310. 2. The SOLVD Investigators (1991) Effect of enalapril on survival in patients with reduced left ventricular ejection fraction and congestive heart failure. N Engl J Med 325:296–302. 3. Mancini DM, Walter G, Reichek N et al. (1992) Contribution of skeletal muscle atrophy to exercise intolerance and altered muscle metabolism in heart failure. Circulation 85:1364– 1373. 4. Anker SD, Swan JW, Volterrani M et al. (1997) The influence of muscle mass, strength, fatiguability and blood flow on exercise capacity in cachectic and non-cachectic patients with chronic heart failure. Eur Heart J 18:259–269.
5. Volterrani M, Clark AL, Ludman PF et al. (1994) Determinants of exercise capacity in chronic heart failure. Eur Heart J 15:801–809. 6. Levine B, Kalman J, Mayer L, Fillit HM, Packer M (1990) Elevated circulating levels of tumour necrosis factor in severe chronic heart failure. N Engl J Med 323:236–241. 7. McMurray J, Abdullah I, Dargie HJ, Shapiro D (1991) Increased concentrations of tumour necrosis factor in ‘cachectic’ patients with severe chronic heart failure. Br Heart J 66:356–358. 8. Anker SD, Egerer KR, Volk H-D, Kox WJ, Poole-Wilson PA, Coats AJS (1997) Elevated soluble CD14 receptors and altered cytokines in chronic heart failure. Am J Cardiol 79:1426–1430. 9. Bozkurt B, Kribbs S, Clubb FJ Jr et al. (1998) Pathophysiologically relevant concentrations of tumour necrosis factor- promote progressive left ventricular dysfunction and remodelling in rats. Circulation 97:1382–1391. 10. Finkel MS, Oddis CV, Jacob TD, Watkins SC, Hattler BG, Simmons RL (1992) Negative inotropic effects of cytokines on the heart mediated by nitric oxide. Science 257:387–389. 11. Krown KA, Page MT, Nguyen C et al. (1996) Tumour necrosis factor alpha-induced apoptosis in cardiac myocytes; involvement of the sphingolipid signalling cascade in cardiac cell death. J Clin Invest 98:2854–2865. Agnoletti L, Curello S, Bachetti T et al. (1999) Serum from patients with severe heart failure downregulates eNOS and is proapoptotic: role of tumour necrosis factor-alpha. Circulation 100: 1983–1991. 13. Anker SD, Clark AL, Kemp M et al. (1997) Tumour necrosis factor and steroid metabolism in chronic heart failure: possible relation to muscle wasting. J Am Coll Cardiol 30:997–1001. 14. Deswal A, Bozkurt B, Yukihiro S et al. (1999) Safety and efficacy of a soluble P75 tumour necrosis factor receptor (Enbrel, etanercept) in patients with advanced heart failure. Circulation 99:3224–3226. 15. Bozkurt B, Torre-Amione G, Deswal A, Soran OZ, Whitmore J, Mann DL (1999) Regression of left ventricular remodelling in chronic heart failure after treatment with ENBREL (Etanercept, p75 TNF Receptor Fc Fusion Protein) (abstr). Circulation 100 (Suppl):1–645. 16. Anker SD, Chua TP, Ponikowski P et al. (1997) Hormonal changes and catabolic/anabolic imbalance in chronic heart failure and their importance for cardiac cachexia. Circulation 96: 526–534. 17. Torre-Amione G, Kapadia S, Benedict C, Oral H, Young JB, Mann DL (1996) Proinflammatory cytokine levels in patients with depressed left ventricular ejection fraction: a report from the Studies of Left Ventricular Dysfunction (SOLVD). J Am Coll Cardiol 27:1201–1206. 18. Suffredini AF, Fromm RE, Parker MM et al. (1989) The cardiovascular response of normal humans to the administration of endotoxin. N Engl J Med 321:280–287. 19. Hegewish S, Weh HJ, Hossfeld DK (1990) TNF-induced cardiomyopathy. Lancet 2:294–295. 20. Millar AB, Foley NM, Singer M, Meager A, Johnson NM, Rook GA (1989) Tumour necrosis factor in bronchopulmonary secretions of patients with adult respiratory distress syndrome. Lancet 2:712–714. 21. Anker SD, Ponikowski PP, Clark AL et al. (1999) Cytokines and neurohormones relating to body composition alterations in the wasting syndrome of chronic heart failure. Eur Heart J 20:683–693. 22. Anker SD, Chua TP, Ponikowski P et al. (1997) Hormonal changes and catabolic/anabolic imbalance in chronic heart failure and their importance for cardiac cachexia. Circulation 96: 526–534. 23. Adams V, Jiang H, Yu J et al. (1999) Apoptosis in skeletal myocytes of patients with chronic heart failure is associated with extensive exercise intolerance. J Am Coll Cardiol 33:959–965. 24. Warren RS, Starnes HFJr, Gabrilove JL, Oettgen HF, Brennan MF (1987) The acute metabolic effects of tumour necrosis factor administration in humans. Arch Surg 122:1396–1400. 25. Anker SD, Volterrani M, Egerer KR et al. (1998) Tumour necrosis factor alpha as a predictor of impaired leg blood flow in patients with chronic heart failure. Q J Med 91:199–203.
86 / Cicoira et al. 26. Tracey KJ, Vlassara H, Cerami A (1989) Cachectin/tumour necrosis factor. Lancet 1:1122–1126. 27. Yoshizumi M, Perrella, Burnett JC, Lee ME (1993) Tumour necrosis factor downregulates an endothelial nitric oxide synthase mRNA by shortening its half-life. Circ Res 73:205–209. 28. Feldman AM, Combes A, Wagner D et al. (2000) The role to tumour necrosis factor in the pathophysiology of heart failure. J Am Coll Cardiol 35:537–544. 29. Torre-Amione G, Bozkurt B, Deswal A, Mann DL (1999) An overview of tumour necrosis factor alpha and the failing human heart. Curr Opin Cardiol 14:206–210.
CYTOKINE, Vol. 15, No. 2 (21 July, 2001: 80–86) 30. Kubota T, Bounoutas GS, Miyagishima M et al. (2000) Soluble tumour necrosis factor receptor abrogates myocardial inflammation but not hypertrophy in cytokine-induced cardiomyopathy. Circulation 101:2518–2525. 31. Chua TP, Ponikowski P, Harrington D et al. (1997) Clinical correlates and prognostic significance of the ventilatory response to exercise in chronic heart failure. J Am Coll Cardiol 29: 1585–1590. 32. Jones DA, Round JM, Edwards RHT, Grindrod SR, Tofts PS (1983) Size and composition of the calf and quadriceps in Duchenne muscular dystrophy. J Neurol Sci 60:307–322.