- Email: [email protected]
Plasma endothelin following cardiac arrest: differences between survivors and non-survivors
RESUSCITATION
ELSEVIER SCIENCE IRELAND
Resuscitation 27 ( 1994) I 17- 122
Plasma endothelin following cardiac arrest: differences between survivors and non-survivors William G. Haynes*a, D. Wayne Hamerb, Colin E. Robertsonb, David J. Webb” *Department of Medicine. University of Edinburgh, Western General Hospital, Edinburgh EH4 2XV, UK bDepartment of Accident and Emergency Medicine, Royal Infirmary of Edinburgh, Lauriston Place. Edinburgh EH3 7EB, UK
(Received 20 November 1993; revision received 6 December 1993;accepted 8
December
1993)
Abstract Cardiac arrest is associated with major metabolic disturbances, including severe hypoxia and large increases in circulating catecholamines, both of which are known to stimulate generation of the potent endothelium-derived vasoconstrictor peptide endothelin-1. We have, therefore, examined plasma immunoreactive endothelin concentrations following cardiac arrest. Blood was sampled at IO-min intervals from a central venous catheter inserted at onset of resuscitation in 38 patients (13 female; mean age, 67 years) presenting with cardiac arrest to the Accident and Emergency Department at the Royal Infirmary of Edinburgh. Plasma immunoreactive endothelin concentrations (mean * S.D.) in patients following cardiac arrest (5.4 f 2.3 pg/ml) were no different from those in healthy subjects (5.1 f 1.2 pg/ml). There was no significant difference between endothelin concentrations at presentation in survivors and non-survivors of cardiac arrest. However, non-survivors had a significant fall in endothelin concentrations with time from onset of resuscitation from 5.4 f 2.2 pg/ml to 3.5 f 1.8 pg/ml (P = 0.002), while survivors had a nonsignificant increase in concentrations. On multiple regression analysis there was a significant association between higher plasma endothelin concentration and survival (r = 0.37; P = 0.009). The failure of plasma endothelin to increase after cardiac arrest is unexpected. Although the fall in plasma endothelin with time in non-survivors may reflect the adverse physiological milieu that occurs during cardiac arrest, it is also possible that low endothelin concentrations contribute to the poor prognosis in this condition. Key words: Resuscitation; Ventricular fibrillation; Vasoconstrictor peptides; Catecholamines
Asystole; Ischaemic heart disease; Endothelin;
1. Introduction The vascular endothelium, which forms the inner lining of all blood vessels, produces both vasodilator agents, such as nitric oxide and pro* Corresponding author. 0 1994 0300-9572/94/$07.00 SSDI 0300-9572(93)00741-J
Endothelium;
stacyclin [I], and vasoconstrictors, such as angiotensin II [2] and endothelin-1 [3]. Endothelin-1 is a potent vasoconstrictor and pressor peptide with a uniquely prolonged action in both animals [3] and man [4,5]. Endothelin-1 is a member of a family of related isoforms, endothelin-1, -2 and -3, and endothelin-1 is the major isofortn produced within
Elsevier Science Ireland Ltd. All rights reserved.
118
the vasculature [6]. Although very high concentrations have been measured in patients on haemodialysis [7], sufficient to exert effects on vascular tone [8], in most circumstances endothelin-1 appears to be a locally acting autocrine and paracrine rather than a circulating hormone. Plasma concentrations of endothelin are likely to represent overspill from much higher concentrations at the interface between endothelium and vascular smooth muscle cell, similar to the situation for catecholamine release from nerve terminals. Increased circulating concentrations of endothelin are found in diseases associated with regional vasoconstriction, including Prinzmetal’s angina [9], Raynaud’s disease [lo] and cardiac failure [11,12]; in the last, plasma endothelin correlates with the severity of pulmonary hypertension. Plasma endothelin is also elevated in cardiogenic [13,14] and septic shock [15,16], possibly as a response to maintain arterial homeostatic pressure. Cardiac arrest results in circulatory standstill and, even with adequate resuscitation, arterial blood pressure, oxygenation and pH are greatly reduced. During cardiac arrest there is a marked increase in sympathetic discharge, producing substantially higher plasma catecholamine concentrations than recorded in any other condition [ 171. We hypothesised that plasma endothelin would increase during cardiac arrest, not only because factors that stimulate endothelin-1 production, such as catecholamines [3,18] and hypoxia [19], increase in cardiac arrest, whatever the cause, but also because impaired renal function would decrease clearance of endothelin-1 [20]. We have, therefore, measured plasma concentrations of immunoreactive endothelin in patients following cardiac arrest, and assessed their relationship with arterial blood oxygenation and survival. 2. Methods 2.1. Patients and protocol All patients presenting with cardiac arrest to the Accident and Emergency Department at the Royal Infirmary of Edinburgh between 0900 and 1700 h in the period October 1991 to February 1992 were
W.G. Haynes et al. / Resusciiarion 27 (1994) 117-122
prospectively enrolled into this study, which was approved by the local Ethics Review Committee. The Accident and Emergency Department treats approximately 400 cardiac arrests each year. Resuscitation was carried out strictly according to the current UK Resuscitation Council guidelines [21,22]. Where indicated, patients were intubated, received DC countershock, and standard cardiac resuscitation drugs in doses directed by treatment protocols. A mechanical compression/ventilation device (ThumperTM) was used in all patients as an aid to cardiopulmonary resuscitation. Soon after the onset of resuscitation a central venous catheter was inserted for administration of drugs. Blood was withdrawn from this line at lomin intervals while circulatory standstill persisted. In a subset of patients, a simultaneous arterial sample was obtained on one occasion for blood gas analysis. For each sample, 9 ml of blood was added to 1 ml ethylenediamine-tetra-acetate (EDTA potassium salt; final concentration 10 mmol/l). Each sample was separated at 4°C (within 10 min) and then stored at -20°C for 24 h before being transferred to a -70°C freezer. 2.2. Analytical procedures Plasma immunoreactive endothelin was measured by radioimmunoassay [23]. SepPak C 18 silica columns (Waters Associates, Milford, MA) were equilibrated by washing with methanol (5 ml), distilled water (5 ml) and then 4% acetic acid (5 ml). Each 2-ml plasma sample was diluted with 3 ml of 4% acetic acid and loaded onto a column; these were washed with 25% ethanol (3 ml), and eluted with 4% acetic acid in 86% ethanol (2 x 1 ml). The eluates were evaporated under nitrogen in a waterbath at 37°C and reconstituted in borate buffer of pH 8.4 (2 ml). Duplicate extracted samples and standards containing l-48 pg/ml of endothelin-1 (each 200 ~1) were incubated with rabbit polyclonal antibody raised against endothelin-1 (ITS Production B.V., Wijchen, Netherlands; in 100 ~1 distilled water) and [‘251]endothelin-1 (ITS; in 100 ~1 distilled water). After vortexing, tubes were incubated for 18 hours at 4°C. Donkey anti-rabbit gamma globulin bound on solid phase (ITS; 100 ~1) was added to all tubes,
W.G. Haynes et al. /Resuscitation 27 (1994)
117-122
except to the total count tubes, and tubes were incubated for 30 min at room temperature. After adding distilled water (1 ml), tubes were centrifuged for 15 min at 2000 x g at room temperin the ature. The amount of radioactivity antibody-bound fraction was determined by gamma counting for 2 min. The recovery of added endothelin-1 was 84%. Intra- and inter-assay coefficients of variation were 2.4% (n = 6) and 4.2% (n = 5), respectively. The sensitivity of this assay is 1 pg/ml endothelin. Cross reactivities of the assay with endothelin- 1, endothelin-2, endothelin-3 and proendothetin-1 are 100, 52, 96 and 7%, respectively. The normal range (mean f 2 SD.) for plasma immunoreactive endothelin with this assay, derived from blood obtained from 19 healthy control subjects (six female; age, 20-38 year) maintained recumbent for 30 min, is 2.7-7.5 pg/ml (mean = 5.1). 2.3. Statistical analysis Statistical analysis was performed using Student’s paired and unpaired t-tests, together with simple and multiple regression analysis of plasma immunoreactive endothelin against time from resuscitation, arterial hydrogen ion concentration and oxygen tension (Pao2) in survivors and non-survivors, taking significance at the 5% level. Results are expressed as mean i SD., with ranges provided where relevant. 3. Results During the study period, 38 patients fulfilled the entry criterion; 13 were female and 25 were male with mean age 67 years (range, 49-91). Fifteen patients left the resuscitation room alive and 23 died. The mean time from arrest to commencement of resuscitation was significantly shorter in survivors (5 f 7 min) than in non-survivors (10 f 8 min; P = 0.03). The primary documented rhythm was ventricular fibrillation in 20, asystole in 14 and electromechanical dissociation in four. All 15 of the survivors had ventricular fibrillation. All patients received intravenous adrenaline (2.1 f 1 mg), with non-survivors (2.9 f 1 mg) tending to receive higher doses than survivors
119
(1.3 f 0.9 mg; NS). Non-survivors also received more atropine (0.6 f 0.2 mg) than survivors (0.2 f 0.2 mg; P = 0.01). The mean time from onset of resuscitation to obtaining the first sample for endothelin assay was 20 f 14 min in survivors, and 25 f 12 min in non-survivors (P= 0.27), with between one and four samples taken per patient (mean = 2.1). The mean plasma concentration of endothelin was 5.4 i 2.3 pg/ml in the first sample obtained from all patients presenting with cardiac arrest. This was not significantly different from our normal range. Endothelin concentrations in the first sample from survivors (5.9 f 2.6 pg/ml) were not significantly different from those in non-survivors (5.1 f 2.1 pg/ml). However, plasma endothelin fell significantly between the first (5.4 i 2.2 pg/ml) and last (3.5 f 1.8 pg/ml; P = 0.002) samples in those non-survivors who had serial samples taken, but tended to increase in those survivors who had serial samples taken (from 5.6 f 2.4 pg/ml to 7.0 f 2.8 pg/ml; P = 0.16; Fig. 1). In addition, while there was no significant correlation between endothelin concentrations and time from institution of resuscitation in survivors (r = 0.10; P = 0.68), there was a significant negative correlation in non-survivors (r = -0.45; P = 0.002; Fig. 1). Nineteen patients had simultaneous samples obtained for arterial blood gas analysis, with a mean time to arterial sampling of 19 f 11 min in survivors and 26 f 10 min in non-survivors (not significantly different; P = 0.15). Arterial P,o, was not significantly different between survivors (mean = 21 f 16 kPa) and non-survivors (mean = 20 f 22 kPa; P = 0.89) and Paoz did not correlate with plasma endothelin concentration in either survivors (r = 0.13; P = 0.74) or nonsurvivors (r = 0.09; P = 0.81). Hydrogen ion concentration was significantly (P = 0.003) lower in survivors (mean = 55 f 18 mmol/l) than nonsurvivors (mean = 87 f 22 mmol/l). There was a positive correlation between the degree of acidosis and endothelin concentration in the survivors (r = 0.75; P = 0.02), but not in the non-survivors (r = 0.26;P = 0.46; Fig. 2). In a multiple regression analysis of endothelin concentrations in all samples against outcome,
120
W.G. Haynes et al. /Resuscitation 27 (1994) 117-122
A: Survivor
lmm cardiac arrest
B: Noksurvivors
12 -I
‘2
I
I
0
10
20
30
40
50
Time fmm resuscitation
60
70
80
fmm cardiac arrest
1
I
0
I
I
0
10
20
30
(min)
0
1
40
Time from resuscitation
So
60
I
70
(min)
Fig. 1. Plasma immunoreactive endothelin concentrations plotted against time from institution of resuscitation in individual survivors (A) and non-survivors(B) from cardiac arrest. Endothelin concentrations fall with time in non-survivors (P = 0.002). but not survivors (P= 0.16), of cardiac arrest. Correlation between plasma endothelin and time is also shown and is significant for non-survivors (r = -0.45; P = 0.002), but not for survivors(r = 0.10;P = 0.68).
time from resuscitation, arterial hydrogen ion concentration and Paoz, there was a significant positive association with outcome (r = 0.37; P = 0.0009), but not with the other factors.
10
91
p ,a7 .$ 6. 15 $ 4.
I b
:. 1. 0' 30
40
50
60 Hydrogen
70
80
90
100
ton
Fig. 2. Correlation between plasma immunoreactive endothelin concentration and arterial hydrogen ion concentration in survivors (r = 0.75; P = 0.02) from cardiac arrest.
4.Discussion In this study, plasma immunoreactive endothelin was not elevated in patients with cardiac arrest. This is an unexpected finding, given the severe metabolic disturbance that occurs following cardiac arrest. This includes severe hypoxaemia, metabolic acidosis and massive release of catecholamines [ 171; factors which stimulate generation of endothelin-1 [3,18,19]. Further, it might be expected that plasma endothelin concentrations would increase due to decreased renal clearance during circulatory standstill and resuscitation. It is unlikely that our negative findings could be accounted for by the difference in ages between the patients and the control group, because plasma endothelin concentrations do not appear to be affected by age [24]. Plasma endothelin may fail to rise early after cardiac arrest because endothelin-1 is not stored within endothelial cells and de novo synthesis of the peptide takes several hours to occur. This is certainly the case for stimulation of generation of endothelin-1 in vitro by hypoxia and catechol-
121
W.G. Haynes et al. /Resuscitation 27 (1994) 117-122
amines [3,18,19]. However, plasma endothelin can increase within minutes in vivo in man, in response to stimuli such as orthostasis and the cold pressor test [6]. Generation of endothelin-1 by isolated endothelial cells is stimulated by shear stress [25]. Therefore, endothelin concentrations may fail to increase because of decreased exposure of endothelial cells to shear stress during cardiac arrest, particularly pulsatile shear stress. In addition, the adverse physiological conditions that occur during cardiac arrest may prevent DNA transcription, protein production from mRNA or the activity of proendothelin-1 converting enzyme. Such conditions may help to explain why endothelin concentrations fall with time in non-survivors. Although such factors may account for the lack of elevation of endothelin-1 concentrations early in resuscitation, they do not explain why endothelin concentrations are not increased later in the course of the arrest. It is possible that other mechanisms come into play to prevent increases in endothelin generation. These may include an increase in the generation of nitric oxide, and of Aand C-type natriuretic peptides, as well as rapid destabilisation of the mRNA for endothelin-1 [6,261. Although there was no difference in plasma endothelin concentrations at presentation between survivors and non-survivors, there was a significant positive association between outcome and plasma endothelin concentrations in multiple regression analysis. This relation with outcome may help to explain the associations we observed with time and hydrogen ion concentrations in the non-survivors and survivors, respectively. As opposed to blood flow to the heart and brain, blood flow to peripheral tissues is very poor during cardiopulmonary resuscitation. This leads to local tissue hypoxia, even in the presence of apparently adequate arterial P,o,. This in turn causes a local metabolic acidosis, which may stimulate production of endothelin- 1, which will tend to accumulate in pooled blood in these tissues. The positive correlation of endothelin concentrations with hydrogen ion only in survivors, may therefore be most readily explained by higher peripheral tissue blood flows during resuscitation in these patients, with greater venous return of peripheral
blood, containing high concentrations of endothelin- 1. The failure of plasma endothelin to correlate with hydrogen ion in non-survivors is not explained by survivors having a longer delay to arterial sampling, allowing more time for generation of endothelin, because survivors tended to have a shorter time from resuscitation to arterial sampling than non-survivors. Neither can the association be attributed to a greater degree of acidosis in survivors, reflecting a more potent stimulus for generation of endothelin, because this group had significantly lower hydrogen ion concentrations. Because plasma endothelin concentrations are not elevated at presentation in patients with cardiac arrest, it is likely that endothelin has no major role as a circulating pressor hormone in this condition. However, the failure of circulating endothelin concentrations to increase does not exclude a role for endothelin in the cardiovascular response to cardiac arrest. Sustained endothelin generation may occur without an increase in circulating endothelin, because generation is directed away from the lumen of the blood vessel. Here, experimental evidence may be helpful, using specific receptor antagonists or inhibitors of proendothelin- 1 converting enzyme in animal models to determine whether local generation of endothelin contributes to the restoration of vascular tone following cardiac arrest. During cardiac arrest and cardiopulmonary resuscitation, vasopressor agents such as adrenaline are used to increase systemic vascular resistance in order to improve myocardial and cerebral blood flow [27]. As endothelin-1 has potent long lasting vasoconstrictor, pressor and positive inotropic effects [6], high endothelin concentrations may be beneficial during cardiac arrest. Although the correlation of plasma endothelin with outcome may reflect the adverse physiological milieu that occurs during cardiac arrest, it is also possible that the fall in endothelin concentrations in non-survivors contributes to their poor prognosis. 5. Acknowledgement This study was supported by a grant from the Scottish Home and Health Department.
122
W.G. Haynes et al. /Resuscitation
6. References 1 Vane JR, binggird EA, Botting RM. Regulatory functions of the vascular endothelium. N Engl J Med 1990; 323: 27-36. 2 Webb DJ, Cockcroft JR. Circulating and tissue reninangiotensin systems and the role of the endothelium. In: Warren JB, editor. The endothelium: an introduction to current research. New York: Wiley-Liss, 1990; 65-80. 3 Yanagisawa M, Kurihara H, Kimura S et al. A novel potent vasoconstrictor peptide produced by vascular endothelial cells. Nature 1988; 332: 41 I-415. 4 Clarke JG, Benjamin N, Larkin SW, Webb DJ, Davies GJ, Maseri, A. Endothelin is a potent long-lasting vasoconstrictor in men. Am J Physiol 1989; 257: H2033-H2035. 5 Vierhapper H, Wagner 0, Nowotny P, Waldhausl W. Effect of endothelin-1 in man. Circulation 1990; 81: 1415-1418. 6 Haynes WG, Webb DJ. The endothelin family of peptides: local hormones with diverse roles in health and disease? Clin Sci 1993; 84: 485-500. 7 Warrens AN, Cassidy MJD, Takahashi K, Ghatei MA, Bloom SR. Endothelin in renal failure. Nephrol Dial Transplant 1990; 5: 418-422. 8 Webb DJ, Cockcroft JR. Plasma immunoreactive endothelin in uraemia. Lancet 1989; ii: 1211. 9 Toyo-oka T, Aizawa T, Suzuki N et al. Increased plasma level of endothelin-I and coronary spasm induction in patients with vasospastic angina pectoris. Circulation 1991; 83: 476-483. 10 Zamora MR, O’Brien RF, Rutherford RB, Weil JV. Serum endothelin-I concentrations and cold provocation in primary Raynaud’s phenomenon. Lancet 1990; 336: 1144-1147. 11 Cody RJ, Haas GJ, Binkley PF, Capers Q, Kelly R. Plasma endothelin correlates with the extent of pulmonary hypertension in patients with chronic congestive heart failure. Circulation 1992; 85: 504-509. 12 Stewart D, Cernacek P, Costello K, Rouleau JL. Elevated endothelin-I in heart failure and loss of normal response to postural change. Circulation 1992; 85: 510-517. 13 Cernacek P, Stewart D. Immunoreactive endothelin in human plasma: marked elevations in patients with cardiogenie shock. Biochem Biophys Res Commun 1989; 161: 562-567. 14 Yasuda M, Kohno M, Tahara A et al. Circulating
15
16
17
18
19
20
21
22
23
24
25
26
27
27 (1994)
117-122
immunoreactive endothelin in ischaemic heart disease. Am Heart J 1990; 119: 801-806. Weitzberg E, Lundberg JM, Rudehill A. Elevated plasma levels of endothehn in patients with sepsis syndrome. Circ Shock 1991; 33: 222-227. Sanai L, Grant IS, Webb DJ. The role of endothelin-I in the adult respiratory distress syndrome. Proceedings of the 6th European Congress on Intensive Care Medicine, 1992. Little RA, Frayn KN, Randall PE et al. Plasma catecholamines in patients with acute myocardial infarction and cardiac arrest. Q J Med 1985; 54: 133-140. Prasad MR, Jones RM, Kreutzer DL. Release of endothelin from cultured bovine endothelial cells. J Mel Cell Cardiol 1991; 23: 655-658. Kourembanas S, Marsden PA, McQuillan LP, Failer DV. Hypoxia induces endothelin gene expression and secretion in cultured human endothelium. J Clin Invest 1991; 88: 1054-1057. Kohno, M., Murakawa, K., Yasunari, K. et al. Prolonged blood pressure elevation after endothelin administration in bilaterally nephrectomized rats. Metabolism 1989; 38, 712-713. Chamberlain DA. Advanced life support: revised recommendations of the Resuscitation Council (UK). Br Med J 1989; 299: 446-448. Marsden AK. Basic life support: revised recommendations of the Resuscitation Council (UK). Br Med J 1989; 299: 442-445. Florijn KW, Derkx FHM, Visser W et al. Plasma immunoreactive endothelin-I in pregnant women with and without pre-eclampsia. J Cardiovasc Pharmacol 1991; 17 Suppl 7: 5446-5448. Lerman A, Edwards BS, Hallett JW, Heublein DM, Sandberg SM, Burnett JC. Circulating and tissue endothelin immunoreactivity in advanced atherosclerosis. N Engl J Med 1991; 325: 997-1001. Yoshizumi M, Kurihara H, Sugiyama T et al. Haemodynamic shear stress stimulates endothelin production by cultured endothelial cells. Biochem Biophys Res Commun 1989; 161: 859-864. Haynes WG, Davenport AP, Webb DJ. Endothehn: progress in physiology and pharmacology. Trends Pharmacol Sci 1993; 14: 225-228. Lindner KH, Koster R. Vasopressor drugs during cardioresuscitation. Resuscitation 1992; 24: pulmonary 147-153.
ELSEVIER SCIENCE IRELAND
Resuscitation 27 ( 1994) I 17- 122
Plasma endothelin following cardiac arrest: differences between survivors and non-survivors William G. Haynes*a, D. Wayne Hamerb, Colin E. Robertsonb, David J. Webb” *Department of Medicine. University of Edinburgh, Western General Hospital, Edinburgh EH4 2XV, UK bDepartment of Accident and Emergency Medicine, Royal Infirmary of Edinburgh, Lauriston Place. Edinburgh EH3 7EB, UK
(Received 20 November 1993; revision received 6 December 1993;accepted 8
December
1993)
Abstract Cardiac arrest is associated with major metabolic disturbances, including severe hypoxia and large increases in circulating catecholamines, both of which are known to stimulate generation of the potent endothelium-derived vasoconstrictor peptide endothelin-1. We have, therefore, examined plasma immunoreactive endothelin concentrations following cardiac arrest. Blood was sampled at IO-min intervals from a central venous catheter inserted at onset of resuscitation in 38 patients (13 female; mean age, 67 years) presenting with cardiac arrest to the Accident and Emergency Department at the Royal Infirmary of Edinburgh. Plasma immunoreactive endothelin concentrations (mean * S.D.) in patients following cardiac arrest (5.4 f 2.3 pg/ml) were no different from those in healthy subjects (5.1 f 1.2 pg/ml). There was no significant difference between endothelin concentrations at presentation in survivors and non-survivors of cardiac arrest. However, non-survivors had a significant fall in endothelin concentrations with time from onset of resuscitation from 5.4 f 2.2 pg/ml to 3.5 f 1.8 pg/ml (P = 0.002), while survivors had a nonsignificant increase in concentrations. On multiple regression analysis there was a significant association between higher plasma endothelin concentration and survival (r = 0.37; P = 0.009). The failure of plasma endothelin to increase after cardiac arrest is unexpected. Although the fall in plasma endothelin with time in non-survivors may reflect the adverse physiological milieu that occurs during cardiac arrest, it is also possible that low endothelin concentrations contribute to the poor prognosis in this condition. Key words: Resuscitation; Ventricular fibrillation; Vasoconstrictor peptides; Catecholamines
Asystole; Ischaemic heart disease; Endothelin;
1. Introduction The vascular endothelium, which forms the inner lining of all blood vessels, produces both vasodilator agents, such as nitric oxide and pro* Corresponding author. 0 1994 0300-9572/94/$07.00 SSDI 0300-9572(93)00741-J
Endothelium;
stacyclin [I], and vasoconstrictors, such as angiotensin II [2] and endothelin-1 [3]. Endothelin-1 is a potent vasoconstrictor and pressor peptide with a uniquely prolonged action in both animals [3] and man [4,5]. Endothelin-1 is a member of a family of related isoforms, endothelin-1, -2 and -3, and endothelin-1 is the major isofortn produced within
Elsevier Science Ireland Ltd. All rights reserved.
118
the vasculature [6]. Although very high concentrations have been measured in patients on haemodialysis [7], sufficient to exert effects on vascular tone [8], in most circumstances endothelin-1 appears to be a locally acting autocrine and paracrine rather than a circulating hormone. Plasma concentrations of endothelin are likely to represent overspill from much higher concentrations at the interface between endothelium and vascular smooth muscle cell, similar to the situation for catecholamine release from nerve terminals. Increased circulating concentrations of endothelin are found in diseases associated with regional vasoconstriction, including Prinzmetal’s angina [9], Raynaud’s disease [lo] and cardiac failure [11,12]; in the last, plasma endothelin correlates with the severity of pulmonary hypertension. Plasma endothelin is also elevated in cardiogenic [13,14] and septic shock [15,16], possibly as a response to maintain arterial homeostatic pressure. Cardiac arrest results in circulatory standstill and, even with adequate resuscitation, arterial blood pressure, oxygenation and pH are greatly reduced. During cardiac arrest there is a marked increase in sympathetic discharge, producing substantially higher plasma catecholamine concentrations than recorded in any other condition [ 171. We hypothesised that plasma endothelin would increase during cardiac arrest, not only because factors that stimulate endothelin-1 production, such as catecholamines [3,18] and hypoxia [19], increase in cardiac arrest, whatever the cause, but also because impaired renal function would decrease clearance of endothelin-1 [20]. We have, therefore, measured plasma concentrations of immunoreactive endothelin in patients following cardiac arrest, and assessed their relationship with arterial blood oxygenation and survival. 2. Methods 2.1. Patients and protocol All patients presenting with cardiac arrest to the Accident and Emergency Department at the Royal Infirmary of Edinburgh between 0900 and 1700 h in the period October 1991 to February 1992 were
W.G. Haynes et al. / Resusciiarion 27 (1994) 117-122
prospectively enrolled into this study, which was approved by the local Ethics Review Committee. The Accident and Emergency Department treats approximately 400 cardiac arrests each year. Resuscitation was carried out strictly according to the current UK Resuscitation Council guidelines [21,22]. Where indicated, patients were intubated, received DC countershock, and standard cardiac resuscitation drugs in doses directed by treatment protocols. A mechanical compression/ventilation device (ThumperTM) was used in all patients as an aid to cardiopulmonary resuscitation. Soon after the onset of resuscitation a central venous catheter was inserted for administration of drugs. Blood was withdrawn from this line at lomin intervals while circulatory standstill persisted. In a subset of patients, a simultaneous arterial sample was obtained on one occasion for blood gas analysis. For each sample, 9 ml of blood was added to 1 ml ethylenediamine-tetra-acetate (EDTA potassium salt; final concentration 10 mmol/l). Each sample was separated at 4°C (within 10 min) and then stored at -20°C for 24 h before being transferred to a -70°C freezer. 2.2. Analytical procedures Plasma immunoreactive endothelin was measured by radioimmunoassay [23]. SepPak C 18 silica columns (Waters Associates, Milford, MA) were equilibrated by washing with methanol (5 ml), distilled water (5 ml) and then 4% acetic acid (5 ml). Each 2-ml plasma sample was diluted with 3 ml of 4% acetic acid and loaded onto a column; these were washed with 25% ethanol (3 ml), and eluted with 4% acetic acid in 86% ethanol (2 x 1 ml). The eluates were evaporated under nitrogen in a waterbath at 37°C and reconstituted in borate buffer of pH 8.4 (2 ml). Duplicate extracted samples and standards containing l-48 pg/ml of endothelin-1 (each 200 ~1) were incubated with rabbit polyclonal antibody raised against endothelin-1 (ITS Production B.V., Wijchen, Netherlands; in 100 ~1 distilled water) and [‘251]endothelin-1 (ITS; in 100 ~1 distilled water). After vortexing, tubes were incubated for 18 hours at 4°C. Donkey anti-rabbit gamma globulin bound on solid phase (ITS; 100 ~1) was added to all tubes,
W.G. Haynes et al. /Resuscitation 27 (1994)
117-122
except to the total count tubes, and tubes were incubated for 30 min at room temperature. After adding distilled water (1 ml), tubes were centrifuged for 15 min at 2000 x g at room temperin the ature. The amount of radioactivity antibody-bound fraction was determined by gamma counting for 2 min. The recovery of added endothelin-1 was 84%. Intra- and inter-assay coefficients of variation were 2.4% (n = 6) and 4.2% (n = 5), respectively. The sensitivity of this assay is 1 pg/ml endothelin. Cross reactivities of the assay with endothelin- 1, endothelin-2, endothelin-3 and proendothetin-1 are 100, 52, 96 and 7%, respectively. The normal range (mean f 2 SD.) for plasma immunoreactive endothelin with this assay, derived from blood obtained from 19 healthy control subjects (six female; age, 20-38 year) maintained recumbent for 30 min, is 2.7-7.5 pg/ml (mean = 5.1). 2.3. Statistical analysis Statistical analysis was performed using Student’s paired and unpaired t-tests, together with simple and multiple regression analysis of plasma immunoreactive endothelin against time from resuscitation, arterial hydrogen ion concentration and oxygen tension (Pao2) in survivors and non-survivors, taking significance at the 5% level. Results are expressed as mean i SD., with ranges provided where relevant. 3. Results During the study period, 38 patients fulfilled the entry criterion; 13 were female and 25 were male with mean age 67 years (range, 49-91). Fifteen patients left the resuscitation room alive and 23 died. The mean time from arrest to commencement of resuscitation was significantly shorter in survivors (5 f 7 min) than in non-survivors (10 f 8 min; P = 0.03). The primary documented rhythm was ventricular fibrillation in 20, asystole in 14 and electromechanical dissociation in four. All 15 of the survivors had ventricular fibrillation. All patients received intravenous adrenaline (2.1 f 1 mg), with non-survivors (2.9 f 1 mg) tending to receive higher doses than survivors
119
(1.3 f 0.9 mg; NS). Non-survivors also received more atropine (0.6 f 0.2 mg) than survivors (0.2 f 0.2 mg; P = 0.01). The mean time from onset of resuscitation to obtaining the first sample for endothelin assay was 20 f 14 min in survivors, and 25 f 12 min in non-survivors (P= 0.27), with between one and four samples taken per patient (mean = 2.1). The mean plasma concentration of endothelin was 5.4 i 2.3 pg/ml in the first sample obtained from all patients presenting with cardiac arrest. This was not significantly different from our normal range. Endothelin concentrations in the first sample from survivors (5.9 f 2.6 pg/ml) were not significantly different from those in non-survivors (5.1 f 2.1 pg/ml). However, plasma endothelin fell significantly between the first (5.4 i 2.2 pg/ml) and last (3.5 f 1.8 pg/ml; P = 0.002) samples in those non-survivors who had serial samples taken, but tended to increase in those survivors who had serial samples taken (from 5.6 f 2.4 pg/ml to 7.0 f 2.8 pg/ml; P = 0.16; Fig. 1). In addition, while there was no significant correlation between endothelin concentrations and time from institution of resuscitation in survivors (r = 0.10; P = 0.68), there was a significant negative correlation in non-survivors (r = -0.45; P = 0.002; Fig. 1). Nineteen patients had simultaneous samples obtained for arterial blood gas analysis, with a mean time to arterial sampling of 19 f 11 min in survivors and 26 f 10 min in non-survivors (not significantly different; P = 0.15). Arterial P,o, was not significantly different between survivors (mean = 21 f 16 kPa) and non-survivors (mean = 20 f 22 kPa; P = 0.89) and Paoz did not correlate with plasma endothelin concentration in either survivors (r = 0.13; P = 0.74) or nonsurvivors (r = 0.09; P = 0.81). Hydrogen ion concentration was significantly (P = 0.003) lower in survivors (mean = 55 f 18 mmol/l) than nonsurvivors (mean = 87 f 22 mmol/l). There was a positive correlation between the degree of acidosis and endothelin concentration in the survivors (r = 0.75; P = 0.02), but not in the non-survivors (r = 0.26;P = 0.46; Fig. 2). In a multiple regression analysis of endothelin concentrations in all samples against outcome,
120
W.G. Haynes et al. /Resuscitation 27 (1994) 117-122
A: Survivor
lmm cardiac arrest
B: Noksurvivors
12 -I
‘2
I
I
0
10
20
30
40
50
Time fmm resuscitation
60
70
80
fmm cardiac arrest
1
I
0
I
I
0
10
20
30
(min)
0
1
40
Time from resuscitation
So
60
I
70
(min)
Fig. 1. Plasma immunoreactive endothelin concentrations plotted against time from institution of resuscitation in individual survivors (A) and non-survivors(B) from cardiac arrest. Endothelin concentrations fall with time in non-survivors (P = 0.002). but not survivors (P= 0.16), of cardiac arrest. Correlation between plasma endothelin and time is also shown and is significant for non-survivors (r = -0.45; P = 0.002), but not for survivors(r = 0.10;P = 0.68).
time from resuscitation, arterial hydrogen ion concentration and Paoz, there was a significant positive association with outcome (r = 0.37; P = 0.0009), but not with the other factors.
10
91
p ,a7 .$ 6. 15 $ 4.
I b
:. 1. 0' 30
40
50
60 Hydrogen
70
80
90
100
ton
Fig. 2. Correlation between plasma immunoreactive endothelin concentration and arterial hydrogen ion concentration in survivors (r = 0.75; P = 0.02) from cardiac arrest.
4.Discussion In this study, plasma immunoreactive endothelin was not elevated in patients with cardiac arrest. This is an unexpected finding, given the severe metabolic disturbance that occurs following cardiac arrest. This includes severe hypoxaemia, metabolic acidosis and massive release of catecholamines [ 171; factors which stimulate generation of endothelin-1 [3,18,19]. Further, it might be expected that plasma endothelin concentrations would increase due to decreased renal clearance during circulatory standstill and resuscitation. It is unlikely that our negative findings could be accounted for by the difference in ages between the patients and the control group, because plasma endothelin concentrations do not appear to be affected by age [24]. Plasma endothelin may fail to rise early after cardiac arrest because endothelin-1 is not stored within endothelial cells and de novo synthesis of the peptide takes several hours to occur. This is certainly the case for stimulation of generation of endothelin-1 in vitro by hypoxia and catechol-
121
W.G. Haynes et al. /Resuscitation 27 (1994) 117-122
amines [3,18,19]. However, plasma endothelin can increase within minutes in vivo in man, in response to stimuli such as orthostasis and the cold pressor test [6]. Generation of endothelin-1 by isolated endothelial cells is stimulated by shear stress [25]. Therefore, endothelin concentrations may fail to increase because of decreased exposure of endothelial cells to shear stress during cardiac arrest, particularly pulsatile shear stress. In addition, the adverse physiological conditions that occur during cardiac arrest may prevent DNA transcription, protein production from mRNA or the activity of proendothelin-1 converting enzyme. Such conditions may help to explain why endothelin concentrations fall with time in non-survivors. Although such factors may account for the lack of elevation of endothelin-1 concentrations early in resuscitation, they do not explain why endothelin concentrations are not increased later in the course of the arrest. It is possible that other mechanisms come into play to prevent increases in endothelin generation. These may include an increase in the generation of nitric oxide, and of Aand C-type natriuretic peptides, as well as rapid destabilisation of the mRNA for endothelin-1 [6,261. Although there was no difference in plasma endothelin concentrations at presentation between survivors and non-survivors, there was a significant positive association between outcome and plasma endothelin concentrations in multiple regression analysis. This relation with outcome may help to explain the associations we observed with time and hydrogen ion concentrations in the non-survivors and survivors, respectively. As opposed to blood flow to the heart and brain, blood flow to peripheral tissues is very poor during cardiopulmonary resuscitation. This leads to local tissue hypoxia, even in the presence of apparently adequate arterial P,o,. This in turn causes a local metabolic acidosis, which may stimulate production of endothelin- 1, which will tend to accumulate in pooled blood in these tissues. The positive correlation of endothelin concentrations with hydrogen ion only in survivors, may therefore be most readily explained by higher peripheral tissue blood flows during resuscitation in these patients, with greater venous return of peripheral
blood, containing high concentrations of endothelin- 1. The failure of plasma endothelin to correlate with hydrogen ion in non-survivors is not explained by survivors having a longer delay to arterial sampling, allowing more time for generation of endothelin, because survivors tended to have a shorter time from resuscitation to arterial sampling than non-survivors. Neither can the association be attributed to a greater degree of acidosis in survivors, reflecting a more potent stimulus for generation of endothelin, because this group had significantly lower hydrogen ion concentrations. Because plasma endothelin concentrations are not elevated at presentation in patients with cardiac arrest, it is likely that endothelin has no major role as a circulating pressor hormone in this condition. However, the failure of circulating endothelin concentrations to increase does not exclude a role for endothelin in the cardiovascular response to cardiac arrest. Sustained endothelin generation may occur without an increase in circulating endothelin, because generation is directed away from the lumen of the blood vessel. Here, experimental evidence may be helpful, using specific receptor antagonists or inhibitors of proendothelin- 1 converting enzyme in animal models to determine whether local generation of endothelin contributes to the restoration of vascular tone following cardiac arrest. During cardiac arrest and cardiopulmonary resuscitation, vasopressor agents such as adrenaline are used to increase systemic vascular resistance in order to improve myocardial and cerebral blood flow [27]. As endothelin-1 has potent long lasting vasoconstrictor, pressor and positive inotropic effects [6], high endothelin concentrations may be beneficial during cardiac arrest. Although the correlation of plasma endothelin with outcome may reflect the adverse physiological milieu that occurs during cardiac arrest, it is also possible that the fall in endothelin concentrations in non-survivors contributes to their poor prognosis. 5. Acknowledgement This study was supported by a grant from the Scottish Home and Health Department.
122
W.G. Haynes et al. /Resuscitation
6. References 1 Vane JR, binggird EA, Botting RM. Regulatory functions of the vascular endothelium. N Engl J Med 1990; 323: 27-36. 2 Webb DJ, Cockcroft JR. Circulating and tissue reninangiotensin systems and the role of the endothelium. In: Warren JB, editor. The endothelium: an introduction to current research. New York: Wiley-Liss, 1990; 65-80. 3 Yanagisawa M, Kurihara H, Kimura S et al. A novel potent vasoconstrictor peptide produced by vascular endothelial cells. Nature 1988; 332: 41 I-415. 4 Clarke JG, Benjamin N, Larkin SW, Webb DJ, Davies GJ, Maseri, A. Endothelin is a potent long-lasting vasoconstrictor in men. Am J Physiol 1989; 257: H2033-H2035. 5 Vierhapper H, Wagner 0, Nowotny P, Waldhausl W. Effect of endothelin-1 in man. Circulation 1990; 81: 1415-1418. 6 Haynes WG, Webb DJ. The endothelin family of peptides: local hormones with diverse roles in health and disease? Clin Sci 1993; 84: 485-500. 7 Warrens AN, Cassidy MJD, Takahashi K, Ghatei MA, Bloom SR. Endothelin in renal failure. Nephrol Dial Transplant 1990; 5: 418-422. 8 Webb DJ, Cockcroft JR. Plasma immunoreactive endothelin in uraemia. Lancet 1989; ii: 1211. 9 Toyo-oka T, Aizawa T, Suzuki N et al. Increased plasma level of endothelin-I and coronary spasm induction in patients with vasospastic angina pectoris. Circulation 1991; 83: 476-483. 10 Zamora MR, O’Brien RF, Rutherford RB, Weil JV. Serum endothelin-I concentrations and cold provocation in primary Raynaud’s phenomenon. Lancet 1990; 336: 1144-1147. 11 Cody RJ, Haas GJ, Binkley PF, Capers Q, Kelly R. Plasma endothelin correlates with the extent of pulmonary hypertension in patients with chronic congestive heart failure. Circulation 1992; 85: 504-509. 12 Stewart D, Cernacek P, Costello K, Rouleau JL. Elevated endothelin-I in heart failure and loss of normal response to postural change. Circulation 1992; 85: 510-517. 13 Cernacek P, Stewart D. Immunoreactive endothelin in human plasma: marked elevations in patients with cardiogenie shock. Biochem Biophys Res Commun 1989; 161: 562-567. 14 Yasuda M, Kohno M, Tahara A et al. Circulating
15
16
17
18
19
20
21
22
23
24
25
26
27
27 (1994)
117-122
immunoreactive endothelin in ischaemic heart disease. Am Heart J 1990; 119: 801-806. Weitzberg E, Lundberg JM, Rudehill A. Elevated plasma levels of endothehn in patients with sepsis syndrome. Circ Shock 1991; 33: 222-227. Sanai L, Grant IS, Webb DJ. The role of endothelin-I in the adult respiratory distress syndrome. Proceedings of the 6th European Congress on Intensive Care Medicine, 1992. Little RA, Frayn KN, Randall PE et al. Plasma catecholamines in patients with acute myocardial infarction and cardiac arrest. Q J Med 1985; 54: 133-140. Prasad MR, Jones RM, Kreutzer DL. Release of endothelin from cultured bovine endothelial cells. J Mel Cell Cardiol 1991; 23: 655-658. Kourembanas S, Marsden PA, McQuillan LP, Failer DV. Hypoxia induces endothelin gene expression and secretion in cultured human endothelium. J Clin Invest 1991; 88: 1054-1057. Kohno, M., Murakawa, K., Yasunari, K. et al. Prolonged blood pressure elevation after endothelin administration in bilaterally nephrectomized rats. Metabolism 1989; 38, 712-713. Chamberlain DA. Advanced life support: revised recommendations of the Resuscitation Council (UK). Br Med J 1989; 299: 446-448. Marsden AK. Basic life support: revised recommendations of the Resuscitation Council (UK). Br Med J 1989; 299: 442-445. Florijn KW, Derkx FHM, Visser W et al. Plasma immunoreactive endothelin-I in pregnant women with and without pre-eclampsia. J Cardiovasc Pharmacol 1991; 17 Suppl 7: 5446-5448. Lerman A, Edwards BS, Hallett JW, Heublein DM, Sandberg SM, Burnett JC. Circulating and tissue endothelin immunoreactivity in advanced atherosclerosis. N Engl J Med 1991; 325: 997-1001. Yoshizumi M, Kurihara H, Sugiyama T et al. Haemodynamic shear stress stimulates endothelin production by cultured endothelial cells. Biochem Biophys Res Commun 1989; 161: 859-864. Haynes WG, Davenport AP, Webb DJ. Endothehn: progress in physiology and pharmacology. Trends Pharmacol Sci 1993; 14: 225-228. Lindner KH, Koster R. Vasopressor drugs during cardioresuscitation. Resuscitation 1992; 24: pulmonary 147-153.