Clinical and Experimental Pharmacology and Physiology (1992) 19,733-743
REVIEW BRADYCARDIA DURING REVERSIBLE HYPOVOLAEMIC SHOCK: ASSOCIATED NEURAL REFLEX MECHANISMS AND CLINICAL IMPLICATIONS Niels H. Secher, John Jacobsen, Daniel B. Friedman and Steen Matzen Department of Anaesthesia, Rigshospitalet 2034, University of Copenhagen, Copenhagen, Denmark (Received I 9 May 1992)
SUMMARY 1. Heart rate response to reversible central hypovolaemia can be divided into three stages. In the first stage (corresponding to a reduction of the blood volume by approximately 15%) a modest increase in heart rate ( < 100 beats/ min) and total peripheral resistance compensate for the blood loss, and a near normal arterial blood pressure prevails (preshock). During the second stage, a reduction of the central blood volume by approximately 30% results in a decrease in heart rate, total peripheral resistance and blood pressure due to activation of unmyelinated vagal afferents (C-fibres) from the left ventricle. In the third stage, blood pressure falls further as haemorrhage continues and tachycardia (>120 beats/min) is manifest. This stage may proceed into irreversible shock with death from cardiac arrest probably related to the formation of free oxygen radicals. 2. Recognition of the vasodepressor-cardioinhibitory reaction to a reduced circulating blood volume is important and suggests the need for immediate treatment with volume expansion in critically ill patients.
Key words: Bainbridge reflex, baroreceptors, Bezold-Jarisch reflex, blood pressure, brain ischaemia, cardiopulmonary receptors, haemorrhage, heart rate, vagal C-fibres, vasovagal syncope.
INTRODUCTION The monitoring of heart rate (HR) and blood pressure in patients who are critically ill or undergoing surgery has a long tradition (Cushing 1903). Despite this vast clinical experience, the HR response to hypovolaemia
remains incompletely understood. In textbook descriptions of hypovolaemic shock the characteristic deviation of HR is stated to be tachycardia, and bradycardia is mentioned only when shock is irre-
Correspondence: Niels H. Secher, Department of Anaesthesia, Rigshospitalet 2034, University of Copenhagen, Blegdamsvej 9, DK-2100 Copenhagen 8,Denmark. Present address for D. B. Friedman: Harry S. Moss Heart Center, University of Texas Southwestern Medical Center, Harry Hines Blvd. 75235-9034, Dallas, TX 5323, USA. Present address for J. Jacobsen: Department of Anaesthesia, Herlev Hospital, Herlev Ringvej 75, DK-2730 Herlev, Denmark. Present address for S. Matzen: Department of Surgery RT, Rigshospitalet 2102, Blegdamsvej 9, DK-2100 Copenhagen $3, Denmark.
734
N. H. Secher et al.
versible. Likewise, when the cardiovascular responses to procedures performed in order to reduce central blood volume have been reviewed, the hypotensive bradycardic phase has not been included (Mark & Mancia 1983; Rowel1 1986). However, this conclusion is reached only by excluding the simultaneous decrease in HR and mean arterial pressure (MAP) that takes place with a reduction of the central blood volume by approximately 30% (Secher & Bie 1985). When attempts have been made to define a role for bradycardia during hypovolaemic shock (Sjostrand 1976; Bishop et al. 1983; Abboud 1989; van Leeuwen et al. 1989; Schadt & Ludbrook 1991), tachycardia is considered to be the normal response while bradycardia is confined to severe haemorrhage. Confining bradycardia to severe haemorrhage introduces a discrepancy between experimental findings and the wellestablished clinical experience that patients in severe shock are tachycardic. A concomitant decrease in HR and MAP is mentioned in textbooks under headings such as ‘vasovagal syncope’ or ‘fainting’, reflecting how the early experience was described (Edholm 1952) without considering that hypovolaemic shock may be among the most clinically relevant causes for the reaction. As a result, relative bradycardia during haemorrhage is not recognized by the clinician, and bleeding patients with abnormal or low HR are often examined for heart disease, myxoedema or head trauma before even overt haemorrhage is diagnosed and accepted as causal. Most importantly, the diagnosis of haemorrhage may be delayed and the irreversible stage of hypovolaemic shock entered beforeuppropriate measures are taken to stop the bleeding and institute fluid therapy. The purpose of this paper is to review the HR responses to a reduced central blood volume emphasizing three well defined stages: normotension with moderate tachycardia, moderate hypotension with bradycardia, and a stage with severe hypotension and manifest tachycardia. Further, an attempt is made to explain the HR responses to haemorrhage by a model that includes neural reflex mechanisms.
STAGES OF HYPOVOLAEMIC SHOCK Studies involving haemorrhage in the rat (Sjostrand 1976; Skoog et al. 1985; Darlington et al. 1986; Haggendal 1986; Morgan et al. 1988; Victor et al. 1989), cat (Oberg & White 1970a; Oberg & Thordn 1972a; Elam et al. 1985), dog (Meek & Eyster 1921; Ebert et al. 1962; Morita & Vatner 1985), rabbit (Morita et al. 1988) and pig (Jacobsen et al. 1990) have shown that the initial slight increase in HR is
most often followed by a reduction in HR and then tachycardia as hypotension becomes more severe or prolonged (SjBstrand 1973; Horton et al. 1984; Haggendal 1986; Koyama et al. 1988) with the transition to the irreversible stage of shock (Jacobsen et al. 1990). An exception is rats of the Swedish germ-free strain, which always show tachycardia (Castenfors & Sjostrand 1972). In humans, several methods have been employed to induce a transient decrease in the central blood volume of sufficient magnitude to elicit presyncopal symptoms (nausea, light-headedness, heat, paleness, sweating) with associated bradycardia and hypotension. The most commonly applied models include lower body negative pressure (Epstein et al. 1968; Murray et al. 1968; Baylis et al. 1978; Sander-Jensen et al. 1988), head-up tilt (reverse Trendelenburgposition; Asmussen et al. 1938; Brigden et al. 1950; Epstein et al. 1968; Davies et al. 1976; Bergenwald et al. 1977; Secher et al. 1984; Sander-Jensen et al. 1986a; Matzen et al. 1989, 1990, 1991, 1992), positive pressure breathing (Ernsting 1966), venous congestion of the legs with or without venosection (Barcroft et al. 1944; Barcroft & Edholm 1945; Warren et al. 1945; Bearn et al. 1951; Sander-Jensen et al. 1987), spinal or epidural anaesthesia and the lordotic posture (Brigden Howarth & Sharpey-Schafer 1950; Arndt et al. 1985; SanderJensen et al. 1989; Jacobsen et al. 1992). HR seldom exceeds 100 beatslmin during the initial response to these manoeuvres and MAP may be elevated. During presyncope, HR and MAP decrease. The typical finding is that HR decreases toward the resting value at the time when the experiment has to be terminated. However, if termination of the intervention is delayed, extreme bradycardia (e.g. 1-2 beatslmin) may occur (Sander-Jensen et al. 1986a). A low HR has also been demonstrated in volunteers after bleeding approximately 1 L (Ebert et al. 1941; Wallace & Sharpey-Schafer 1941; Barcroft et al. 1944; Shenkin et al. 1944; Barcroft & Edholm 1945; Price et al. 1966; Bergenwald et al. 1975, 1977), in patients with a blood loss of approximately 2 L (Hoffman 1972; Hyun et al. 1972; Jansen 1978; Secher et al. 1984; Sander-Jensen et al. 1986b; Barriot & Riou 1987), and even during haemorrhage in anaesthetized patients (Rwsgaard & Secher 1986). The finding that patients appear to have a larger blood loss than volunteers at the time when a low HR appears reflects the fact that fluid therapy, albeit inadequate, delays the response in the patients. Patients in hypovolaemic shock may also present with tachycardia following the bradycardia stage, and it is then usually associated with a lower MAP (Hoffman 1972). Table 1 presents a comparison of 34 consecutive bleeding, hypotensive
735
Bradycardia during haemorrhage patients with either a low ( < 100 beats/ min) or a high (> I00 beats/min) HR (Jacobsen & Secher 1992). The patients presenting with tachycardia had approximately a 60% larger blood loss than those with a low HR. They also had a lower MAP. All patients who had a low HR when haemorrhage was recognized recovered. In the tachycardic group, six of 21 patients Stage o f reversible shock I II 111
125
250
died, either during operation (three patients) or during intensive care treatment due to multiple organ failure. These findings support the hypothesis (Fig. 1) that tachycardia during hypovolaemic shock represents a more severe degree of shock than bradycardia in accordance with Atkinson et al. (1987): ‘what was formerly known as primary shock is now termed “vasovagal collapse”, which is characterized by a slow pulse, unlike the tachycardia associated with true (secondary) shock’. Thus, as with the experience gained from studies in animals, tachycardia (>100 beats/min) during haemorrhage in humans may represent a transition to an irreversible stage of shock that is possibly related to oxygen radical-induced lipid peroxidation (Hall 1992). In terms of the stages of central volume depletion, it may be relevant to include the first circulatory changes observed after the withdrawal of approximately 500-700 mL of blood that results in moderate tachycardia and peripheral vasoconstriction. This first stage could be considered ‘preshock‘ as the patient or subject has a near normal MAP and no symptoms. The second stage represents the simultaneous decrease in HR and MAP, and manifest hypotension is associated with tachycardia in the third stage. That the three stages follow each other during progressive depletion of central blood volume is supported by the finding that the succession may be reversed by volume expansion. The hypotensive and bradycardic phase (second stage) reverts to normotension with a moderately elevated HR before the resting HR is reached during expansion of the circulating blood volume (Barcroft & Edholm 1945; Bearn et al. 1951; Edholm 1952; Epstein et al. 1968; Bergenwald et al.
1 0 1 0 2 0 3 0 4 Q 5 0 6 0
Blood IOSS
(%)
AFFERENT N E R V E ACTIVITY Cardiopulmonary Myelinated Unrnyelinated Arterial Baroreceptors Brain ischaemia
Fig. 1. Model for heart rate (beats/min) changes taking place during haemorrhage dividing the responses into three stages. In the first stage (preshock) an increase in H R ( 0 )i s associated with a normal or even increasing arterial blood pressure (mmHg). A small decrease in M A P (0)together with a further increase in H R appear with the onset of vague presyncopal symptoms before a decrease in HR takes place and the patients become ill (second stage). In the third stage, severe hypotension is associated with tachycardia before the shock becomes irreversible, which may be related to brain ischaemia. During volume loading the HR responses in the first and second stages of shock are reversed (i.e. volume expansion results in an increase in HR before the resting value is re-established). If resuscitation from the third stage of shock is not immediate, HR is maintained above 100 beats/min. Also shown are the proposed nerve activities in myelinated and unmyelinated cardiopulmonary afferents, afferents from arterial baroreceptors, and from brain ischaemia regulating the change in heart rate during haemorrhage. Receptor-type domination is indicated by a double line. With continuous activity in unmyelinated cardiac afferents, their H R and MAP-lowering effect escapes. However, unmyelinated cardiac afferents may be of importance for maintaining peripheral circulation in the third stage of shock.
Table 1. Clinical characteristics (median with range) for 34 consecutive patients in circulatory shock due to haemorrhage with either a low ( < l o 0 beats/min; n = 13) or a high (> 100 beats/ min; n = 21) heart rate when the diagnosis was established Heart rate (beatsimin)
79
(60-95)
Age (years) 55 (23-81) Blood pressure 60 (43-73) (mmHg) Blood loss (L) 2.3 (1.5-3.7) Volume replacement Blood products (L) 2.0 (0.9-2.5) Crystalloids (L) 2.7 (1.0-4.1) Survived 131 13
129
(110-160’)
53
(26-92)
48 (0-70)* 3.6 (2.0-4.6)** 3.4 (1.0-4.5)** 2.5 (0.5-4.5) 15/2It
*P
REVIEW BRADYCARDIA DURING REVERSIBLE HYPOVOLAEMIC SHOCK: ASSOCIATED NEURAL REFLEX MECHANISMS AND CLINICAL IMPLICATIONS Niels H. Secher, John Jacobsen, Daniel B. Friedman and Steen Matzen Department of Anaesthesia, Rigshospitalet 2034, University of Copenhagen, Copenhagen, Denmark (Received I 9 May 1992)
SUMMARY 1. Heart rate response to reversible central hypovolaemia can be divided into three stages. In the first stage (corresponding to a reduction of the blood volume by approximately 15%) a modest increase in heart rate ( < 100 beats/ min) and total peripheral resistance compensate for the blood loss, and a near normal arterial blood pressure prevails (preshock). During the second stage, a reduction of the central blood volume by approximately 30% results in a decrease in heart rate, total peripheral resistance and blood pressure due to activation of unmyelinated vagal afferents (C-fibres) from the left ventricle. In the third stage, blood pressure falls further as haemorrhage continues and tachycardia (>120 beats/min) is manifest. This stage may proceed into irreversible shock with death from cardiac arrest probably related to the formation of free oxygen radicals. 2. Recognition of the vasodepressor-cardioinhibitory reaction to a reduced circulating blood volume is important and suggests the need for immediate treatment with volume expansion in critically ill patients.
Key words: Bainbridge reflex, baroreceptors, Bezold-Jarisch reflex, blood pressure, brain ischaemia, cardiopulmonary receptors, haemorrhage, heart rate, vagal C-fibres, vasovagal syncope.
INTRODUCTION The monitoring of heart rate (HR) and blood pressure in patients who are critically ill or undergoing surgery has a long tradition (Cushing 1903). Despite this vast clinical experience, the HR response to hypovolaemia
remains incompletely understood. In textbook descriptions of hypovolaemic shock the characteristic deviation of HR is stated to be tachycardia, and bradycardia is mentioned only when shock is irre-
Correspondence: Niels H. Secher, Department of Anaesthesia, Rigshospitalet 2034, University of Copenhagen, Blegdamsvej 9, DK-2100 Copenhagen 8,Denmark. Present address for D. B. Friedman: Harry S. Moss Heart Center, University of Texas Southwestern Medical Center, Harry Hines Blvd. 75235-9034, Dallas, TX 5323, USA. Present address for J. Jacobsen: Department of Anaesthesia, Herlev Hospital, Herlev Ringvej 75, DK-2730 Herlev, Denmark. Present address for S. Matzen: Department of Surgery RT, Rigshospitalet 2102, Blegdamsvej 9, DK-2100 Copenhagen $3, Denmark.
734
N. H. Secher et al.
versible. Likewise, when the cardiovascular responses to procedures performed in order to reduce central blood volume have been reviewed, the hypotensive bradycardic phase has not been included (Mark & Mancia 1983; Rowel1 1986). However, this conclusion is reached only by excluding the simultaneous decrease in HR and mean arterial pressure (MAP) that takes place with a reduction of the central blood volume by approximately 30% (Secher & Bie 1985). When attempts have been made to define a role for bradycardia during hypovolaemic shock (Sjostrand 1976; Bishop et al. 1983; Abboud 1989; van Leeuwen et al. 1989; Schadt & Ludbrook 1991), tachycardia is considered to be the normal response while bradycardia is confined to severe haemorrhage. Confining bradycardia to severe haemorrhage introduces a discrepancy between experimental findings and the wellestablished clinical experience that patients in severe shock are tachycardic. A concomitant decrease in HR and MAP is mentioned in textbooks under headings such as ‘vasovagal syncope’ or ‘fainting’, reflecting how the early experience was described (Edholm 1952) without considering that hypovolaemic shock may be among the most clinically relevant causes for the reaction. As a result, relative bradycardia during haemorrhage is not recognized by the clinician, and bleeding patients with abnormal or low HR are often examined for heart disease, myxoedema or head trauma before even overt haemorrhage is diagnosed and accepted as causal. Most importantly, the diagnosis of haemorrhage may be delayed and the irreversible stage of hypovolaemic shock entered beforeuppropriate measures are taken to stop the bleeding and institute fluid therapy. The purpose of this paper is to review the HR responses to a reduced central blood volume emphasizing three well defined stages: normotension with moderate tachycardia, moderate hypotension with bradycardia, and a stage with severe hypotension and manifest tachycardia. Further, an attempt is made to explain the HR responses to haemorrhage by a model that includes neural reflex mechanisms.
STAGES OF HYPOVOLAEMIC SHOCK Studies involving haemorrhage in the rat (Sjostrand 1976; Skoog et al. 1985; Darlington et al. 1986; Haggendal 1986; Morgan et al. 1988; Victor et al. 1989), cat (Oberg & White 1970a; Oberg & Thordn 1972a; Elam et al. 1985), dog (Meek & Eyster 1921; Ebert et al. 1962; Morita & Vatner 1985), rabbit (Morita et al. 1988) and pig (Jacobsen et al. 1990) have shown that the initial slight increase in HR is
most often followed by a reduction in HR and then tachycardia as hypotension becomes more severe or prolonged (SjBstrand 1973; Horton et al. 1984; Haggendal 1986; Koyama et al. 1988) with the transition to the irreversible stage of shock (Jacobsen et al. 1990). An exception is rats of the Swedish germ-free strain, which always show tachycardia (Castenfors & Sjostrand 1972). In humans, several methods have been employed to induce a transient decrease in the central blood volume of sufficient magnitude to elicit presyncopal symptoms (nausea, light-headedness, heat, paleness, sweating) with associated bradycardia and hypotension. The most commonly applied models include lower body negative pressure (Epstein et al. 1968; Murray et al. 1968; Baylis et al. 1978; Sander-Jensen et al. 1988), head-up tilt (reverse Trendelenburgposition; Asmussen et al. 1938; Brigden et al. 1950; Epstein et al. 1968; Davies et al. 1976; Bergenwald et al. 1977; Secher et al. 1984; Sander-Jensen et al. 1986a; Matzen et al. 1989, 1990, 1991, 1992), positive pressure breathing (Ernsting 1966), venous congestion of the legs with or without venosection (Barcroft et al. 1944; Barcroft & Edholm 1945; Warren et al. 1945; Bearn et al. 1951; Sander-Jensen et al. 1987), spinal or epidural anaesthesia and the lordotic posture (Brigden Howarth & Sharpey-Schafer 1950; Arndt et al. 1985; SanderJensen et al. 1989; Jacobsen et al. 1992). HR seldom exceeds 100 beatslmin during the initial response to these manoeuvres and MAP may be elevated. During presyncope, HR and MAP decrease. The typical finding is that HR decreases toward the resting value at the time when the experiment has to be terminated. However, if termination of the intervention is delayed, extreme bradycardia (e.g. 1-2 beatslmin) may occur (Sander-Jensen et al. 1986a). A low HR has also been demonstrated in volunteers after bleeding approximately 1 L (Ebert et al. 1941; Wallace & Sharpey-Schafer 1941; Barcroft et al. 1944; Shenkin et al. 1944; Barcroft & Edholm 1945; Price et al. 1966; Bergenwald et al. 1975, 1977), in patients with a blood loss of approximately 2 L (Hoffman 1972; Hyun et al. 1972; Jansen 1978; Secher et al. 1984; Sander-Jensen et al. 1986b; Barriot & Riou 1987), and even during haemorrhage in anaesthetized patients (Rwsgaard & Secher 1986). The finding that patients appear to have a larger blood loss than volunteers at the time when a low HR appears reflects the fact that fluid therapy, albeit inadequate, delays the response in the patients. Patients in hypovolaemic shock may also present with tachycardia following the bradycardia stage, and it is then usually associated with a lower MAP (Hoffman 1972). Table 1 presents a comparison of 34 consecutive bleeding, hypotensive
735
Bradycardia during haemorrhage patients with either a low ( < 100 beats/ min) or a high (> I00 beats/min) HR (Jacobsen & Secher 1992). The patients presenting with tachycardia had approximately a 60% larger blood loss than those with a low HR. They also had a lower MAP. All patients who had a low HR when haemorrhage was recognized recovered. In the tachycardic group, six of 21 patients Stage o f reversible shock I II 111
125
250
died, either during operation (three patients) or during intensive care treatment due to multiple organ failure. These findings support the hypothesis (Fig. 1) that tachycardia during hypovolaemic shock represents a more severe degree of shock than bradycardia in accordance with Atkinson et al. (1987): ‘what was formerly known as primary shock is now termed “vasovagal collapse”, which is characterized by a slow pulse, unlike the tachycardia associated with true (secondary) shock’. Thus, as with the experience gained from studies in animals, tachycardia (>100 beats/min) during haemorrhage in humans may represent a transition to an irreversible stage of shock that is possibly related to oxygen radical-induced lipid peroxidation (Hall 1992). In terms of the stages of central volume depletion, it may be relevant to include the first circulatory changes observed after the withdrawal of approximately 500-700 mL of blood that results in moderate tachycardia and peripheral vasoconstriction. This first stage could be considered ‘preshock‘ as the patient or subject has a near normal MAP and no symptoms. The second stage represents the simultaneous decrease in HR and MAP, and manifest hypotension is associated with tachycardia in the third stage. That the three stages follow each other during progressive depletion of central blood volume is supported by the finding that the succession may be reversed by volume expansion. The hypotensive and bradycardic phase (second stage) reverts to normotension with a moderately elevated HR before the resting HR is reached during expansion of the circulating blood volume (Barcroft & Edholm 1945; Bearn et al. 1951; Edholm 1952; Epstein et al. 1968; Bergenwald et al.
1 0 1 0 2 0 3 0 4 Q 5 0 6 0
Blood IOSS
(%)
AFFERENT N E R V E ACTIVITY Cardiopulmonary Myelinated Unrnyelinated Arterial Baroreceptors Brain ischaemia
Fig. 1. Model for heart rate (beats/min) changes taking place during haemorrhage dividing the responses into three stages. In the first stage (preshock) an increase in H R ( 0 )i s associated with a normal or even increasing arterial blood pressure (mmHg). A small decrease in M A P (0)together with a further increase in H R appear with the onset of vague presyncopal symptoms before a decrease in HR takes place and the patients become ill (second stage). In the third stage, severe hypotension is associated with tachycardia before the shock becomes irreversible, which may be related to brain ischaemia. During volume loading the HR responses in the first and second stages of shock are reversed (i.e. volume expansion results in an increase in HR before the resting value is re-established). If resuscitation from the third stage of shock is not immediate, HR is maintained above 100 beats/min. Also shown are the proposed nerve activities in myelinated and unmyelinated cardiopulmonary afferents, afferents from arterial baroreceptors, and from brain ischaemia regulating the change in heart rate during haemorrhage. Receptor-type domination is indicated by a double line. With continuous activity in unmyelinated cardiac afferents, their H R and MAP-lowering effect escapes. However, unmyelinated cardiac afferents may be of importance for maintaining peripheral circulation in the third stage of shock.
Table 1. Clinical characteristics (median with range) for 34 consecutive patients in circulatory shock due to haemorrhage with either a low ( < l o 0 beats/min; n = 13) or a high (> 100 beats/ min; n = 21) heart rate when the diagnosis was established Heart rate (beatsimin)
79
(60-95)
Age (years) 55 (23-81) Blood pressure 60 (43-73) (mmHg) Blood loss (L) 2.3 (1.5-3.7) Volume replacement Blood products (L) 2.0 (0.9-2.5) Crystalloids (L) 2.7 (1.0-4.1) Survived 131 13
129
(110-160’)
53
(26-92)
48 (0-70)* 3.6 (2.0-4.6)** 3.4 (1.0-4.5)** 2.5 (0.5-4.5) 15/2It
*P
