SUMMARY1. In unanaesthetized rabbits, haemorrhage was simulated by inflating a cuff placed round the inferior vena cava so that cardiac output fell at a constant rate of 8 % of its resting value per minute. The circulatory responses were measured after injections into the fourth ventricle of saline vehicle, selective opioid antagonists, selective opioid agonists, and agonist-antagonist mixtures. Three sets of experiments were done to determine if a specific subtype of opiate receptor within the central nervous system is responsible for the circulatory decompensation that occurs during simulated haemorrhage.2. In six rabbits the effects of ascending doses of the antagonists naloxone (,uZselective), Mr 2266 (K-and ,t-selective), ICI 174864 (6-selective) and nor-binaltorphimine (K-selective) were tested. In three rabbits the effects of the antagonist naloxone, the agonists HTyr-D-Ala-Gly-MePhe-NH(CHi)2OH (DAGO, ,u-selective), U 50488H (K-selective), and [D-Pen2, D-Pen5]-enkephalin (DPDPE, 6-selective), and combinations of these agonists with naloxone were tested. In four rabbits the doserelated effects of DAGO on respiratory, as well as circulatory, functions were examined.3. After injecting saline vehicle, the circulatory response to simulated haemorrhage had two phases. During the first phase, systemic vascular conductance fell, heart rate rose, and mean arterial pressure fell by only -10 mmHg. A second, decompensatory, phase began when cardiac output had fallen to -50 % of its resting level. At this point, there was an abrupt rise in systemic vascular conductance and a fall in mean arterial pressure to < 40 mmHg.4. The lower range of doses of naloxone (3-30 nmol), Mr 2266 (10-100 nmol), ICI 174864 (10-30 nmol), and all doses of nor-binaltorphimine (1-100 nmol), were without effect on the circulatory response to simulated haemorrhage. Higher doses of naloxone (30-100 nmol), Mr 2266 (100-300 nmol) and ICI 174864 (30-100 nmol) abolished the decompensatory phase. The relative order of antagonist potency was ICI 174864 > naloxone > Mr 2266 > nor-binaltorphimine.5. In the second set of experiments, the critical dose of naloxone necessary to prevent circulatory decompensation during simulated haemorrhage was 30-150 1.S 74: 16 R. G. EVANS, J. LUDBROOK AND A. F. VANLEEUWEN nmol. The d-agonist DPDPE (50 nmol) did not affect the haemodynamic response to simulated haemorrhage, but it did block the effect of naloxone on the response. The u-agonist DAGO (30 nmol) and the K-agonist U 50488H (300 nmol) prevented circulatory decompensation during simulated haemorrhage, whether given alone or in combination with the critical dose of naloxone, but DAGO (30 mmol) also raised arterial pressure, lowered systemic vascular conductance and slowed respiration.6. In the third set of experiments we found that lower doses of DAGO (100-300 pmol) would also prevent circulatory decompensation during simulated haemorrhage. These doses did not affect resting arterial pressure or systemic vascular conductance, and had insignificant effects on res...

1. We have made a within-rabbit comparison of the effects of four general anaesthetic regimens on the haemodynamic response to acute reduction in central blood volume and on baroreflex control of heart rate. 2. Acute haemorrhage was simulated by gradually inflating a cuff on the inferior vena cava in order to cause cardiac output to fall at a constant rate of 8.5%/min while the responses of systemic vascular resistance, arterial pressure and heart rate were measured. The full range of the baroreceptor-heart rate reflex was elicited by inflating aortic and vena caval cuffs. These indices of circulatory control were repeatedly measured within five protocols, to which each rabbit was exposed in randomized order. 3. In each protocol the rabbit was first studied unanaesthetized. Then a small dose of thiopentone sodium was given (16 mg/kg). In the four main protocols the rabbit was then intubated and ventilated, first with 100% oxygen and then with 50% nitrous oxide, during administration of one of four anaesthetic agents. These were halothane (2.0 and 2.5%), ketamine (2.5 mg/kg per min), propofol (0.83 and 1.25 mg/kg per min) and alfentanil (2.5 and 3.33 micrograms/kg per min). In a sham protocol the effects of 100% oxygen, then those of 50 and 75% nitrous oxide, were studied while the rabbit remained conscious. 4. In unanaesthetized rabbits, in the presence or absence of nitrous oxide, the normal biphasic haemodynamic response to simulated haemorrhage occurred. The first, vasoconstrictor, phase was attenuated by halothane, ketamine and propofol, so that arterial pressure fell more steeply than normal. Not only was the vasoconstrictor phase unaffected by alfentanil but it was extended, so that arterial pressure remained at a normal level even when cardiac output had fallen by 59%. This effect of alfentanil appeared to be mediated centrally, since it could be reproduced by injecting small doses (1.5-7.5 micrograms) into the fourth ventricle. All four anaesthetic agents and nitrous oxide attenuated the baroreceptor control of heart rate. The effect was least with nitrous oxide and alfentanil, greatest with halothane.

Information has come forward recently from several sources which provides new insights into the mechanisms that underlie the haemodynamic responses to acute blood loss. In unanaesthetised animals and human volunteers there are two distinct phases to these responses. At first, the engagement of baroreflexes results in a progressive rise in sympathetic vasoconstrictor drive and peripheral resistance, and the maintenance of arterial blood pressure at a near-normal level. When about one-third of blood volume has been lost, reflex sympathetic drive is switched off, and peripheral resistance and blood pressure fall abruptly to low levels despite a burst of vasopressin release. Research in conscious animals has now shown that the onset of this decompensatory phase is triggered by a signal from the heart, which activates an endogenous opioid mechanism in the brain. Activation of this mechanism can be prevented by administering a selective δ-receptor antagonist, or selective μ-receptor agonists (including alfentanil). It has not yet been established that this endogenous opioid mechanism is responsible for the decompensatory phase of acute blood loss in man, nor that it can be prevented or reversed by selective opioid agonists or antagonists.

We have studied in eight rabbits the cardiovascular effects of midazolam, propofol and alfentanil with graded hypoxia. Central blood volume was reduced progressively by gradual inflation of a thoracic vena cava cuff so that cardiac index (CI) decreased at a constant rate. Under control conditions the haemodynamic response was biphasic. During phase I, mean arterial pressure (MAP) was maintained by a progressive decrease in systemic vascular conductance (SVCI). When CI had declined to a critical level, phase II occurred with an abrupt increase in SVCI and decrease in MAP. Phase I was prolonged by hypoxia, alfentanil and midazolam, but the effects were not additive. Phase I was shortened by propofol and this effect increased with hypoxia. The gradient of the SVCI response in phase I was also reduced by propofol > midazolam, but not by alfentanil. The occurrence of phase II was less frequent during alfentanil infusion than midazolam and propofol with all of the inspired gas mixtures. Thus the opioid was protective against circulatory collapse with hypoxia and simulated hypovalaemia.