Cognitive decline in the elderly after surgery and anaesthesia: results from the Oxford Project to Investigate Memory and Ageing (OPTIMA) cohort

Concerns have been raised about the effects on cognition of anaesthesia for surgery, especially in elderly people. We recorded cognitive decline in a cohort of 394 people (198 women) with median (IQR) age at recruitment of 72.6 (66.6–77.8) years, of whom 109 had moderate or major surgery during a median (IQR) follow‐up of 4.1 (2.0–7.6) years. Cognitive decline was more rapid in people who on recruitment were: older, p = 0.0003; male, p = 0.027; had worse cognition, p < 0.0001; or carried the ε4 allele of apoliprotein E (APOEε4), p = 0.008; and after an operation if cognitive impairment was already diagnosed, p = 0.0001. Cognitive decline appears to accelerate after surgery in elderly patients diagnosed with cognitive impairment, but not other elderly patients.

☛ CPD available at http://www.learnataagbi.org

The first intravenous anaesthetic: how well was it managed and its potential realized?

Our speciality commonly traces its origin to a demonstration of the inhalation of ether by a patient undergoing surgery in Boston in 1846. Less well known is the demonstration of the i.v. injection of opium with alcohol into a dog in Oxford in 1656, leading to anaesthesia followed by full long-term recovery. After gaining i.v. access, a mixture of opium and alcohol was injected, resulting in a brief period of anaesthesia. After a period during which the dog was kept moving to assist recovery, a full recovery was made. Details from this momentous experiment allow us to compare the technique used with modern management. It is important to consider why there was a failure to translate the results into clinical practice and nearly 200 yr of potentially pain-free surgery. Possible factors include lack of equipment for i.v. access, lack of understanding of dose-response effects, and a climate of scientific discovery rather than clinical application. Given the current interest in total i.v. anaesthesia, it seems appropriate to identify its origins well before those of inhalation anaesthesia.

Human pulmonary vascular responses to hypoxia and hypercapnia

The ability of alveolar gas composition to influence pulmonary vascular tone has been appreciated for over 50 years. In particular, it has been proposed that both O2 and CO2 could play a role in the matching of perfusion to ventilation within the lung, improving the overall efficiency of gas exchange. A wide variety of experimental approaches has been used to investigate pulmonary vascular effects of the respiratory gases in a range of mammalian species. In this article, we review experiments performed in healthy humans, identify particular difficulties in the interpretation of such experiments, and discuss possible approaches to future study.

Release by hypoxia of a soluble vasoconstrictor from rabbit small pulmonary arteries

BACKGROUND

Soluble pulmonary vasoconstrictors released in response to hypoxia have been reported in pig and rat preparations, but not in rabbit preparations.

METHODS

We used myography to evaluate the contribution of a soluble factor to constriction in rabbit small pulmonary arteries (external diameter 300-475 microm) exposed to 45 min hypoxia (PO(2)=9 mm Hg).

RESULTS

Hypoxia produced gradually intensifying constriction. Return to euoxia (PO(2)=145 mm Hg) for 30 min relaxed only approximately 30% of the constriction, whereas elution of the myograph bath yielded full relaxation. Reapplication of the eluent gradually restored the constriction to its pre-elution level over a 30-min period.

CONCLUSIONS

In this closed system, a soluble factor contributes substantially to hypoxic pulmonary vasoconstriction.

Inhibition of active sodium absorption leads to a net liquid secretion into in vivo rabbit lung at two levels of alveolar hypoxia

Active sodium transport across alveolar epithelium is known to contribute to the resolution of pulmonary oedema. We have attempted to assess whether sodium transport is essential to prevent liquid accumulation in healthy pulmonary alveoli exposed to mild hypoxia, and whether its contribution to liquid absorption differs between mild and moderate levels of hypoxia. In twenty-four anaesthetized adult rabbits we used direct bronchial cannulation to measure liquid movement from the liquid-filled left lung over 3.5 h. Half of the rabbits were studied at a level of mixed venous (and alveolar) oxygen partial pressure, PVO2, of 6.5 kPa and half at 4.5 kPa. PVO2 was altered by changing the inspired oxygen fraction in the ventilated right lung. Alveolar hydrostatic pressure was 0.3 kPa. In each group of 12, six animals with inhibitors of sodium transport in the isosmotic instillate were compared with six controls. We have shown an alveolar liquid secretion (approximately 0.6 microl min(-1) (kg body weight)(-1)) in the presence of inhibitors of active transport and an absorption (approximately 4 microl min(-1) (kg body weight)(-1)) in controls. Changing PVO2 had no influence on these movements. We conclude that, in this model of pulmonary oedema, active sodium transport appears to be essential for prevention of alveolar liquid accumulation via secretion. Furthermore, the contribution of active sodium transport to liquid absorption remains constant at oxygen tensions between 4.5 and 6.5 kPa.

Effects of 8 h of isocapnic hypoxia with and without muscarinic blockade on ventilation and heart rate in humans

This study examined the role of muscarinic parasympathetic mechanisms in generating the progressive increases in ventilation (V(E)) and heart rate previously reported with 8 h exposures to hypoxia. The sensitivities of V(E) (G(p)) and heart rate (G(HR)) to acute variations in hypoxia, and V(E) and heart rate during acute hyperoxia were assessed in 10 subjects before and after two 8 h exposures to isocapnic hypoxia (end-tidal P(O2) = 50 mmHg). The responses were measured during muscarinic blockade with glycopyrrolate (0.015 mg kg(-1)) and without glycopyrrolate, as a control. There were significant increases in G(p) (P < 0.01) and V(E) during hyperoxia (P < 0.01) following hypoxic exposure, but these were unaffected by glycopyrrolate. G(HR) increased significantly by 0.29 +/- 0.08 beats min(-1) %(-1) (mean +/- S.E.M.) following exposure to hypoxia under control conditions, but only non-significantly by 0.10 +/- 0.08 beats min(-1) %(-1) with glycopyrrolate. This difference was significant. Changes in heart rate during hyperoxia were slight and inconclusive. We conclude that muscarinic mechanisms play little role in the progressive ventilatory changes that occur over 8 h of hypoxia, but that they do mediate much of the progressive increase in heart rate. Experimental Physiology (2001) 86.4, 529-538.

Respiratory control in humans after 8 h of lowered arterial PO2, hemodilution, or carboxyhemoglobinemia

In humans exposed to 8 h of isocapnic hypoxia, there is a progressive increase in ventilation that is associated with an increase in the ventilatory sensitivity to acute hypoxia. To determine the relative roles of lowered arterial PO2 and oxygen content in generating these changes, the acute hypoxic ventilatory response was determined in 11 subjects after four 8-h exposures: 1) protocol IH (isocapnic hypoxia), in which end-tidal PO2 was held at 55 Torr and end-tidal PCO2 was maintained at the preexposure value; 2) protocol PB (phlebotomy), in which 500 ml of venous blood were withdrawn; 3) protocol CO, in which carboxyhemoglobin was maintained at 10% by controlled carbon monoxide inhalation; and 4) protocol C as a control. Both hypoxic sensitivity and ventilation in the absence of hypoxia increased significantly after protocol IH (P < 0.001 and P < 0.005, respectively, ANOVA) but not after the other three protocols. This indicates that it is the reduction in arterial PO2 that is primarily important in generating the increase in the acute hypoxic ventilatory response in prolonged hypoxia. The associated reduction in arterial oxygen content is unlikely to play an important role.

Inhibition of nitric oxide synthesis augments pulmonary oedema in isolated perfused rabbit lung

The role of nitric oxide (NO) in precipitating pulmonary oedema in acute lung injury remains unclear. We have investigated the mechanism of involvement of NO in the maintenance of liquid balance in the isolated rabbit lung. Thirty pairs of lungs were perfused with colloid for up to 6 h, during which pulmonary vascular resistance (PVR) and capillary pressure (PCP) were measured frequently, and time to gain 5 g in weight (t5) was recorded. Four protocols with different perfusate additives were studied: (i) none (control, n = 11); (ii) 10 mmol NG-nitro-L-arginine methyl ester (L-NAME) (n = 6); (iii) 10 mmol L-NAME with 100 mumol lodoxamide, an inhibitor of mast cell degranulation (n = 7); (iv) 10 mmol L-NAME with 10 mumol 8-bromo-3',5'-cyclic guanosine monophosphate (8Br-cGMP), an analogue of cGMP that may reduce vascular permeability by relaxing contractile elements in endothelial cells (n = 6). Neither PVR nor PCP differed between protocols. L-NAME markedly reduced t5 from 248 (27) min (mean (SEM)) in protocol (i) to 144 (5) min in protocol (ii) (P < 0.05). Both lodoxamide (t5 = 178 (7) min) and 8Br-cGMP (t5 = 204 (10) min) substantially corrected the effect of L-NAME (P < 0.005). Results suggest that maintenance of a low permeability by NO may involve mast cell stabilization and endothelial cell relaxation.

Cardiovascular effects of 8 h of isocapnic hypoxia with and without beta-blockade in humans

This study seeks to confirm the progressive changes in cardiac output and heart rate previously reported with 8 h exposures to constant hypoxia, and to examine the role of sympathetic mechanisms in generating these changes. Responses of ten subjects to four 8 h protocols were compared: (1) air breathing with placebo; (2) isocapnic hypoxia (end-tidal PO2 = 50 mm Hg) with placebo; (3) isocapnic hypoxia with beta-blockade; and (4) air breathing with beta -blockade. Regular measurements of heart rate and cardiac output (using ultrasonography and N2O rebreathing techniques) were made with subjects seated in the upright position. The sensitivity of heart rate to rapid variations in hypoxia (GHR) and heart rate in the absence of hypoxia were measured at times 0, 4 and 8 h. No significant progressive effect of hypoxia on cardiac output was detected. There was a gradual rise in heart rate with hypoxia of 11+/-2 beats min(-1) in the placebo protocol and of 10+/-2 beats min(-1) in the beta-blockade protocol over 8 h, compared to the air breathing protocols. The rise in heart rate was progressive (P<0.001) and accompanied by progressive increases in both GHR (P<0.001) and heart rate measured in the absence of hypoxia (P<0.05). No significant effect of beta-blockade was detected on any of these progressive changes. We conclude that sympathetic mechanisms that act via beta -receptors play little role in the progressive changes in heart rate observed over 8 h of moderate hypoxia.

Effects of desferrioxamine on serum erythropoietin and ventilatory sensitivity to hypoxia in humans

In cell culture, hypoxia stabilizes a transcriptional complex called hypoxia-inducible factor-1 (HIF-1) that increases erythropoietin (Epo) formation. One hallmark of HIF-1 responses is that they can be induced by iron chelation. The first aim of this study was to examine whether an infusion of desferrioxamine (DFO) increased serum Epo in humans. If so, this might provide a paradigm for identifying other HIF-1 responses in humans. Consequently a second aim was to determine whether an infusion of DFO would mimic prolonged hypoxia and increase the acute hypoxic ventilatory response (AHVR). Sixteen volunteers undertook two protocols: 1) continuous infusion of DFO over 8 h and 2) control. Epo and AHVR were measured at fixed times during and after the protocols. The results show that 1) compared with control, Epo increased in most subjects at 8 h [52.8 +/- 57.7 vs. 6.9 +/- 2.5 (SD) mIU/ml, for DFO = 4 g/70 kg body wt, P < 0.05] and 12 h (63.7 +/- 76.3 vs. 7.3 +/- 2.5 mIU/ml, P < 0.001) after the start of DFO administration and 2) DFO had no significant effect on AHVR. We conclude that, whereas infusions of DFO mimic hypoxia by increasing Epo, they do not mimic prolonged hypoxia by augmenting AHVR.

Effects of somatostatin on the control of breathing in humans

1. Somatostatin depresses the ventilatory response to hypoxia (AHVR). This study sought to determine whether somatostatin also reduced the peripheral chemoreflex sensitivity to hypercapnia, and if so, whether this was related to the reduction in AHVR. 2. Nine subjects completed the study. AHVR and the ventilatory responses to hypercapnia under both hyperoxic and hypoxic conditions were assessed both without and with an infusion of somatostatin (0.5 BsBs5mgBs5 h-1). Peripheral (fast) and central (slow) responses to hypercapnia were distingushed by use of a multi-frequency binary sequence input in end-tidal PCO2 (PET,CO2) that included 13 steps into and out of hypercapnia. 3. The acute ventilatory response to a reduction in end-tidal PO2 (PET,O2) from 100 to 50 Torr (at a PET, CO2 of +1.5-2.0 Torr above normal) was reduced from (mean +/- s.e.m. ) 16.4 +/- 3.3 to 9.5 +/- 3.2 l min-1 (P < 0.005, Student's t test) by somatostatin. The magnitude of the ensuing hypoxic ventilatory decline was unaltered (8.8 +/- 2.7 l min-1 in control vs. 8.0 +/- 2. 9 l min-1 with somatostatin). 4. The peripheral chemoreflex sensitivity to CO2 in hypoxia was reduced from 2.42 +/- 0.36 to 1.18 +/- 0.20 l min-1 Torr-1 (P < 0.005) with somatostatin. The reduction under hyperoxic conditions from 0.75 +/- 0.34 to 0.49 +/- 0.09 l min-1 Torr-1 did not reach significance. Central chemoreflex sensitivity to CO2 was unchanged. Changes in peripheral chemoreflex sensitivity to CO2 in hypoxia correlated with changes in AHVR. 5. We conclude that peripheral chemoreflex sensitivity to CO2 is reduced by somatostatin, probably via the same mechanism as that by which somatostatin exerts its effects on AHVR.

Effects of subanaesthetic sevoflurane on ventilation. 2: Response to acute and sustained hypoxia in humans

We have determined the influence of 0.1 minimum alveolar concentration (MAC) of sevoflurane on the acute ventilatory response to hypoxia (AHVR), hypoxic ventilatory decline (HVD) and the magnitude of the rapid decline in ventilation on relief of sustained hypoxia (the off-response) in eight healthy adult volunteers. The following design was used with and without 0.1 MAC of sevoflurane: end-tidal PO2 was maintained at 13.3 kPa for 5 min, at 7.9 kPa for 20 min and at 13.3 kPa for 5 min. End-tidal PCO2 was held constant throughout at 1.3 kPa above the subject's normal value. A dynamic end-tidal forcing system was used to generate these gas changes. Sevoflurane reduced AHVR from 14.5 (SEM 1.2) to 11.6 (1.6) litre min-1, and the off-response at cessation of hypoxia from 7.1 (1.1) to 6.3 (1.4) litre min-1. The magnitude of HVD was slightly increased by sevoflurane from 8.2 (1.1) to 10.6 (2.8) litre min-1. None of these changes was significant (ANOVA). These results suggest that 0.1 MAC of sevoflurane had very little effect on the AHVR, and that it did not markedly alter the processes underlying HVD during sustained hypoxia.

Effects of subanaesthetic sevoflurane on ventilation. 1: Response to acute and sustained hypercapnia in humans

We have determined the influence of 0.1 minimum alveolar concentration (MAC) of sevoflurane on ventilation, the acute ventilatory response to a step change in end-tidal carbon dioxide and the ventilatory response to sustained hypercapnia in 10 healthy adult volunteers. Subjects undertook a preliminary 10-min period of breathing air without sevoflurane to determine their normal ventilation and natural end-tidal PCO2. This 10-min period was repeated while breathing 0.1 MAC of sevoflurane. Subjects then undertook two procedures: end-tidal PO2 was maintained at 13.3 kPa and end-tidal PCO2 at 1.3 kPa above the subject's normal value for 30 min of data collection, first with and then without 0.1 MAC of sevoflurane. A dynamic end-tidal forcing system was used to generate these gas profiles. Sevoflurane did not significantly change ventilation: 10.1 (SEM 1.0) litre min-1 without sevoflurane, 9.6 (0.9) litre min-1 with sevoflurane. The response to acute hypercapnia was also unchanged: mean carbon dioxide response slopes were 20.2 (2.7) litre min-1 kPa-1 without sevoflurane and 18.8 (2.7) litre min-1 kPa-1 with sevoflurane. Sustained hypercapnia caused a significant gradual increase in ventilation and tidal volume over time and significant gradual reduction in inspiratory and expiratory times. Sevoflurane did not affect these trends during sustained hypercapnia. These results suggest that 0.1 MAC of sevoflurane does not significantly affect the acute ventilatory response to hypercapnia and does not modify the progressive changes in ventilation and pattern of breathing that occur with sustained hypercapnia.

Ventilatory effects of 8 h of isocapnic hypoxia with and without beta-blockade in humans

This study investigated whether changing sympathetic activity, acting via beta-receptors, might induce the progressive ventilatory changes observed in response to prolonged hypoxia. The responses of 10 human subjects to four 8-h protocols were compared: 1) isocapnic hypoxia (end-tidal PO2 = 50 Torr) plus 80-mg doses of oral propranolol; 2) isocapnic hypoxia, as in protocol 1, with oral placebo; 3) air breathing with propranolol; and 4) air breathing with placebo. Exposures were conducted in a chamber designed to maintain end-tidal gases constant by computer control. Ventilation (VE) was measured at regular intervals throughout. Additionally, the subjects' ventilatory hypoxic sensitivity and their residual VE during hyperoxia (5 min) were assessed at 0, 4, and 8 h by using a dynamic end-tidal forcing technique. beta-Blockade did not significantly alter either the rise in VE seen during 8 h of isocapnic hypoxia or the changes observed in the acute hypoxic ventilatory response and residual VE in hyperoxia over that period. The results do not provide evidence that changes in sympathetic activity acting via beta-receptors play a role in the mediation of ventilatory changes observed during 8 h of isocapnic hypoxia.

Effects of dopamine and domperidone on ventilatory sensitivity to hypoxia after 8 h of isocapnic hypoxia

Acclimatization to altitude involves an increase in the acute hypoxic ventilatory response (AHVR). Because low-dose dopamine decreases AHVR and domperidone increases AHVR, the increase in AHVR at altitude may be generated by a decrease in peripheral dopaminergic activity. The AHVR of nine subjects was determined with and without a prior period of 8 h of isocapnic hypoxia under each of three pharmacological conditions: 1) control, with no drug administered; 2) dopamine (3 microg. min-1. kg-1); and 3) domperidone (Motilin, 40 mg). AHVR increased after hypoxia (P </= 0. 001). Dopamine decreased (P </= 0.01), and domperidone increased (P </= 0.005) AHVR. The effect of both drugs on AHVR appeared larger after hypoxia, an observation supported by a significant interaction between prior hypoxia and drug in the analysis of variance (P </= 0. 05). Although the increased effect of domperidone after hypoxia of 0. 40 l. min-1. %saturation-1 [95% confidence interval (CI) -0.11 to 0. 92 l. min-1. %-1] did not reach significance, the lower limit for this confidence interval suggests that little of the increase in AHVR after sustained hypoxia was brought about by a decrease in peripheral dopaminergic inhibition.

Effects of haloperidol on ventilation during isocapnic hypoxia in humans

Exposure to isocapnic hypoxia produces an abrupt increase in ventilation [acute hypoxic ventilatory response (AHVR)], which is followed by a subsequent decline [hypoxic ventilatory depression or decline (HVD)]. In cats, both anesthetized and awake, haloperidol has been reported to increase AHVR and almost entirely abolish HVD. To investigate whether this occurs in humans, the ventilatory responses of 15 healthy young volunteers to 20 min of isocapnic hypoxia (end-tidal PO2 = 50 Torr) were assessed at 1, 2, and 4.5 h after placebo (control) and after oral haloperidol (Seranace, 0.05 mg/kg) on different days. Three subjects were unable to complete the study because of akathisia. AHVR was significantly greater with haloperidol compared with control (P < 0.01, analysis of variance). However, no significant change in HVD was found [control HVD = 9.3 +/- 1.6 (SD) l/min, haloperidol HVD = 9.9 +/- 2.1 l/min; P = not significant, analysis of variance]. We conclude that combined central and peripheral dopamine-receptor antagonism in humans with haloperidol produces a similar pattern of change to that reported previously with the peripheral antagonist domperidone. We have been unable to show in humans a decrease in HVD by the centrally acting drug as observed in cats.

Time course of the human pulmonary vascular response to 8 hours of isocapnic hypoxia

To examine the hypothesis that the human pulmonary vascular response to hypoxia has a component with a slow time course, we measured pulmonary vascular resistance (PVR) in six healthy adult males during 8 h of isocapnic hypoxia. A balloon-tipped pulmonary artery catheter with thermistor was introduced via a forearm vein and used to derive PVR. The subjects were seated in a chamber in which the oxygen and carbon dioxide concentrations were adjusted to maintain an end-tidal Po2 of 50 Torr and an end-tidal Pco2 equal to the subject's normal prehypoxic value. PVR was measured before and at 0.5-h intervals during 8 h of hypoxia, the following 3 h of isocapnic euoxia (end-tidal Po2 100 Torr), and a subsequent 1-h reexposure to hypoxia. PVR rose from 1.23 +/- 0.26 (SE) Torr-min.1(-1) under euoxia [time (t) = 0] to 1.77 +/- 0.21 Torr.min.1(-1) at t = 0.5 h, reached a maximum at 2 h (2.91 +/- 0.33 Torr.min.1(-1)), and remained fairly constant between 2 and 8 h. Restoration of euoxia at 8 h led to a reduction in PVR with a slow component. Reexposure to hypoxia at 11 h resulted in a greater increase in PVR than at 1 h. Systemic vascular resistance had a similar slow component to its response, falling from 18.6 +/- 1.3 Torr.min.1(-1) at t = 0 to 17.3 +/- 1.4 Torr.min.1(-1) at t = 0.5 h, 14.4 +/- 0.6 Torr.min.1(-1) at t = 4 h, and 13.8 +/- 0.8 Torr.min.1(-1) at t = 8 h. The human pulmonary and systemic vascular responses to hypoxia extend over at least several hours.

Influence of 0.2 minimum alveolar concentration of enflurane on the ventilatory response to sustained hypoxia in humans

To determine the influence of 0.2 minimum alveolar concentration (MAC) of enflurane on the time course of ventilation during sustained hypoxia, we studied 10 healthy adult volunteers with and without enflurane. The following design was used: end-tidal Po2 was maintained at 13.3 kPa for 8 min, at 6.7 kPa for 20 min and at 13.3 kPa for 8 min. End-tidal Pco2 was held constant throughout at 0.67 kPa above the subject's natural value. Control experiments were conducted with no hypoxia imposed. During the experiment subjects breathed via a mouthpiece from an automated gas mixing system which controlled end-tidal values. Enflurane reduced baseline (euoxic) ventilation from 20.9 (SEM 2.0) litre min-1 to 10.1 (1.0) litre min-1 (ANOVA, P < 0.001). Enflurane reduced the acute ventilatory response to hypoxia (AHVR) from 20.1 (3.3) litre min-1 to 5.0 (1.3) litre min-1 (ANOVA, P < 0.01), and the ventilatory off-response at cessation of hypoxia from 11.7 (2.4) litre min-1 to 1.8 (0.5) litre min-1 (ANOVA, P < 0.02). There was no significant difference in hypoxic ventilatory decline (HVD) without and with enflurane (8.9 (2.4) litre min-1 vs 5.5 (1.1) litre min-1; ANOVA, ns). These results confirm that 0.2 MAC of enflurane suppressed the acute ventilatory response to hypoxia, but had no significant effect on the subsequent ventilatory decline during sustained hypoxia.

Time course of hypoxic pulmonary vasoconstriction: a rabbit model of regional hypoxia

There is disagreement in the literature about the time required for hypoxic constriction of pulmonary vessels to reach its full intensity. Some studies suggest that only minutes are required, others that several hours are needed. We examined the time course over 6 h of changes in pulmonary shunt (as a fraction of cardiac output) following induction of unilateral hypoxia by collapse or liquid filling of the left lung in 47 anesthetized rabbits. The time course was examined at four degrees of lung inflation: during collapse and at airway pressures of 0.3 kPa, 0.6 kPa, and 0.9 kPa. The respective volumes (mean +/- SD) of the liquid-filled lung were estimated to be 6.4 +/- 1.0, 12.8 +/- 2.5, and 15.8 +/- 1.6 ml/kg body weight (BW). During sustained hypoxia (the period from 150 to 360 min after inducing hypoxia), shunt declined at a slow linear rate of 2.37 x 10(-4)/min, which was independent of lung inflation (p = 0.65 analysis of variance [ANOVA]) and significantly different from zero (p < 0.001). The stability of cardiac output in this animal model, as measured sequentially by thermodilution, was confirmed in a further 20 animals. The experiments provide evidence for a slow intensification of blood-flow diversion at a rate that does not depend upon the degree of lung inflation. Whether this change is a feature of hypoxic constriction itself, or some modulation of it, remains unclear.

Comparison of the effects of sub-hypnotic concentrations of propofol and halothane on the acute ventilatory response to hypoxia

To compare the effects of sub-anaesthetic concentrations of propofol and halothane on the respiratory control system, we have studied the acute ventilatory response to isocapnic hypoxia (AHVR) in 12 adults with and without three different concentrations of propofol and halothane. Target doses for propofol were 0, 0.05, 0.1 and 0.2 of the effective plasma concentration (EC50 = 8.1 micrograms ml-1). Target doses for halothane were 0, 0.05, 0.1 and 0.2 minimum alveolar concentration (MAC = 0.77%). The doses achieved experimentally were 0.01, 0.06, 0.13 and 0.26 of the EC50 for propofol and 0, 0.05, 0.11 and 0.20 MAC for halothane. During the experiment subjects breathed via a mouthpiece from an end-tidal forcing system. End-tidal PO2 (PE'O2) was held at 13.3 kPa for 5 min, and then at 6.7 kPa for 5 min. End-tidal PCO2 (PE'CO2) was held constant at 0.13-0.27 kPa greater than the subject's natural level throughout. The mean values for AHVR with propofol were: 12.8 (SEM 2.4) litre min-1 (0.01 EC50), 10.0 (1.9) litre min-1 (0.06 EC50), 9.8 (2.3) litre min-1 (0.13 EC50) and 4.9 (1.2) litre min-1 (0.26 EC50). The values for AHVR with halothane were: 11.9 (2.4) litre min-1 (0 MAC), 7.8 (1.6) litre min-1 (0.05 MAC), 5.9 (1.2) litre min-1 (0.11 MAC) and 3.2 (1.6) litre min-1 (0.2 MAC). The decline in AHVR with increasing dose for both drugs was statistically significant (ANOVA, P < 0.001); there was no significant difference between the two drugs with respect to this decline. Normoxic ventilation with propofol declined from 13.2 (1.6) litre min-1 (0.01 EC50) to 8.3 (0.9 litre min-1 (0.26 EC50), and with halothane declined from 13.5 (2.0) litre min-1 (0 MAC) to 11.8 (1.6) litre min-1 (0.2 MAC). This was significant for both drugs (ANOVA, P < 0.001).

Dependence of pulmonary venous admixture on inspired oxygen fraction and time during regional hypoxia in the rabbit

In order to examine the value of assuming constant pulmonary venous admixture with respect to changes in inspired oxygen fraction (FIO2) and time during sustained unilateral hypoxia, we studied venous admixture for 6 h in 27 anaesthetized rabbits in which the left lung was filled with liquid, isosmotic with plasma. In one group of 10 rabbits the right lung was ventilated for 6 h with FIO2 = 1; in a second group of 10 the right lung was ventilated with FIO2 = 1 for 2.5 h and then with FIO2 = 0.3 for 3.5 h. A third group was similarly studied by changing from FIO2 = 1 to FIO2 = 0.5. We found that hypoxic pulmonary vasoconstriction continued to intensify over 3 h. At 3-6 h, with FIO2 = 0.3, venous admixture (0.32 (SEM 0.03)) was higher than baseline (0.13 (0.01), t = 0 min during bilateral oxygenation) by twice the elevation above baseline of the venous admixture (0.22 (0.01)) in the group with FIO2 = 1. The finding of a marked increase in venous admixture with decreasing FIO2 is discussed in relation to current models of hypoxic pulmonary vasoconstriction.

An assessment of central-peripheral ventilatory chemoreflex interaction using acid and bicarbonate infusions in humans

1. The object of this study was to investigate the effect of central chemoreceptor stimulation on the ventilatory responses to peripheral chemoreceptor stimulation. 2. The level of central chemoreceptor stimulation was varied by performing experiments at two different levels of end-tidal CO2 pressure (PCO2). Variations in peripheral chemoreceptor stimulus were achieved by varying arterial pH (at constant end-tidal PCO2) and by varying end-tidal O2 pressure (PO2). 3. Two protocols were each performed on six human subjects. In one protocol ventilatory measurements were made during eucapnia, when the arterial pH was lowered from 7.4 to 7.3. The variation in pH was achieved by the progressive infusion of acid (0.1 M HCl). In the other protocol ventilatory measurements were made during hypercapnia, when the arterial pH was increased from 7.3 to 7.4. The variation in pH was achieved by the progressive infusion of 1.26% NaHCO3. In each protocol ventilatory responses were measured during euoxia (end-tidal PO2, 100 Torr), hypoxia (end-tidal PO2, 50 Torr) and hyperoxia (end-tidal PO2, 300 Torr), with end-tidal PCO2 held constant. 4. The increase in ventilatory sensitivity to arterial pH induced by hypoxia (50 Torr) was not significantly different between protocols (acid protocol, -104 +/- 31 l min-1 (pH unit)-1 vs. bicarbonate protocol, -60 +/- 44 l min-1 (pH unit)-1; mean +/- S.E.M.; not significant (n.s.)). The ventilatory sensitivity to hypoxia at an arterial pH of 7.35 was not significantly different between protocols (acid protocol, 14.7 +/- 3.3 l min-1 vs. bicarbonate protocol, 15.6 +/- 2.4 l min-1; mean +/- S.E.M.; n.s.). The results provide no evidence to suggest that peripheral chemoreflex ventilatory responses are modulated by central chemoreceptor stimulation.

Active transport in the alveolar epithelium of the adult lung: vestigial or vital?

Active secretion by mammalian fetal pulmonary alveolar epithelium is well recognized, as is the role of the adult epithelium in the secretion of surfactant. Recent studies have demonstrated active absorption by adult epithelium involving two sodium-dependent pathways. This finding has focused attention on how poorly we understand both the disposition of alveolar liquid and the physiological role of surfactant. In this paper we review the evidence that the adult mammalian alveolar epithelium absorbs solutes by active transport, and we assess the physiological importance of the resulting liquid movements.

Effect of lung inflation on active and passive liquid clearance from in vivo rabbit lung

Active sodium transport contributes to liquid clearance from the alveoli. We hypothesized that the magnitude of active transport of alveolar liquid depends on the extent to which the alveolar epithelium is stretched and, consequently, on the degree of alveolar inflation. In a study on 38 adult rabbits, the left lung was filled in vivo with a solution of glucose (10 mmol/l) made isosmotic with plasma, using sodium chloride, and held at a constant airway pressure of 3, 6, or 9 cmH2O for 6 h. Alveolar liquid clearance was measured directly as a flow into a left main bronchial catheter. Control animals were compared with animals in which active epithelial sodium transport was inhibited by adding amiloride and phloridzin (both 1 mmol/l) to the instillate. At low inflation, active sodium transport reversed a secretion of liquid into the alveoli; at high inflation, active sodium transport made little or no contribution to transepithelial flow. Hydraulic conductance of the left lung was 1.57 microliters.min-1.cmH2O-1.kg body wt-1. The experiments suggest that pulmonary inflation renders active liquid clearance ineffective.

Effect of low-dose enflurane on the ventilatory response to hypoxia in humans

To investigate the effects of enflurane on the control of breathing we have studied the ventilatory responses to isocapnic hypoxia in 12 adults with and without sedation with enflurane. Design 1 consisted of three steps into hypoxia (PE' O2 = 6.7 kPa), each lasting 3 min, separated by periods of euoxia lasting 5 min (PE' O2 = 13.3 kPa). Design 1 was repeated four times in each subject on the same day in random order: with carrier gas (control) and with 0.04 MAC, 0.07 MAC and 0.13 MAC of end-tidal enflurane concentrations. Design 2 consisted of 20-min exposures to hypoxia with and without 0.07 MAC of enflurane. Each exposure was preceded and followed by 5 min of euoxia. End-tidal PCO2 was held constant at 0.13-0.27 kPa greater than the resting level throughout both designs. Mean (SEM) ventilatory responses to hypoxia for design 1 were: 8.2 (1.3) litre min-1 (control), 6.6 (1.4) litre min-1 (0.04 MAC), 5.7 (1.1) litre min-1 (0.07 MAC) and 3.7 (0.5) litre min-1 (0.13 MAC) (P < 0.001). For design 2, enflurane produced a 15% reduction in resting ventilation (P < 0.01), a 40% decrease in the acute ventilatory response to hypoxia (P < 0.01) and a 32% reduction in ventilatory decline (ns) which occurred during sustained hypoxia.

Effects of midazolam and flumazenil on ventilation during sustained hypoxia in humans

The purpose of this study was to investigate whether increases in gamma-aminobutyric acid (GABA) in the brain stem underlie the ventilatory decline observed during hypoxia in man. The ventilatory responses to sustained isocapnic hypoxia were studied in six adult male subjects on three separate days in three pharmacological conditions: (1) without any drug administration; (2) during infusion of midazolam (a drug which potentiates the effect of GABA); and (3) during infusion of flumazenil (a benzodiazepine antagonist). On each experimental day, the following protocol was repeated three times: end-tidal PO2 was held at 100 Torr for 10 min, then at 50 Torr for 20 min and finally at 100 Torr for 5 min. End-tidal PCO2 was held constant throughout. Responses in the three pharmacological conditions were similar. We conclude that neither potentiation of GABA transmission (midazolam) nor antagonism of this potentiation (flumazenil) greatly affect the decline in ventilation which occurs during extended exposure to hypoxia.

Cardiac sympathetic nerve stimulation enhances cardiovascular performance during hyperkalaemia in the anaesthetized pig

A rapid increase in arterial plasma potassium concentration to values seen during intense exercise depresses cardiac function at rest. Increasing the cardiac concentration of noradrenaline by right-sided sympathetic stimulation in eleven anaesthetized pigs significantly augmented cardiovascular performance during hyperkalaemia, while electrical pacing of the right atrium at equivalent rates to sympathetic stimulation afforded no protection against the deleterious effects of hyperkalaemia. We conclude that the inotropic effect of sympathetic activation may be important in sustaining cardiac function during exercise-induced hyperkalaemia.

Intense slow hypoxic pulmonary vasoconstriction in gas-filled and liquid-filled lungs: an in vivo study in the rabbit

To examine the hypothesis that hypoxic pulmonary vasoconstriction may have a slower time course and greater intensity than is currently recognized, experiments were conducted in twelve anaesthetized rabbits subjected to unilateral lung hypoxia for 6 h. Endobronchial cannulation was used to maintain apnoea of one lung at constant airway pressure whilst inflating the lung with nitrogen or liquid. The second lung was ventilated with oxygen to maintain normocapnia and oxygen transfer. A pulmonary ventilated with oxygen to maintain normocapnia and oxygen transfer. A pulmonary artery catheter was introduced non-invasively. Pulmonary shunt was derived from mixed venous and arterial blood gas parameters. Pulmonary artery pressure was monitored continuously and cardiac output was estimated from oxygen uptake measurements before and after 6 h unilateral hypoxia. The experiments show that a rapid phase of hypoxic pulmonary vasoconstriction is followed by a slow phase which develops over hours. The slow phase is associated with a massive blood flow diversion from the hypoxic lung, such that pulmonary shunt after 6 h unilateral hypoxia is indistinguishable from baseline shunt during bilateral ventilation with oxygen. The response is reversible, but with a similarly slow time course. Results from nitrogen and liquid filling of the lung are similar. These findings are consistent with early experiments by Dirken and Heemstra in 1948 (Quart F Exp Physiol 34, 193-211), and challenge the prevailing notion that hypoxic pulmonary vasoconstriction is always a rapid and relatively weak physiological response to hypoxia.

Left ventricular regional function and relaxation: effects of inotropic stimulation

The effects of increasing doses of two inotropes, isoproterenol and calcium chloride (CaCl2), on left ventricular regional myocardial function and isovolumic relaxation were studied in six anesthetized sheep. After baseline data, CaCl2 was given as intravenous boluses to yield doses of 10 mg/kg, 20 mg/kg, 40 mg/kg, 80 mg/kg, and 160 mg/kg. After a second series of baseline data were obtained, constant infusions of isoproterenol were begun with doses of 0.025 micrograms/kg/min, 0.05 micrograms/kg/min, 0.1 micrograms/kg/min, 0.2 micrograms/kg/min, and 0.4 micrograms/kg/min. During each stage of the protocol with both inotropes, data were recorded during acute constriction of the descending thoracic aorta. Left ventricular relaxation was assessed by analysis of peak negative left ventricular (LV) dP/dt and the time constant of isovolumic left ventricular relaxation (Trelax). Regional myocardial function showed little change in either apical or basal segments until high doses of the inotropes. Peak negative LV dP/dt significantly changed from baseline (775 +/- 60 mmHg/s) with 0.2 micrograms/kg/min (1780 +/- 400 mmHg/s, P < 0.05 v baseline) and 0.4 micrograms/kg/min (2,220 +/- 380 mmHg/s, P < 0.05 v baseline) of isoproterenol, and was unchanged by CaCl2. Trelax was significantly decreased by all doses of isoproterenol, whereas only one dose of CaCl2 decreased Trelax. Trelax was increased with afterloading and this effect was altered by isoproterenol. It is concluded that isoproterenol hastens, whereas CaCl2 does not alter, left ventricular relaxation. This may reflect beta-adrenergic modulation of calcium fluxes during isovolumic relaxation.

Femoral arteriovenous extracorporeal carbon dioxide elimination using low blood flow

BACKGROUND AND METHODS

Conventional extracorporeal CO2 removal systems require blood flow rates of 1 to 2.5 L/min in the extracorporeal circuit. We hypothesized that standard hemofiltration equipment can be combined with a high-performance extracorporeal lung to achieve high rates of CO2 removal at lower blood flow rates. To test this hypothesis, we performed experiments on nine sheep to examine the extent to which CO2 elimination can be achieved at blood flow rates less than 600 mL/min using a 5-m2 hollow fiber membrane lung with countercurrent gas flow, combined with a hemofiltration blood pump, and connected to femoral arterial and venous hemodialysis catheters.

RESULTS

CO2 eliminations of 130 to 180 mL/min at standard temperature and pressure were achieved with blood flow rates in the range 470 to 600 mL/min. With a pumpless artery-to-vein shunt, up to 90 mL/min of CO2 at standard temperature and pressure was eliminated. However, in this mode, the resistance of the access catheters and tubing was the main factor limiting CO2 elimination.

CONCLUSIONS

Standard hemofiltration equipment may be combined with a hollow fiber membrane lung to remove the equivalent of a high proportion of the basal metabolic CO2 production of an adult human at low blood flow rates. Use of this technology would bring extracorporeal CO2 removal within the budget and capability of more ICUs.

An assessment of central-peripheral ventilatory chemoreflex interaction in humans

The independence of the central and peripheral chemoreflexes has been tested in humans. Acute metabolic acidosis generated by a prior bout of brief, hard exercise was used to stimulate primarily the peripheral chemoreceptors, and respiratory acidosis generated by inhaled CO2 was used to stimulate both central and peripheral chemoreceptors. Seven healthy young men were studied. Ventilation and arterial pH, PCO2 and PO2 were recorded. Peripheral chemoreflex sensitivity to hypoxia during acute metabolic acidosis was repeatedly determined by measuring ventilation in euoxia (PETO2 = 100 Torr) and hypoxia (PETO2 = 50 Torr) as the subject recovered from exercise-induced acidosis. Peripheral chemoreflex sensitivity to hypoxia during CO2 inhalation was repeatedly determined by measuring ventilation in euoxia and hypoxia at two levels of hypercapnia (PETCO2 = 45 Torr and PETCO2 = 50 Torr). The ventilatory sensitivity to hypoxia at matched arterial pH values was not significantly different between conditions of high (CO2 inhalation) and low (metabolic acidosis) central chemoreceptor activity. We therefore conclude that interaction between central and peripheral chemoreflexes was non-significant in all subjects.

Effect of potassium on ventilation in the rhesus monkey

Increasing the concentration of arterial plasma K+ to 6-8 mM increased ventilation in two sedated analgesic-treated rhesus monkeys who had their end-tidal CO2 held constant during euoxia (arterial oxygen pressure, Pa,O2, ca 100 Torr) and hypoxia (Pa,O2, ca 40 Torr). During euoxia and hypoxia, hyperkalaemia increased ventilation up to 40 and 250%, respectively. This effect was reduced in euoxia and virtually abolished in hypoxia following an abrupt switch to 100% oxygen. Thus the ventilatory response of this primate to hyperkalaemia is at least as sensitive as that of the cat and if hypoxia is added the two stimuli generate a powerful drive to breathing.

Effects of dopamine and domperidone on ventilation during isocapnic hypoxia in humans

In order to investigate the role of dopamine in the ventilatory response to sustained, isocapnic hypoxia six subjects were studied three times in each of three pharmacological conditions: (1) in the absence of any drug administration, (2) during i.v. infusion of dopamine (3 micrograms.kg-1.min-1), and (3) after pretreatment with domperidone. Otherwise the experimental protocol was identical on each day and consisted of holding the subjects' end-tidal PO2 at 100 Torr for 10 min, then 50 Torr for 20 min and finally at 100 Torr again for 5 min. End-tidal PCO2 was held constant 2-3 Torr above normal throughout the experiment. Domperidone increased, and dopamine decreased the magnitudes of both the fast on- and off-responses, but neither drug affected the magnitude of the hypoxic ventilatory decline (HVD). The results of this study suggests: (1) that a peripheral dopaminergic mechanism is not involved in the genesis of HVD, and (2) the peripheral chemoreflex may be modulated peripherally to produce HVD.

Effect of a single inflation of the lungs on oxygenation during total extracorporeal carbon dioxide removal in experimental respiratory distress syndrome

Respiratory distress syndrome (RDS) was modelled in rabbits using pulmonary lavage to remove surfactant. The stability of the resulting pressure-volume hysteresis of the lungs in vivo was studied with the aid of whole-body plethysmography during apnoeic oxygenation made possible by total extracorporeal carbon dioxide removal. Systemic oxygen delivery was measured as a function of the constant airway pressure during apnoea. In 6 subjects a single brief inflation of the lungs to 3.5 kPa resulted in a doubling of both expired lung volume (volume above functional residual capacity) and arterial oxygen partial pressure at an airway pressure of 0.65 kPa. These rises were well maintained for 40 min following the inflation. In a further 6 subjects with RDS single inflations permitted optimum systemic oxygen transport to occur at the low airway pressure of 0.3 kPa, similar to the optimum airway pressure in 6 healthy control subjects. Where pressure-volume hysteresis is present in RDS it can be exploited during apnoeic oxygenation, and probably during high frequency ventilation, to improve oxygenation by the use of infrequent single inflations of the lungs.

Peak airway pressure during high frequency jet ventilation: theory and measurement

A mathematical model has been developed to predict the peak airway pressure attainable during jet ventilation. The theory assumes inviscid and incompressible flow and agrees closely with experimental results using bench models of simple jet systems and systems using a tracheal tube designed for jet ventilation. The results of a previous published study also show good agreement with the predicted results.

Haemolysis during in vitro CO2 removal from human blood using a membrane lung

Haemolysis of human blood has been examined in vitro as a function of pH in the range 7.2-8.0. The hydrogen ion concentration of freshly donated blood from 11 donors was manipulated in 42 experiments, entirely by altering the carbon dioxide fraction of air with which the blood was equilibrated using a membrane lung. In contrast to the known alkalaemic haemolysis which occurs in canine blood, we observed no correlation between plasma haemoglobin concentrations and blood pH. We conclude that alkalaemic haemolysis is unlikely to complicate the clinical application of extracorporeal carbon dioxide removal in the management of acute respiratory failure.

Asystole with convulsion following a subanesthetic dose of propofol plus fentanyl

A case is presented in which asystole and convulsions occurred after an attempted induction of anaesthesia with propofol and fentanyl. The case suggests that a history of syncope may be associated with unusual susceptibility to the bradycardic effects of propofol.

Rebreathing during spontaneous and controlled ventilation with T-piece breathing systems: a general solution

A general solution is presented to the problem of finding the degree of rebreathing generated by T-piece breathing systems. The solution is applicable to any ventilatory waveform, dead space volume and tidal volume and is identical for spontaneous and controlled ventilation for any given ventilatory waveform. The method is graphical and its use and understanding require no mathematical skills. However, if an analytical form of the ventilatory waveform is known, the method is easily extended by use of calculus to obtain a precise analytical solution.

A randomized comparison of total extracorporeal CO2 removal with conventional mechanical ventilation in experimental hyaline membrane disease

Apnoeic oxygenation (AO) combined with extracorporeal CO2 removal (ECCO2R), using venovenous perfusion across a membrane area of 0.1 m2 has been shown to be feasible in six healthy anaesthetized rabbits. In a further twelve rabbits, ECCO2R has been randomly compared with conventional mechanical ventilation (CMV) following saline lavage to induce respiratory failure. Blood gases were maintained for up to 6 h within the same range (PaO2 = 8-20 kPa, PaCO2 = 4-6 kPa) in two groups of six by varying airway pressures and the oxygen fraction delivered either to the membrane lung (ECCO2R group) or to the ventilator (CMV group). The influence of single hourly sustained inflations (SI) on oxygenation was studied. ECCO2R subjects remained stable and survived. CMV subjects deteriorated and had 80% mortality. Hyaline membranes were absent from ECCO2R subjects and present in all CMV subjects. The response to SI suggests that a lung volume recruitment is maintained during AO for up to 1 h but is ineffective during CMV.