Cerebral haemodynamic and electrocortical CO2 reactivity in pigs anaesthetized with fentanyl, nitrous oxide and pancuronium

How is depth of anaesthesia assessed in experimental pigs? A scoping review

BACKGROUND

Despite the large number of pigs involved in translational studies, no gold standard depth of anaesthesia indicators are available. We undertook a scoping review to investigate and summarize the evidence that sustains or contradicts the use of depth of anaesthesia indicators in this species.

METHODS

Medline, Embase and CAB abstract were searched up to September 22nd 2022. No limits were set for time, language and study type. Only original articles of in vivo studies using pigs or minipigs undergoing general anaesthesia were included. The depth of anaesthesia indicators reported in the selected papers were divided in two categories: A, indicators purposely investigated as method to assess depth of anaesthesia; B, indicators reported but not investigated as method to assess depth of anaesthesia.

RESULTS

Out of 13792 papers found, 105 were included after the screening process. Category A: 17 depth of anaesthesia indicators were found in 19 papers. Studies were conducted using inhalant anaesthetics as the main anaesthetic agent in the majority of the cases (13/19 = 68.4%), while 3/19 (15.8%) used propofol. The most investigated depth of anaesthesia indicators were bispectral index (8/19 = 42.1%) and spectral edge frequency 95% (5/19 = 26.3%). Contrasting results about the specific usefulness of each depth of anaesthesia indicators were reported. Category B: 23 depth of anaesthesia indicators were found in 92 papers. The most reported depth of anaesthesia indicators were: motor response following a stimulus (37/92 = 40.2%), depth of anaesthesia scores (21/92 = 23.3%), bispectral index (16/92 = 17.8%) and spectral edge frequency 95% (9/92 = 9.8%).

CONCLUSION

Results highlight the lack of scientifically valid and reliable indicators to ensure adequate depth of anaesthesia in pigs.

Correlation of Sedline-generated variables and clinical signs with anaesthetic depth in experimental pigs receiving propofol

Most of currently available electroencephalographic (EEG)-based tools to assess depth of anaesthesia have not been studied or have been judged unreliable in pigs. Our primary aim was to investigate the dose-effect relationship between increasing propofol dose and variables generated by the EEG-based depth of anaesthesia monitor Sedline in pigs. A secondary aim was to compare the anaesthetic doses with clinical outcomes commonly used to assess depth of anaesthesia in this species. Sixteen juvenile pigs were included. Propofol infusion was administered at 10 mg kg-1 h-1, increased by 10 mg kg-1 h-1 every 15 minutes, and stopped when an EEG Suppression ratio >80% was reached. Patient state index, suppression ratio, left and right spectral edge frequency 95%, and outcomes from commonly used clinical methods to assess depth of anaesthesia in pigs were recorded. The best pharmacodynamic model was assessed for Patient state index, suppression ratio, left and right spectral edge frequency 95% in response to propofol administration. The decrease of Patient state index best fitted to an inhibitory double-sigmoid model (including a plateau phase). The increase of suppression ratio fitted a typical sigmoid Emax model. No relevant relationship could be identified between spectral edge frequency 95% values and propofol administration. A large variability in clinical outcomes was observed among pigs, such that they did not provide a reliable evaluation of propofol dose. The relationship between propofol dose and Patient state index/suppression ratio described in the present study can be used for prediction in future investigations. The evaluation of depth of anaesthesia based on common clinical outcomes was not reliable.

Influence of moderate and profound hyperventilation on cerebral blood flow, oxygenation and metabolism

OBJECTIVE

The aim of the present study was to examine the impact of moderate and profound hyperventilation on regional cerebral blood flow (rCBF), oxygenation and metabolism.

MATERIALS AND METHODS

Twelve anesthetized pigs were subjected to moderate (mHV) and profound (pHV) hyperventilation (target arterial pO(2): 30 and 20 mmHg, respectively) for 30 min each, after baseline normoventilation (BL) for 1 h. Local cerebral extracellular fluid (ECF) concentrations of glucose, lactate, pyruvate and glutamate as well as brain tissue oxygenation (p(ti)O(2)) were monitored using microdialysis and a Licox oxygen sensor, respectively. In nine pigs, regional cerebral blood flow (rCBF) was also continuously measured via a thermal diffusion system.

RESULTS

Both moderate and profound hyperventilation resulted in a significant decrease in rCBF (BL: 37.9+/-4.3 ml/100 g/min; mHV: 29.4+/-3.6 ml/100 g/min; pHV: 23.6+/-4.7 ml/100 g/min; p<0.05) and p(ti)O(2) (BL: 22.7+/-4.1 mmHg; mHV: 18.9+/-4.9 mmHg; pHV: 13.0+/-2.2 mmHg; p<0.05). A p(ti)O(2) decrease below the critical threshold of 10 mmHg was induced in three animals by moderate hyperventilation and in five animals by profound hyperventilation. Furthermore, significant increases in lactate (BL: 1.06+/-0.18 mmol/l; mHV: 1.36+/-0.20 mmol/l; pHV: 1.67+/-0.17 mmol/l; p<0.005), pyruvate (BL: 46.4+/-7.8 micromol/l; mHV: 58.0+/-10.3 micromol/l; pHV: 66.1+/-12.7 micromol/l; p<0.05), and lactate/glucose ratio were observed during hyperventilation. (Data are presented as mean+/-S.E.M.)

CONCLUSIONS

Both moderate and profound hyperventilation may result in insufficient regional oxygen supply and anaerobic metabolism, even in the uninjured brain. Therefore, the use of hyperventilation cannot be considered as a safe procedure and should either be avoided or used with extreme caution.

Cerebral vascular reactivity response to anaesthetic induction with propofol in patients with intracranial space-occupying lesions and vascular malformations

BACKGROUND AND OBJECTIVE

In clinical trials, autoregulation and carbon dioxide reactivity are preserved during propofol anaesthesia. Paradoxical increases of blood flow velocity during induction of anaesthesia could be demonstrated in patients with brain tumours. This study evaluates the effects of propofol on cerebral blood flow velocity in patients undergoing surgery for brain tumours and vascular malformations.

METHODS

Changes in cerebral blood flow velocity after the administration of propofol were assessed using bilateral 2 MHz transcranial Doppler probes in 47 patients undergoing surgery for brain tumours and in 22 patients undergoing surgery for aneurysms and angiomas.

RESULTS

Flow reduction after propofol was slightly less pronounced on the side of the tumour; in patients with cerebrovascular lesions, no difference between the two sides was detectable. After the administration of propofol a flow increase was present on the side of the tumour in 2 patients. In 3 patients with angiomas, the flow decrease after the administration of propofol was less pronounced on the side of the angioma. Neither observation gave statistical proof of abnormality.

CONCLUSIONS

The flow changes after propofol may give a hint of cerebrovascular reactivity. Further investigations should focus on combined measurements of cerebral autoregulation and carbon dioxide reactivity and should focus on patients with impaired consciousness to test for reliability.

Clearance of fentanyl, alfentanil, methohexitone, thiopentone and ketamine in relation to estimated hepatic blood flow in several animal species: application to prediction of clearance in man

We have used estimated hepatic blood flow (Qhep) as an aid to evaluate clearance (CL) values in animals and to predict clearance in man of five anaesthetic agents: fentanyl, alfentanil, methohexitone, thiopentone and ketamine. The disposition of methohexitone was determined in rats and that of ketamine in rats, rabbits and pigs. Further data were compiled from the literature and supplemented experimentally as needed. Allometric interspecies scaling, according to three different methods, was used to estimate blood clearance and unbound clearance (CLu) in man. The results of scaling according to the three different methods were evaluated in relation to estimated hepatic extraction ratio (CL/Qhep) of the drugs. In most animals the clearance of the drugs were comparable with or lower than estimated Qhep. However, ketamine showed extensive extrahepatic clearance in rabbits. Prediction of clearance in man was successful by at least one method for all five drugs, while prediction of CLu generally failed. Estimates of CL/Qhep gave no indication as to the choice of the best method. Volume of distribution at steady state could be predicted for alfentanil, thiopentone and ketamine. Comparison of clearance with Qhep should be used to evaluate clearance data in animals, however estimation of hepatic extraction ratios appears to be of little use for allometric scaling. The use of ketamine as an anaesthetic agent in rabbits is questionable, while the use of fentanyl in pigs, methohexitone in rats and ketamine in rats and pigs is well supported by the pharmacokinetic data.

Electrocortical brain activity during hypoxia and hypotension in anesthetized newborn lambs

Blood gas and blood pressure disturbances do influence cerebral blood flow in newborns. To what extent cerebral blood flow changes affect electrocortical brain activity remains uncertain. We studied the effect of severe hypoxia and hemorrhagic hypotension on carotid artery blood flow and electrocortical brain activity in newborn anesthetized lambs. During hypoxia carotid artery blood flow increased significantly, whereas electrocortical brain activity remained unchanged. The hemorrhagic hypotension study showed that the lower limit of the autoregulatory ability of the cerebral vascular bed was 60 mmHg. Electrocortical brain activity however remained stable until mean aortic pressure had dropped below 30 mmHg, carotid artery blood flow below 10.6 ml/kg/min, and cerebral oxygen delivery below 1.4 ml/kg/min.

Effects of sevoflurane on intracranial pressure, cerebral blood flow and cerebral metabolism. A dose-response study in patients subjected to craniotomy for cerebral tumours

BACKGROUND

Studies concerning the cerebrovascular effects of sevoflurane in patients with space-occupying lesions are few. This study was carried out as a dose-response study comparing the effects of increasing sevoflurane concentration (1.5% (0.7 MAC) to 2.5% (1.3 MAC)) on cerebral blood flow (CBF), intracranial pressure (ICP), cerebrovascular resistance (CVR), metabolic rate of oxygen (CMRO2) and CO2-reactivity in patients subjected to craniotomy for supratentorial brain tumours.

METHODS

Anaesthesia was induced with propofol/fentanyl/atracurium and maintained with 1.5% sevoflurane in air/oxygen at normocapnia. Blood pressure was maintained constant by ephedrine. In group 1 (n = 10), the patients received continuously 1.5% sevoflurane. Subdural ICP, CBF and CMRO2 were measured twice at 30-min intervals. In group 2 (n = 10), sevoflurane concentration was increased from 1.5% to 2.5% after CBF1. CBF2 was measured after 20 min during 2.5% sevoflurane. Finally, CO2-reactivity was studied in both groups.

RESULTS

In group 1, no time-dependent alterations in CBF, CVR, ICP and CMRO2 were found. In group 2, an increase in sevoflurane from 1.5% to 2.5% resulted in an increase in CBF from 29 +/- 10 to 34 +/- 12 ml 100 g-1 min-1 and a decrease in CVR from 2.7 +/- 0.9 to 2.3 +/- 1.2 mmHg ml-1 min 100 g (P < 0.05), while ICP and CMRO2 were unchanged. CO2-reactivity was maintained at 1.5% and 2.5% sevoflurane.

CONCLUSION

Sevoflurane is a cerebral vasodilator in patients with cerebral tumours. Sevoflurane increases CBF and decreases CVR in a dose-dependent manner. CO2-reactivity is preserved during 1.5% and 2.5% sevoflurane.

In vivo estimation of cerebral blood flow, oxygen consumption and glucose metabolism in the pig by [15O]water injection, [15O]oxygen inhalation and dual injections of [18F]fluorodeoxyglucose

There is a need for suitable non-primate laboratory animals for studies of brain function by positron emission tomography (PET). To provide a comparative index of the circulatory physiology of the pig, we have applied novel PET tracer methodology to seven anaesthetized pigs, and measured cerebral regional oxygen consumption (CMR[O2]), cerebral blood flow (CBF), and cerebral glucose metabolism (CMR[glc]). Blood flow and flow-metabolism couple were estimated for selected cerebral regions of interest. We found an average hemispheric CMR(O2) of 171 +/- 18 micromol/100 cm3/min. Individual hemispheric CBF measurements varied between 33 and 41 ml/100 cm3/min, with an average of 37 +/- 3 ml/100 cm3/min at an average PaCO2 of 4.3 +/- 0.9 kPa. The blood flow dependency on arterial PCO2 was calculated from the results of the carbon dioxide response in two pigs in which the CBF measurements obeyed the equation CBF (ml/100 cm3/min) = 8.9 PaCO2 (kPa). In each pig, CMR(glc) was studied twice with a double-injection FDG method. In the first session, the values of CMR(glc) averaged 27 +/- 3 and 23 +/- 4 micromol/100 cm3/min, estimated by multilinear and linear regression analysis, respectively. In the second session, the corresponding averages were 27 +/- 3 and 24 +/- 3 micromol/100 cm3/min, respectively. The average oxygen extraction fraction was 0.46 +/- 0.09 and the oxygen-glucose ratio was 6.1 +/- 0.8. The findings indicate that the pig is suitable for PET studies of cerebral blood flow, cerebral oxygen consumption and glucose metabolism.

Pharmacokinetic-pharmacodynamic modelling of the EEG effect of alfentanil in rats

The purpose of the present investigation was to develop a methodology for quantification of the concentration-pharmacological effect relationship of alfentanil in individual rats by using effect parameters derived from quantitative EEG analysis. In particular, the role of the opioid-induced side effects of proconvulsant activity, hypothermia, and respiratory depression was investigated. Amplitudes in the 0.5- to 4.5-Hz frequency band of the EEG power spectrum were used as a descriptor of the effect. The pharmacokinetics and pharmacodynamics of alfentanil were determined after intravenous administration of 2000 micrograms/kg during 20 min. The administration of alfentanil induced paroxysmal seizure patterns in the EEG which made meaningful analysis of the EEG effect impossible. A constant infusion of midazolam (5.5 mg/kg/hr) prevented alfentanil-induced seizures. When no precautions were taken to control body temperature, analysis of the concentration-EEG effect relationship was complicated by proteresis due to alfentanil-induced hypothermia. This proteresis disappeared when body temperature was stabilized at 37.5 degrees-38.5 degrees C with isothermal pads. Alfentanil-induced respiratory depression was managed successfully by artificial ventilation. Adequateness of artificial ventilation was ascertained by monitoring of arterial pO2, pCO2, and pH. When these opioid side effects were controlled, the pharmacokinetics were most adequately described by a biexponential function. The values of the pharmacokinetic parameters were (mean +/- SE, n = 7); clearance = 53 +/- 6 ml.min/kg, volume of distribution = 1.19 +/- 0.19 l/kg and terminal half-life = 24.5 +/- 2.3 min. By pharmacokinetic-pharmacodynamic modelling, the individual concentration-effect relationships of alfentanil were derived, which were successfully characterized by the sigmoidal Emax pharmacodynamic model. The values of the pharmacodynamic parameters were (mean +/- SE, n = 7): E0 = 57 +/- 4 microV, Emax = 95 +/- 17 microV, EC50 = 202 +/- 55 ng/ml and Hill factor = 1.53 +/- 0.20. No delay was observed between alfentanil concentration and effect. The results of the present study show that when side effects are controlled adequately, the concentration-EEG effect relationship of alfentanil can be characterized in individual rats using amplitudes in the 0.5- to 4.5-Hz frequency band of the EEG as a measure of the pharmacological response. The described methodology can be very useful for detailed studies into the pharmacodynamics of (new) synthetic opioids in vivo.

Changes in the cerebral arteriovenous oxygen content difference by surgical incision are similar during sevoflurane and isoflurane anaesthesia

PURPOSE

To investigate changes of cerebral arteriovenous oxygen content difference (AVDO2) induced by surgical incision and to determine carbon dioxide (CO2) reactivity of the cerebral circulation during sevoflurane and isoflurane anaesthesia.

METHODS

Twenty-one ASA 1-2 patients undergoing elective surgery for supratentorial tumours were randomly allocated to receive either 1.3 MAC sevoflurane/N2O anaesthesia (n = 10) or equi-MAC isoflurane/N2O anaesthesia (n = 11). Before and after incision, haemodynamic measurements and AVDO2 determinations were performed. After opening the dura, AVDO2 was determined before and after the respiration rate was increased by 50%.

RESULTS

Incision produced an increase in mean arterial pressure from 69 +/- 11 to 97 +/- 22 mmHg (mean +/- SD) (P < 0.0005) and from 71 +/- 6 to 89 +/- 12 mmHg (P < 0.0001) in the sevoflurane and isoflurane groups, respectively, whereas the heart rate increased from 60 +/- 9 to 72 +/- 8 bpm (P < 0.001) and from 65 +/- 6 to 70 +/- 7 bpm (P < 0.001), respectively. Arterial carbon dioxide tension (PaCO2) was increased from 33.6 +/- 2.3 to 34.6 +/- 1.8 mmHg (P < 0.05) with incision in the sevoflurane group. The AVDO2 was decreased from 6.5 +/- 1.6 to 5.3 +/- 1.6 vol% (P < 0.0005) in the sevoflurane group and from 6.7 +/- 1.1 to 6.0 +/- 1.1 vol% (P < 0.01) in the isoflurane group. The % change of AVDO2 was larger in the sevoflurane group than in the isoflurane group (-18.3 +/- 8.4% vs -9.1 +/- 9.0%; P < 0.05) but no difference remained after the post-incisional AVDO2 value of the sevoflurane group was corrected for pre-incisional PaCO2. Carbon dioxide reactivity, calculated as the percent change in AVDO2 per mmHg change in PaCO2, was 6.1 +/- 3.0%.mmHg-1 in the sevoflurane group and 5.9 +/- 2.4%.mmHg-1 in the isoflurane group (P = NS).

CONCLUSIONS

Sevoflurane and isoflurane are associated with similar impairment of cerebral flow-metabolism coupling at incision, while CO2 reactivity is maintained during both anaesthetics.

Metabolic, ventilatory and circulatory effects of doxapram in anaesthetized pigs during normoxia and hypoxia

BACKGROUND

In a previous clinical study doxapram was found to improve ventilatory efficacy postoperatively, presumably via effects on hypoxic pulmonary vasoconstriction (HPV). The present study was designed to see whether doxapram induced any changes of arterial oxygenation and pulmonary vascular resistance during normoxia or hypoxia and whether the changes were influenced by the anaesthetic agents.

METHODS

Seventeen piglets were anaesthetized by combinations of either midazolam + fentanyl + pancuronium + pentobarbital (TIVA, n = 9), or by midazolam + fentanyl + pancuronium + halothane, 0.5% in end-tidal gas (Hal, n = 8). Analyses of expired gas and mixed venous and arterial blood in combination with determinations of central blood flow and pressures allowed for calculations of standard metabolic, ventilatory and circulatory data. Values were obtained at normoventilation using normoxic (FIO2 = 0.3) and hypoxic (FIO2 = 0.08) gas mixtures at calculated doxapram plasma concentrations of 1, 2 and 4 micrograms.ml-1.

RESULTS

With few exceptions doxapram administration affected the investigated variables only moderately during normoxia. In group Hal, PVR and SVR showed a biphasic raise (P < 0.05), CO fell (P < 0.05-P > or = 0.05) and C (a-v)O2 rose (P < 0.05). In group TIVA, PaO2 fell (P < 0.01-0.05) despite unchanged PVR, CO and VD/VT. Hypoxia affected a moderate increase in PVR in group TIVA (P < 0.05), which was slightly lower at the lowest and highest plasma levels of doxapram (P < 0.05). In group Hal, the induction of hypoxia induced a more pronounced rise in PVR (P < 0.05) which showed a biphasic response to increasing dose levels of doxapram, the lowest dose affecting a further rise (P < 0.05) and the highest a reduction to values below hypoxia control levels (P < 0.05). Pronounced differences between the two groups with respect to values for metabolic and circulatory variables make the interpretation of data difficult.

CONCLUSIONS

Doxapram administration to anaesthetized animals did not induce any effects indicative of augmentation of the HPV response.

Cerebral uptake of morphine in the pig calculated from arterio-venous plasma concentration gradients: an alternative to tissue microdialysis

The aim of this study was to characterize the reversible cerebral uptake of morphine in the pig by measuring the changing arterio-venous plasma concentration gradient over the brain. Seven pigs were anaesthetized by continuous infusions of ketamine and pancuronium and ventilated with oxygen in nitrous oxide. During and after 5-min intravenous infusions of morphine hydrochloride, blood samples were drawn from a central artery and from the internal jugular vein. Concomitantly, cerebral blood flow (CBF) was repeatedly measured as clearance of 133Xe from the brain after intracarotid injection. Plasma concentrations of morphine and, in samples from two animals, morphine glucuronides were assayed by high-performance liquid chromatography. Drug flux (Jnet) from arterial blood to brain was calculated from the arterio-venous plasma concentration gradients, the blood:plasma concentration ratio and CBF. Uptake of morphine from arterial blood to brain was very rapid, with a maximal Jnet typically at 3 min after the beginning of the infusion. The initial cerebral extraction of morphine was close to 50%. When the arterial and jugular venous concentration curves crossed, 1-5 min after the end of the infusion, the initially rapid uptake of morphine changed into a slow and steady release. The cerebral extraction of morphine glucuronides was comparable to that of morphine, however, Jnet was lower due to lower plasma concentrations at time of maximal extraction. The findings demonstrate how the cerebral uptake and release of morphine and its metabolites can be studied with a method that is entirely non-invasive to the brain and permits very flexible sampling. Uptake and release of drug is observed directly and need not be inferred from cerebral concentration curves.

A porcine model for evaluation of cerebral haemodynamics and metabolism during increased intracranial pressure

In patients with severe head injuries raised intracranial pressure (ICP) constitutes the most important cause of mortality. Several new therapies for increased ICP have recently been suggested and it is of importance to study the physiological effects of these treatments in animal experiments during steady state conditions. A porcine model for evaluation of cerebral haemodynamics and metabolism during increased ICP is presented. Intracranial hypertension was induced by inflation of two tonometric gastric balloons placed extradurally covering a major part of the parietooccipital region bilaterally. The distribution of the blood flow supplied by the carotid artery used for the cerebral blood flow (CBF) measurements was studied by intraarterial (i.a.) injection of 99mTc-HMPAO. The measurements showed that following ligation of the external carotid and the occipital artery no accumulation of tracer substance occurred in extracranial tissues during normal or increased ICP. Cerebral physiological variables (CBF, Cavo2, and ICP) were measured 5, 20 and 60 min after induction of intracranial hypertension. The results confirm that the experimental situation gives a reproducible increase in ICP (25-28 mm Hg) and that the physiological variables remain stable during the period of intracranial hypertension. We conclude that the model simulates the effects of an acute intracranial focal mass and is well suited for the evaluation of different pharmacological therapies of increased ICP.

The effect of thiopental on cerebral blood flow, and its relation to plasma concentration, during simulated induction of anaesthesia in a porcine model

The reversible effect of an induction dose of thiopental on the cerebral blood flow (CBF) was characterized by repeated 133Xe washout measurements during stable physiological conditions in anaesthetized pigs. A thiopental effect corresponding to induction of light and transient anaesthesia was confirmed by electroencephalography (EEG). The concentration (arterial plasma) -effect (-% CBF) relationship of thiopental was estimated using a sigmoidal Emax model. The injection caused a rapid 36 +/- 4.5% (mean +/- s.d.) drop in CBF, with return to baseline by 80 min. According to the pharmacodynamic model, the maximal effect of thiopental (Emax) in this experimental set-up was a 58% lowering of the CBF and the concentration at half-maximal effect (EC50) was 25 micrograms.ml-1. This study provides a complete characterization of the effect of thiopental on the CBF, including the time-course and concentration-effect relationship. A comparison to limited data in the literature suggests that the findings in the pigs constitute a fair approximation of the action of thiopental during the clinical induction of anaesthesia.

Low-dose midazolam antagonizes cerebral metabolic stimulation by ketamine in the pig

In order to test the hypothesis that low-dose midazolam reduces excitatory cerebral symptoms by attenuating ketamine-induced increases in the cerebral metabolic rate for oxygen (CMRO2), we compared the cerebral effects of a combination of an anesthetic dose of ketamine hydrochloride (10.0 mg.kg-1 i.v.) and a subanaesthetic dose of midazolam maleate (0.25 mg.kg-1 i.v., n = 6; or 0.10 mg.kg-1 i.v., n = 6) with results recently obtained with ketamine (10.0 mg.kg-1 i.v.) in normoventilated pigs anaesthetized with fentanyl, nitrous oxide and pancuronium. Cerebral blood flow (CBF) was measured with the intra-arterial 133Xe clearance technique, and CMRo2 was calculated from CBF and the cerebral arteriovenous oxygen content difference (CaVO2). The CMRO2 did not increase significantly. In contrast, the maximal increase in cerebral CaVo2 (by 56-59% at 10 min; P < 0.01) was similar to that induced by ketamine, since CBF was more depressed (by 35-45% at 1 min: P < 0.001) by ketamine-midazolam than by ketamine only. Midazolam was found to increase CVR (P < 0.01) and further depress CBF (P < 0.01), and to antagonize the ketamine-induced increase in CMRO2 (P < 0.05). Ketamine-induced effects on mean arterial pressure (MAP) and spectral electroencephalographic (EEG) voltage were not significantly altered by midazolam. The pharmacokinetics of ketamine, as measured during an 80-min period, were not affected by the concomitant administration of midazolam. We propose that a ketamine-midazolam combination comprising a low-dose fraction (1/100-1/40) of midazolam is superior to ketamine alone for anaesthetic use.