Differential effects of intravenous anaesthetic agents on the response of rat mesenteric microcirculation in vivo after haemorrhage{dagger}

Z. L. S. Brookes, N. J. Brown and C. S. Reilly*

Section of Surgical and Anaesthetic Sciences, Division of Clinical Sciences, University of Sheffield, Royal Hallamshire Hospital, Sheffield S10 2JF, UK*Corresponding author

{dagger}This work was performed in the Section of Surgical and Anaesthetic Sciences and funded by a British Journal of Anaesthesia Project Grant.

Accepted for publication: September 25, 2001


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Background. The differential effects of i.v. anaesthesia on the response of the mesenteric microcirculation after haemorrhage in vivo are previously unexplored.

Methods. Male Wistar rats (n=56) were anaesthetized intravenously either with propofol and fentanyl (propofol/fentanyl), ketamine or thiopental. A tracheostomy and carotid cannulation were performed and the mesentery surgically prepared for observation of the microcirculation using fluorescent in vivo microscopy. Animals were allocated to one of three groups: control, haemorrhage or haemorrhage re-infusion.

Results. After haemorrhage, the response of the microcirculation differed during propofol/fentanyl, ketamine and thiopental anaesthesia. During propofol/fentanyl anaesthesia there was constriction of arterioles (–16.7 (3.9)%), venules (–5.9 (1.7)) and capillaries (–16.3 (2.8)) (n=12). During ketamine and thiopental anaesthesia both constriction and dilation was observed. After haemorrhage and re-infusion, macromolecular leak occurred from venules during propofol/fentanyl and thiopental anaesthesia (P<0.05), but not during ketamine anaesthesia.

Conclusion. In summary, i.v. anaesthetic agents differentially alter the response of the mesenteric microcirculation to haemorrhage.

Br J Anaesth 2002; 88: 255–63

Keywords: anaesthetics, i.v.; complications, haemorrhage; blood, mesenteric microcirculation; measurement techniques, in vivo microscopy


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The physiological response to haemorrhage involves tachycardia and generalized vasoconstriction.1 The mesenteric microcirculation is particularly prone to intense vasoconstriction due to the dense supply of sympathetic nerves.2 Thus, following haemorrhage the bowel is particularly susceptible to the secondary complications associated with haemorrhage, such as ischaemia.3 The differential effects of intravenously administered anaesthetic agents on the response of the rat mesenteric microcirculation to haemorrhage have not previously been investigated. However, in dogs, the use of the i.v. agent propofol has been associated with low intestinal pH after haemorrhage, indicating ischaemia.4

It is well known that propofol and thiopental act centrally to inhibit the sympathetic nervous system, whereas ketamine has sympathomimetic actions.57 It may be hypothesized, therefore, that these anaesthetic agents will also indirectly alter the response of sympathetically innervated arterioles (>18 µm) and venules (>30 µm) to haemorrhage.1 Propofol, ketamine and thiopental have also been shown to possess direct locally mediated vasodilator properties, a response that appears to be mediated via inhibition of L-type voltage gated Ca2+ channels in vitro.811 Previous studies performed in our laboratory have demonstrated differential responses to propofol/fentanyl, ketamine and thiopental anaesthesia with respect to the magnitude and incidence of small arteriolar (<25 µm) and venular (<30 µm) dilation.12 These smaller vessels are influenced primarily by local control factors and may also respond differentially to haemorrhage during i.v. anaesthesia.13

Nitric oxide is a potent vasodilator and may also be affected by anaesthesia. Propofol has been shown to cause the release of nitric oxide by induction of nitric oxide synthase (NOS).14 Nitric oxide is involved in maintaining blood flow in the mesenteric microcirculation after haemorrhage.15 In contrast, vasodilation in response to the nitric oxide-donor sodium nitroprusside (SNP) appears to be reduced by propofol and this may be a mechanism by which propofol could potentiate vasoconstriction in response to haemorrhage.16 Therefore, SNP will be used to determine how the response of the mesenteric microcirculation to nitric oxide is altered by i.v. anaesthesia.

In clinical practice, haemorrhage is usually treated by the re-infusion of fluid, which is associated with increased leakage of macromolecules in the gastrointestinal tract.17 The presence of macromolecular leak is indicative of damage to the endothelium and compromised vessel integrity.18 19 Macromolecular leak may pose a significant problem in the gastrointestinal tract, as this may be associated with the translocation of bacteria and the development of sepsis.20 The effects of i.v. anaesthetic agents on macromolecular leak after haemorrhage have not previously been investigated.

The aims of this study, therefore, are to determine the effects of haemorrhage on the mesenteric microcirculation during propofol and fentanyl, ketamine and thiopental anaesthesia. In vivo microscopy has been used to study small arterioles, venules and capillaries and determine changes in vessel diameter and macromolecular leak.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Animals
Male Wistar rats weighing between 150 and 220 g (n=56) were obtained from the University of Sheffield Field Laboratories. All procedures were performed under the Home Office Animal Procedures Act (1986), Project Licence number 50/01252.

Anaesthesia and monitoring
Rats were placed in a restrainer and a paediatric butterfly needle inserted into the ventral tail vein to allow access for the induction and maintenance of anaesthesia with either: propofol (Diprivan, Zeneca Ltd, Macclesfield, UK) and fentanyl (Fentanyl Citrate, Janssen-Cilag Ltd, High Wycome, UK) (propofol 10–30 mg kg–1 induction, 10–50 mg kg–1 h–1 maintenance, fentanyl 6–8 µg kg–1 induction, 20–30 µg kg–1 h–1 maintenance); ketamine (Ketalar, Parke-Davis, Eastleigh, UK) (30–60 mg kg–1 induction 50–90 mg kg–1 h–1 maintenance); or thiopental (Intra-Vital Sodium, Rhone-Poulenc Rourer, West Malina, UK) (30 mg kg–1 induction, 40–90 mg kg–1 h–1 maintenance). Anaesthesia was induced at a steady rate over a 5-min period and maintained via ‘step-down’ continuous i.v. infusion (Graseby 3200 syringe pumps, Watford, UK). The doses used in this study were derived from doses described in the literature and refined in a series of preliminary studies. The infusion rate was adjusted to maintain a stable light plane of anaesthesia based upon the Lumb and Wynn Jones criteria.21 Each animal exhibited a sluggish pedal withdrawal reflex, stable systemic arterial pressure, regular respiration rate and constricted pupils.

During anaesthesia, air containing 30% oxygen was introduced into a chamber surrounding the animal (1–2 litre min–1) to ensure blood gases remained within the physiological range. Animals were placed on a warming pad to maintain body temperature between 36 to 38°C, which was monitored by an oesophageal thermistor probe and thermometer (Fluke, Washington, USA).

Surgical preparation
A tracheostomy was performed and a Portex cannula inserted to preserve the airway. A silicone cannula containing heparinized saline (100 Units) was inserted into the left carotid artery to allow continuous monitoring of cardiovascular variables via a pressure transducer and physiograph (Micro-Med, Louisville, USA).

The mesenteric surgical preparation was based on methods previously described.22 In brief, a midline abdominal incision was made and the proximal end of the ileum and the adjoining mesentery were exteriorized. The tissue was kept moist with regular applications of warm saline and care was taken during handling to prevent physical damage. The area of interest was approximately 16–20 cm proximal to the ileocaecal valve (in a retrograde direction from the large intestine) and included 3–4 mesenteric windows. The tissue of interest was gently placed over the abdomen and the remaining intestine and mesentery were returned to the abdomen, which was then closed using a suture. Five stay sutures were then positioned at equidistant lengths along the avascular plane of the exteriorized ileum, with care taken not to perforate the small intestine. The animal was placed on its left side and the sutures used to firmly hold the mesentery flat on a microscope slide mounted on pillars on a Perspex board. The small intestine and mesentery were surrounded by gauze, moistened with warm (37°C) saline and covered with an impermeable membrane (Saran Wrap, Dow Chemical Co., UK) to prevent dehydration.

In vivo fluorescent microscopy
The animal and Perspex board, with attached warming pad, were transferred to the modified stage of an Optiphot-2 Nikon fluorescent microscope (Nikon Ltd, Kingston, Surrey, UK). This was equipped with a tungsten lamp for transmitted light and a mercury arc lamp for epi-illumination fluorescent light microscopy. A filter cube interposed into the path of a mercury arc lamp allowed blue light (450–490 nm) to be selected for epi-illumination of the mesenteric microcirculation. Fluorescein isothiocyanate (FITC) conjugated to bovine serum albumin (BSA) (FITC-BSA) is retained in the vasculature and epi-illumination with blue light results in green fluorescence, which allows the mesenteric microcirculation to be clearly visualized.18 Images of the preparation were monitored using a CCD camera (Hitachi, UK), displayed on a high-resolution monitor (Sony, PVM-1443) and recorded by video (SLV-593, Sony, UK) onto tape (E-180 SX, JVC, Japan) for later off-line analysis. This included the addition of time and date information (DTG1, NG Systems, UK).

An area of the mesenteric microcirculation was recorded using epi-illumination to include an arteriole (10–25 µm), venule (15–30 µm) and ‘capillary’ (4–10 µm). These smallest vessels defined as ‘capillaries’ may have also included some pre-capillary arterioles and post-capillary venules. All vessels were located within the sub-mucosal mesenteric border.

Experimental protocol
Animals were allocated into one of the following experimental groups. 1. Control (C) (propofol, n=6, ketamine, n=6, thiopental, n=6). 2. Haemorrhage (H) (propofol, n=6, ketamine, n=7, thiopental, n=7). 3. Haemorrhage re-infusion (H–R) (propofol, n=6, ketamine, n=6, thiopental, n=6). C animals received i.v. anaesthesia alone. In the H group during anaesthesia, haemorrhage was induced by removing approximately 10% of the total body blood volume (actual volume=0.75 ml 100 g–1) at a constant rate over a 10-min period (approx. 0.1 ml min–1100 g–1). No re-infusion was administered to the H group. However, in the H-R group, 30-min haemorrhage was followed by re-infusion of sterile saline and half the amount of blood removed during haemorrhage (saline: blood=2:1) at a steady rate over 10 min.

During the 30-min stabilization period (t=–30 to 0 min) areas were selected to assess microcirculatory variables and at t=0 min FITC-BSA 0.25 ml 100 g–1 was administered systemically via the carotid artery. This was followed by a 30-min baseline period (t=0–30 min) and an experimental period of 90 min (t=30–130 min) in all groups. In both the H and H–R groups, between t=30 and t=40 min, 10% haemorrhage was induced. In the H–R group, 30 min after haemorrhage, blood and saline were re-infused (t=70–80 min). In all groups at t=135 min, 1 ml of 10–5 M SNP (Sigma-Aldrich, UK) in saline was applied topically to the mesentery preparation to induce maximal dilation.23 Variables were then recorded immediately (t=135 min) and 15 min after application (t=150 min).

At the end of the experiment, 1 ml of blood was removed from all animals via a single use arterial blood sampling system (Concorde, Hythe, UK) for immediate analysis of blood gases (PO2, PCO2 and pH) (CIBA Corning, Medfield, USA). This sample was also used to determine haematocrit. A further 1 ml of blood was removed for later analysis of plasma anaesthetic concentration using standard high-pressure liquid chromatography and gas chromatography techniques.

Measurements
Temperature, cardiovascular and microcirculatory variables were measured over 30 s every 10 min for the entire 150-min experimental duration (t=0–150 min). Temperature, heart rate, systolic arterial pressure (SAP) and diastolic arterial blood pressure (DAP) were recorded on-line. Diameters of arterioles, venules and capillaries and macromolecular leak were assessed off-line using a Viglen IV/25 personal computer (Viglen, Middlesex, UK) and an image analysis software package (Image Pro Plus, Media Cyberkinetics, Silver Spring, Maryland, USA). External vessel diameter (including lumen and vessel walls) was measured in microns (µm). Macromolecular leak was assessed using the arbitrary grey scale to quantify interstitial FITC-BSA fluorescence. This was determined from the mean of three distinct areas (3 mm2) adjacent to each arteriole, venule and capillary. Data for both variables are subsequently presented as a percentage change from t=30 min.

Statistics
All data were expressed as mean (SEM). Vessel diameters were expressed as a percentage change from t=30 min. After ANOVA, the Kruskall–Wallis and Mann–Whitney U-tests for non-parametric data were used to determine differences between the experimental groups for each anaesthetic agent. Within groups, variation was assessed using the Kruskall–Wallis test and the Wilcoxon test for paired data. All results were considered significantly different at P<0.05. A commercial software package was employed to determine r2 values for the purposes of regression analysis and the relevant statistical tests performed using Minitab Inc., UK.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Cardiovascular variables
In C animals, heart rate generally remained stable between t=0 min and t=130 min. SAP and DAP declined gradually throughout the experiment (Fig. 1AC). Between t=0 and t=130 min, SAP decreased by 27% (propofol/fentanyl), 7% (ketamine) and 13% (thiopental). After haemorrhage, a tachycardia was observed. During ketamine anaesthesia, a small increase in heart rate was observed from 433.1 (13.4) beats min-1 (t=30 min) to 460.8 (16.8) (t=40 min) (and H–R, n=13). Significant increases were observed from 434.0 (7.7) to 549.0 (33.3) during propofol/fentanyl anaesthesia (P<0.05, n=12) and from 389.5 (10.7) to 421.7 (10.3) during thiopental anaesthesia (n=13, P<0.05). Haemorrhage caused a significant decrease in SAP during propofol/fentanyl, ketamine and thiopental anaesthesia (Fig. 1AC). DAP followed the same pattern. In H animals, the decrease in SAP and DAP was maintained for the remainder of the experiment.



View larger version (29K):
[in this window]
[in a new window]
 
Fig 1 Systolic arterial pressure (SAP, mm Hg) in control (C), haemorrhage (H) and haemorrhage re-infusion (H–R) groups of animals during propofol/fentanyl (A), ketamine (B) and thiopental (C) anaesthesia. Haemorrhage occurred between 30 and 40 min and re-infusion between 70 and 80 min. *P<0.05 significantly different from pre-haemorrhage (t=30 min). #P<0.05 significantly different from C. $P<0.05 significantly different from t=70 min (before re-infusion).

 
In H–R animals, re-infusion caused a transient (10–20 min) increase in cardiovascular variables close to pre-haemorrhage and control values (t=80, 90), as demonstrated with SAP (Fig. 1AC).

Vessel diameter
Prior to experimental manipulation (t=30 min), vessel diameters (µm) in all groups of animals (C, H and H–R) were 17.1 (0.9) (arterioles), 21.0 (1.3) (venules) and 8.0 (0.7) (capillaries) during propofol/fentanyl anaesthesia (n=18), 12.6 (0.5) (arterioles), 16.4 (0.8) (venules) and 7.5 (0.4) (capillaries) during ketamine anaesthesia (n=19) and 13.9 (0.6) (arterioles), 16.1 (0.9) (venules) and 7.5 (0.4) (capillaries) during thiopental anaesthesia (n=19). The diameters of arterioles and venules were greater during propofol/fentanyl anaesthesia, when compared with ketamine and thiopental (P<0.05).

In C animals, vessel diameters remained stable between t=0 and t=130 min with mean diameter changes for arterioles of –0.9(0.4)%, venules 1.4(0.5)% and capillaries –0.4(0.2)%, relative to baseline.

During propofol/fentanyl anaesthesia, after haemorrhage (t=40 min), there was constriction of arterioles (–16.7 (3.9)), venules (–5.9 (1.7)) and capillaries (–16.3 (2.8), n=12) (Fig. 2). At t=60 min the constrictor response was potentiated in arterioles (–29.2 (8.2)) and maintained in venules (–4.9 (3.7)) and capillaries (–16.4 (3.4)) (Fig. 2).



View larger version (17K):
[in this window]
[in a new window]
 
Fig 2 Percentage change in the diameters of arterioles, venules and capillaries in response to haemorrhage during propofol/fentanyl anaesthesia.

 
During ketamine anaesthesia, after haemorrhage (t=40 min) both constriction and dilation of arterioles (–11.3 (2.0); n=11, 22.0, n=2), venules (–5.6 (1.4); n=10, 10.5; n=3) and capillaries (–7.5 (1.6); n=6, 24.9 (6.5); n=7) was observed. At t=60 min these responses tended to diminish as vessel diameters returned towards values observed during the baseline period. During thiopental anaesthesia, after haemorrhage both constriction and dilation of arterioles (–20.3 (3.3); n=11, 3.9; n=2), venules (–7.9 (2.5); n=8, 3.7 (1.3); n=5) and capillaries (–12.8 (3.3); n=9, 4.7; n=2) was again observed.

The relationship between changes in arteriolar diameter in response to haemorrhage with the corresponding changes in venules and capillaries is shown in Figure 3 (A and B) for all animals (H and H–R) at t=40 min. Constriction of arterioles was generally accompanied by a smaller constriction of venules with all three agents, and this relationship was significant for propofol/fentanyl (r2=0.45, P<0.01), but not ketamine and thiopental (r2=0.18 and 0.08 respectively). However, arteriolar constriction was accompanied by a similar degree of capillary constriction during propofol/fentanyl anaesthesia (r2=0.54, P<0.01) and to a lesser extent during thiopental (NS). During ketamine anaesthesia arteriolar constriction tended to be accompanied by capillary dilation (r2=0.27, P<0.05). In all animals a similar pattern was still present at t=60 min (H and H–R before re-infusion).



View larger version (16K):
[in this window]
[in a new window]
 
Fig 3 Percentage change in arteriolar diameter (x-axis) against the percentage change in diameter of (A) venules and (B) capillaries (y-axis) after haemorrhage (t=40 min) for H and H–R animals receiving either propofol/fentanyl, ketamine or thiopental anaesthesia.

 
In H–R animals, with all three anaesthetic regimens re-infusion caused a significant change in vessel diameters towards pre-haemorrhage values (P<0.05, t=80, 90). However, this change was transient and within 10–20 min vessel diameters returned to post haemorrhage values.

Systolic arterial pressure versus vessel diameter
To demonstrate the variation in vessel diameter response after haemorrhage SAP vs percentage change in diameter of arterioles, venules and capillaries for the three anaesthetic agents are compared in Figure 4AC. In general, there was a constrictor response in all vessel types with decreasing SAP. However, a dilatory response was seen in arterioles during ketamine anaesthesia when SAP was maintained (NS) and also during thiopental anaesthesia when SAP was low (r2=0.42, P<0.05). In capillaries, dilation was seen as SAP decreased with ketamine (r2=0.81, P<0.001) and with thiopental (r2=0.41, P<0.05). No dilatory responses were found with propofol anaesthesia. In venules, a relationship was identified between the decrease in SAP and the percentage change in diameter during ketamine (r2=0.39, P<0.05) and thiopental (r2=0.42, P<0.05) but not during propofol/fentanyl anaesthesia.



View larger version (18K):
[in this window]
[in a new window]
 
Fig 4 Systolic arterial pressure (SAP, mm Hg) vs percentage change in vessel diameter after haemorrhage (t=40 min) for (A) arterioles, (B) venules and (C) capillaries during propofol/fentanyl (n=12, H and H–R), ketamine (n=13, H and H–R) and thiopental (n=13, H and H–R) anaesthesia. A negative value indicates vessel constriction.

 
Sodium nitroprusside
In C animals, SNP caused significant dilation of arterioles venules and capillaries during ketamine and thiopental, but not propofol/fentanyl anaesthesia at t=135 min (Table 1, P<0.05). The greatest magnitude of response was observed in arterioles. Responses in arterioles were greater after haemorrhage during propofol/fentanyl and ketamine anaesthesia than in control animals (Table 1, P<0.05). Responses in arterioles were greater in H–R animals during thiopental anaesthesia than in C animals (Table 1, P<0.05).


View this table:
[in this window]
[in a new window]
 
Table 1 Percentage change in arteriolar, venular and capillary diameter 5 min after topical application of sodium nitroprusside in control (C), haemorrhage (H) and haemorrhage re-infusion (H–R) groups during propofol/fentanyl (n=18), ketamine (n=19) and thiopental (n=19) administration. *P<0.05 significantly different from the end of the experimental study period, #P<0.05 significantly different from C animals
 
Macromolecular leak
In C animals arterioles, venules and capillaries exhibited no significant increase in macromolecular leak between t=0 and t=130 min during propofol/fentanyl ketamine or thiopental anaesthesia (Fig. 5AC). Haemorrhage did not induce a significant increase in macromolecular leak with any anaesthetic agent (Fig. 5AC). Re-infusion after haemorrhage caused a significant increase in macromolecular leak from venules during propofol/fentanyl and thiopental, but not ketamine anaesthesia (Fig. 5AC).



View larger version (24K):
[in this window]
[in a new window]
 
Fig 5 Percentage change in macromolecular leak in control (C), haemorrhage (H) and haemorrhage re-infusion (H–R) groups of animals during (A) propofol/fentanyl, (B) ketamine and (C) thiopental anaesthesia. Haemorrhage occurred between 30 and 40 min and reinfusion between 70 and 80 min. *P<0.05 significantly different from pre-haemorrhage (t=30 min). #P<0.05 significantly different from C.

 
Blood gases
In C animals, PO2, PCO2 and pH were within the normal physiological range at the end of the experimental protocol during propofol/fentanyl, ketamine and thiopental anaesthesia (Table 2). After haemorrhage, however, pH was significantly lower in H than C animals during propofol/fentanyl anaesthesia (P<0.05). Haematocrit was lower in H than in C animals during ketamine anaesthesia (Table 2, P<0.05).


View this table:
[in this window]
[in a new window]
 
Table 2 Blood gases and haematocrit at the end of the experimental period, including PO2 (kPa), PCO2 (kPa) and pH in control (C), haemorrhage (H) and haemorrhage re-infusion (H–R) groups during propofol/fentanyl (n=18), ketamine (n=19) and thiopental (n=19) administration. #P<0.05 significantly different from C animals
 
Plasma anaesthetic concentration
Plasma anaesthetic concentrations (µg ml–1) were not significantly different between C, H and H–R groups for each anaesthetic agent. Propofol concentrations were 6.39 (1.08) (C), 6.69 (0.92) (H) and 6.53 (0.62) (H–R). Plasma ketamine concentrations were 0.75 (0.12) (C), 0.85 (0.14) (H) and 0.87 (0.12) (H–R). Thiopental concentrations were 62.7 (2.9) (C), 63.5 (2.4) (H) and 61.1 (2.9) (H–R).


    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
This study has demonstrated for the first time that propofol/fentanyl, ketamine and thiopental differentially influence the response of the mesenteric microcirculation after haemorrhage. During ketamine anaesthesia, when SAP was maintained, dilation of mesenteric arterioles and venules was accompanied by constriction of capillaries. However, as SAP decreased, constriction of arterioles and venules was accompanied by dilation of capillaries. Different responses were observed during thiopental anaesthesia. At higher values of SAP there was constriction of arterioles and capillaries accompanied by dilation of venules, but, at lower pressures, arterioles and capillaries tended towards dilation and venules towards constriction. During propofol/fentanyl anaesthesia, SAP decreased in all animals and constriction of mesenteric arterioles, venules and capillaries was observed after haemorrhage. There was no evidence of a change in response at lower arterial pressures, as seen with thiopental.

Sympathetic nerves supply mesenteric microvessels to the level of terminal arterioles (>20 µm) and small collecting venules (>30 µm), thus, in the present study, constriction of vessels with similar diameters may have resulted from increased stimulation of the sympathetic nervous system after haemorrhage.1 During propofol/fentanyl anaesthesia, the magnitude of constriction observed in arterioles and venules was greater at lower arterial pressures, accompanied by tachycardia, and probably associated with greater stimulation of the baroreceptor reflex.1324

After haemorrhage, vasoconstriction of mesenteric vessels is a potential problem, increasing the risk of tissue ischaemia and gastrointestinal damage. In the current study, vasoconstriction was greatest during propofol/fentanyl anaesthesia. A greater decrease in systemic pH was observed after haemorrhage during propofol/fentanyl anaesthesia, compared with the other anaesthetic agents. This observation does not directly confirm mesenteric ischaemia, but indicates a generalized reduction in tissue perfusion and the occurrence of anaerobic respiration.25 26 The vasoconstriction observed after haemorrhage during propofol/fentanyl anaesthesia could be an explanation for the decrease in local pH reported in previous studies.4

During thiopental anaesthesia, the degree of constriction of arterioles and capillaries appeared to decrease as SAP reduced, in contrast to the propofol/fentanyl study. This may reflect the fact that fewer animals had low SAP but could also show that the mechanisms maintaining arterial pressure were better preserved during thiopental anaesthesia. Lower arterial pressures after haemorrhage may have increased stimulation of the sympathetic nervous system and constriction of the mesenteric feeding arterioles (not measured), accompanied by reduced constriction and/or dilation of arterioles and capillaries to maintain adequate tissue perfusion. Mediators of vasodilation are probably not produced by the connective tissue of the mesentery, thus the observed vasodilation was probably myogenic in origin.24 The vasodilator responses observed during thiopental anaesthesia may have reduced tissue hypoxia, and in agreement with this, systemic pH was maintained after haemorrhage. However, the reduction in both tachycardia and the vessel response to decreased SAP indicated a degree of baroreceptor reflex inhibition of sympathetic outflow by thiopental, in agreement with previous studies.6 27

During ketamine anaesthesia, dilation of arterioles and venules and constriction of capillaries occurred when arterial pressure was maintained. This was presumably due to sympathomimetic effects of ketamine, maintaining systemic arterial pressure after haemorrhage, with reduced sympathetic stimulation required to maintain SAP.7 The vasodilator responses may have been myogenic in origin. For example, a decrease in intravascular pressure resulting from the decrease in systemic pressure and reinforced by constriction of larger arterial vessels may well result in myogenic dilation. The microcirculatory response to haemorrhage was altered at lower arterial pressures, with constriction of arterioles and venules accompanied by dilation of capillaries. Dilation of capillaries, when blood flow was decreased due to upstream vasoconstriction of arterioles, may have allowed the maintenance of local tissue perfusion. In the present study, this was confirmed by the maintenance of systemic pH with ketamine anaesthesia.26 These observations are in agreement with previous studies in skeletal muscle, demonstrating well-maintained capillary perfusion with ketamine anaesthesia.28 Constriction of arterioles and venules may have also maintained hydrostatic pressure and prevented fluid loss into the mesenteric extravascular space and hence, the abdomen. Interestingly, after haemorrhage fluid absorption appeared to be greater during ketamine anaesthesia,29 as indicated in the present study by the greater decrease in haematocrit. Nevertheless, it would be expected that changes in capillary diameter are passive and their response should reflect upstream changes in blood flow. This was true during propofol/fentanyl and thiopental anaesthesia. Their response during ketamine anaesthesia however, confirms that perhaps some of the vessels described as ‘capillaries’ in this study were indeed precapillary arterioles, as these possess {alpha}2 adrenoreceptors and thus may dilate in response to haemorrhage.30 Dilation of precapillary arterioles would also have contributed to the maintenance of capillary perfusion previously described.

The infusion rate of each of the three agents was continually adjusted to achieve the appropriate plane of anaesthesia. The resulting plasma concentrations for each agent did not differ between the study groups suggesting that a constant level was achieved. The concentrations were within the range previously described for these agents in rats. Fentanyl was added to the propofol infusion protocol after our preliminary studies demonstrated that the dose of propofol alone required to maintain a suitable depth of anaesthesia for this study produced cardiovascular instability.

In control animals anaesthetized with thiopental, topical application of sodium nitroprusside (SNP) induced vasodilation in arterioles and capillaries, with a reduced response observed in venules. During propofol/fentanyl anaesthesia this response was inhibited, as arterioles and capillaries did not dilate in response to SNP. Thus, propofol/fentanyl anaesthesia may be inhibiting the response of the microcirculation to nitric oxide.8 Alternatively, the absence of dilation may have been due to the vasodilator effects of propofol resulting in maximal vessel diameter.8 Interestingly, our study demonstrated that arteriolar and venular diameters were greatest during propofol/fentanyl anaesthesia. Inhibition of the response to nitric oxide however, would confirm previous studies in the rat aorta in vitro whereby propofol was shown to inhibit vasodilation in response to SNP.16 In the mesentery, nitric oxide is released in response to haemorrhage and assists with the maintenance of tissue perfusion.15 31 32 Propofol may inhibit the response to nitric oxide in vitro and this mechanism could explain the observed increase in constriction after haemorrhage during propofol/fentanyl anaesthesia.16 This hypothesis could not be confirmed in the present study as, after haemorrhage, SNP caused greater dilation of arterioles and capillaries than in control animals, indicating that the response of the mesenteric microcirculation to nitric oxide after haemorrhage is not inhibited by propofol/fentanyl anaesthesia.

This study has also demonstrated that macromolecular leak occurs in response to 10% haemorrhage when followed by re-infusion during propofol/fentanyl and thiopental anaesthesia, indicating compromised vessel integrity and damage of the mesenteric microcirculation.15 Macro molecular leak has previously been reported in the mesentery, after ischaemia reperfusion injury.19 Macro molecular leak was not observed in response to haemorrhage and re-infusion, during ketamine anaesthesia, perhaps indicating activation of a protective mechanism. It has previously been suggested that ketamine prevents leukocyte-endothelial interactions, which may be responsible for tissue and endothelial-cell damage.33

In conclusion, propofol/fentanyl, ketamine and thiopental have differential effects on the mesenteric microcirculation in response to haemorrhage, which relate to the decrease in systemic arterial pressure. The responses observed indicate that propofol/fentanyl promotes possible detrimental responses after haemorrhage, including greater decreases in SAP, systemic pH and increased mesenteric vasoconstriction. Ketamine, however, appears to exhibit a number of beneficial effects, including the maintenance of SAP, and the prevention of macromolecular leak after haemorrhage and re-infusion. In the present study therefore, ketamine appears to be the most suitable i.v. anaesthetic agent for use during haemorrhage.


    Acknowledgements
 
This work was supported by a British Journal of Anaesthesia Project Grant. The authors thank Mr Peter Henderson for his technical assistance with the HPLC and GC techniques.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Furness JB, Marshall JM. Correlation of the directly observed responses of mesenteric vessels of the rat to nerve stimulation and noradrenaline with the distribution of adrenergic nerves. J Physiol 1974; 239: 75–88[ISI][Medline]

2 Flint LM, Cryer HM, Simpson CJ, Harris PP. Microcirculatory norepinephrine constrictor response in haemorrhagic shock. Surgery 1984; 96: 240–7[ISI][Medline]

3 Kitajima T, Tani K, Yamaguchi Y, et al. Reperfused rat gut elaborates PAF that chemoattracts and primes neutrophils. Digestion 1995; 56: 111–16[ISI][Medline]

4 Hong S, Chon J, Choi J, Moon S. The comparative effects of dobutamine on the recovery of gastric intramucosal pH during hemorrhagic shock. Proc Twelfth World Cong Anesthesiol 2000 (abstract)

5 Krassioukov AV, Gelb AW, Weaver LC. Actions of propofol on central sympathetic mechanisms controlling blood pressure. Can J Anaesth 1993; 40: 761–9[Abstract]

6 Ebert TJ, Kanitz DD, Kampine JP. Inhibition of sympathetic neural outflow during thiopental anesthesia in humans. Anesth Analg 1990; 71: 319–26[Abstract]

7 Ivankovich AD, Miletich DJ, Reimann C, Albreich RF, Zahad B. Cardiovascular effects of centrally administered ketamine in goats. Anesth Analg 1974; 53: 924–33[ISI][Medline]

8 Holzmann A, Schmidt H, Gebhardt MM, Martin E. Propofol-induced alterations in the microcirculation of hamster striated muscle. Br J Anaesth 1995; 75: 452–6[Abstract/Free Full Text]

9 Longnecker DE, Miller FN, Harris DE. Small artery and vein response to ketamine HCl in the bat wing. Anesth Analg 1974; 53: 64–8[ISI][Medline]

10 Wada DR, Harashima H, Ebling WF, Osaki EW, Stanski DR. Effects of thiopental on regional blood flows in the rat. Anesthesiol 1996; 84: 596–604

11 Hirota K, Lambert DG. I.V. anaesthetic agents inhibit dihydropyridine binding to L-type voltage sensitive Ca2+ channels in rat cerebrocortical membranes. Br J Anaesth 1996; 77: 248–53[Abstract/Free Full Text]

12 Brookes ZLS, Brown NJ, Reilly CS. Intravenous anaesthesia and the rat microcirculation: the dorsal microcirculatory chamber. Br J Anaesth 2000; 85: 901–3[Abstract/Free Full Text]

13 Marshall JM. The influence of the sympathetic nervous system on individual vessels of the microcirculation of skeletal muscle of the rat. J Physiol 1982; 332: 169–86[ISI][Medline]

14 Boillot A, Laurant P, Berthelot A, Barale F. Effects on propofol on vascular reactivity in isolated aortae from normotensive and hypertensive rats. Br J Anaesth 1999; 83: 622–9[Abstract/Free Full Text]

15 Koch MA, Hassler EH, Schadt JC. Influence of nitric oxide on the hemodynamc response to hemorrhage in conscious rabbits. Am J Physiol 1995; 268: R171–82

16 Miyawaki I, Nakamura K, Terasako K, Toda H, Kakuyama M, Mori K. Modification of endothelium-dependent relaxation by propofol, ketamine, and midazolam. Anesth Analg 1995; 81: 474–9[Abstract]

17 Russell DH, Baretto JC, Klemm K, Miller TA. Hemorrrhagic shock increases gut macromolecular permeability in the rat. Shock 1995; 4: 50–5[ISI][Medline]

18 Miller FN, Joshua IG, Anderson GL. Quantification of vasodilator induced macromolecular leakage by in vivo microscopy. Microvas Res 1982; 24: 56–67[ISI]

19 Johnston B, Gaboury JP, Suematsu M, Kubes P. Nitric oxide inhibits microvascular protein leakage by leukocyte adhesion-independent and adhesion-dependent inflammatory mediators. Microcirculation 1999; 6: 153–62[ISI][Medline]

20 Wilder-Baker JW, Deitch EA, Berg RD, Specian RD. Haemorrhagic shock induces bacterial translocation from the gut. J Trauma 1988; 28: 896–906[ISI][Medline]

21 Lumb WV, Wynn Jones E, eds. In: Veterinary Anaesthesia. Lea and Febiger: Philadelphia, Pennsylvania, 1973

22 Zweifach BW, Lowenstein B, Chambers R. Response of blood capillaries to acute hemorrhage in the rat. Am J Physiol 1944; 142: 80–93[Free Full Text]

23 Asher EF, Alsip NL, Zhang PY, Harris PD. Prostaglandin related microvascular dilation in pentobarbital and etomidate anesthetised rats. Anesthesiol 1992; 76: 271–8[ISI][Medline]

24 Hébert MT, Marshall JM. Direct observations of the effects of baroreceptor stimulation on mesenteric circulation of the rat. J Physiol 1988; 400: 29–44[Abstract]

25 Seyde WC, McGowan L, Lund N, Duling B, Longnecker DE. Effects of anesthetics on regional hemodynamics in normovolemic and hemorrhaged rats. Am J Physiol 1985; 249: H164–73[Abstract/Free Full Text]

26 Evans TW, Smithies M. Organ dysfunction. Br Med J 1999; 318: 1606–9[Free Full Text]

27 Scheiffer GJ, Ten Voorde BJ, Karamarker JM, Ras HH, De Lange JJ. Effects of thiopentone, etomidate and propofol on beat to beat cardiovascular signals in man. Anaesthesia 1993; 48: 849–55[ISI][Medline]

28 Gustafsson U, Sjöberg F, Lewis DH, Thorborg P. Influence of pentobarbital, propofol and ketamine on skeletal muscle capillary perfusion during haemorrhage: a comparative study in the rabbit. Int J Microcirc 1995; 15: 163–9[ISI][Medline]

29 Lundvall J, Hillman J. Fluid transfer from skeletal muscle to blood during haemorrhage. Importance of beta-adrenergic vascular mechanisms. Acta Physiol Scand 1978; 102: 450–8[ISI][Medline]

30 Faber JE. In situ analysis of {alpha}-adrenoceptors on arterial and venular smooth muscle in rat skeletal muscle microcirculation. Circ Res 1988; 62: 37–50[Abstract]

31 Ekelund U, Mellander S. Endogenous nitric oxide release as a physiological regulator of vascular tone in cat skeletal muscle during haemorrhage. Acta Physiol Scand 1996; 157: 471–9[ISI][Medline]

32 Bertuglia A, Calantuoni A. Venular oscillatory flow during hemorrhagic shock and NO inhibition in hamster cheek pouch microcirculation. Microvasc Res 1997; 54: 233–42[ISI][Medline]

33 Schmidt H, Ebeling D, Bauer H, et al. Ketamine attenuates endotoxin induced leukocyte adherence in rat mesenteric venules. Crit Care Med 1995; 23: 2008–14[ISI][Medline]