LABORATORY INVESTIGATIONS Concentration-dependent inotropic effects of halothane, isoflurane and sevoflurane on rat ventricular myocytes

British Journal of Anaesthesia 82 (5): 723 30 (1999) LABORATORY INVESTIGATIONS Concentration-dependent inotropic effects of halothane, isoflurane and sevoflurane on rat ventricular myocytes L. A. Davies, D. L. Hamilton, P. M. Hopkins 1, M. R. Boyett and S. M. Harrison* School of Biomedical Sciences, University of Leeds, Leeds LS2 9NQ, UK Present address: 1 Academic Unit of Anaesthesia, University of Leeds, Leeds LS2 9NQ, UK *Corresponding author We have described the concentration-dependent inotropic effects of halothane, isoflurane and sevoflurane on rat ventricular cells and investigated the role of the sarcoplasmic reticulum (SR) in these inotropic actions. Single ventricular myocytes, isolated from rat hearts, were stimulated electrically at 1 Hz and contractions recorded optically. Cells were exposed to a range of concentrations of halothane, isoflurane or sevoflurane for a period of 1 min to determine the concentration-dependency of their inotropic actions. For each anaesthetic, the peak negative inotropic action was determined early during an exposure, and sustained negative inotropic action was measured at steady-state just before wash-off. In some experiments, cells were equilibrated with ryanodine 1 mol litre 1 to investigate the role of the SR in these intropic effects. Halothane caused a concentration-dependent initial increase in contractions (to mean 130 (SEM 28)% at 10 mmol litre 1 ) followed by rapid onset of a negative inotropic effect (K 0.5 0.34 mmol litre 1 for peak effect; K 0.5 0.46 mmol litre 1 for sustained effect). Exposure to isoflurane induced a small potentiation of contractions in some cells, followed by a concentration-dependent decrease in contraction in all cells (K 0.5 0.85 mmol litre 1 for peak effect; K 0.5 1.92 mmol litre 1 for sustained effect); contractions recovered partially during a 1-min exposure. On wash-off, contractions were increased transiently above control. Sevoflurane caused a large initial decrease in contraction which then returned rapidly towards control (K 0.5 0.2 mmol litre 1 for peak effect; K 0.5 2.57 mmol litre 1 for sustained effect). In common with isoflurane, removal of sevoflurane caused a transient increase in contractions above control. After exposure to ryanodine, the positive inotropic effects of halothane and isoflurane did not occur, and recovery of contractions during exposure to isoflurane and sevoflurane was abolished as was the transient increase in contractions seen on wash-off, indicating that these effects were mediated via the SR. Halothane had the most potent sustained negative inotropic effect but there was little difference between the negative inotropic effects of isoflurane and sevoflurane at clinically relevant concentrations. At higher concentrations, sevoflurane caused a less potent negative inotropic effect than isoflurane. The SR plays a major role in the effects of all three anaesthetics. One possible mechanism underlying the initial potentiation of contraction by halothane (and isoflurane) may be sensitization of the Ca 2 -induced Ca 2 -release process of the SR. Br J Anaesth 1999; 82: 723 30 Keywords: heart, inotropism; anaesthetics volatile, sevoflurane; anaesthetics volatile, halothane; anaesthetics volatile, isoflurane; pharmacology, ryanodine; heart, myocytes; rat Accepted for publication: December 16, 1998 Presented in part to the Physiological Society meeting, Bristol, UK, 1997 British Journal of Anaesthesia

Davies et al. The commonly used volatile anaesthetics, halothane, isoflurane and sevoflurane, in addition to producing unconsciousness, have adverse side effects, one of which is a potent negative inotropic effect on the heart. 1 4 It is generally thought that halothane and enflurane (at eqi-anaesthetic concentrations) have greater depressant effects on the heart than isoflurane. 56 However, much less is known about the inotropic actions of sevoflurane. Sevoflurane has been reported to have either equivalent 3 4 or less depressant effects 5 than isoflurane in working heart preparations and isolated tissue. The precise mechanisms underlying the cardiac depressant effects of these agents have still to be elucidated although it is possible that myocardial depression is caused by the combined effects of a reduction in (i) transsarcolemmal L-type Ca 2 current (I Ca ), 7 12 (ii) sarcoplasmic reticulum (SR) Ca 2 content 13 and (iii) the sensitivity of the myofilaments to Ca 2. 21415 The contribution of each of these mechanisms to the negative inotropic effect of halothane, isoflurane and sevoflurane still needs to be established. In this study we have focused on the role of the SR in the inotropic effects of these agents. Previous studies have shown that initial exposure to halothane leads to a transient increase in contractions 16 18 before the well described negative inotropic effect of halothane becomes evident. This response, which has not been observed with isoflurane, 18 is similar to that seen during exposure to a low concentration (e.g. 0.1 0.25 mmol litre 1 ) of caffeine. 19 It has been suggested that a low concentration of caffeine leads to sensitization of the Ca 2 -induced Ca 2 -release (CICR) mechanism of the SR such that a greater fraction of SR Ca 2 content is released by any given I 19 Ca (the normal trigger for SR Ca 2 release). If this is the case with halothane, it may be expected that inhibiting SR function (e.g. with ryanodine) 20 21 would block the initial positive inotropic effect of halothane. The aims of this study were to describe the positive and negative inotropic effects of halothane, isoflurane and sevoflurane in intact electrically stimulated single ventricular myocytes over a range of concentrations which included those expected in the clinical setting. A further aim was to investigate the role of the SR in the positive and negative inotropic effects of halothane, isoflurane and sevoflurane. These data allow direct comparison of the inotropic effects of these agents and give further insight into the mechanisms contributing to their depressant action on the heart. Materials and methods Cell isolation The technique used to prepare rat ventricular myocytes has been described previously. 22 Briefly, rats (250 300 g in weight) were killed (under Home Office licence) by a blow to the head followed by cervical dislocation, and the heart was excised rapidly into an isolation solution (see below for composition), supplemented with CaCl 2 750 mol litre 1 and equilibrated with 100% oxygen. The heart was perfused via the coronary arteries with the above solution for 4 min to flush the heart of blood, followed by 4 min of perfusion with the isolation solution to which Na 2 EGTA 100 mol litre 1 was added. The heart was then perfused for 9 min with the isolation solution supplemented with collagenase 1 mg ml 1 (type 1, Worthington Biochemical Corp., NJ, USA) and protease 0.1 mg ml 1 (Sigma, type XIV) after which the ventricles were cut from the heart, chopped finely and shaken in enzyme solution (to which 1% BSA was added) for 5-min intervals. Dissociated cells were harvested by filtration at the end of each 5-min digestion and the remaining tissue returned for further enzyme treatment. The dissociated cells were centrifuged at 30 g for 40 s and resuspended in a physiological salt solution (see below) containing CaCl 2 1 mmol litre 1 and stored at 4 C until required. The yield of viable cells (i.e. those with rod-shaped morphology and clear striations) was typically 60 70%. Solutions The isolation solution was composed of (in mmol litre 1 ): NaCl 130; KCl 5.4; MgCl 2 1.4; NaH 2 PO 4 0.4; HEPES 5; glucose 10; taurine 20; creatine 10; ph 7.3 at 24 C. After dissociation, cells were stored in and subsequently perfused with a physiological salt solution of the following composition (in mmol litre 1 ): NaCl 140; KCl 5.4; MgCl 2 1.2; NaH 2 PO 4 0.4; HEPES 5; glucose 10; CaCl 2 1; ph 7.4 at 30 C. Various concentrations of halothane (0.01 10 mmol litre 1 ), isoflurane (0.01 10 mmol litre 1 ) and sevoflurane (0.01 4 mmol litre 1 ) were delivered from stock solutions made up in dimethyl sulphoxide. These encompass the MAC 50 values for humans and rats: halothane 0.19/0.27 mmol litre 1 ; isoflurane 0.27/0.31 mmol litre 1 ; and sevoflurane 0.3/0.35 mmol litre 1. After dilution of the stock solutions, the final concentration of dimethyl sulphoxide in the superfusate never exceeded 0.75%, a concentration that had no significant effect on contractions (not shown). Ryanodine (Calbiochem, Nottingham, UK) was delivered from a stock solution of 1 mmol litre 1 in water and was used at a final concentration of 1 mol litre 1. Unless otherwise stated, all solution constituents were from Sigma (Poole, UK). Recording cell length Freshly dissociated cells were transferred to a small tissue chamber (volume 0.1 ml) attached to the stage of an inverted microscope (Nikon Diaphot). The cells were allowed to settle for several minutes onto the glass bottom of the chamber before being superfused at a rate of approximately 3mlmin 1 with the physiological salt solution. Solutions were delivered to the experimental chamber by magnetic drive gear metering pumps (Micropump, Concord, CA, USA) and solution level and temperature were maintained by feedback circuits. 23 724

Volatile anaesthetics and cardiac contraction Fig 1 Inotropic effects of a range of concentrations of halothane. Slow time base recordings of cell length are shown. Contraction results in a shortening of the cell, that is a downward deflection. Halothane 0.1, 0.5, 1 and 10 mmol litre 1 was applied for 1 min as shown by the bar. Data recorded from four different cells. Cells with clear striation patterns and normal rod-shaped morphology were selected for experimentation and were stimulated electrically at a frequency of 1 Hz (stimulus duration, 2 ms) via two platinum electrodes situated in the sides of the chamber. Cell length was recorded continuously using an optical system based on a photodiode array 24 and displayed on a chart recorder (Gould 2600S). A sample and hold circuit 25 was used to display active shortening of a cell during each contraction (twitch shortening) on the chart recorder. This circuit has the effect of excluding changes in resting cell length, although the time course of the twitch is recorded faithfully. Cell length and twitch shortening were recorded using a pulse code modulator (Neuro-Corder DR-890, Neuro Data Instruments Corp., NY, USA) coupled to a standard VHS video recorder. Statistical analysis and curve-fitting Data are presented as mean (SEM) and statistical comparisons were carried out using the Student s paired or unpaired t tests as appropriate. SigmaPlot software (Jandel Scientific) was used to create curve fits (least-squares fits) to the experimental data shown in Figures 2, 4 and 6 using the following equation: Y/Y max [X] n /([X] n K n d ) where [X] represents an anaesthetic concentration, n is the Hill coefficient and K n d is the concentration at which the half maximal effect was observed (referred to subsequently as K 0.5 ). Results Halothane Figure 1 illustrates slow time base records of cell length before, during and after application of a range of concentrations of halothane. Halothane initially potentiated Fig 2 Dose-dependence of the positive and negative inotropic effects of halothane. A: The greatest increase in contraction observed on application of halothane plotted against halothane concentration. B: The greatest negative inotropic effect of halothane early during exposure (peak effect) and the negative inotropic effect observed just before wash-off of halothane (sustained effect) plotted against halothane concentration. Data are expressed as mean (SEM) percentage of control for 6 49 cells. The solid lines represent the best fit to each data set using the equation described in the Methods. The broken lines represent the MAC 50 values for rats (R) and humans (H) in terms of an aqueous concentration 30 and the horizontal bar, the range of arterial halothane concentrations seen in a group of patients. 31 contractions and this was followed by a negative inotropic effect which was maintained for the duration of the exposure. 4171826 In most cases there was a further transient depression of contractions on wash-off of halothane after which contractions recovered to control values. The positive and negative inotropic effects tended to be greater at higher concentrations. Figure 2A shows the concentrationdependency of the initial positive inotropic effect of halothane; at 0.01 mmol litre 1, this effect was very small but at higher concentrations the magnitude of this effect was enhanced such that at 10 mmol litre 1, contraction was increased by 130 (28)% compared with controls. Figure 2B shows the effect of a range of concentrations of halothane on the decrease in contraction. Two measurements were made during exposure to halothane: (i) the peak negative inotropic response which occurred early during exposure and (ii) the sustained negative inotropic effect just 725

Davies et al. Table 1 Mean (SEM) K 0.5 values for negative and positive inotropic actions of halothane, isoflurane and sevoflurane (see text for further details) Peak negative Sustained negative Positive inotropic inotropic effect inotropic effect effect on wash-off (mmol litre 1 ) (mmol litre 1 ) (mmol litre 1 ) Halothane 0.34 (0.07) 0.46 (0.05) Not present Isoflurane 0.85 (0.23) 1.89 (0.19) 0.86 (0.24) Sevoflurane 0.20 (0.03) 2.57 (0.7) 0.22 (0.05) Fig 3 Inotropic effects of a range of concentrations of isoflurane. Slow time base recordings of cell length are shown before, during and after a 1-min exposure (as shown by the bar) to isoflurane 0.3, 1, 3 and 10 mmol litre 1. Data recorded from three different cells. before wash-off. Mean data at each concentration were fitted with a typical dose response curve (see Materials and methods) yielding a K 0.5 of 0.34 (0.07) mmol litre 1 for the peak effect and 0.46 (0.05) mmol litre 1 for the sustained effect (see Table 1). This suggests that during exposure to halothane, contraction recovers slightly (this can be seen in the trace during exposure to halothane 1 mmol litre 1 shown in Fig. 1; see also Fig. 7A). Isoflurane Figure 3 shows slow time base records of contractions during 1-min exposure to isoflurane 0.3, 1, 3 and 10 mmol litre 1. Application of isoflurane occasionally led to a small potentiation of contractions (see examples at 0.3 and 10 mmol litre 1 in Fig. 3). This effect was much less marked than that seen with halothane and in most cases it was difficult to determine if any positive inotropic effect occurred given that a small degree of beat-to-beat variability in the size of contractions was observed under control conditions. Early during exposure to isoflurane, contractions diminished but then recovered partially towards control values. This effect was most apparent at lower isoflurane concentrations (see Fig. 3). On wash-off, contractions increased transiently above control before declining. Again, the inotropic effects of isoflurane tended to be greater at higher concentrations. Fig 4 Dose-dependence of the negative and positive inotropic effects of isoflurane. A: Decrease in contraction observed early during exposure to isoflurane (peak effect) and just before wash-off of isoflurane (sustained effect) plotted against isoflurane concentration. B: The greatest increase in contraction (expressed as a percentage of control) observed following wash-off of isoflurane plotted against isoflurane concentration. Data are expressed as mean (SEM) percentage of control for 2 37 cells. The solid lines represent the best fit to each data set using the equation described in the Methods. The broken lines represent the MAC 50 values for rats (R) and humans (H) in terms of an aqueous concentration. 30 Figure 4A shows mean data describing the peak (measured shortly after exposure) and sustained (measured just before wash-off) decreases in contraction induced by a range of concentrations of isoflurane. Curve fits to these data gave K 0.5 values of 0.85 (0.23) and 1.89 (0.19) mmol litre 1 (see Table 1) for peak and sustained decreases in contraction, respectively. Figure 4A illustrates that at low concentrations ( 0.3 mmol litre 1 ) the initial depression of contractions was almost fully reversible during a 1-min exposure and only at higher concentrations ( 0.3 mmol litre 1 ) did depression in contractility persist. Unlike halothane, washoff of isoflurane led to a transient increase in contractions above control. Figure 4B shows the concentrationdependency of the increase in contractions seen on wash-off of isoflurane. At low concentrations ( 0.1 mmol litre 1 ), this effect was small but it increased at higher concentrations. The K 0.5 for this response was 0.86 (0.24) mmol litre 1, a value very similar to that seen for the peak negative inotropic effect of isoflurane (see Table 1). 726

Volatile anaesthetics and cardiac contraction Fig 5 Inotropic effects of a range of concentrations of sevoflurane. Slow time base recordings of cell length are shown before, during and after a 1-min exposure (as shown by the bar) to sevoflurane 0.03, 0.25, 1.5 and 4 mmol litre 1. Data recorded from three different cells. Fig 6 Dose-dependence of the negative and positive inotropic effects of sevoflurane. A: Decrease in contraction observed early during exposure to sevoflurane (peak effect) and just before wash-off of sevoflurane (sustained effect) plotted against sevoflurane concentration. B: The greatest increase in contraction (expressed as a percentage of control) observed following wash-off of sevoflurane plotted against sevoflurane concentration. Data are expressed as mean (SEM) percentage of control for 3 11 cells. The solid lines represent the best fit to each data set using the equation described in the Methods. The broken lines represent the MAC 50 values for rats (R) and humans (H) in terms of an aqueous concentration. 30 Sevoflurane Figures 5 and 6 show the results of similar experiments with sevoflurane. Figure 5 shows that when sevoflurane was first applied there was no potentiation of contractions (cf. halothane, Fig. 1) but, as with isoflurane, there was a marked initial depression in contractions followed by recovery towards control. At higher concentrations, the decrease in contractions was more sustained during continued exposure to sevoflurane. The extent to which contractions recovered during exposure is shown in Figure 6A. The concentration response curve describing the peak negative inotropic effect had a lower slope and K 0.5 than that for the sustained effect; K 0.5 for the peak effect was 0.20 (0.03) mmol litre 1 but an order of magnitude higher for the sustained effect (2.57 (0.70) mmol litre 1 ) (see Table 1). Thus at low concentrations ( 0.1 mmol litre 1 ), after an initial depressant effect, sevoflurane had no significant sustained negative inotropic effect. On wash-off of sevoflurane, contractions increased dramatically before returning to control (Fig. 5). This positive inotropic effect was greater following wash-off of higher concentrations of sevoflurane; K 0.5 of the effect was 0.22 (0.05) mmol litre 1 (Fig. 6B), a value very similar to the K 0.5 of the initial negative inotropic effect (0.20 mmol litre 1 ; Table 1). This response was apparent even at concentrations that had little or no sustained negative inotropic effect. For example, at 0.1 mmol litre 1, these data suggest that there is little or no sustained decrease in contraction during exposure to sevoflurane, but a transient increase of approximately 30% of control on removal of sevoflurane. The role of the SR Experiments were carried out to determine the role of the SR in the positive and negative inotropic effects of the three anaesthetics. The upper trace in Figure 7A shows a typical response to application of halothane 0.5 mmol litre 1 : an initial positive inotropic effect followed by a sustained depression in contractions which returned to control when halothane was removed. In nine cells, early during application of halothane 0.5 mmol litre 1, contractions were reduced to 25 (4)% of control. There was a small but significant recovery of contractions (to 28 (4)%; P 0.011, paired t test) during exposure to halothane 0.5 mmol litre 1. Cells were then exposed to ryanodine 1 mol litre 1 to inhibit the SR. 20 21 After equilibration with ryanodine, cell contraction declined to 13 (2)% (n 26) of control as the SR Ca 2 load declined (not shown). When a new steady state was reached, the cell was exposed again to the same concentration of halothane. The lower trace in Figure 7A shows that, after equilibration with ryanodine, the initial positive inotropic effect of halothane was abolished, suggesting it is mediated via the SR. In nine cells following ryanodine treatment, halothane depressed contractions to 23 (3)%, a similar depression in proportionate terms to that observed under control conditions and there was no significant recovery of contractions during exposure to halothane (P 0.43). These data suggest that in the presence of halothane, depression of Ca 2 entry 727

Davies et al. following ryanodine treatment, the extent of recovery of contractions during exposure to sevoflurane was greatly reduced compared with control (peak and sustained negative inotropic effects were 33 (4)% and 42 (3)%, respectively) but recovery of contractions during sevoflurane exposure was still significant (P 0.018, paired t test). The increase in contractions on wash-off of sevoflurane was abolished by ryanodine. All ryanodine sensitive responses to halothane, isoflurane and sevoflurane are presumed to be mediated via the SR. Fig 7 Effects of ryanodine on the inotropic effects of halothane, isoflurane and sevoflurane. The upper traces of each panel show slow time base recordings of change in cell length before, during and after a 1-min exposure to halothane 0.5 mmol litre 1 (A), isoflurane 1 mmol litre 1 (B) and sevoflurane 0.6 mmol litre 1 (C), and the lower traces in each panel show the effects of exposure to the same concentration of each anaesthetic after equilibration of the cell with ryanodine 1 mol litre 1. rather than the SR determines the extent of the sustained negative inotropic effect. Similar experiments were carried out with cells exposed to isoflurane 1 mmol litre 1 (Fig. 7B). In seven cells, the peak negative inotropic effect of isoflurane reduced contractions to 62 (5)% of control but, before washout, contractions had recovered to 75 (8)% of control (P 0.026, paired t test). The lower panel of Figure 7B shows that ryanodine abolished both the recovery of contractions that took place during exposure to isoflurane (peak and sustained negative inotropic effects of isoflurane were 38 (7)% and 41 (8)%, respectively; P 0.5, Wilcoxon signed rank test) and the transient increase in contractions on wash-off. Unlike halothane, the sustained negative inotropic effect of isoflurane was significantly greater in the presence of ryanodine, suggesting that the SR is capable of ameliorating the negative inotropic effect of isoflurane under control conditions. In 10 control cells, on exposure to sevoflurane 0.6 mmol litre 1, contraction was initially decreased to 30 (3)% but then recovered to 59 (5)% of control before washoff (Fig. 7C; P 0.001, paired t test). In the same cells Discussion The aim of this study was to quantitatively define and compare the concentration-dependence of the positive and negative inotropic effects of halothane, isoflurane and sevoflurane, to determine their relative potencies and investigate the role of the SR in these inotropic effects. In previous studies, papillary muscles 2 4 27 or Purkinje fibres 6 were used and in most cases a limited range of anaesthetic concentrations were studied. Our experiments were carried out in single ventricular myocytes and between seven and nine concentrations of halothane, isoflurane and sevoflurane were used to define the concentration-dependency of the positive and negative inotropic effects of these agents. Single ventricular myocytes are an ideal preparation for investigating the mechanisms involved in the inotropic actions of anaesthetics, because the anaesthetic equilibrates rapidly throughout the entire preparation and therefore minimizes any problems associated with diffusion delays which could occur in multicellular preparations. Furthermore, single cells appear to respond to inotropic interventions in qualitatively the same way as multicellular preparations. However, these experiments were carried out at 30 C on rat ventricular myocytes stimulated at 1 Hz and as there are species differences in action potential configuration and Ca 2 handling, it is possible that quantitatively different results may be found in other mammalian species. Comparison of inotropic effects Figures 2B, 4Aand 6A show concentration response curves describing the sustained negative inotropic effect of the three volatile anaesthetics. These concentration response curves showed that contraction was reduced to 50% of control by halothane 0.5 mmol litre 1, isoflurane 1.9 mmol litre 1 and sevoflurane 2.6 mmol litre 1. At higher concentrations, between 1 and 10 mmol litre 1 (which for the most part is in excess of the usual clinical range) halothane had the most potent action followed by isoflurane and then sevoflurane, with isoflurane and sevoflurane having similar potencies. These results are consistent with studies carried out in whole heart preparations 3 5 and isolated papillary muscles. 4 At lower concentrations (less than 1 mmol litre 1, i.e. those within the usual clinical range), halothane had the most potent effect followed by sevoflurane and then isoflurane. 728

Volatile anaesthetics and cardiac contraction With isoflurane and sevoflurane, after initial depression, contractions recovered towards control. At concentrations of 0.1 mmol litre 1 or less, after an initial decrease, contractions returned to close to control values (see Figs 4A, 6A) although the initial depression of contraction was greater with sevoflurane. Removal of sevoflurane caused a much larger positive inotropic effect than that seen on wash-off of isoflurane. Table 1 illustrates that the K 0.5 values of the peak negative and positive inotropic actions for both isoflurane and sevoflurane were similar. As such it is tempting to speculate that the mechanism(s) which contributes to the peak negative inotropic response is reversed on washout of the anaesthetic such that a positive inotropic response is induced. Role of the SR in the inotropic effects The role of the SR in the positive and negative inotropic effects of halothane, isoflurane and sevoflurane was elucidated from experiments in which ryanodine was used to disable the normal function of the SR. Contraction remaining in the presence of ryanodine reflects mainly Ca 2 entry 28 (predominately via I Ca ). In these experiments, ryanodine reduced contractions to 13 (2)% of control, illustrating that the SR makes a major contribution to the intracellular Ca 2 transient which underlies contraction in rat ventricular myocytes. Figure 1 illustrates that the initial application of halothane caused a large transient increase in the size of contractions (by 130 (28)% with halothane 10 mmol litre 1 ), similar to that observed previously. 16 18 However, Wheeler and colleagues 18 stated that isoflurane did not exhibit this behaviour, but in our study, a much smaller but discernible initial positive inotropic effect was seen in some cells with very stable and reproducible contractions under control conditions. However, this effect was absent or difficult to detect in cells exposed to sevoflurane. Ryanodine abolished the initial increase in contractions observed on exposure to halothane suggesting that this was mediated via the SR. This initial positive inotropic effect was very similar to that seen on exposure of rat ventricular myocytes to a low concentration of caffeine (e.g. 0.25 mmol litre 1 ). 19 For caffeine, O Neill and Eisner 19 suggested that its initial positive inotropic effect may result from sensitization of the CICR mechanism of the SR such that for any given Ca 2 trigger (i.e. I Ca ) a greater than normal Ca 2 release is induced from the SR. The sensitivity of the halothaneinduced positive inotropy to ryanodine supports the possibility that halothane has a similar action on the mechanism of CICR, most probably at the level of the SR Ca 2 -release channel (ryanodine receptor). Furthermore, wash-off of halothane led to a further decrease in contraction before recovery towards control, a behaviour similar to that observed on wash-off of caffeine 0.25 mmol litre 1. 19 If halothane was sensitizing CICR, and this effect was reversed rapidly when halothane was removed, this would reduce the sensitivity of the CICR mechanism such that SR Ca 2 Table 2 Comparison of the negative inotropic effects of halothane, isoflurane and sevoflurane at equi-anaesthetic concentrations. MAC 50 values (mmol litre 1 ) in terms of aqueous concentration are those for the rat. 30 Values for % inhibition of contraction expected at 1 and 2 MAC 50 for halothane, isoflurane and sevoflurane were interpolated from Figures 2B, 4Aand 6A % Inhibition % Inhibition at 1 MAC 50 at 2 MAC 50 MAC 50 Peak Sustained Peak Sustained Halothane 0.27 43.0 34.6 63.4 54.6 Isoflurane 0.31 29.1 6.3 43.0 15.7 Sevoflurane 0.35 58.2 12.7 67.9 22.3 release would be reduced and consequently contractions would also decrease. Sensitization of CICR by halothane may be undesirable as initial exposure to halothane was occasionally associated with spontaneous contractions which could be potentially arrhythmogenic (see Fig. 1, halothane 10 mmol litre 1 trace). It should be noted that the initial positive inotropic effect of halothane could, alternatively, be the result of a rapid and substantial leak of Ca 2 from the SR. 13 29 Ryanodine abolished any small increase in contraction seen on initial application of isoflurane, abolished or greatly reduced recovery of contractions during exposure to isoflurane or sevoflurane, respectively, and abolished the increase in contractions on wash-off of isoflurane and sevoflurane. The sensitivity of these effects to ryanodine suggests that they are mediated via the SR although these results do not provide definitive information on the mechanisms responsible for the inotropic actions. One possibility underlying recovery of contraction during exposure to isoflurane or sevoflurane is that the Ca 2 content of the SR increases, which would also contribute to the greater contractions observed immediately after wash-off, before equilibrium is re-established. However, further experiments are needed to determine if this is the case. In the case of isoflurane, the percentage decrease in contraction was greater in the presence of ryanodine. This suggests that in this case, the SR limits the sustained negative inotropic effect. In the presence of ryanodine, all three anaesthetics depressed contraction this is likely to have been the result of a decrease in I Ca and/or a reduction in myofilament Ca 2 sensitivity. Extent of negative inotropy at clinically relevant concentrations It is important to consider the extent to which these three anaesthetics may decrease cardiac contractility under clinical conditions. The broken lines in Figures 2B, 4Aand 6A represent the MAC 50 values for both humans (H) and rats (R) in terms of aqueous concentration 30 for each anaesthetic. For comparison, the MAC 50 values of the rat have been used to interpolate values for % inhibition of contraction (both peak and sustained effects) for halothane, isoflurane and sevoflurane at aqueous concentrations equivalent to 1 and 2 the MAC 50 value (see Table 2). In 729

Davies et al. Figure 2B, the horizontal bar represents a range of halothane concentrations in arterial blood from patients undergoing surgery. 31 This shows that the range of concentrations of halothane found during surgery span the MAC 50 value and reach considerably higher levels. When considering both the influence of patient age on the efficacy of volatile anaesthetics, 32 and the higher arterial concentrations reached during induction, it is possible that the range shown in Figure 2B is still an underestimate of the upper limit of clinically relevant concentrations of halothane. Bearing this in mind, Figure 2B shows that the negative inotropic effect of halothane during surgery is expected to be considerable. Furthermore, from the same stand point, the MAC 50 values for isoflurane and sevoflurane will be spanned above and below by clinically relevant concentrations (information on which is currently unavailable). Therefore, Figures 4A and 6A show that, even though sevoflurane and isoflurane have smaller depressant effects on contraction than halothane, during surgery both agents may lead to a considerable decrease in cardiac contractility and possibly hypotension. Acknowledgements This work was supported by a British Journal of Anaesthesia project grant and the British Heart Foundation. We are grateful to Luke Blumler, Andy O Brien and Dave Johannson for expert technical assistance. References 1 Sonntag H, Donath U, Hillebrand W, Merin RG, Radke J. Left ventricular function in conscious man and during halothane anesthesia. Anesthesiology 1978; 48: 320 4 2 Housmans PR, Murat I. Comparative effects of halothane, enflurane and isoflurane at eqipotent anesthetic concentrations on isolated ventricular myocardium of the ferret I. Contractility. Anesthesiology 1988; 69: 451 63 3 Graf BM, Vincenzi MN, Bosnjak ZJ, Stowe DF. The comparative effects of equimolar sevoflurane and isoflurane in isolated hearts. Anesth Analg 1995; 81: 1026 32 4 Hanouz J-L, Vivien B, Gueugniaud P-Y, Lecarpentier Y, Coriat P, Riou B. Comparison of the effects of sevoflurane, isoflurane and halothane on rat myocardium. Br J Anaesth 1998; 80: 621 7 5 Skeehan TM, Schuler HG, Riley JL. Comparison of the alteration of cardiac function by sevoflurane, isoflurane and halothane in the isolated working rat heart. J Cardiothorac Vasc Anesth 1995; 9: 706 12 6 Stowe DF, Sprung J, Turner LA, Kampine JP, Bosnjak ZJ. Differential effects of halothane and isoflurane on contractile force and calcium transients in cardiac Purkinje fibers. Anesthesiology 1994; 80: 1360 8 7 Ikemoto Y, Yatani A, Arimura H, Yoshitake J. Reduction of the slow inward current of isolated rat ventricular cells by thiamylal and halothane. Acta Anaesthesiol Scand 1985; 29: 583 6 8 Terrar DA, Victory JGG. Effects of halothane on membrane currents associated with contraction in myocytes isolated from guinea-pig ventricle. Br J Pharmacol 1988; 94: 500 8 9 Bosnjak ZJ, Supan FD, Rusch NJ. The effects of halothane, enflurane and isoflurane on calcium current in isolated canine ventricular cells. Anesthesiology 1991; 74: 340 5 10 Eskinder H, Rusch NJ, Supan FD, Kampine JP, Bosnjak ZJ. The effects of volatile anesthetics on L- and T- type calcium channel currents in canine cardiac Purkinje cells. Anesthesiology 1991; 74: 919 26 11 Puttick RM, Terrar DA. Effects of propofol and enflurane on action potentials, membrane currents and contraction of guinea-pig isolated ventricular myocytes. Br J Pharmacol 1992; 107: 559 65 12 Pancrazio JJ. Halothane and isoflurane preferentially depress a slowly inactivating component of Ca 2 channel current in guinea-pig myocytes. J Physiol 1996; 494: 91 103 13 Connelly TJ, Coronado R. Activation of the Ca 2 release channel of cardiac sarcoplasmic reticulum by volatile anesthetics. Anesthesiology 1994; 81: 459 69 14 Pask HT, England PJ, Prys-Roberts C. Effects of inhalational anesthetic agents on isolated bovine cardiac myfibrillar ATPase. J Mol Cell Cardiol 1981; 13: 293 301 15 Murat I, Ventura-Clapier R, Vassort G. Halothane, enflurane and isoflurane decrease calcium sensitivity and maximal force in detergent-treated rat cardiac fibers. Anesthesiology 1988; 69: 892 9 16 Luk H-N, Lin C-I, Chang C-L, Lee A-R. Differential inotropic effects of halothane and isoflurane in dog ventricular tissues. Eur J Pharmacol 1987; 136: 409 13 17 Robinson M, Harrison SM, Winlow W, Hopkins PM, Boyett MR. The effect of halothane on intracellular calcium and contraction in ventricular cells isolated from rat hearts. J Physiol 1993; 473: 110P 18 Wheeler DM, Rice RT, DuBell WH, Spurgeon HA. Initial contractile response of isolated rat heart cells to halothane, enflurane, and isoflurane. Anesthesiology 1997; 86: 137 46 19 O Neill SC, Eisner DA. A mechanism for the effects of caffeine on Ca 2 release during diastole and systole in isolated rat ventricular myocytes. J Physiol 1990; 430: 519 36 20 Bers DM. Ryanodine and the calcium content of cardiac SR assessed by caffeine and rapid cooling contractures. Am J Physiol 1987; 253: C408 15 21 Hansford RG, Lakatta EG. Ryanodine releases calcium from sarcoplasmic reticulum in calcium-tolerant rat cardiac myocytes. J Physiol 1987; 390: 453 67 22 Harrison SM, McCall E, Boyett MR. The relationship between contraction and intracellular sodium in rat and guinea-pig ventricular myocytes. J Physiol 1992; 449: 517 50 23 Cannell MB, Lederer WJ. A novel experimental chamber for singlecell voltage-clamp and patch-clamp applications with low electrical noise and excellent temperature and flow control. Pflügers Arch 1986; 406: 536 9 24 Boyett MR, Levi AJ. A simple electronic circuit for determining the twitch force and resting force of small heart muscle preparations. Pflügers Arch 1987; 410: 340 1 25 Boyett MR, Moore M, Jewell BR, Montgomery RAP, Kirby MS, Orchard CH. An improved apparatus for the optical recording of contraction of single heart cells. Pflügers Arch 1988; 413: 197 205 26 Wheeler DM, Rice RT, Lakatta EG. The action of halothane on spontaneous contractile waves and stimulated contractions in isolated rat and dog heart cells. Anesthesiology 1990; 72: 911 20 27 Bosnjak ZJ, Aggarwal A, Turner LA, Kampine JM, Kampine JP. Differential effects of halothane, enflurane and isoflurane on Ca 2 transients and papillary muscle tension in guinea pigs. Anesthesiology 1992; 76: 123 31 28 Bers DM. Ca influx and sarcoplasmic reticulum Ca release in cardiac muscle activation during postrest recovery. Am J Physiol 1985; 248: H366 81 29 Wheeler DM, Katz A, Rice RT, Hansford RG. Volatile anesthetic effects on sarcoplasmic reticulum Ca content and sarcolemmal Ca flux in isolated rat cardiac cell suspensions. Anesthesiology 1994; 80: 372 82 30 Franks NP, Lieb WR. Temperature dependence of the potency of volatile general anesthetics. Anesthesiology 1996; 84: 716 20 31 Davies DN, Steward A, Allott PR, Mapleson WW. A comparison of arterial and arterialized venous concentrations of halothane. Br J Anaesth 1972; 44: 548 50 32 Katoh T, Ikeda K. Minimum alveolar concentration of sevoflurane in children. Br J Anaesth 1992; 68: 139 41 730