Little is known about how general anaesthetics affect cardiac tissue that has been subjected to hypothermia. Such knowledge may be important when these agents are used in patients undergoing cardiac surgery performed during cold cardioplegia. Hypothermia has for instance been associated with intracellular calcium overload [1,2] which has been associated with the induction of life-threatening arrhythmias [1-5]. Bjørnstad et al. using conventional microelectrode techniques demonstrated that nisoldipine, a dihydropyridine calcium channel blocker, prevented the electrophysiological alterations normally associated with a reduction in myocardial temperature, i.e. increased action potential duration and maximum force . Identification of anaesthetic agents that might help to ameliorate the genesis of hypothermiainduced arrhythmias and provide myocardial protection during cooling and following rewarming after cold-cardioplegic arrest might be useful.
Diazepam is a commonly employed agent in modern anaesthesia. Gershon, utilizing the patch-clamp technique, concluded that diazepam and other benzodiazepines block dihydropyridine-sensitive voltagedependent Ca2+-channels . Thus, if diazepam produces calcium-channel blockade, it might also be expected to ameliorate the electrophysiological alterations produced by hypothermia as has been demonstrated using nisoldipine.. The purpose of this study was therefore to determine how diazepam affects the electrophysiological and mechanical response of cardiac tissue to hypothermia.
Guinea pigs weighing 250-350 g were anaesthetized by intraperitoneal injection of pentobarbitone 150 mg. The hearts were then rapidly removed and placed into cold oxygenated Tyrode's solution while perfused through a cannula inserted into the aorta. The composition of the Tyrode's solution employed was (in mM): K+, 5.4; Na+, 153; Ca2+, 2.7; Mg2+, 1.05; Cl−, 162.9; HEPES, 1.04; and glucose, 11. This solution was continuously gassed with 95% O2 and 5% CO2 and the pH of the solution was adjusted to 7.4. The animals utilized in the study were procured, maintained, and killed in accordance with the guidelines established by the Norwegian Ministry of Agriculture and adhered to by all Norwegian academic institutions utilizing animals in research.
Ventricular papillary muscle preparations, no thicker than 1 mm were dissected and mounted horizontally in a 5 mL tissue chamber while superfused with Tyrode's solution (5 mL min−1). One end of each preparations was connected to a fixed hook located on the bottom of the superfusion chamber while the other end was attached to a hook on a miniature force transducer (AE 801, SensorNor a.s., Horten, Norway). The preparations were stretched to their physiological resting length (i.e. the point at which they displayed the maximum amplitude of developed force) while stimulated at a fixed pacing cycle length (500 ms) using constant current, 1 ms long stimuli (S8800BS stimulator, Grass Instrument Co. Quincy Mass., USA) delivered through a pair of platinum electrodes placed near the preparation. Transmembrane electrical activity was recorded using conventional microelectrode techniques employing 3 M KCl filled glass electrodes having resistances in the range of 5-20 MΩ.
Action potentials and developed force were displayed on a storage oscilloscope while simultaneously connected to, analysed and stored on a personal computer employing a 16-bit, 100 kHz data acquisition card (AT-MIO-16X), and data acquisition termination board (BNNC-2080, National Instruments, Austin, Texas, USA). A computer program for determining the values of diastolic membrane potential, action potential amplitude, maximum rate of depolarization (dV/dtmax), action potential durations at 10, 50 and 90% repolarization (APD10, APD50 and APD90) and developed force was created by A. Danielsen (Kirkenes, Norway) using LabView 4.0 (National Instruments, Austin, Texas, USA). The temperature in the bath was continuously monitored using a thermocouple connected to a digital multimeter (model 75, John Fluke Mfg. Co. Inc., Everett, Washington, USA). A temperature control module (heater + circulator) (C1, Haake, Karlsruhe, Germany) was used to control the temperature of the superfusate.
After the preparation was suspended in the organ bath, it was allowed to stabilize for 60 min at 37°C. The pacing cycle length was 500 ms throughout all the experiments. Initial measurements were recorded at 37°C, and after that either no drug, 1 μM nisoldipine, or one of four concentrations (0.1, 1.0, 10, or 100 μM) of diazepam was added. Thus, six different experimental groups were created (n=6 per group). The time of initial exposure to either no drug, nisoldipine or diazepam at 37°C in each experiment was 30 min. During this 30-min period, measurements were acquired at 5-min intervals. The temperature was then reduced in all groups to 27°C for a period of 15 min. Each preparation was then allowed to equilibrate at 27°C for an additional 15 min. Finally, each preparation was rewarmed to 37°C and then re-exposed to drug-free Tyrode's solution for 30 min while recordings were obtained every 5 min.
Diazepam was obtained from Roche (Norsk Medisinaldepot, 4X 207) and dissolved in propylene glycol (Norsk Medisinaldepot, 5M044/2) to produce a concentration of diazepam 0.1 mM. This solution was further diluted in Tyrode's solution to produce the final concentrations of utilized diazepam (0.1, 1.0, 10 and 100 μM). Nisoldipine (SmithKline Beecham Pharmaceuticals, Milan, Italy) was dissolved in ethylene glycol (Norsk Medisinaldepot, 4B039/1) to produce a 1 mM nisoldipine stock solution, and this was further diluted in Tyrode's solution to yield a final concentration of 1 μM.
Values have been expressed throughout the study as mean ± SEM. We performed a two-way Anova for multivariate data analysis. Then, when indicated, paired Student's t-tests were used to compare changes within each group. Comparisons of results between the two groups (treatment versus control) were made using Student's t-tests. In all cases, a P-value of 0.05 or less was considered significant.
Effects of nisoldipine and diazepam at 37°C
At the end of the initial equilibration period at 37°C, the mean values for developed force in all groups were expressed as 100%; and, further changes were expressed as a percent of these values (Table 1). During the first 30-min period in the no drug, control group there was a significant time-dependent decrease in developed force. In this control group, developed force was reduced by 20% from 1.0±0.7 mN to 0.8±0.7 mN, P<0.05(Fig. 1a). Nisoldipine (1 μM) produced a significantly greater fall in developed force over this equivalent period of time so that developed force decreased by 68% at the end of 30 min compared with the initial base-line value (from 1±0.4 mN to 0.32±0.15 mN, P<0.05) (Table 1, Fig. 1a). This represented a 48% greater decrease in developed than that observed in the time-matched control group at 30 min of superfusion (Fig. 2). All concentrations of diazepam examined produced an apparent biphasic (positive followed by negative) inotropic effect. In all the diazepam groups, developed force was initially increased from its baseline values and then reduced. As indicated in Fig. 2 these changes in developed force appear to be concentration dependent. The maximal increase in developed force was reached approximately 10 min following exposure to 0.1, 1.0 and 10 μM diazepam; while in the 100 μM diazepam group increases in developed force were only observed during the first 5 min of exposure. Only 100 μM diazepam significantly decreased developed force, so that the 30 min value in the diazepam group was 57% of its baseline value(0.6±0.17 mN vs. 0.36±0.18 mN, P<0.05). This diazepam-induced reduction in developed force was 23% greater than the decrease in developed force observed in the time-matched no-drug, control group at the end of 30 min (Fig. 2).
During this same initial 30-min period at 37°C in the control, no-treatment group, there was no change in APD90 or in any other action potential characteristics (Fig 1b, Table 1). However, nisoldipine (1 μM) shortened APD50 from 167±7.5 ms to 135±8.1 ms, P<0.05 (Fig. 1b). While the lowest concentration of diazepam (0.1 μM) had no effect on APD90 the three remaining concentrations of diazepam (1.0, 10 and 100 μM) shortened APD90 by an average of 5, 6 and 14 ms, respectively. Thus, diazepam appeared to shorten APD90 in a concentration-dependent manner. Only was the shortening produced by 100 μM diazepam significantly less than its base-line value (154±3.1 ms vs. 168±7.0 ms, P<0.05) (Fig. 1b).
Effects of cooling in the presence of nisoldipine and diazepam
Reducing the temperature of the superfusate from 37 to 27°C significantly increased developed force from the initial base-line value in the control, no-drug group (0.93±0.7 mN vs. 1.42±0.84 mN, P<0.05) (Table 1, Fig. 1a). In the nisoldipine group, developed force was decreased compared with its base-line value at the end of the period of exposure to 27°C (1.1±0.4 mN vs. 0.39±0.43 mN) (Table 1, Fig. 1a). Thus, as previously reported , 1 μM nisoldipine prevented the hypothermia-induced increase in developed force. The three lowest concentrations of diazepam (0.1, 1.0 and 10 μM) failed to produce any apparent difference in the increase in developed force produced by cooling from 37 to 27°C(Table 1). In the group exposed to the highest concentration of diazepam (100 μM), however, developed force failed to significantly change from the base-line value when the temperature was reduced from 37 to 27°C (0.58±0.17 mN vs. 0.53±0.47 mN) (Table 1, Fig. 1a). Thus, the value of developed force in the diazepam group was not different from the base-line value at 37°C and was significantly lower than that in the no-drug, control group at 27°C (0.53±0.47 mN vs. 1.42±0.84 mN,P<0.05) (Fig. 1a).
The increase in developed force produced by cooling in the control group from 37 to 27°C was associated with an increase in APD90 increased from 168±6.7 ms to 253±10.9 ms, P<0.05 (Fig. 1b). In contrast, in the 1 μM nisoldipine group cooling only lengthened APD90 by 2 ms (from a base-line value of 167±7.5 ms to 169±6.3 ms at the end of the period of exposure to 27°C (Fig. 1b). Diazepam, in an apparent concentration-dependent manner, blunted the increase in action potential duration produced by cooling the superfusate from 37 to 27°C. This effect was significant (P<0.06) only in the group exposed to 100 μM diazepam so that in the control group APD90 was 253±10.9 ms while in the 100 μM diazepam group it was 229±9.1 ms at the end of the period of exposure to 27°C superfusate (Fig. 1b).
The effects of time and temperature alone, 1 μM nisoldipine and 100 μM diazepam are summarized in Fig. 3 in which recordings of transmembrane potential and force development are depicted.
Effects of rewarming and wash-out of nisoldipine and diazepam
After rewarming and wash-out of drugs, developed force remained reduced in all of the experimental groups when compared with their base-line values. But, only in the nisoldipine group was it significantly lower than that in the no-drug, control group (0.19±0.15 mN vs. 0.4±0.22 mN, P<0.05) (Table 1, Fig. 1a).
Rewarming and wash-out shortened APD90 back towards base-line values in all the groups, except in the 1.0 μM nisoldipine and 100 μM diazepam groups. In these groups, APD90 values remained significantly shorter compared with those in the control group (158±1.2 ms and 111±12.7 ms, respectively, vs. 168±3.4 ms, P<0.05) (Fig. 1b).
In the present study we compared the electromechanical effects of four concentrations of diazepam with those produced by 1 μM nisoldipine and time alone at 37 and 27°C. As previously demonstrated ; nisoldipine, a cardioselective dihydropyridine calcium channel blocker , reduced the hypothermia-induced increases in action potential duration and developed force. These effects of nisoldipine were thus in close agreement with the results previously reported  under similar experimental conditions. With regards to the effects of diazepam, the main findings were:
- At 37°C, 30 min of exposure to 100 μM diazepam reduced developed force while lower concentrations ranging from 0.1 to 10 μM failed to produce significant effects. This decrease in developed force was associated with a shortening of APD following 100 μM diazepam.
- At 27°C, 100 μM diazepam reduced the increase in developed force and action potential duration associated with cooling, as did 1 μM nisoldipine. The magnitude of the diazepam effects were not as great as those produced by nisoldipine.
Hypothermia has been associated with intracellular calcium overload [1,2]. Increased intracellular calcium concentration leads to an increase in force development. Thus, calcium channel blockers have been used by some for the management of this phenomena; although their clinical use in the setting of hypothermia has not proved to be beneficial . In the present study calcium channel blockade produced by 100 μM diazepam and 1 μM nisoldipine reduced these hypothermia induced changes in isolated guinea pig ventricular papillary muscle preparations.
In the present study diazepam appeared to act as a calcium channel blocker. However, the mechanism responsible for its positive inotropic effect at 37°C remains unclear. Nisoldipine did not produce a similar effect and thus this effect of diazepam is not believed to be due to direct effects on Ca2+-channels. However, our data support those of our previous report  and those of others . Namely that at 37°C diazepam has a biphasic effect (positive followed by negative) on inotropy. Others have explained this effect of diazepam as being due to a transient release of endogenous catecholamines from sympathetic nerve terminals as the initial positive inotropic effect is reversed by pretreatment with propanolol[8,10].
Although diazepam did indeed reduce, in the present study, the hypothermia induced changes, the concentration required to do so was much greater than that expected to be obtained in clinical practice. Diazepam serum concentration after intravenous (i.v.) administration have been reported to vary in the range of 1600-1700 ng mL−1, i.e. approximately 60 μM [11,12]. However, diazepam is highly bound to plasma proteins (95-98%) [13,14]. The free fraction of diazepam, therefore, should be expected to range from 1.2 to 3 μM. This active diazepam concentration is considerably less than that which produced an effect in our study and that has been demonstrated to directly block calcium channels . The results of our study thus suggest that clinically relevant concentrations of diazepam are unlikely to produce any significant effects on cardiac electromechanical activity in vivo during normothermia as well as during hypothermia. These conclusions are supported by the recent findings of another study that demonstrated the ineffectiveness of midazolam, another benzodiazepine derivative structurally related to diazepam, to alter cardiac inotropy at clinically relevant concentrations .
During the performance of this investigation A. L. Melnikov was supported by a grant from the Norwegian Foreign Ministry and the Council for International University Cooperation(Bergen, Norway). Additional support for the study was received in the form of a grant to K. G. Helgesen from the Fund for medical Research in Finnmark. The authors thank A. Danielsen for the development of the program used for data analysis and E. P. Popova for technical assistance.
1 Liu B, Wohlfart B, Johansson B. Effects of low temperature on contraction in papillary muscles from rabbit, rat and hedgehog. Cryobiology
2 Piwnica A, Menasche P. Role of calcium blockers in protecting the myocardium in cardiac surgery. Therapie
3 Angelakos E, Torres J. Cardiovascular physiology under hypothermia. Int Anesth Clin
4 Zagorski W, Czaplick S, Szepietowski J, Stanowski E. Electrocardiogram in deep experimental hypothermia in dogs. Pol Med J
5 Kearns J, Murnaghan M. Ventricular fibrillation during hypothermia. (Abstract) J Physiol (London)
6 Bjørnstad H, Lathrop D, Refsum H. Prevention of some hypothermia induced changes by calcium channel blockade. Cardiovasc Res
7Gershon E. Effect of benzodiazepine ligands on Ca2+ channel currents in Xenopus oocytes injected with rat heart RNA. J Basic Clin Physiol Pharmacol
8 Melnikov A, Helgesen K, Lathrop D. Cardiac effects of diazepam are altered by hypothermia. Clin Exp Cardiol
9 Lathrop D, Valle-A J, Millard R et al.
Comparative electrophysiological and coronary hemodynamic effects of diltiazem, nisoldipine, and verapamil on myocardial tissue. Am J Cardiol
10 Leeuwin R, Zeegers A, van Wilgenburg H. Actions of benzodiazepines on the inotropy of the perfused rat heart. Arch Int Pharmacodyn
11 Hillestad L, Hansen T, Melsom H, Drivenes A. Diazepam metabolism in normal man 1. Serum concentrations and clinical effects after intravenous, intramuscular, and oral administration. Clin Pharmacol Therapeutics
12 Lundgren S. Serum concentration and drug effect after intravenous and rectal administration of diazepam. Anesth Prog
13 Johnson R, Schenker S, Roberts R, Desmond P, Wilkinson G. Plasma binding of benzodiazepines in humans. J Pharm Sci
14Ingum J, Pettersen G, Sager G, Morland J. Relationship between unbound plasma concentrations and various psychomotor and subjective effects after intakes of diazepam and flunitrazepam. Int Clin Psychopharmacology
15 Gelissen H, Epema A, Henning R, Krijnen H, Jennis P, den Hartog A. Inotropic effects of propofol, thiopental, midazolam, etomidates and ketamine on isolated human atrial muscle. Anesthesiology