In Vitro  Effects of Propofol and Volatile Agents on Pharmacologically Induced Chloride Channel Myotonia

After approval from the Institutional Animal Care and Use Committee of the University of Minnesota (Minneapolis, Minnesota), 12 castrated male Yorkshire crossbreed swine with a mean weight of 85.8 ± 8.8 kg were pretreated with an intramuscular injection of a mixture of 250 mg tiletamine and 250 mg zolazepam (Telazol, Fort Dodge Animal Health, Fort Dodge, IA). After anesthesia with 5–7 mg/kg IV thiopental (Gensia Pharmaceuticals, Inc., Irvine, CA), their tracheae were intubated and lungs mechanically ventilated with an air-oxygen mixture. Shortly thereafter, rectus abdominis muscle biopsies were obtained (within 15 min) and samples placed in preoxygenated Krebs-Ringer solution (gassed with carbogen, 95% O²and 5% CO²) and then immediately transferred to a dissecting dish (with continuous gassing). Using a dissecting microscope, the muscle biopsies were dissected into small muscle bundles of intact fibers from tendon to tendon.10,11Next, they were transferred to in vitro , temperature-controlled (37 ± 0.1°C) muscle experimental chambers (containing Krebs-Ringer solution bubbled continuously with carbogen); each bundle was independently stimulated with supramaximal electrical field stimulation pulses of 1 ms duration at a frequency 0.1 Hz, as previously described.12,13To identify the optimal muscle lengths, length-tension relationships were derived (bundles were stretched) according to the standardized protocol for in vitro  contracture testing for MH susceptibility.14An average of 23 ± 8 muscle bundles was obtained from each pig for these experiments. After maintaining a stable baseline for approximately 10-15 min, muscle bundles were pharmacologically challenged according to the study protocol (see below in the section Investigation Protocols). Each muscle bundle length was measured after optimal lengths had been determined (as the distance between the suture ties that hold the muscle in the tissue bath). Muscle mass was determined as the wet weight of the specimen after removing it from the bath and cutting it at the sutures. Next, the specimens were patted dry and weighed immediately. Cross-sectional areas were estimated as follows15: Area (mm²) = muscle mass (mg)/(fiber length [mm]× muscle density [mg/mm³]) using 1.06 mg/mm³as the density of skeletal muscle.16 

The investigated and administered drugs were 9-AC (Sigma-Aldrich GmbH, Steinheim, Germany), a specific voltage-gated chloride channel blocker17; propofol in Intralipid (Disoprivan 1%; Astra Zeneca, Plankstadt, Germany); pure propofol (pharmaceutical sample; B. Braun, Melsungen, Germany); Intralipid (Intralipid 10%; Fresenius Kabi, Stans, Switzerland); dimethyl sulfoxide (DMSO; Fluka, Buchs, Switzerland); halothane (Sigma-Aldrich, St. Louis, MO); isoflurane (MINRAD Inc., Bethlehem, PA); sevoflurane (Abbott Laboratories, North Chicago, IL); acetazolamide (Bedford Laboratories, Bedford, OH); lidocaine (ICN Biochemicals Inc., Aurora, OH), and procainamide (Sigma-Aldrich GmbH, Buchs, Switzerland). The Krebs-Ringer solution consisted of 118.1 mm NaCl, 3.4 mm KCl, 1.2 mm KH²PO⁴, 1.0 mm MgSO⁴-(7H²O), 11.0 mm d-glucose, 25.0 mm NaHCO³, and 2.5 mm CaCl²-(2H²O). The chloride-free Krebs-Ringer solution consisted of 118.1 mm NaNO³, 3.4 mm KNO³, 1.2 mm KH²PO⁴, 1.0 mm MgSO⁴-(7H²O), 11.0 mm d-glucose, 25.0 mm NaHCO³, and 2.5 mm CaSO⁴-(2H²O).

Dose–response curves were established for 9-AC in DMSO (n = 5), propofol in Intralipid (n = 5), propofol in DMSO (n = 5), acetazolamide (n = 5), lidocaine (n = 5), and procainamide (n = 5) at concentrations of 4, 8, 16, 32, 64, 128, 256, and 512 μm. DMSO (n = 5) and Intralipid alone (n = 5) were each tested according to the volume used in the groups where they served as the solvent. The volatile anesthetics halothane (n = 4), isoflurane (n = 4), and sevoflurane (n = 4) were tested at 0.25, 0.5, 1.0, 2.0, and 4.0 vol% up to the maximum dose allowable (4.5, 5.8, and 8.0 vol% for halothane, isoflurane, and sevoflurane, respectively) by a given vaporizer (Dräger, Drägerwerk AG, Lübeck, Germany).13Before entry into the flowmeters, the concentration of the volatile agents was continuously sampled by an inline servo gas monitor (Datex Ohmeda 5530 Agent Monitor, Datex-Ohmeda, Inc., Louisville, CO).

For the various pharmacologic groups, muscle bundles were pretreated with 9-AC (64 μm) for 10 min,18followed by a challenge of increasing concentrations of propofol in Intralipid (n = 11), propofol in DMSO (n = 10), Intralipid (n = 10), DMSO (n = 10), halothane (n = 8), isoflurane (n = 8), sevoflurane (n = 8), acetazolamide (n = 10), lidocaine (n = 10), or procainamide (n = 10) in the concentrations according to the study protocol. In addition, we tested the effects of caffeine (n = 4) at 0.5 mm and 1.0 mm, respectively. Control muscle bundles were either untreated (n = 30) or bathed in 9-AC alone (n = 19). To investigate a direct effect of chloride, additional muscle bundles were incubated in a chloride-free Krebs-Ringer solution for 40 min. Thereafter, the muscle bundles were exposed to increasing concentrations (every 10 min) of propofol in Intralipid (n = 12) or Intralipid alone (n = 12). As appropriate controls, muscle bundles were incubated in chloride-free Krebs-Ringer solution (n = 12) or normal Krebs-Ringer solution (n = 12).

The following variables, representing skeletal muscle contractile function, were recorded: Peak forces, areas under the force curves (AUCs), contraction times, half-relaxation times, and baseline forces (preloads). As such, all values were normalized to relative baseline forces and then expressed as percents of change. AUCs were measured for a period of 9.5 s after muscle stimulation (fig. 1). We waited for 8 min after each treatment to achieve full effects of a given agent and then collected the data during the subsequent 60 s (i.e. , six twitch responses per collection period); this data then was used for later analyses. All data collection was automated employing LabView (ADEPT, Adept Scientific, Hertfordshire, United Kingdom). For relative anesthetic comparisons, concentrations of the volatile agents were calculated as minimum alveolar concentration (MAC) equivalents.

For measurement with respect to time, data were analyzed using two-way ANOVA with repeated measures, and one-way ANOVA were used for determining differences between drug groups at each time point. Bonferroni post hoc  tests were used to identify intergroup differences. To estimate relative changes, the appropriate control groups were used. All data are presented in this report as mean ± SEM, except the muscle characteristics data, which is given in mean ± SD; a P < 0.05 was considered statistically significant. All statistical analyses were performed using a Prism software package (GraphPad Software, La Jolla, CA).

Length, masses, or cross-sectional areas of the muscle bundles used in the various treatment groups did not differ between groups. In the dose–response experiments, the mean values of the muscle bundles' characteristics were 30.5 ± 4.0 mm, 179.3 ± 42.0 mg, and 5.58 ± 1.3 mm²(see tables 1 and 2, Supplemental Digital Content 1, which list the characteristics of the specific muscle bundles of the various treatment groups used in this study, https://links.lww.com/ALN/A537). All control muscle bundles with no treatment elicited both significant decreases in peak forces and AUCs over time (P < 0.001). All administered drugs led to significantly altered values of peak forces, AUCs, contraction times, and half relaxation times, as summarized in table 1. As compared to time controls, 9-AC led to significantly increased AUCs, starting at concentrations of 16 μm (P < 0.05); the dose response curve was sigmoidal in shape (see tables 3–7, Supplemental Digital Content 1, https://links.lww.com/ALN/A537, which are tables providing the entire data set of the dose–response experiments using normal skeletal muscle bundles. The different treatment groups and the indices of skeletal muscle performance, including peak force [mN], AUC [g*s], contraction time [s], half relaxation time [s], and baseline tension [mN] are provided).

In the experiments with induced myotonia, the mean values of the muscle bundles' characteristics were 31.9 ± 4.1 mm, 191.8 ± 67.0 mg, and 5.73 ± 2.0 mm². The chloride channel blocker 9-AC (64 μm) led to significantly increased peak forces for up to 50 min after its administration, as compared with controls (P < 0.05). The same treatment significantly increased AUCs during the entire study period as compared with controls, starting at 10 min after administration (fig. 2; P < 0.001). There were no differences in contraction times and half relaxation times between the two groups. Ninety minutes after 9-AC administration (64 μm) the baseline force values were significantly higher (P < 0.01). After incubation in chloride-free Krebs-Ringer solution, the muscle bundles showed increased peak forces and AUCs during the entire study period (fig. 2; P < 0.05). In addition, half relaxation times began to increase significantly 80 min after incubation in the chloride-free medium (P < 0.05). Contraction times and baseline force values were not different for muscle samples in either normal or chloride-free Krebs-Ringer solutions.

The myotonic reactions, indicated by the increased values of AUCs after treatments with 9-AC, were reversed by the subsequent administrations of either propofol in DMSO or propofol in Intralipid, starting at 64 μm (P < 0.05) and 128 μm (P < 0.01), respectively (fig. 3A). This was not the case with either Intralipid or DMSO administration alone. Challenges with halothane, isoflurane, and sevoflurane increased the myotonic reactions, starting at 5.4 (P < 0.001), 0.21 (P < 0.01), and 0.5 MAC (P < 0.05), respectively (fig. 3B). Interestingly, the higher concentrations of the inhalational anesthetic agents appeared to somewhat reverse the elevated myotonic reaction. While isoflurane started to decrease myotonia at 0.83 MAC, sevoflurane and halothane only did so at their maximum dosages (4.0 MAC and 6.1 MAC) (fig. 3B). Importantly, myotonia induced by chloride-free Krebs-Ringer solution was reversed by the administration of propofol in Intralipid at 128 μm (fig. 4A; P < 0.05). Similarly, the sodium channel blockers lidocaine and procainamide both decreased the myotonic effects significantly at concentrations of 8 μm (P < 0.001) and 256 μm (P < 0.05), respectively (fig. 4B). However, the administration of lidocaine increased peak forces at 128 μm and 256 μm concentrations in the same setting. It was observed that treatment with the chloride channel activator acetazolamide had no effects on the induced myotonia (fig. 4B). In contrast, caffeine administration significantly increased the myotonic reactions of exposed muscle bundles (P < 0.05). The effects of these agents on the contractile features of the 9-AC–pretreated muscles are summarized in table 2. (See tables 8–12, Supplemental Digital Content 1, https://links.lww.com/ALN/A537, which are tables listing the entire data set of the dose response experiments using myotonic [9-AC–pretreated] skeletal muscle bundles. The different treatment groups and the indices of skeletal muscle performance, including peak force [mN], AUC [g*s], contraction time [s], half relaxation time [s] and baseline tension [mN] are provided.)

We have investigated the in vitro  effects of propofol and various volatile anesthetics on pharmacologically induced chloride channel myotonia. The primary findings of this study were that propofol administration reversed the pharmacologically induced myotonia, whereas the volatile agents halothane, isoflurane, and sevoflurane were associated with aggravated myotonic reactions.

Under normal physiologic conditions, the influx of chloride is considered to rapidly stabilize the sarcolemmal membrane potential after a depolarization; e.g. , induced action potential within the muscle fiber membrane. Clinically, in patients with myotonia congenita, impaired chloride conductance because of altered chloride channels results in membrane hyperexcitability, i.e. , the myogenic production of repetitive action potentials, which in turn prolongs the overall contractions within their skeletal muscles. In the experiments presented here, we observed that in these isolated muscle bundles we could reduce fiber chloride conductances, either by pharmacologically blocking specific chloride conductance with 9-AC18or by lowering extracellular chloride concentration. As expected, these treatments led to disturbed repolarizations with afterdepolarizations, and thus repetitive aftercontractions (myotonia).19 

Importantly, it was observed here that propofol dramatically reversed these drug-induced myotonic reactions at concentrations above 64 μm. This effect was considered part as a result of the activity of propofol itself; i.e. , because neither Intralipid nor DMSO alone altered the relative degrees of induced myotonia. Consistent with these findings, there are several previous reports citing the direct myorelaxant effects of propofol on skeletal muscle.20–22Thus, in our study, to determine whether voltage-gated chloride channels were primarily involved in these responses, we tested the relative effects of the voltage-gated chloride channel activator acetazolamide.23,24We observed that the treatment with acetazolamide induced minimal or no such antimyotonic effects in such in vitro  experiments. This may suggest that chloride anions are not primarily involved in these antimyotonic responses. Furthermore, propofol administration resulted in the same efficacies in resolving the myotonic reactions in a chloride-free environment.

In general, it is considered that human skeletal muscle sodium channels are blocked by propofol in a voltage-dependent manner.25,26This action on the voltage-gated sodium channels may have been responsible for the antimyotonic effects of propofol that were observed in the present studies. Therefore, we also tested the relative effect of the sodium channel inhibitors lidocaine and procainamide. Interestingly, these two agents were associated with reversals of the myotonic reactions in our experiments as well.27,28This may be further evidence that the antimyotonic actions of propofol are mainly a result of modulations of voltage-gated sodium channels within the sarcolemmal membranes of skeletal muscle.

We observed here that the volatile anesthetic agents halothane, isoflurane, and sevoflurane each significantly increased the myotonia induced by 9-AC. Of potential clinical interest was the present observation that both isoflurane and sevoflurane administration induced enhanced myotonic responses at clinically relevant doses: 0.21 MAC and 0.5 MAC, respectively. This may in part have been because of the expected increase in the intracellular release of calcium into skeletal muscle cells after exposure to inhalational agents.29More specifically, like caffeine, volatile anesthetics are known to act agonistically on the skeletal muscle type 1 ryanodine receptor and lead to increased calcium release from the sarcoplasmic reticulum. This phenomenon was supported by our findings that both caffeine and volatile agents were associated with increased myotonic reactions in our experiments; i.e. , the sarcoplasmic release of calcium through ryanodine receptors may have been at least partially involved in these responses. In addition, the known abilities of isoflurane to inhibit the voltage-gated sodium channels of skeletal muscle could explain the favorable effect of isoflurane on 9-AC–induced myotonia at higher concentrations; in addition, isoflurane administration has been associated with decreased peak force values at concentrations above 1.7 MAC.30 

We consider here that our experiments support the view that in muscles that elicit hyperexcitability, volatile anesthetics may secondarily increase overall muscle metabolism and could therefore be potentially dangerous; i.e. , patients affected by myotonic syndromes may elicit aggravated myotonia causing hypermetabolic events (when extreme, this may even mimic episodes of malignant hyperthermia).31,32However, to date there are no comparative clinical or in vivo  animal studies on these issues, and we therefore have to rely on occasional and often anecdotal case reports. It is of interest to note that in some of these reports, propofol in myotonic patients is not without observed side effects; there have been variable responses, marked sensitivity to its depressant effects, and even precipitation of myotonia reported with its use.8,33 

There are several potential limitations of our experimental approaches. First, in these isolated muscle bundles there is no vascular perfusion, and thus if diameters are too large, adequate oxygenation of centralized fibers may be limited. In addition, it is not exactly clear how in vitro  test bath concentrations of administered agents apply to blood concentrations when employed in vivo . Nevertheless, it has been proposed that in experiments in which volatile anesthetics are administered for the in vitro  contracture testing for the determination of MH susceptibility, such responses will occur at clinically relevant concentrations. Therefore, we postulate that the concentrations of volatile anesthetics used in this study may also be considered to be clinically relevant. It is less clear whether the results with propofol administration should be interpreted somewhat differently. It may be proposed that higher concentrations of propofol would likely be needed to simulate the situation in vivo , given that propofol had no effect at clinically relevant concentrations and even showed a reversal of chloride channel–induced myopathy at high concentrations. Finally, half relaxation times represent the established marker for myotonia.19There are publications describing changes in relaxation times with smaller changes in twitch amplitudes or AUCs in humans with myotonia.34,35In our study model, we were not able to significantly increase the half relaxation times of the muscle bundles by using 9-AC, irrespective of dosage. However, in some of the aforementioned studies the biggest differences between normal and myotonic patients were in the second half of the relaxation phases. Consistent with these results of the indices of skeletal muscle performances observed in our experimental model, AUC seemed to be the most sensitive marker for an elicited myotonic reaction. Furthermore, if we considered here changes in half relaxation times as the primary measures of elicited myotonia, the primary findings would remain the same. Although 9-AC did not alter half relaxation times, propofol was associated with shortening of half relaxation times, whereas the volatile agents isoflurane and halothane increased half relaxation times.

In conclusion, according to our in vitro  data, general anesthesia with propofol seems to be the better choice for use in a patient who may have myotonic activity. However, we emphasize that it is difficult to conclude from these in vitro  studies employing swine skeletal muscle bundles that one can assume relative safety of these agents in patients. Again, we have not studied the relative effects of these protocols on muscles of myotonic patients directly. Regardless of the chosen anesthetic and the likely preferred choices, the anesthesiologist must consider responses in individual patients as the anesthetic proceeds.

The authors would like to thank Allison Dwileski, B.S. (Administrative Assistant, Department of Anesthesia, University Hospital, University of Basel, Switzerland) and Monica Mahre, B.S. (Lab Services Coordinator, Department of Surgery, University of Minnesota, Minneapolis, Minnesota) for their editorial assistance and help with manuscript submission.