Effects of different dosages esketamine on cardiac conduction and heterogeneity of Cx43: the epicardial mapping in guinea pigs.

Background: Esketamine is favored in clinical settings. Relative to other anesthetics it preserves protective airway reflexes, maintains spontaneous respiration, stabilizes hemodynamics, and alleviates neuropathic pain. This study sought to evaluate the cardiac safety of esketamine at 3 sub-anesthetic gradient concentrations.
Methods: We examined the cardiac electrophysiological effects of esketamine with infusion rates of 0.125, 0.25, and 0.5 mg·kg-1·h-1. Short-term studies were performed in ventricular myocytes using patch-clamp techniques and optically mapped Langendorff-perfused guinea-pig hearts. Long-term studies were performed using Langendorff-perfused guinea-pig hearts and electrically mapping the receipt of the infusion for 3 hours.
Results: Esketamine changed the action potential (AP) morphology of cardiomyocytes. Notably, it increased the resting membrane potential (RMP), attenuated the amplitude of action potential (APA), reduced the maximum upstroke velocity (Vmax), and shortened the action potential duration (APD) at 30% to 70%, which led to relatively prolonged monophasic action potentials (MAP) triangulation in G0.25 and G0.5. All the effects were partially eluted. Optical mapping demonstrated almost equal and heterogeneous conduction. G0.125 resulted in an increased heart rate (HR) accompanied by a shortened APD. No detectable arrhythmia was observed at the cycle lengths (CLs) used. Long-term electrical mapping demonstrated the dose-dependent deceleration of the Vmax and APA, but only prolonged the AP parameters in G0.5. Left-ventricular isochronal conduction maps revealed the conduction heterogeneities at G0.5, and conduction velocity (CV) was increased in G0.125 and G0.25. None of these effects were reversed on drug washout. Electrocardiogram (ECG) traces revealed an accelerated HR with the associated curtailment of QT intervals in G0.125; HRs were decreased in G0.25 and G0.5; the PR intervals and QRS duration differed between G0.125 and G0.25, G0.5, which elicited electrical alternans. Connexin43 (Cx43) expression were significantly decreased in G0.125, G0.25 and G0.5. Conclusions: These data provide a basic electrophysiology for esketamine. Specifically, we found that (I) various methods of esketamine infusion had different effects on cardiac conduction at different dosages; (II) the heterogeneous expression of Cx43 is associated with spatially dispersed conduction; and (III) potential cardiac risks should be considered for high-risk patients receiving continuous esketamine infusions of high dosages. 2022 Annals of Translational Medicine. All rights reserved.

Introduction

Under the concept of “Rational Perioperative Opioid Management,” multi-mode analgesia is becoming more and more important. Ketamine is a racemic mixture in 1:1 of R-ketamine and S-ketamine, and acts as an antagonist of the non-competitive N-methyl-D-aspartate (NMDA)-receptor, which has dissociative anesthetic and analgesic properties (1). Due to the ketamine-induced increase in sympathetic tone, it is a good field anesthetic, as its rapid action and does not cause hypotension or respiratory depression. Historically, the use of ketamine has been limited due to concerns related to its associated psychomimetic effects; it places patients in a state of consciousness to which they are not previously accustomed. Recently, ketamine and esketamine were approved by the Food and Drug Administration as antidepressants for treatment-resistant bipolar depression patients (2). Both ketamine and esketamine have shown gradually recovered in perioperative acute pain management and multimodal analgesia (3).

Esketamine is favored in clinical settings. Relative to other anesthetics, it preserves protective airway reflexes, maintains spontaneous respiration, stabilizes hemodynamics, and alleviates neuropathic pain. The best studied administration route of ketamine is intravenous. Ketamine is a valuable adjunct in the induction and maintenance of anesthesia in patients with unstable hemodynamics and significant pulmonary comorbidities. It can be used as both a sedative and analgesic agent without compromising a patient’s respiratory status. In relation to Esketamine’s new introductions, such as the antidepressant effects, the reducing suicide ideations and the reduction of dose-dependent dissociative properties of ketamine, safety concerns arise in a number of patient populations, particularly those with widely prevalent comorbidities, like morbid obesity, diabetes mellitus, cardiovascular diseases (e.g., hypertension, cardiac insufficiency, myocardial infarction, and cardiac arrhythmias) and depression-associated comorbidities (e.g., epilepsy and stroke). Safety concerns also arise in relation to concomitant medications, the aging population, and possible adverse reactions from the administered therapies.

In the clinical setting, ketamine acts both directly on cardiac action and activates the autonomic nervous system, which leads to the complex hemodynamic effects. Ketamine has also been shown to produce direct negative inotropic outcomes in heart preparations (4), isolated hearts (5), intact animals (6), or in the clinical trials (7). Ketamine concentration dependently lengthened the RR interval, slowed ventricular conduction velocity (CV), and stretched the period of ventricular effective refractory without induced arrhythmia (8); however, there have been a report of transient bradycardia and hypotension associated with the rapid administration of ketamine at an analgesic dose (9). In guinea-pig myocardium, clinically relevant concentrations of ketamine decreased ischemia-induced action potential (AP) shortening and spontaneous reperfusion-induced ventricular arrhythmias (10). Changes in both CV and refractoriness have been shown to promote arrhythmias of the re-entrant mechanism (11). It has been reported that NMDA-receptor activation induces ventricular electric remodeling and facilitates ventricular arrhythmias (12), reduces heart rate variability (HRV) and enhances atrial fibrillation (AF) inducibility by degrading gap junction proteins (13). Research has examined whether esketamine affects contractility (4-6); however, its electrophysiologic effects on ventricles have only been poorly documented (8). Thus, we conducted the present study first to examine the effects of esketamine on cardiac conduction and susceptibility to arrhythmia, and second to preliminarily explore the potential molecular mechanisms by observing the expression level and distribution of Cx43. We present the following article in accordance with the ARRIVE reporting checklist (available at https://atm.amegroups.com/article/view/10.21037/atm-22-2614/rc).

Methods

Ethical approval

All the procedures were approved by the Institutional Animals Ethics Committee of The Third Affiliated Hospital, Guizhou Medical University (No. 2021A010), Guizhou, China, and conducted according to the national guidelines under which the institution operates, and the National Institutes of Health (NIH) Guidelines for the Care and Use of Laboratory Animals (8th edition). A study protocol was prepared before the study without registration.

Animals

Male guinea pigs (weighing 250–320 g) were purchased from The Experimental Animal Center of Guizhou Medical University. The guinea pigs were housed at 23 ℃ and 60% humidity on a 12/12 h light/dark cycle with free access to chow and water.

Single-cell electrophysiological recordings

Single ventricular myocytes were enzymatically dissociated from the hearts of the adult guinea pigs as described previously (14). In short, the hearts were perfused with Tyrode’s solution (containing 5.4 mM of KCl, 135 mM of NaCl, 10 mM of HEPES, 0.33 mM of NaH2PO4, 10 mM of glucose, 1 mM of MgCl2, 1.8 mM of CaCl2, and pH 7.4 with NaOH) for 5 min retrogradely to pump out the residual blood. After 5 min of Ca2+-free modified Tyrode’s solution perfusion, the solution was switched to 1 containing bovine serum albumin (BSA) fraction (1 mg/mL, Sigma) and collagenase type II (0.52 mg/mL, Worthington Biochemical Corporation), and the hearts were removed from the perfusion apparatus after 40 min. Myocytes were obtained from the free transmural left-ventricular (LV) wall. The myocytes were kept in a KB solution, which was composed of 120 mM of KOH, 120 mM of L-glutamic acid, 10 mM of KCl, 10 mM of Taurine, 10 mM of KH2PO4, 10 mM of HEPES, 10 mM of glucose, 0.5 mM of K-EGTA, 0.5 mM of EGTA, 1.8 mM of MgSO4·7H2O, 0.2% BSA, and pH 7.4 with KOH. The cells were then calcified to 1.8 mM and underwent electrophysiological recordings within 6–8 h of isolation. A MultiClamp 700A amplifier, 1550A digitizer, and pCLAMP 10.6 software (Molecular devices, USA) were used to record the APs. The pipette solution contained 130 mM of KCl, 10 mM of NaCl, 0.5 mM of MgCl2, 5 mM of Mg-ATP, 0.5 mM of EGTA, 10 mM of HEPES, 0.4 mM of Tris GTP, and pH 7.2 with KOH. The external solution contained 135 mM of NaCl, 5.4 mM of KCl, 10 mM of HEPES, 0.33 mM of NaH2PO4, 10 mM of glucose, 1.8 mM of CaCl2, 1 mM of MgCl2, and pH 7.4 with NaOH. Patch electrodes were pulled with resistances of 3–5 MΩ using a P-97 micropipette puller (Sutter Instruments, USA). The APs were evoked in the current-clamp mode at a rate of 1 Hz with an amplitude of 1,000–3,000 pA current pulse for a duration of 2 ms applied via patch electrodes. The collected data were analyzed using Clampfit software (Version 10, Molecular Devices, USA).

Langendorff-perfused isolated hearts

The guinea pigs (male, weighing 250–320 g) were humanely sacrificed by inhaling 8% sevoflurane vaporized with 60% oxygen. After thoracotomy, the hearts were rapidly excised and then mounted onto a Langendorff-perfusion system, after which they were perfused with the modified Tyrode’s solution (containing 119 mM of NaCl, 4 mM of KCl, 1 mM of MgCl2·6 mM of H2O, 25 mM of NaHCO3, 1.2 mM of KH2PO4, 1.8 mM of CaCl2, and 10 mM of glucose), equilibrated with 95% O2–5% CO2, pH =7.35~7.45, at a flow rate of 8 mL/min at 37 ℃. The hearts were monitored and sustained for 20 min before the experimental procedures commenced.

Ex-vivo heart preparations for optical mapping

After the Langendorff-perfused hearts reached stability, the contraction artefacts were minimized using 10 µM of blebbistatin (Tocris Bioscience, MN, USA). Dye loading was aided by pre-perfusion with pluronic F127 (20% w/v in DMSO). Rh237 (Vm indicator) and Rhod2-AM (Cai indicator) were perfused to enable the measurement of the simultaneous membrane potential and Ca2+ at 37 ℃. Epifuorescence was acquired simultaneously using high-speed cameras (Integrated Imaging Systems; Cairn Research, Kent, England). Action potential duration at 90% (APD90) repolarization and CV were studied using a dynamic pacing protocol. Ventricular tachyarrhythmias (VTA) inducibility was defined as the ability to provoke Ventricular Tachycardia/Ventricular Fibrillation (VT/VF) with the dynamic pacing protocol and/or programmed extra-stimuli. To analyze the optical mapping signals and generate the isochronal maps, the data were semi-automatically processed using ElectroMap software.

Stimulation protocol

The S1S2 protocol was used to assess arrhythmogenicity and identify re-entrant substrates (15). The electrical recordings were made during programmed stimulations that involved the application of the premature extra-stimulus (S2) after a train of 7 regular (S1) pulses. The protocol was performed both before and after the drugs were delivered. The stimuli were generated by an isolated constant voltage/current stimulator (MyoPacer EP Field Stimulator, IonOptix, Milton, MA, USA), and delivered onto the epicardial apex at a 2 ms pulse width and an amplitude 1.5× the diastolic voltage threshold with a platinum electrode. The heart was paced at basic cycle lengths (BCL) from 200 to 170 ms with consecutive 10 ms decrements of 2 ms in duration for 50 times. The optical mapping recording was performed at the last 2 s in each episode.

Ex-vivo heart preparations for electrical mapping

The guinea-pig hearts were randomly divided into 4 groups (n=6 per group). The control group (Gcontrol) did not receive esketamine (No. 210425BL, Jiangsu Hengrui Pharmaceutical Co., Jiangsu, China). In the remaining 3 groups, the hearts were treated with the following different doses: 0.125 mg·kg−1·h−1 (the 0.125 group or G0.125), 0.25 mg·kg−1·h−1 (the 0.25 group or G0.25), and 0.5 mg·kg−1·h−1 (the 0.5 group or G0.5). The drugs were administered with Tyrode’s solution. After their successful preparation, the hearts were perfused on the Langendorff apparatus. Immediately after stabilization, the experimental protocol of continuous perfusion for 3 h was commenced. Electrocardiograms (ECGs) were continuously recorded by 2 ECG electrodes (IH-SR, Hugo Sachs Elektronic, GER), positioned on the right atrium and left ventricle, respectively. A 64-channel multi-electrical array (MEA) mapping system (EMS64-USB-1003, MappingLab Ltd., UK) was employed to record the extracellular potential of the LV epicardium. EMapRecord 5.0 software (MappingLab Ltd., UK) was used to monitor and calculate the CV from the known distance between the recording points and the difference in timing. The data were stored for the subsequent off-line analysis.

Western blot and immunofluorescence detection

Western blot

The expressions of the Cx43 proteins in each group were detected by Western blot. Cardiac tissues were dissected from the left ventricle at the end of the ex-vivo electrophysiological studies for protein quantification (n=6 per group). The ventricle myocardium tissues were homogenized after 200 µL of the specific cocktail proportion of proteinlysate and protease inhibitor were added at 4 ℃ and centrifuged at 12,000 rpm for 15 min. The sample proteins were quantified by BCA colorimetric assays, and phosphate buffered solution (PBS) was then used to dilute the samples to the same concentration. The samples were boiled at 100 ℃ for 10 min, and 12% separating gel and 5% loading gel were used for separating by sodium dodecyl-sulfate polyacrylamide gel electrophoresis. Electrophoresis was performed in the electrophoretic buffer, and the target bands were then then transferred to polyvinylidene fluoride membranes by semi-dry transfer (10V, 15 min). The membranes were blocked with 5% BSA and incubated in tris buffered saline with tween (TBST) for 2 hours. They were washed with TBST and incubated overnight with the following primary antibodies at 4 ℃: anti-connexin 43 antibody (1:1,000, Sigma, USA) and anti-GAPDH antibody (1:500, Abcam, Cambridge, UK). The next day, after washing with TBST, the membranes were incubated at room temperature for 1 hour with a second antibody (rabbit anti-goat, 1:12,000, Beijing Boosen Biotechnology Co., Ltd., Beijing, China), which was labeled with horseradish peroxidase. They were then washed with TBST again, and ECL hypersensitive reagent was added to the membranes. Next, an automatic gel imaging analysis system was used to visualize (ZF-158, American Bio-rad Company).

Single immuno-labeling Cx43 immunofluorescence

The paraffin sections were made to detect Cx43 using immunofluorescence as previously described in detail (16). The apex cordis tissues were fixed, dehydrated, paraffin embedded, and sectioned. Cryostat sections, 8 µm thick, were mounted on glass slides overnight at 40 ℃. Rehydration and heat-induced epitope retrieval were then performed. The sections were subsequently incubated with a goat-anti-rabbit Cx43 (8 h at 4 ℃; 1:2,000, Bioss, Beijing, China). After being washed 3 times in PBS, the sections were incubated with IgG H&L FITC (1:200, Abcam, Cambridge, England). After nuclear staining with DAPI [Mounting Medium, antifading (with DAPI), Solarbio Inc., China], 5 random fields of each slide were separately observed and examined with a microscope illuminated by a sequential scanning mode at 405 and 488 nm (40×, Olympus IX83, Olympus Life Science Inverted Microscopes, Japan), and the fluorescence intensity was determined using ImageJ software.

Statistical analysis

SPSS 17.0 software was used for the data analysis (SPSS, Chicago, IL, USA). Clampfit 10.6 (Molecular Devices, USA), OriginPro 8.0 (Origin Lab, USA), and Adobe Illustrator 10.0 (Adobe, USA) were used to analyze the patch-clamp recorded data. All the data are expressed as the means ± standard errors of the means. A one-way analysis of variance followed by a multiple-comparison test was used to evaluate the multiple test treatments. A P value <0.05 was considered statistically significant. In the figures, the designations for the P values are as follows: *, P<0.05, **, P<0.01, and ***, P<0.001 respectively.

Results

Patch-clamp studies of single-cell APD

We examined the effects of different doses on the AP parameters in single guinea-pig ventricular myocytes. At a stimulation frequency of 1 Hz, the action potential duration (APD), resting membrane potential (RMP), maximum upstroke velocity (Vmax), amplitude of action potential (APA), and APD at 30%, 50%, 70%, and 90% (APD30, APD50, APD70, APD90) repolarization were recorded via the bath perfusion system. The results showed esketamine changed the AP morphology of the cardiomyocytes (see ) and increased the RMP of the cardiomyocytes (see ), attenuated the APA (see ), reduced the Vmax (see ) and shortened the APD30 ~APD70 (see ) and there are no significant changes between groups in APD90 (see ), which led to no changes in early repolarization phase (see ) but a relatively prolonged late repolarization phase (see ) and MAP triangulation in G0.25 and G0.5 (see ). All the effects were especially significant in G0.5 and were then partially eluted.

Figure 1

Effects of different dosages of esketamine on the AP in single guinea-pig ventricular myocytes. (A) APs recorded; (B-H) corresponding measured (B) RMP, (C) APA, (D) Vmax, (E) APD30, (F) APD50, (G) APD70, (H) APD90; (I) APD30–60, (J) APD60–90 and (K) MAP triangulation (APD30–90). (Six cardiocytes per group; *, P<0.05, **, P<0.01 and ***, P<0.001). AP, action potential; RMP, resting membrane potential; APA, amplitude of action potential; APD30, action potential duration at 30%; APD50, action potential duration at 50%; APD70, action potential duration at 70%; APD90, action potential duration at 90%; APD30–60, action potential duration from 30% to 60%; APD60–90, action potential duration from 60% to 90%; APD30–90, action potential duration from 30% to 90%; MAP, monophasic action potentials.

Optical mapping studies and assessments of arrhythmic tendency in ex-vivo heart preparations

We then investigated the effects of different dosages of esketamine with continuous intravenous infusion optical fluorescence mapping in sinus-paced Langendorff-perfused hearts using the recording and ECG monitoring configurations (see ). AP initiation and conduction were measured by voltage mapping (see ), and we found an increase in HR in G0.125, but a decrease in HR in G0.25 and G0.5 [see (i and iii)] with more heterogeneous and almost equal conduction [see (iv)]. Additionally, optical potentiometric mapping revealed that G0.125 shortened APD90 [see (ii and v)]. Finally, voltage dye optical mapping studies during ventricular pacing were used to assess susceptibility of the isolated perfused hearts to the induction of arrhythmias before and after drug administration [see (ii)], but no arrhythmic effects were detected at the CLs used.

Figure 2

Optical mapping of voltage RH237 and Rhod-2 in isolated Langendorff-perfused hearts. (A) ECG monitoring and optical mapping configuration; (B) (i) AP initiation mapping, (ii) maps of APD90; corresponding measured, (iii) HR, (iv) conduction, and (v) APD90 averaged over the field of view velocities; (C) (i) a premature AP was induced on extra-stimulus application S2 immediately after termination of the effective refractory period in the preceding S1 beat; the upper and the lower MAP were recorded, respectively, and (ii) voltage RH237 optical mapping of ventricular arrhythmic tendency in isolated Langendorff-perfused hearts. (This procedure was repeated on 5 hearts; **, P<0.01 and ***, P<0.001). ECG, electrocardiogram; AP, action potential; HR, heart rate; CV, conduction velocity; APD90, action potential duration at 90%; LV, left ventricle; RV, right ventricle; LA, left atrium; RA, right atrium; BL, base line; PCL, pacing cycle lengths.

Ex-vivo heart preparations of electrical mapping and electrocardiographic studies

Next, we explored the long-term effects of 3 different dosages of esketamine in whole isolated hearts with continuous infusion on electrophysiological activation and recovery. summarizes the sites of the ECG and MEA recordings in the electrical mapping. shows that the APs recorded from the defined LV [see (ii and iii)] in the presence of esketamine. LV isochronal conduction maps [see (i)] were observed in the presence of esketamine revealed there was no significant difference in RMP between groups [see (ii)], a synergistic reduction of the APA [see (iii)], and a clear dose-depended deceleration of the Vmax [see (iv)]. Compared with Gcontrol, APD30 was increased from G0.125 to G0.5 [see (v)]; APD50, APD70 and especially APD90 of G0.125 were decreased, the parameters of G0.25 were basically the same as those of the Gcontrol; every stage in G0.5 were significantly increased which resulted the APD was only prolonged in G0.5 [see (vi-ix)]; Compared with Gcontrol, APD30–60 [see (x)] was significantly shortened. there was no significant difference between each group in APD60–90 [see (xi)] and also APD30–90 was significantly shortened in G0.125 and G0.25 [see (xii)]. Compared with G0.125 and G0.25, APD30–60 [see (x)] and APD30–90 [see (xii)] were significantly prolonged in G0.5. Notably, there were conduction heterogeneities in G0.5, and the CV was markedly increased in G0.125 and G0.25 [see (i)]. These effects were not all reversed on drug washout. The quantification of the corresponding ECG traces [see (i–iv)] showed that G0.125 accelerated the HR [see (i) and (i)] with an associated curtailment of the QT intervals [see (ii)], but both G0.25 and G0.5 exhibited the opposite effects. The PR interval [see (iii)], and the QRS durations [see (iv)] differed significantly between G0.125 and G0.25, G0.5, which elicited electrical alternans. Together, these results suggested that G0.5 exerted an inhibitory effect on the HR and CV, and prolonged cardiac repolarization.

Figure 3

MEA mapping and ECG studies in isolated Langendorff-perfused guinea-pig hearts following continuous perfusion for 3 h. (A) Schematic summary of MEA and ECG recording configuration; (B) (i) successive LV isochronal maps, and (ii) quantified conduction velocities for the conditions; (C) (i) representative APs recorded from the defined LV, (ii) RMP, (iii) APA, (iv) Vmax, (v) APD30, (vi) APD50, (vii) APD70, (viii) APD90, (ix) APD, (x) APD30–60, (xi) APD60–90 and (xii) APD30–­90, MAP triangulation; (D) ECG studies, ECG records obtained as described above, followed by drug washout, (i) Gcontrol, (ii) G0.125, (iii) G0.25 and (iv) G0.5; (E) ECG corresponding measured (i) HR, (ii) QT intervals, (iii) PR intervals and (iv) QRS durations. (Six hearts per group; *, P<0.05, **, P<0.01 and ***, P<0.001). MEA, multi-electrical array; ECG, electrocardiogram; AP, action potential; RMP, resting membrane potential; APA, amplitude of action potential; APD, action potential duration; APD30, action potential duration at 30%; APD50, action potential duration at 50%; APD70, action potential duration at 70%; APD90, action potential duration at 90%; APD30–60, action potential duration from 30% to 60%; APD60–90, action potential duration from 60% to 90%; APD30–90, action potential duration from 30% to 90%; LV, left ventricle; HR, heart rate; RV, right ventricle; LA, left atrium; RA, right atrium.

Immunofluorescence detection and Western blot

The expression of Cx43 in the left ventricle was detected by immunofluorescence [see (i)], compared to the Gcontrol, the fluorescence intensity of Cx43 was meaningfully declined in G0.125, G0.25, and G0.5 [see (ii)]. The quantified expression of Cx43 detected by western blot [see (i)] is consistent with the IF [see (ii)].

Figure 4

Immunofluorescence and confocal laser scanning microscopy of Cx43 (A), localization (arrows) at the membrane of neighboring cells [detail A (i), and merged image]. High-intensity specific immunoreactive signals of Cx43 were clearly identified and regularly distributed in Gcontrol compared to the other groups; with the duration of continuous infusion Cx43 signals decreased and were heterogeneously distributed A (ii). The expression levels of Cx43 were examined by Western blot (B). Cx43 expression were significantly decreased in G0.125, G0.25 and G0.5, compared with G0.125 and G0.25, G0.5 was significantly decreased. (Six hearts per group; **, P<0.01 and ***, P<0.001).

Discussion

The main and original results of the present study are as follows: (I) no arrhythmogenesis spontaneously occurred during and after the clinically relevant concentrations of esketamine were administered; (II) arrhythmia could not be induced by pacing; (III) the prolonged infusion of high dosages of esketamine altered the cardiac electrophysiological effects and caused visible conduction heterogeneities; and (IV) heterogeneous Cx43 expression is associated with spatially dispersed conduction.

The use of opioids for analgesia might cause hyperalgesia and patients can develop a tolerance to opioids, both of which are partially attributed to NMDA-receptor activation. Preventive and post-operative analgesia with NMDAR antagonists can block the acute tolerance of opioids and reduce the development of neuropathic pain (17). A study had reported that a low-dose intravenous infusion of ketamine can be used in acute and chronic post-operative pain management (18). Despite these advantages, the side effects of ketamine, which include nightmares and delusions, limit its routine use. However, some research has suggested that the acute psychoactive effects, including the dissociative, psychotomimetic, and ‘mystical’ effects, experienced during or shortly after a sub-anesthetic dosage of ketamine administration, may be associated with its therapeutic benefits. Esketamine has twice the analgesic effect of racemic ketamine, has fewer side effects, such as hallucinations, results in a faster recovery, and is able to lower the sevoflurane MAC value (19). Esketamine also has sympathomimetic properties that cause less cardiorespiratory depression than other sedatives, which had been reported that propofol combined with esketamine exhibited a good safety profile and high reliability, was more conducive to the hemodynamics stability, alleviating surgical stress and inflammatory responses (20). Theoretically, esketamine possesses some non-opioid analgesic properties that aim to reduce opioids consumption among patients undergoing surgery, improve recovery, and alleviate opioid-related side effects; it had been reported that esketamine effectively countered remifentanil-induced respiratory depression which attributed to an increase in remifentanil-reduced ventilatory CO2 chemosensitivity (21). Thus, the use of esketamine in perioperative pain management is promising.

Despite its wide clinical use in perioperative patients, minimal data indicate that single doses of intravenous esketamine or ketamine use during cesarean section delivery may not affect the breastfed infant or lactation (22), and no instances of clinically apparent liver injury have as yet been reported with long term use of esketamine (23). There also has very few studies examined the effects of esketamine on cardiac electrophysiology. Based on drug instructions and reference materials, the required dosages of esketamine are usually 0.1–0.5 mg·kg−1·h−1(approximate mean: 0.4 mg·kg−1·h−1) (24,25) with or without bolus. It has been reported that in most critically ill patients suffering from COVID-19 induced acute respiratory distress syndrome (ARDS), the prescribed sedation levels are only achievable with several combinations of sedatives, and a mean dose of 0.86 (±0.76) mg·kg−1·h−1 of esketamine as an NMDAR inhibitor was required (26). In our study, we examined 1 control group and 3 dosing groups (i.e., G0.125, G0.25, and G0.5) that received continuous intravenous doses at wide concentration ranges that encompassed the reported therapeutic levels.

The cardiac AP arises and spreads throughout the myocardium as a consequence of the highly organized spatial and temporal expression of ion channels conducting Na+, Ca2+ or K+ currents, which when changed, may induce cardiac arrhythmias. The effects of a drug on the AP parameters can predict its effects on each ion current. In this study, we first investigated the electrophysiology of esketamine on APs in the ventricular myocytes of guinea pigs. Similar to previous studies on ketamine, we found that esketamine produced decreases in the Vmax, APA, and APD50, and increases in the RMP and APD90, which indicated that it exerted a frequency-independent inhibitory action on the cardiac fast inward sodium current (INa) (27,28) in G0.25 and G0.5. As inward rectifier potassium currents (IK1) maintain RMP and modulate the phase-3 repolarization of the APD in cardiomyocytes (29,30), the short-term cellular electrophysiological assays with the plateau of the RMP shifted to more positive values. However, RMP was not observed to change in isolated hearts. This may be because drugs are more highly dispersed in individual cells than dense tissues.

Additionally, the effects of esketamine on AP repolarization were somewhat complicated, and depended on the infusion rate and treatment period. In the in-vitro cell experiments, the APD30 and the APD50 decreased in G0.25 and G0.5, which accelerated the repolarization, while the APD70 gradually recovered in G0.125, and the APD90 changed insignificantly in all groups. Optical mapping perfuses with voltage-sensitive dyes and either biochemical or mechanical contractile uncoupling, which may also interfere with the wave shape of normal APs (31). We found that only the APD90 decreased in G0.125. Next, we found that in the isolated hearts with long-term esketamine continual perfusion, the APD30 was prolonged in all treatment groups. With higher concentrations and longer exposure times, the APD90 was obviously prolonged in G0.5 and shortened in G0.125. Our previous studies have also demonstrated that dexmedetomidine can prolong APD50 and APD90 which maybe the mechanism of slowing HR (32), and sevoflurane alleviates reperfusion arrhythmia induced by myocardial ischemia-reperfusion through the shortening of MAPD90 in isolated rat hearts (33).

There are reports that some drugs barely affect the APD90 in isolated guinea-pig papillary muscles (4,5), but prolong the QT intervals in humans (34), which indicate that the APD90 alone is not reliable for evaluating the inhibitory effects of compounds on ion currents. MAP triangulation, which is defined as the slowing of final repolarization, has been shown to be a strong predictor of early afterdepolarizations (EADs) or proarrhythmia. The cell study showed the plateau drooped, which resulted in the AP becoming more triangular. Additionally, the APD30 shortened, and the largest prolongation was limited to the APD90. Conversely, at the level of the whole heart, the final AP was prolonged in the high-dosage group that showed markedly changed the plateau. Generally, a shortened or lengthened APD is only proarrhythmic when combined with instability, triangulation, or reverse use-dependence but can become antiarrhythmic when ≥1 of these are absent. The drug induced the prolongation of APD with a square AP is antiarrhythmic (35).

Notably, repolarization disturbance limitation, abnormal automaticity (36) and conduction disturbances (37) can lead to monomorphic arrhythmias (including VT) (38). Using optical voltage mapping and electrical mapping, we also demonstrated that ventricular conduction in response to esketamine is complicated. As reflected by the prolongation of the QRS duration in G0.5 electrocardiographically, the intraventricular conduction was clearly compromised. Additionally, the cardiac conduction heterogeneity increased. Gap junction expression and function can also influence cardiac conduction through ion channel availability (39). The effects of the G0.5 sustained infusion are similar to those previously reported for ketamine slowed CV, and prolonged refractoriness without changing anisotropy or increasing dispersion of refractoriness, which resulted in the significant antiarrhythmic effects of ketamine (8).

Cx43 is known to be mainly expressed in ventricular cardiomyocytes. We observed that Cx43 in the membranous fraction was reduced in the treatment groups. Intriguingly, the effects of long-term low- and medium-dosages of esketamine on the CV and ECG parameters were largely opposite to the changes in Cx43 expression. Notably, the level of connexin proteins are related to intercellular coupling through many complicated factors, including trafficking, assembly, and posttranslational modification. It has been reported that the inhibition of cardiac gap junctional communication by the lipophilic drugs (such as heptanol, octanol, oleic acid or palmitoleic acid) is their ability to dissolve in the lipid bilayer that lead to impairment of the transcellular gap junction channels (40,41), which might explained the relevantly effects of esketamine on Cx43 expressions. The heart has a conduction reserve, and a medium and homogeneous decrease of 50% in Cx43 expression by itself does not always lead to conduction impairment (42). The enhanced sympathomimetic activities compatible with the increased vascular contraction and raising HR were common in clinical observations (43) and in-vivo studies (44). The Langendorff-perfusion system itself is separate from neural and humoral control. The NMDA-receptor is widely expressed in the heart and is involved in the onset of arrhythmias, and research has shown that its inhibition could improve Cx43 degradation (45). Previous studies have shown that ketamine-induced cardiovascular responses are centrally mediated (46,47). A previous study showed that ketamine does not produce a pressor response in dogs by epidural anesthesia (48). There are also reports that due to the release of or response to synaptic transmitter substances were more effectively inhibited in sympathetic ganglia than at sympathetic neuromuscular junctions (49), ketamine can potentiate responses to both noradrenaline and sympathetic nerve stimulation in vitro (50). These results need to be further explored.

Conclusions

The main and original results from this study are as follows: (I) various methods of esketamine infusion showed a far cry from effects on cardiac conduction with different dosages; and (II) high-risk patients using continuous infusions of high dosages of esketamine should be concerned about potential cardiac risks.

Limitations of the present study

The present study did not explore the association between ion channels, NMDA-receptors, and Cx43. Further, parallel clinical investigations on humans need to be conducted to explore the blood or plasma concentrations which correlated to effective concentration and to corresponding dosages for clinical pharmacokinetic properties of esketamine.

The article’s supplementary files as