Gender Differences in Electrophysiology

1. Abstract

Cardiovascular disease is one of the main causes of mortality globally. Cardiac arrhythmias are a cardiovascular disease of growing concern due to an increasing prevalence and burden on health and social care organisations. Established research into cardiac electrophysiological gender differences has implied that pro-arrhythmic behaviour may be introduced with different genders and their interaction with specific anti-arrhythmic drugs, but these investigations into this area of research remains limited. Therefore, the aim of this study was to a) develop a computational model of gender differences in electrophysiology, to then b) use the model to evaluate differences in response to anti-arrhythmic drugs. Isoprenaline was implemented to induce pro-arrhythmic behaviour in each model so that anti-arrhythmic drug parameters could be applied for this investigation. Computational simulations using a model of rat ventricular myocyte electrophysiology showed key differences in action potential duration and [Ca²⁺]_SR between genders with class III and class IV anti-arrhythmics. The female action potential duration and [Ca²⁺]_SR increased by 30.4% and 1.76%, respectively. However, with class IV anti-arrhythmic drug application, the action potential duration was shortened in both genders, in females by 5.1% and 1.7% in the male model, and there was a pronounced decrease in [Ca²⁺]_SR in females by 11.2%. A pro-longed action potential duration promotes sarcoplasmic reticulum loading, and with the addition of an increased [Ca²⁺]_SR, the occurrence of spontaneous Ca²⁺ release resulting in an arrhythmic event is more probable. Therefore, this study recognises the unsuitability of class III anti-arrhythmic drugs for females. Whereas, males with class III anti-arrhythmic drug application resulting in arrhythmic behaviour is improbable and is expected to be effective in preventing re-entrant arrhythmias in males.

2. Introduction

2.1 Background

Cardiovascular disease (CVD) is a chronic condition that has considerable impacts worldwide. All CVDs can be equally as fatal, but the increased occurrence of cardiac fibrillation and sudden death indicates an escalating problem with cardiac arrhythmias^(1-3). In 2019 CVD took 8.9 million lives globally, hence why CVD is still current and a growing concern⁽⁴⁾. Across the population women tend to have fewer cases of CVD compared to males, but have higher mortality rate and a worse prognosis after an acute event⁽⁵⁾. This suggests that the severity and susceptibility of CVD differs between genders. Electrophysiological gender differences may play a role in the CVD outcome for males and females. Therefore, further research into this area will develop our understanding to enhance treatment with accurate drug application and improve CVD prevention strategies to reduce the economic, health and social care strain^(6,7).

2.2 Cardiac electrophysiology and excitation-contraction coupling

Cardiac electrophysiology is determined by fundamental ion channels, which are critical for ion permeability and gating for the maintenance of the heart’s contractility and regular rhythm^(8,9). Calcium (Ca²⁺) is integral in the heart’s physiological and electrical mechanics by converting electrical stimulation (action potential, AP) into a physical contraction by a process known as cardiac excitation-coupling^(10,11). Ca²⁺ handling is mediated by: the increase in Ca²⁺ entering the cell via L-type Ca²⁺ channels, then Ca²⁺ diffusion through the ryanodine receptor type 2 (RyR2) channels which induces a phenomenon called Ca²⁺ induced Ca²⁺ release. Following this, the increase in free intracellular calcium within the cytoplasmic space results in Ca²⁺ binding to the troponin C for contraction initiation and finally, Ca²⁺ is then transported out of the cytosol either back into the SR or out of the cell via the SR Ca²⁺ ATPase (SERCA), sarcolemmal Na⁺/Ca²⁺ exchanger or the sarcolemmal Ca²⁺ pump (I_CaP)^(11-14). Additionally, the role of potassium ions (K+) is also crucial for the initiation of repolarisation. The delayed rectifier K⁺ currents (I_Kr and I_Ks) are activated during the AP plateau at depolarisation voltages to control this phase and the inward rectifier current I_K1, maintains the resting membrane potential (E_m)^(15,16).

2.3 Cardiac remodelling in disease

CVDs induce cardiac remodelling, which can be structural or electrical in form and occur from the sub-cellular to the organ-level. These changes lead to the development of life-threatening ventricular arrhythmias⁽¹⁷⁾. Ion current remodelling, specifically remodelling of Ca²⁺ handling ion channels can promote ventricular arrhythmias, such as the generation of early after depolarisation (EADs) and delayed after depolarisations (DADs), which induce premature beats increasing re-entrant probability^(17,18). AP prolongation is an additional hallmark of heart failure (HF) that can cause arrhythmias due to an increased time for calcium loading and a higher probability of spontaneous Ca²⁺ release. Triggers for cardiac remodelling appear to differ between genders, observation with females showed that triggers occur more progressively, and therefore results in a slower CVD diagnosis and treatment application⁽¹⁹⁾. Additionally, there is a low female representation at clinical trials, and therefore the consideration female specific electrophysiology is overlooked, which is likely to result in inappropriate treatment^(20,21). Further research into gender differences in cardiac remodelling could potentially result in more timely and appropriate treatment to reduce deaths from fatal arrhythmic events.

2.4 Cardiac arrhythmias and anti-arrhythmic treatment

Arrhythmias are heart rate abnormalities. A variation of cardiac arrhythmias can occur either from the formation of an abnormal impulse (ectopic foci) or conduction disturbances (re-entry)^(22,23). Re-entrant arrhythmias arise from altered electrical conductions, whereby areas of previously excited tissue are re-excited continuously outside of the normal heart rhythm⁽²⁴⁾. The generation of a non re-entrant arrhythmia from abnormal impulses is either down to automaticity which is a spontaneous impulse or because of a multiple impulses occurring from a prior trigger⁽²⁵⁾. To target arrhythmias anti-arrhythmic (AA) drugs have been developed and categorised into the Vaughan-Williams series. Of note to this study, class III AA drugs are K⁺ channel blockers which aim to slow repolarisation and increase the refractory period, whereas Class IV AA drugs block the Ca²⁺ channels aiming to reduce contractility by reducing Ca²⁺ load, with a secondary effect resulting in the reduction of spontaneous Ca²⁺ release⁽²⁶⁾.

2.5 Established cardiac gender differences

Experimental studies have been carried out using different animal species allowing notable electrophysiological gender differences to be identified. The I_Kr , I_Ks and I_Ca,L currents between genders differand a prolonged QT_C interval was discovered in adult females by Bazett in 1920^(27-37).In other research, differences in cardiac electrophysiology between genders have been previously recognised to have an influence on arrhythmic events^(29,37). Prior research indicates that the understanding of cardiac gender differences is advancing, but research into the effect of AA drugs with these gender electrophysiological differences remains limited. Therefore, this is an important new area of research to assess the risk AA drugs may pose with males and females.

2.6 Computational modelling

Pioneering experimental investigations into the movement of ions during electrical activity to determine the AP was published in a sequence of papers in 1952 by Alan Hodgkin and Alan Huxley^(38-42). The majority of cardiac electrophysiology models are largely based on this research. Hodgkin and Huxley identified three currents that contribute to the AP; the sodium current,I_Na, the potassium current I_K, and the leak current, I_1. The use of the voltage clamp techniques resulted in relating the ion current membrane capacitance and membrane potential changes in the following equation^(38-42). Further investigations by Noble (1962) modified the original equations to describe action and pace-maker potentials of the Purkinje fibres of the heart⁽⁴³⁾. With formulated channel equations and later findings from Reuter (1967) identifying the inward calcium current into Purkinje fibres, biophysically detailed cardiac computational models were developed⁽⁴⁴⁾. Furthermore, laboratory rats have been used as model organisms for important cardiac investigations of human disease^(45,46). Their comparability to humans and low expense makes them appropriate to use in cardiac experimental investigations^(47-49). Thus, computational modelling of rat cardiac physiology is well established from the range using this approach and is an alternative method for research into cardiology without ethical constraints^(50-53). Computational models are useful tools for cardiac research, providing ideal platforms for hypothesis testing and quantification of mechanisms not permitted by traditional experimental approaches.

2.7 Project aims

This study aimed to develop gender-specific models of rat ventricular electrophysiology and use the newly developed models to investigate differences in pro-arrhythmic activity and anti-arrhythmic drugs between sexes. Such computational investigations may provide insight into whether gender differences in electrophysiology necessitate different anti-arrhythmic treatment strategies and whether certain classes of AA drugs are more or less suitable for each sex. It was hypothesised that AA drugs inducing APD prolongation and alterations to Ca²⁺ handling concentrations would be unsuitable for females, but suitable for males, from previously established gender differences (pro-longed QT_C interval and the enhanced risk of spontaneous Ca²⁺ release.

3. Methodology

3.1 Model used

The Mathematic model of rat ventricular electrophysiology and Ca²⁺ handling used to investigate gender differences with anti-arrhythmic drugs was developed by Stevenson-Cocks et al.⁽⁵⁴⁾. Previous electrophysiology models of rat ventricular myocytes produced by Pandit et al. (2001, 2003)were hybridised with a model produced by Gattoni et al.^(50-52). Figure 1 shows the Pandit et al. (2001) model for the investigation of ionic mechanisms behind transmural heterogeneity of action potentials. This hybridisation with additional Ca²⁺ handling framework from Colman (2019) shown in Figure 2. created the model for this investigation of gender differences with anti-arrhythmic drugs⁽⁵⁵⁾.

3.2. Development of gender variants

The majority of experimental data collected in the development of the original Pandit rat model (and subsequent models since) was derived from male animals (Table.1). Therefore, by default the baseline model is male-specific without any data implemented. To produce a female specific model from the original code an array of widely reported data from relevant literature was collated relating gender differing membrane currents and fluxes. Relevant changes shown in Table.2 were incorporated into the code to produce female orientated model.

3.3. Pro-arrhythmic conditions

Isoprenaline (ISO)is widely used in cardiac research to investigate the effects of beta-adrenergic stimulation on cardiac physiology. Through activation of adenylate cyclase, ISO causes an increase in cAMP and activation of protein kinase A (PKA), ISO ultimately causes increased ICaL open probability and SERCA activity by PKA-dependent phosphorylation ^(57-61). The repolarising current I_to and I_ss also increase in response to ISO, leading to a shortened action potential duration (APD)^(54,62). Parameters implemented are shown in Table 3.

3.4. Anti-arrhythmic drug stimulation

Class III and Class IV AA drugs were investigated and data from relevant literature was implemented into each gender variant. Key experimental findings was inhibition of I_K with class III ant-arrhythmics ₍predominantly of I_Kr compared to I_Ks), which in the rat ventricular myocyte model are represented by the steady-state outward potassium current (I_ss). Therefore, a percentage inhibition of I_ss was implemented in the model. Inhibition of the inward rectifier potassium current (I_K1) was additionally widely reported in the literature, however at a smaller percentage. The collated information from the literature was based on a range of type III AA drugs: E403, Amiodarone, Dronedarone, Dofetilide and Solatol. The parameter changes used to simulate the effects of type III AA drugs are shown in Table 4. Class IV AA drugs inhibit I_Ca,L and hERG which encodes the K⁺ current I_Kr. To stimulate these effects, I_Ca,L experimental observations (Table.5). These changes were primarily based on the effects with Verapamil.

3.5. Stimulation protocol

Simulations for each gender model with or without ISO and electrophysiological anti-arrhythmic drug changes were performed at a basic cycle length (BCL) of 250ms until steady-state was achieved (minimum 100 beats).

3.6. Computational aspects

Models were coded in C/C++ format and simulations were performed on a MacBook Air (2018). As the model used is deterministic in nature, model outputs are fully determined by the parameter values and the initial conditions. Therefore, outputs are fixed with no variation, meaning no statistical analysis of results is required to determine significance.

Ca²⁺ handling parameters and associated currents and fluxes are shown in 4. The [Ca2+]i and [Ca2+]SR from the female model was higher in both compared to males. Peak systolic [Ca²⁺]_i was 67.9% higher in the female model than the male model (0.28 vs 0.47 $\mu$M). Additionally, the female peak resting [Ca²⁺]_SR was increased by 40.2% compared to males (695.42 vs 496.17 $\mu$M). Peak I_NaCa and I_Ca,L currents with both gender models showed differences in peak magnitude. Peak I_NaCa magnitude was slightly higher in females by 1.5% compared to males, (-1.34 vs -1.32 pA/pF). The female I_Ca,L magnitude was also higher compared to males by 17.4%, (-11.95 vs -10.18 pA/pF).

4. Results

4.1 Gender differences in electrophysiology and Ca²⁺ handling

The final steady-state APs from control simulations are shown for both gender variants in Figure 3. A clear difference in APD was observed between genders, females had 26.2% longer APD observed compared to males (46.62 vs 36.93 ms).

The final steady-state outward potassium current (I_ss) and the inward rectifier current (I_K1) under control conditions was higher in males by 40% (I_ss) and 13.5% (I_K1), compared to females. (graph not shown).

4.2. Isoprenaline disproportionately promotes SR Ca²⁺ loading in female ventricular myocytes

Simulations with isoprenaline (ISO) application (ISO, Table 3) resulted in differential effects on the electrophysiological and Ca²⁺ handling parameters between genders. ISO in the male model resulted in APD shortening by 14.8% (31.48 vs 36.93ms), whereas in females the APD decreased by 7.8% (42.99 vs 46.62 ms) shown in Figure 5.

Female APD was still overall longer than males under ISO-induced remodelling (42.99 vs 31.48 ms).

No differences in peak I_ss and I_K1 were observed under ISO induced remodelling in male and female models (2.78 vs 2.064 pA/pF) and (1.85 vs 1.63 pA/pF)_, respectively (graphs not shown).

Ca²⁺ handling parameters and associated currents and fluxes are shown in Figure 6. ISO induced large increases in peak systolic [Ca2+]i , from 0.28 to 0.52 $\mu$M in the male and from 0.45 to 0.98 $\mu$M in the female model, representing 85.7% and 117.8% increases, respectively.

ISO with the female model had 88.5% higher [Ca2+]i compared to with males, (0.98 vs 0.52 $\mu$M). ISO induced increased peak systolic [Ca2+]SR, from 496.1 to 795.3 $\mu$M in the male and from 695.4 to 1108.7 $\mu$M in the female model, representing 60.3% and 59.4% increases, respectively. Between genders, ISO with the female model resulted in a 39.4% higher [Ca2+]SR compared to with males, (1108.7 vs 795.3 $\mu$M). The L-type calcium current (ICa,L) (Figure 6) for both gender models with ISO application showed a large increase in peak ICa,L magnitude.

ISO application with males increased peak ICa,L by 60.1% compared to no ISO, (-16.3pA/pF) vs -10.18 pA/pF). ISO application with the female model also resulted in hifher peak ICa,L, (-18.95 pA/pF vs11.95 pA/pF), a 58.6% increase. Females with ISO had the higher peak ICa,L, by 16.2%, compared to males (-16.3 pA/pF vs -18.95 pA/pF). The sodium-calcium exchange current (INaCa) with ISO application also resulted in an increased in peak compared to control simulations.

Males had an increase of 37.1% (-1.81 vs -1.32 pA/pF), whereas females had an increase of 41 % (-1.89 vs -1.34 pA/pF). Females had a slightly higher peak (INaCa), by an increase of 4.4%.

4.3. Action potential duration prolongation by class III anti-arrhythmic drugs may promote triggered activity in female ventricular myocytes

Application of class III AA drug parameters (Table 4) resulted in a prolongation of APD in both gender models (Figure 7.), by 51% in the male model (31.48 ms vs 47.54 ms), and by 30.4% in the female model (56.08 ms vs 42.99 ms). Female APD was still longer than males overall in the presence of class III AA drugs (56.08 ms vs 47.54 ms).

Peak Iss was reduced in both models as a result of class III AA drug application, a decrease of 58.5% in males and 59.7% in females. Additionally, peak IK1 reduced by 20% in both gender variants with class III AA stimulation (graphs not shown).

Ca²⁺ handling parameters and associated currents and fluxes are shown in Figure 8. Both gender variants showed slight increases in [Ca²⁺]_i and [Ca2+]SR compared to ISO models. Class III AA drugs with the male model resulted in a 3.7% higher peak systolic [Ca²⁺]_i compared to ISO (0.54 vs 0.52 $\mu$M), and a similar effect was seen in the female model (1.02 vs 0.98 $\mu$M, a 4.1% increase). Peak systolic [Ca²⁺]_i was 88.9% higher in the female model compared to the male model. The resting [Ca2+]SR increased with class III AA drug application in the male model by 1.31% vs ISO alone (805.7 vs 795.3 $\mu$M) and by 1.75% in the female model (1128.1 vs 1108.7 $\mu$M).

Minor changes were seen with peak I_Ca,L and INaCa with class III AA drug implementation. Peak I_Ca,L decreased by 5.4% (-1.81 vs -1.68 pA/pF) and 3.9% (-1.89 vs -1.79 pA/pF) in the male and female model, respectively, compared to ISO alone. Similarly, peak INaCa decreased by 7.18% the male model (-1.68 vs -1.81 pA/pF) and by 5.3% in the female model (-1.79 vs -1.89 pA/pF).

4.4. Class IV anti-arrhythmic drugs prevent pro-arrhythmic drugs prevent pro-arrhythmic SR Ca2+ loading in males and females

Application of class IV AA drugs resulted in a shortening of the APD in both gender models (Figure 9.), by 1.68% in the male model (30.95 vs 31.38 ms), and by 5.14% in the female model (40.78 vs 42.99 ms). Female APD was still longer than males in the presence of class IV AA drugs and ISO (40.78 ms vs 30.95 ms).

Peak Iss was reduced in both gender models because of class IV AA drug application by 31.8% in males and 33% in females, whereas peak IK1 remained unchanged (graphs not shown). Ca²⁺ handling parameters and associated currents and fluxes are shown in Figure 10.

Both gender variants showed a reduction in [Ca²⁺]_i and [Ca2+]SR compared to with ISO application (Figure 6). Class IV AA drugs with the male model resulted in a 28.85% decrease peak systolic [Ca²⁺]_i compared to ISO (0.37 uM vs 0.52 $\mu$M), and a similar effect was seen in the female model (0.71 vs 0.98 $\mu$M, a 27.6% decrease). Overall, the peak systolic [Ca²⁺]_i was 88.9% higher in the female model. The resting [Ca2+]SR also decreased with class IV AA drugs in both gender variants, in the male model by 11.12% % vs ISO alone (706.95 vs 795.26 $\mu$M) and a similar decrease of 11.20% in the female model (984.53 vs 1108.66 $\mu$M). Class IV AA drug application showed a decreased peak I_Ca,L in both gender variants, males had a decrease of 19.44% compared to ISO alone (-13.14 vs -16.31pA/pF), whereas females had a decrease of 15.65% (-15.90 VS -18.85 pA/pF) shown in Figure 10. Minor changes were observed with peak I_NaCa with class IV drug application_, a reduction of 5.82% in males was observed (-1.78 vs –1.81 pA/pF), and a 5.62% reduction in females (-1.68 vs -1.81 pA/pF).

5. Discussion

In this study, specific gender models of rat ventricular electrophysiology and Ca²⁺handling were developed using prior experimental observations in the literature of AP and ion channel gender differences in ventricular myocytes. Computational simulations revealed stark differences in the suitability of class III and class IV anti-arrhythmic drugs for reducing arrhythmogenesis in either males or females, which could have profound implications for targeting therapeutic interventions depending on gender.

5.1 Gender differences between male and female control models

Control gender simulations identified the electrophysiological gender differences present in the male and female models, which later allowed for comparison with AA drug and ISO application. The final steady-state AP shown in Figure 3. revealed a longer APD was present in females, which corroborated with previous findings in the literature disclosing that a lengthened female QT_C interval was identified^(27,34,85).

Ca²⁺ handling results for control gender models (Figure 4.) showed higher [Ca²⁺]_i and [Ca²⁺]_SR concentrations in females compared to males. Females increased SERCA would result in enhanced Ca²⁺ loading, therefore making SR Ca²⁺ more available for release which would give rise to a higher [Ca²⁺]_i. Additionally, females increased I_Ca,L triggers Ca²⁺ influx into the cell, therefore these two effects cause a greater intracellular Ca²⁺ transient size and enhances SR load⁽⁸⁶⁾. SERCA also accommodates for the removal of [Ca²⁺]_i back into the SR this would additionally explain the higher [Ca²⁺]_SR seen with females^(11,13,86).However, some research findings have reported a higher intracellular Ca²⁺ concentration in male myocytes, Parks and Howlett (2013), Curl et al (2000) and Farrell et al (2010), which contradicts the findings from this study. Nevertheless, the changes implemented in the model to produce the female model variant were informed from experimental observations and so further investigations are required to determine whether there are species and gender differences in Ca²⁺ handling^(87-89). A higher peak I_NaCa with females was observed (Figure.4). Ca²⁺ efflux out of the cell is carried out by the Na⁺/Ca²⁺, SERCA and the sarcolemmal Ca²⁺ pump exchangerto maintain a Ca²⁺ balance at steady-state. Therefore, the I_Ca,L increasing Ca²⁺ concentration would explain the observation of I_NaCa peak, from the Na^+/Ca²⁺ exchanger partly managing Ca²⁺ homeostasis, these findings corroborate findings in Parks and Howlett (2013)⁽⁸⁷⁾. Using computational modelling simulations does possess limitations, the data obtained from the literature for the female model adaptation may lack in validity. Therefore, experimental work should be undertaken to confirm the validity of these findings. However, the use of computational models does allow complex research to be conducted without ethical implications, high expense and a lengthy time frame.

5.2 Gender differences between male and female pro-arrhythmic state models

The application of ISO (Table 3.) to mimic sustained beta-adrenergic stimulation as observed in disease (e.g. heart failure) that are pro-arrhythmic, showed shortening of the APD (Figure 5) as seen with previous studies^(90,91). Between genders, females remained with the longer APD in comparison to males with ISO application, which was consistent with control females (Figure 3).

The ion current I_ss peak magnitude was increased with ISO application in gender variants, males had a higher I_ss peak in comparison to females which corroborates the findings in previous work by Zhu et al (2014)⁽⁹²⁾. The increase in the steady-state outward potassium channel (I_ss) shortens the APD by quicker repolarisation which is in line with the previous finding⁽⁹³⁾. ISO had no effect on I_K1 current when stimulating pro-arrhythmic male and female models.

The effect of ISO on calcium handling (Figure.6) showed marked increases on both gender variants with [Ca²⁺]_i and [Ca2+]S and peak I_NaCa and I_Ca,L. ISO application was identified from the literature to increase maximum I_Ca,L conductance and Jup flux rate. This corroborates with the increased [Ca²⁺]_i and [Ca2+]SR observed, as Ca²⁺ influx into the cell and out of the SR into the cytoplasm via J_up will be upregulated_. This study identified that calcium concentration increased to a greater extent in females, but contradictory research has identified males as having a higher [Ca2+]SR with ISO treatment(94). There may be some discrepancy in these findings from contrasting cellular Ca²⁺ concentration findings between sexes but these findings indicate increased plausibility of spontaneous Ca²⁺release occurring⁽⁹⁵⁾. ISO increased peak I_Ca,L and I_NaCa in both genders to a similar extent. A higher peak I_NaCa current has been associated with the stimulation of arrhythmic behaviour with the addition of ISO in previous literature by Wei et al (2007), identifying that increased I_NaCa would likely heighten depolarisation which would prompt ventricular arrhythmia by after depolarisations, encouraging pro-arrhythmic behaviour with both models⁽⁹⁶⁾.

5.3 Gender difference with Class III Anti-arrhythmic drugs

The class III AA drug parameters implemented (Table.4) resulted in a slight prolongation of the APD in both gender variants (Figure 7.). The increase in APD prolongation was larger in males than females, but overall females APD was longer^(27,34,85). Class III AA drugs delay repolarisation and prolong the refractory to prevent the occurrence of re-entrant arrhythmias, which is an arrhythmic event observed more commonly in young males^(3,22). APD prolongation from class III AA drug application in females resulted in further QT_C prolongation, this has been previously identified to induce torsades de pointes ventricular tachycardia known as drug induced long-QT syndrome^(97,98). This suggests that class III AA drugs are better suited for treatment in males, as QT_C prolongation would not result in adverse effects and re-entrant arrhythmias are reduced. Other studies with similar results indicate that gender should be a factor taken largely into consideration when administering AA drugs to patients, a recommendation for further research is for testing of specific class III AA drugs to identify which are the most likely to trigger drug induced long-QT syndrome in females ^(3,99-101). This could then be applied a healthcare setting and result in risk-based administration of class III AA drugs.

A slight increase in [Ca²⁺]_i and [Ca2+]SR in both genders with class III AA drugs was observed (Figure 6). An increase in [Ca²⁺]_i and [Ca2+]SR in females with class III AA drugs is a concern, the further increase of the SR Ca²⁺ concentration to or above the threshold increases the probability of spontaneous Ca²⁺ release, another finding enhancing the unsafety of using class III AA drugs^(55,102,103). The peak of I_NaCa and I_Ca,L, in both gender models had slightly reduced with class III AA drugs, but no gender differences were identified. However, previous studies using a class III AA drug (Amiodarone) have shown to block the Na⁺/Ca⁺ exchanger and L-type Ca²⁺ channels which would result in a notable reduction of I_NaCa and I_Ca,L currents in both genders^(104-106). The simplistic implementation of drug effects in this study has highlighted some areas for concern, previous electrophysiological differences not recognised in this study(e.g. on I_NaCa and I_Ca,L) might be specific to individual drugs, meaning a comprehensive investigation is required to follow up.

5.4 Gender difference with Class IV Anti-arrhythmic drugs

The class IV AA drug parameters implemented (Table.5) resulted in APD shortening in both genders compared to with ISO (Figure. 9), but a larger APD reduction was observed in females. Limited research has been conducted on the effects of class IV drugs between sexes, but these findings coincide with the aim of class IV AA drugs to control heart rate and contractility, and therefore decreasing the APD⁽²⁶⁾.

Class IV AA drugs did reduce peak I_ss and I_K1 compared to ISO, reducing repolarisation time which reduces the length of the APD (graph not shown). Currently, there are no reported findings on class IV affecting the I_K1 current, therefore further research is essential to confirm these findings are valid.

A key finding shown in Figure 10. was reduction of [Ca²⁺]_i and [Ca2+]SR in both models, although females still had a higher intracellular and SR Ca²⁺ concentration overall compared to males, the reduction seen in [Ca2+]SR discourages spontaneous calcium release and therefore preventing arrhythmic activity, this is due to lower SR Ca²⁺ load away from threshold for spontaneous release^(71,102,103). A decrease in I_Ca,L was observed (Figure 10.) in both genders but to a lesser extent in females compared to males, this was expected due to the non-selective inhibition of L-type channels by class IV AA drugs and a higher I_Ca,Lseen in control females (Figure 4)⁽¹⁰⁷⁾. I_NaCa (Figure 10.) had a slight reeduced peak with class IV AA drugs in both genders compared to ISO, this was as expected as no inhibition of the Na⁺/Ca²⁺ exchanger by class IV anti-arrhythmics occurs (only slow Ca²⁺ channels)^(26,87,108).

5.5 Key findings

Recognition of the pro-longed APD and increased SR calcium concentration with class III AA drug application highlighted that the use of this drug with females is likely to be inappropriate. Problematic behaviour with females is probable with the application of class III AA drugs, due to the promotion of spontaneous Ca²⁺ release. However, this is not a concern with males, due to their naturally shorter APD and lower overall Ca²⁺ concentration observed in gender control models. The pro-longed repolarisation and refractory period reduces the occurrence of re-entrant arrhythmias, which is a type of arrhythmias commonly observed in males, overall implying class III drugs are better suited to males. Class IV drug implementation was shown to be more applicable to females by reducing APD and intracellular and SR Ca²⁺ concentration, therefore preventing the reach of or above threshold promoting spontaneous Ca²⁺ release.

5.6 Limitations of the model and main conclusions

A limitation of this study is the simplistic ISO and AA drug model implementations, both of which could be more comprehensive if informed by a wider range of literature and if specific to individual drugs. Another limitation is that the model is based on rats, not humans. Rats are commonly used in cardiology research meaning there is a wide range of experimental data accessible. But it is important to corroborate findings present in this study with work in higher mammals/and or humans to ascertain whether the proposed mechanisms do translate to the clinical setting. The Ca²⁺ handling model used in this study is a 0D deterministic implementation of a 3D stochastic model, and therefore does not recapitulate spontaneous behaviour which is known to occur physiologically (e.g spontaneous Ca²⁺ release events at high SR calcium loads above 1.1mM, Colman (2019)⁽⁵⁵⁾ . A 3D single cell stochastic model would be required to test spontaneous arrhythmic behaviour upon stimulation to confirm if pro-arrhythmic activity from the simulations ran in this study^(109,110). Single cell models only give an indication of pro-arrhythmic activity, as arrhythmias are inherently tissue scale phenomena, therefore tissue scale modelling would allow the effects of single cell activity on tissue and organ-level behaviour to occur. One such example is quantification of the vulnerable window, which is the spatio-temporal location at which a second impulse can produce a unidirectional propagating wave of excitation and lead to arrhythmic behaviour^(111,112). The vulnerable window is known to be increased by electrical remodelling, as has been implemented in the single cell model used in this study which highlights the importance of progressing to tissue level simulations⁽¹¹³⁾. Preliminary results using the 3D model showed that ISO induces spontaneous Ca²⁺ release in female ventricular myocytes, but not in male, and that these events were not abolished with class III AA drugs (data not shown). However, they were by class IV AA drugs.

This suggests that the protective effect of class III anti-arrhythmics by APD prolongation and refractory period extension against re-entrant arrhythmia development may not prevent against arrhythmogenesis, as spontaneous Ca²⁺ release events may still manifest. Thus, females may be more suited to class IV anti-arrhythmic drugs which act to reduce Ca²⁺ load, whereas males may be better suited to class III anti-arrhythmic drugs which act to prevent re-entrant arrhythmias from refractory period prolongation.

Extension against re-entrant arrhythmia development may not prevent against arrhythmogenesis, as spontaneous Ca²⁺ Despite the limitations, computational modelling provides framework for the investigation of complex, multiscale phenomena such as arrhythmogenesis without the ethical and experimental burden.