WO2008146029A2

WO2008146029A2 - Screening method

Info

Publication number: WO2008146029A2
Application number: PCT/GB2008/001895
Authority: WO
Inventors: Michael J. Curtis; Michael Jonathan Shattock; Andras Farkas
Original assignee: King's College London
Priority date: 2007-06-01
Filing date: 2008-06-02
Publication date: 2008-12-04
Also published as: GB0710510D0; WO2008146029A9

Abstract

The invention provides a method of producing a model of cardiac arrhythmias, by applying agents stimulating both the sympathetic and the parasympathetic receptors, or downstream autonomic signalling pathways, in an isolated heart. The resulting model may be used for screening agents for their propensity to cause and to treat cardiac arrhythmias.

Description

Screening Method

The invention relates to a model of cardiac arrhythmias for use in safety screening of drugs and for testing drugs for use in the treatment of cardiac arrhythmias. In particular, the invention relates to a model of torsades de pointes.

Drug-induced torsades de pointes (TdP) is a very serious cardiac arrhythmia in which the morphology of the QRS complex varies from beat to beat and in which the QT interval is markedly increased. If a therapeutic drug has a tendency (liability) to cause torsades de pointes it is essential that the liability be known before the drug is ever administered to humans. Many drugs have been found to induce TdP in man. Scientists and the pharmaceutical industry continue to seek better animal models for predicting drug-induced torsades de pointes liability (Pugsley MK, Curtis MJ. Safety Pharmacology in focus: new methods developed in the light of the ICH S7B guidance document. J Pharm Tox Methods 54:94-98 2006). Part of the value of the present invention is that it is a new model for this purpose.

The new model is based on the concept of 'autonomic conflict'. The autonomic nervous system has two limbs - the 'stimulatory' sympathetic and the 'inhibitory' parasympathetic nervous systems. Classically, the sympathetic and parasympathetic inputs to the heart are thought of as being exact opposites. The dogma in the prior art is that there is a reciprocal relationship in which sympathetic and parasympathetic influences on the heart are not only in opposition but are activated separately or sequentially. However, it has been recognised in a small number of published studies that this 'alternate' control is not the only pattern of autonomic stimulation. A number of circumstances appear to invoke the simultaneous co-activation of the autonomic inputs to the heart resulting in what the inventors have termed 'autonomic conflict'.

It appears that one strong environmental trigger that can induce autonomic conflict in man is sudden cold-water immersion. This simultaneously activates both sympathetic and parasympathetic inputs to the heart and can generate arrhythmias. These arrhythmias have been seen in ~60% of young healthy male volunteers in response to acute cold-water immersion following the break of breath-hold. (Tipton MJ, Kelleher PC, Golden FS. Supraventricular arrhythmias following breath-hold submersions in cold water. Undersea Hyperb Med. 1994;21(3):305-13).

The inventors recognised that the proposed autonomic conflict generated by cold- immersion may be a very strong pro-arrhythmic trigger. In order to model this in the laboratory, they tested combinations of sympathetic and parasympathetic agonists in isolated rat hearts. With a mild background of sympathetic stimulation, pulses of a parasympathetic agonist (in this case acetylcholine) were shown to induce cardiac arrhythmias including a lethal arrhythmia rarely, if ever, seen in isolated hearts (but not infrequent in man): TdP. The inventors recognised that this model of autonomic conflict could be used to test the liability of drugs to invoke torsades de pointes, and to test the ability of anti -arrhythmic drugs to block this.

Millions of pounds are spent annually by the pharmaceutical industry safety-testing agents for their potential to cause arrhythmias, but at present the available screening methods are inadequate (Pugsley & Curtis 2006). There is a clear need for a medium- throughput screening mechanism for assessing the propensity of pharmaceutical agents to cause arrhythmias. Such a model must be simple to enable the required throughput levels.

According to a first aspect of the invention, there is provided a method of making a model of cardiac arrhythmia comprising simultaneously or sequentially perfusing a mammalian heart with a sympathomimetic agent and a parasympathomimetic agent.

According to a second aspect of the invention, there is provided a method of making a model of cardiac arrhythmia comprising simultaneously or sequentially perfusing an isolated heart with a parasympathomimetic agent and a class III antiarrhythmic agent.

According to a third aspect of the invention, there is provided a method of making a model of cardiac arrhythmia comprising perfusing an isolated heart with a parasympathomimetic agent. The method may comprise continuous perfusion or applying pulses of the parasympathomimetic agent. According to a fourth aspect of the invention, there is provided a method of making a model of cardiac arrhythmia comprising perfusing an isolated heart with an IKs blocker. The heart may also be sequentially or simultaneously perfused with one or both of a sympathomimetic agent and a parasympathomimetic agent.

The term 'cardiac arrhythmia' is used herein to mean an irregularity in the rhythm of the heart beat. In the electrocardiogram (as shown in Figure 5), the P wave represents the atrial signal, and the QRS and T represent ventricular signals. The QT interval, PP interval and PR interval are all intervals on the electrocardiogram of the heart. The QT interval represents the time from the start and the end of electrical activity in the ventricles (the ventricular action potential) for each heart beat. The term sinus rate means the rate of excitatory discharge of the sinus node and this is reflected in the PP interval. Torsades de pointes is identified from the ECG as a ventricular arrhythmia (QRS and T waves not triggered by a preceding P wave) that is statistically linked with a wide QT interval in the ECG immediately preceding the arrhythmia. The two components of the definition mean that torsades de pointes is in strictest terms, a syndrome, although the prior art accepts the term arrhythmia. The cardiac arrhythmia is preferably torsades de pointes. Alternatively, the cardiac arrhythmia is an arrhythmia that means that the heart is more likely to develop TdP, or is primed to develop TdP, if the heart is perfused with another arrhythmia causing drug. Accordingly, the model can be used to identify drugs that are likely to cause harmful arrhythmias, especially TdP, in individuals with an increased susceptibility to develop an arrhythmia.

The heart may be found in an anaesthetised animal or may be an isolated heart. An isolated heart is a model system in which the heart has been surgically removed and is maintained alive and beating by administration of blood, or artificial blood-like water- based solution (the technique of perfusion). When used in drug studies, it is preferably mammalian and preferably non-human. It is more preferably from a small mammal, especially one weighing less than 20kg, more preferably less than 15kgs, even more preferably less than lOkgs, such as a rat, mouse, rabbit, ferret or guinea pig, because it is technically difficult to maintain effective perfusion with larger hearts. When the heart is found in an anaesthetised animal, perfusion of the heart is created by intravenous infusions of the agents used.

The term sympathomimetic agent means an agent that mimics one or more of the effects of the sympathetic nervous system. Sympathomimetic agents differ in mechanism of action and selectivity of action. Naturally-occurring (endogenous) agonists include adrenaline and noradrenaline (known as epinephrine and norepinephrine in some countries including USA). These agonists act directly on alpha-lA, alpha-IB, alpha-ID, alpha -2A, alpha-2B, alpha-2C, beta-1, beta-2 and beta-3 receptors, with varying degrees of potency and selectivity. Synthetic agonists/mimetics also achieve a range of selectivity and potency, and some can act indirectly by affecting synthesis, storage, release, metabolism, neuronal and extraneuronal uptake and/or enzymatic degradation of endogenous or synthetic agonist. Preferred sympathomimetic agents include adrenaline and noradrenaline, sometimes referred to as 'catecholamines'. Other examples of sympathomimetics include alpha receptor agonists (e.g., epinephrine), beta receptor agonists (e.g., isoproterenol), Uptake 1 blockers (e.g. cocaine, desipramine), presynaptic alpha2 receptor blockers (e.g., yohimbine), MAO inhibitors (e.g., imipramine, desipramine) and noradrenaline neuronal release stimulators (e.g., ephedrine, amphetamine, tyramine).

A parasympathomimetic agent is an agent that mimics the effect of the parasympathetic nervous system. Various different classes of parasympathomimetic agents are available. As with adrenergic agents, there are endogenous and synthetic parasympathetic agents, and some are selective and others not, while some are direct acting and others indirect acting. There is one primary naturally occurring agonist, acetylcholine. Examples of synthetic, non-endogenous or indirect acting agents include, bethanechol, carbachol, nicotine, muscarine, pilocarine, donepezil, edrophonium, neostigmine, physostigmine, pyridostigmine, tacrine, cisapride, metoclopramide and sildenafil.

The term class III antiarrhythmic agent is an agent that blocks potassium channels, particularly blocking potassium efflux. Examples of class III antiarrhythmic agents include Class III agents include amiodarone, azimilide, bretylium, clofilium, dofetilide, tedisamil, ibutilide, sematilide, and sotalol. Such agents are also known as Kr blockers.

An DCs blocker is an agent that blocks DCs potassium channels. Examples include chromanol.

The invention relates to the simultaneous or sequential application of sympathomimetic agents and parasympathomimetic agents and to the simultaneous or sequential application of class III antiarrhythmic agents and parasympathomimetic agents. When they are applied sequentially, the second agent may be applied during continued exposure to the first, or during the washout of the first, with single or repeat exposures of the agents to each heart. When single exposures are made, a range of durations of exposure to each agent may be used. With repeat exposures, the durations of exposure to each agent from one exposure to the next may be constant or may be varied. For any regime, either agent may be administered first. In a particular embodiment, a sympathomimetic agent is applied continuously, and pulses of a parasympathomimetic agent are applied to bring about the arrhythmia. In another embodiment, a parasympathomimetic agent is applied continuously, and pulses of a sympathomimetic agent are applied to bring about the arrhythmia. In a further embodiment, a class III anti-arrhythmic agent is applied continuously, and pulses of a parasympathomimetic agent are applied to bring about the arrhythmia. In yet a further embodiment a parasympathomimetic is applied continuously, and pulses of a class DI antiarrhythmic agent are applied to bring about the arrhythmia.

Also provided by the invention is a method for testing for the propensity of a drug to cause an arrhythmia comprising perfusing a cardiac model according to the invention with the drug to be tested and observing the model for the appearance, increase or worsening of the signs of arrhythmia.

Also provided is a method for testing the propensity of a drug to reduce or treat an arrhythmia, comprising perfusing a cardiac model according to the invention with a drug and observing the model for the reduction or removal of signs of arrhythmia. Signs of arrhythmia may include torsades de pointes itself, ventricular fibrillation, tachycardia and premature beats, and surrogate biomarkers such as QT prolongation, QT shape change (e.g., triangulation) and variation in P-P interval.

As mentioned above, the mammal is often a small mammal, such as a guinea pig or a rabbit. Hearts from different mammals have different characteristics. For example, hearts in some species display different types or classes of receptor, especially ion channels than those seen in other species. Rabbits hearts have more IKr potassium channels than DCs potassium channels, as do humans, and accordingly, rabbit hearts are thought to be better models for testing the propensity of drugs, especially drugs affecting potassium channels, to cause arrhythmias in humans. Guinea pig hearts have more IKs potassium channels than HCr potassium channels. Accordingly, the guinea pig heart may provide a better model for testing drugs that might have an effect on long QT syndrome, as mutations in the IKs channel are thought to be involved in that syndrome.The mammal from which the heart is taken may be selected according to the test that is to be performed. For example, the heart may be selected for its similarity to a human heart or may be selected for having more or fewer receptors or channels affected by the drug to be tested. Additionally the heart may be modified in order to make the resulting model closer to the heart of interest, hi particular, the heart may be modified to make its reaction to agents more similar to the reaction likely to be seen in a human or other heart. So, a guinea pig heart treated with the perfusion regime of the first aspect of the invention may be additionally perfused with an IKs blocker, such as chromanol, in order to make it more like a rabbit or human heart. A rabbit heart treated with the perfusion regime of the first aspect of the invention may be additionally perfused with an IKr blocker such as clofilium.

The invention will now be described in detail, by way of example only, with reference to the drawings in which:

Figure 1 shows an example of arrhythmias induced by a protocol of autonomic conflict. Figure 2 shows mean incidence of arrhythmias quantified using an arrhythmia score in 6 hearts during autonomic conflict. Note: catecholamines alone or acetylcholine alone result in a low incidence of arrhythmias

Figure 3 shows the effect of the autonomic conflict protocol on the PR interval of the ECG.

Figure 4 shows the effect of the autonomic conflict protocol on the QT interval of the ECG.

Figure 5 is a schematic of an electrocardiogram (ECG) showing the labelling nomenclature classically used and the relationship of these various phases to stylised ventricular action potentials.

Figure 6 shows an example of arrhythmias induced by the sequential application of cycles of the autonomic conflict protocol.

Figure 7 shows conflict-evoked ventricular fibrillation in guinea pig heart.

Figure 8 shows conflict plus clofilium (IKr blocker)-evoked torsades de pointes in guinea pig heart

Figure 9 shows conflict plus clofilium (IKr blocker)-evoked torsades de pointes in a rabbit heart.

The inventors aimed to precipitate cardiac arrhythmias in an isolated rat heart by bringing about a sympathetic/parasympathetic (autonomic) conflict. The inventors superimposed short 30 second bursts of cholinergic stimulation upon hearts perfused with adrenaline/noradrenaline and observed the hearts to see the effects on the PR interval, the QT interval, the incidence of ventricular arrhythmias. Further models were also developed as described in the examples.

Protocols All experiments were performed in accordance with the United Kingdom Home Office "Guide on the operation of the Animals (Scientific Procedures) Act 1986".

Animals and General Experimental Methods

Male Wistar rats (250-35Og, Bantin and Kingman, UK) were anesthetized with pentobarbital (60 mg/kg IP) and heparinized with 250 U IP sodium heparin to prevent blood clot formation in the coronary vasculature. Hearts were excised and placed in ice-cold Krebs perfusion solution containing (mmol/L) NaCl 118.5, NaHCCβ 25.0, MgSO4 1.2. NaH2PO4 1.2, CaCl 1.4, KCl 4.0, and glucose 11.1. All solutions were filtered (5μm pore size) before use to remove particulate matter. Hearts were then perfused in the Langendorff mode with perfusion solutions delivered at 37 degrees Celsius, pH 7.4 and at a pressure of 73 mmHg. A bipolar ECG was recorded by implanting a silver wire electrode into the epicardial surface of the left ventricle and recording with respect to the stainless steel aortic cannula.

Experimental Protocols

Hearts were perfused using Krebs solution for the first 20 minutes of the protocol, noradrenaline/adrenaline (catecholamines) was then perfused for a further 20 minutes. Following this, the isolated hearts were subjected to a 30 second cycle of acetylcholine, in the maintained presence of the adrenaline/noradrenaline, before being returned to the adrenaline/noradrenaline solution alone (washout) for a further 5 minutes. In some studies, the cycle of exposure to acetylcholine followed by a 5 minute washout period was then repeated a further two times (3 cycles in total).

Arrhythmia determination and ECG analysis

The ECG was recorded using a Powerlab system and arrhythmias were assessed in accordance with the Lambeth Conventions (Walker et al., 1988). Ventricular premature beats (VPBs) were defined as premature QRS complexes occurring independently of a P wave, and hence included individual VPBs, bigeminy and salvos as defined by the Lambeth Conventions. QT intervals at the point of 90% repolarisation (QT₉₀), PR intervals were also measured from the ECG, as previously described (Ridley et al., 1992). The ECG was recorded at a sampling rate of 1 kHz allowing millisecond precision for measurement of ECG intervals.

Rationale for choice of drug concentration

The concentrations of noradrenaline (313 nM) and adrenaline (75 nM) were calculated from an estimate (1 μg min^'1 of noradrenaline and 0.25 μg min^"1 of adrenaline) of the respective amounts needed to restore heart rate to values similar to those encountered in conscious rats (Curtis et al., 1985). The concentration ratio is within the range encountered in man, especially under conditions of stress (Baumgartner et al., 1985; Coplan et al., 1989; Ratge et al., 1986). Ascorbate was added to the adrenaline/noradrenaline solution to protect against auto-oxidation as used previously in related studies in perfused hearts (Hearse & Sutherland, 1999). The rationale for the concentration of acetylcholine (5μm) was based on previous studies used in our laboratory.

Statistical Analysis

Unpaired student t-tests were used in this study. The level of significance was considered as p<0.05. Arrhythmia scoring was also used. Scoring was defined as follows: 1 = AV Block, 2 = Ventricular Premature Beats, 3 = Bigeminy, 4 = Salvo, 5 = Ventricular Tachycardia, 6 = Ventricular Fibrillation, 7 = Torsades de Pointes.

Drugs and Materials

(+/-)-Noradrenaline, (-)-adrenaline, acetylcholine, and ascorbate were all obtained from Sigma Chemicals (UK). All salts were reagent grade chemicals obtained from Sigma Chemicals (UK). Water for preparing perfusion solution was supplied using a reverse osmosis system (USF, Elga Ltd. UK), and had a specific resistance of greater than 18 MΩ .

EXAMPLE 1

Figure 1 shows acetylcholine administered in the presence of a mixture of adrenaline and nor-adrenaline and the trace shows an ECG recording. The top trace shows the entire protocol and the lower traces show expanded regions showing the detail of the arrhythmias induced at the points marked in the top trace. In this example, arrhythmias are characterised by a gradual prolongation of the PR interval and the appearance of complex arrhythmias including torsades de pointes.

Figure 2 shows the mean data from 6 hearts exposed to the simulated autonomic conflict protocol. Figure 2A shows that acetylcholine alone induced a low incidence of arrhythmias. Similarly, catecholamines (adrenaline/noradrenaline) alone did not induce any arrhythmias (Figure 2B). However, the combination of acetylcholine in the maintained presence of the catecholamine was arrhythmogenic (Figure 2B). These arrhythmias were reversed (normal sinus rhythm re-established) during the washout of ACh. Thus, both catecholamines alone and acetylcholine alone are not in themselves profoundly pro-arrhythmic. It is the combination of sympathetic and parasympathetic mimetics that is the strong proarrhythmic trigger.

Figure 3 shows changes in the PR interval of the ECG. Figure 3A shows that acetylcholine alone has a mild non-significant effect on PR interval (ie AV nodal conduction). Catecholamines alone also had no effect on PR interval (Figure 3B). However, the combination of acetylcholine and catecholamines significantly prolonged the PR interval (Figure 3B). In many instances, this resulted in AV block.

Figure 4 shows changes in the QT interval of the ECG. In the rat heart, catecholamines prolong the QT interval (as QT interval and action potential duration in this species is unusually dependent of the underlying Ca transient). Acetylcholine, either alone (Figure 4A) or in combination with catecholamines (Figure 4B), shortened the QT interval slightly.

EXAMPLE 2

Figure 6 shows the effect of multiple (3) cycles of acetylcholine exposure (30 sees) administered in the maintained presence of catecholamines (adrenaline/noradrenaline. In this example, as with the single exposure, catecholamines alone did not induce any arrhythmias. However, the combination of acetylcholine in the maintained presence of the catecholamine was strongly arrhythmogenic. These arrhythmias were reversed (normal sinus rhythm re-established) during the washout of ACh (Wash). This pattern was repeated on the reintroduction of a second and third cycle of ACh exposure in the presence of catecholamines. Thus, it is the co-administration of both sympathetic and parasympathetic agonists that, as before, is strongly arrhythmogenic and this protocol is reproducible when repeated sequentially in the same heart.

EXAMPLE 3

The heart was perfused with catecholamines (75 nM adrenaline plus 313 nM noradrenaline) for 5 min then with added 5 μM ACh for 5 min. The resulting ventricular fibrillation is shown in figure 7.

EXAMPLE 4

This heart was perfused with clofilium 3 μM for 15 min, then with added catecholamines (75 nM adrenaline plus 313 nM noradrenaline) for 5 min then with added 5 μM ACh for 5 min. The conflict protocol with the catecholamines and the ACh was repeated twice more in the continued presence of clofilium 3μM perfusion. The torsades de pointes shown in figure 8 occurred during the 3rd ACh perfusion).

EXAMPLE 5 This heart was perfused first with clofilium (0.3 μM) for 15 min, then with added catecholamines 25 nM adrenaline plus 100 nM noradrenaline) for 5 min then with added 5 μM ACh for 5 minutes. TdP was reached and is shown in figure 9.

Claims

L A method of making a model of cardiac arrhythmia comprising simultaneously or sequentially perfusing a heart with a sympathomimetic agent and a parasympathomimetic agent.

2. A method of making a model of cardiac arrhythmia comprising simultaneously or sequentially perfusing a heart with a parasympathomimetic agent and a class III antiarrhythmic agent.

3. A method of making a model of cardiac arrhythmia comprising perfusing a with a parasympathomimetic agent.

4. A method of making a model of cardiac arrhythmia comprising perfusing an isolated heart with an IKs blocker.

5. A method according to any preceding claim, wherein the cardiac arrhythmia is one or more of ventricular premature beats, ventricular tachycardia, ventricular fibrillation and torsades de pointes.

6. A method according to any preceding claim, wherein the heart is from a non- human mammal

7. A method according to claim 6, wherein the non-human mammal is a small mammal, weighing less than 20kgs.

8. A method according to claim 7, wherein the small mammal is a rabbit, ferret or rodent.

9. A method according to claim 8, wherein the small mammal is a rabbit, ferret, guinea pig, rat or mouse.

10. A method according to any preceding claim, wherein the sympathomimetic agent is adrenaline, noradrenaline, epinephrine, isoproterenol, cocaine, desipramine, yohimbine, imipramine, desipramine, ephedrine, amphetamine, or tyramine.

11. A method according to claim 10, wherein the sympathomimetic agent is adrenaline or noradrenaline.

12. A method according to any preceding claim, wherein the parasympathomimetic agent is acetylcholine, bethanechol, carbachol, nicotine, muscarine, pilocarine, donepezil, edrophonium, neostigmine, physostigmine, pyridostigmine, tacrine, cisapride, metoclopramide or sildenafil.

13. A method according to claim 12, wherein the parasympathomimetic agent is acetylcholine.

14. A method according to any preceding claim, wherein the sympathomimetic agent and parasympathomimetic agent are administered sequentially .

15. A method according to claim 1, wherein the sympathomimetic agonist and parasympathomimetic agent are administered simultaneously.

16. A method according to claim 1, wherein the sympathomimetic agent is administered substantially continuously and at least one pulse of parasympathomimetic agent is applied.

17. A method according to claim 1, wherein the parasympathomimetic agent is administered substantially continuously and at least one pulse of sympathomimetic agent is applied.

18. A method for testing for the propensity of an agent to cause an arrhythmia comprising perfusing a cardiac model produced by the method of any preceding claims with the agent and observing the model for the appearance, increase or worsening of the signs of arrhythmia.

19. A method for testing the propensity of an agent to reduce or treat an arrhythmia, comprising perfusing a cardiac model produced by the method of any of claims 1 to 16 with the agent and observing the model for the exacerbation, reduction or removal of signs of arrhythmia.

20. A cardiac model of cardiac arrhythmias, produced by the method of any one of claims 1 to 17.