Improved Anti-Platelet Compositions and Methods

The present invention relates to compositions for the inhibition of aggregation in platelets which comprise N-acetyl-cysteine and associated methods. In particular the invention relates to compositions comprising a combination both N-acetyl-cysteine and aspirin (acetylsalicylic acid). The compositions are used in the treatment and/or prevention of a number of disease states where anti-platelet action has been found to be effective and whilst the main focus is toward treatment of diabetes, the composition will also be useful in relation to other disease states including coronary artery disease, stroke, peripheral vascular disease and pre-eclampsia. Furthermore, the invention relates to methods of making and using such compositions.

Platelets (also referred to as thrombocytes) are anuclear, disc shaped, metabolically active cell fragments which typically have a diameter in the region of 1 .5-3.0 μm. Platelets circulate in the blood and are involved in the cellular mechanisms of primary haemostasis (coagulation) which leads to the formation of blood clots. When an injury occurs, the formation of a blood clot is initiated by platelets clumping together, this process is often referred to as platelet aggregation. Fibrin, along with other clotting factors, then acts to further bind platelets together eventually resulting in a blood clot. Low levels of platelets, or abnormal low level of function in platelets, increases the likelihood of bleeding, whilst high levels of platelets, or over activity (hyperaggregability) of platelets, may increase the risk of platelet aggregation or blood clotting.

The formation of a blood clot is an extremely important function of the body as it prevents bleeding in the event that an individual is wounded. However, there are instances where platelet aggregation and/or the

formation of blood clots can be problematic. In particular, it is well known that the formation of a blood clot inside an artery can block the normal flow of blood to the tissue that the artery supplies resulting in tissue damage. Examples of this can be seen when a blood clot forms in a coronary artery supplying blood to the heart muscle, which results in a heart attack (acute myocardial infarction or AMI); or a blood clot forms in an artery supplying blood to the brain resulting in a stroke (cerebrovascular accident or CVA).

As well as being problematic in cardiovascular cases, hyperaggregation of platelets has also been implicated in diabetes. A significant proportion (3- 4%) of the UK population suffer from type 2 diabetes mellitus; and this proportion is predicted to rise. The metabolic abnormalities which characterise the disease include hyperglycaemia, an increase in free fatty acids and insulin resistance, each of which provoke molecular mechanisms which result in vascular dysfunction. Notably, the principal cause of morbidity and mortality in sufferers of type 2 diabetes relates to cardiovascular complications.

Oxidative stress has been linked to type 2 diabetes (Houstis et al., 2006). Oxidative stress is found to directly impact platelets (Schaeffer et ah, 1999) by depressing antioxidant status and reducing the synthesis and bioavailability of nitric oxide (NO). It also has an indirect impact on platelet function as it instigates endothelial dysfunction. Damage by oxidants and/or endothelial dysfunction are both associated with enhanced platelet activation. This is further exacerbated as many type 2 diabetes patients exhibit reduced sensitivity to the widely used anti-platelet agent aspirin (acetylsalicylic acid).

Antiplatelet (antithrombotic) agents are drugs which exhibit a lytic or agglutinative action on platelets and which therefore interfere with the

blood's ability to clot. They are often used as prophylactics to prevent blood clots from forming where there is a risk of heart attack or stroke and in patients who have already suffered from the same (or other disorders where excessive clotting is a factor).

Aspirin (acetylsalicylic acid) is a non-steroidal anti-inflammatory drug (NSAID) that has potent antiplatelet action. Initially, aspirin was used in treatment as an analgesic and antipyretic drug until it was discovered to also have anti-inflammatory properties. Aspirin prevents blood from clotting by blocking the production of thromboxane A-2, a chemical that platelets produce that causes them to clump. Aspirin accomplishes this by inhibiting the enzyme cyclo-oxygenase-1 (COX-1 ) that produces thromboxane A-2. However, aspirin blocks only one of the several pathways by which platelet aggregation can occur and therefore platelet aggregation can be stimulated via another pathway thus limiting the effect of the drug as an anti-platelet agent.

As endothelial dysfunction and platelet hyperaggregability are partly due to an imbalance of oxidants and antioxidants it may be thought that antioxidant supplements would provide a level of protection to patients with type 2 diabetes. However, clinical trials have been carried out with antioxidants and Vitamin E in patients with cardiovascular disease and these have failed to demonstrate any clinical benefit (Marchioli et al., 2001 ). One potential explanation is that conventional antioxidants only provide a temporary neutralising of reactive oxygen species (ROS), whereas ROS production is continuous and results in an accumulation of oxidative damage.

However, there are a number of endogenous antioxidant defence mechanisms which prevent oxidative damage to tissues. A principal

defence against oxidative stress in platelets utilises glutathione (Y- glutamylcysteinylglycine; GSH). GSH is an important antioxidant that is synthesised intra-cellularly in high (mM) concentrations thus maintaining a reducing intracellular environment protecting cells from oxidative stress.

GSH is oxidised into its disulphide form (GSSG) during oxidative stress, however this is rapidly reversed by the enzyme glutathione reductase (GR). As GR is constitutively active and induced by oxidative stress GSH is predominantly found in its reduced form. Levels of intracellular GSH are tightly regulated by the enzymes involved in its synthesis, reduction and degradation. GSH biosynthesis requires two enzymatic processes; the first is a rate limiting step involving glutamate-cysteine ligase (GCL); the second utilises glutathione synthetase. In patients with type 2 diabetes GSH levels are depressed in platelets due to oxidative stress and, importantly, the hyperaggregability of platelets in those with type 2 diabetes is associated with low intra-platelet GSH content. It has been found that intravenous administration of GSH can have beneficial effects as part of diabetes management. However, there are a number of problems associated with the use of GSH as a therapeutic, in particular, the tripeptide nature of GSH does not allow for oral administration and it has been found that it is difficult for GSH to cross cell membranes thus limiting clinical effectiveness.

It can therefore be seen that it would be beneficial to provide a therapeutic which has antiplatelet properties. This would be particularly useful for the treatment or prevention of type 2 diabetes and other disease states where platelet hyperaggregebility is a factor.

According to a first aspect of the present invention there is provided a composition which comprises a first component which is acetylsalicylic

acid (aspirin) or pharmaceutically acceptable salts thereof and a second component which is N-acetyl-cysteine (NAC) or pharmaceutically acceptable salts thereof for use as a medicament or prophylactic.

Preferably the composition is for use in treating platelet hyperaggregability, or a disease which is caused by, or associated with, platelet hyperaggregability.

Optionally the composition is for use in the prevention or treatment of one or more selected from the list of;

- Diabetes;

- Coronary artery disease;

- Stroke; and

- Peripheral vascular disease.

A further option is that the composition is for use in the prevention or treatment of pre-eclampsia.

Optionally the composition is for use in the prevention or treatment of one or more selected from the list of;

- cerebrovascular disease in patients with diabetes;

- myocardial infarction in patients with diabetes; and

- vascular disease in the cerebral, coronary or peripheral circulation.

Surprisingly, the inventors have found that a combination therapy of aspirin and NAC reduces platelet hyperaggregability significantly. This is the case even in sufferers of type 2 diabetes, where aspirin alone will often exhibit reduced effectiveness. Accordingly, it is preferred that the composition is for use in treating type 2 diabetes and/or diseases

associated with type 2 diabetes characterised by platelet hyperaggregability.

Preferably the composition further comprises a pharmaceutical carrier or excipient.

Most preferably the composition is formulated for oral delivery.

Preferably N-acetyl-cysteine (NAC) is provided in a daily dosage amount of 500-2000mg. More preferably the NAC is provided in a dosage amount sufficient to achieve a physiological concentration of 10 to 100 μM, preferably 20 to 60 μM, especially around 30 μM in the blood of a patient. Preferably the composition comprises a single dose of NAC able to achieve the desired concentration in a patient.

Preferably acetylsalicylic acid (aspirin) is provided in a daily dosage amount of less than or equal to 75 mg. Suitably the daily dosage of aspirin may be 65 mg or less, optionally 55 mg or less. Lower than conventional dosages of aspirin can be achieved through the combination with NAC.

Preferably the composition is a combined preparation for simultaneous, separate or sequential use in anti-platelet therapy. More preferably it is a composition suitable for simultaneous administration of NAC and aspirin. It is especially preferred that the composition is in a unit dose form such as a pill, tablet or capsule.

A second aspect of the present invention relates to use of a first component which is acetylsalicylic acid (aspirin) or pharmaceutically acceptable salts thereof and a second component which is N-acetyl-

cysteine (NAC) or pharmaceutically acceptable salts thereof for the manufacture of a medicament or prophylactic for treating or preventing platelet hyperaggregability or a disease associated with platelet hyperaggregability.

Suitably the first and second components are combined into a single composition as set out above.

Diseases associated with hyperaggregability include diabetes, coronary artery disease, stroke, peripheral vascular disease and pre-eclampsia, and accordingly the medicament or prophylactic may be for treating or preventing one or more of said diseases. Optionally the medicament or prophylactic may be for treating or preventing one or more of cerebrovascular disease in patients with diabetes, myocardial infarction in patients with diabetes, and vascular disease in the cerebral, coronary or peripheral circulation. In particular, the medicament or prophylactic may be for use in treating type 2 diabetes, and/or diseases associated with type 2 diabetes characterised by platelet hyperaggregability.

The treatment or prophylaxis use may include simultaneous, separate or sequential provision of the components. It is the combination of aspirin and NAC that provides the benefit rather than the dosage regime. However, it is generally preferred that the use is for a simultaneous dosage regime as this is simpler for patients. Accordingly, the use may be to provide a medicament or prophylactic suitable for simultaneous administration of NAC and aspirin to a patient.

Preferably the medicament or prophylactic is in the form of a single dosage unit. Suitably the single dosage unit may be adapted for a single

daily dose, or it may be for multiple daily doses. A single daily dose is generally more convenient for patients and is therefore often preferred.

Optionally the medicament or prophylactic is in the form of a pill/tablet.

Optionally the medicament or prophylactic is in the form of a capsule.

Preferably N-acetyl-cysteine (NAC) is provided in a daily dosage amount of 500-2000mg. Preferably NAC is provided at a dosage sufficient to achieve a physiological concentration of 10 to 100 μM, preferably 20 to 60 μM, especially around 30 μM in the blood of a patient. Preferably the desired concentration is maintained in the blood of the patient for 30 minutes or longer.

Preferably acetylsalicylic acid (aspirin) is provided in a daily dosage amount of less than or equal to 75mg. Suitably the daily dosage of aspirin may be 65 mg or less, optionally 55 mg or less.

In a further aspect the present invention provides a method for treating a patient, comprising the simultaneous, separate or sequential steps, of a) administering a therapeutic amount of aspirin; and b) administering a therapeutic amount of NAC.

Suitably the method involves the administration of a composition as set out above.

Preferably the method comprises simultaneous administration of aspirin and NAC.

The method preferably comprises administering N-acetyl-cysteine (NAC) in a daily dosage amount of 500-2000mg. Suitably the dosage is sufficient to achieve a physiological concentration of 10 to 100 μM, preferably 20 to 60 μM, especially around 30μM in the blood of a patient. Preferably the desired concentration of is maintained for at least 30 minutes.

The method preferably comprises administering acetylsalicylic acid (aspirin) in a daily dosage amount of less than or equal to 75mg. Suitably the daily dosage of aspirin may be 65 mg or less, optionally 55 mg or less.

Suitably the patient is suffering from one of more of diabetes, coronary artery disease, stroke, peripheral vascular disease, and pre-eclampsia.

The method is particularly suitable where the patient is suffering from one or more of cerebrovascular disease in patients with diabetes, myocardial infarction in patients with diabetes, and vascular disease in the cerebral, coronary or peripheral circulation.

The method is especially suited where the patient is suffering from type 2 diabetes and/or a disease associated with type 2 diabetes characterised by platelet hyperaggregability.

In order to provide a better understanding of the present invention further discussion will be provided particularly in relation to certain embodiments. These embodiments should not however be regarded as limiting but should instead be taken as examples to assist with the understanding of the present invention.

The discussion makes reference to a number of figures in which;

Figure 1 is a histogram showing the impact of NAC (1 mM) on endothelial function in an in vitro model of oxidative stress-mediated endothelial dysfunction ( ^*** P<0.05, Dunnett's post-hoc test compared to control after one-way ANOVA; NS = not siginificant) (n=6 for all);

Figure 2 shows the impact of NAC (0, 10, 30 and 100 μM) on antioxidant activity (i.e. the ability of platelets from patients with diabetes to quench the EPR signal) in washed human platelets in Tyrode's buffer (37 ⁰ C; n=6);

Figure 3 shows the ability of NAC and GSH to (a) reduce the spin adduct, 4-oxo-tempo to its spin-silent counterpart, tempone-H and (b) quench the EPR signal generated by Tyrodes buffer with the spin-trap CPH;

Figure 4 shows the effect of NAC (2hr, 37 ⁰ C) on (A) platelet aggregation in response to the agonist, thrombin; (B) basal and activated intraplatelet luminescence in the presence of lucigenin (a measure of oxygen-centred free radical production) and (C) intraplatelet GSDH. In all cases, platelets were obtained from patients with type 2 diabetes;

Figure 5 shows the effect of NAC on platelet derived nitrite, as a marker of nitric oxide production (n=6) - platelets were taken from patients with type 2 diabetes;

Figure 6 shows the effect of NAC incubation (30 mM; 2 h; 37 ⁰ C) on thrombin (1 u ml ^"1 )-induced aggregation in whole blood in vitro with or without aspirin (30 mM). The marginal effect of aspirin alone is small compared to that induced by NAC, in both the vehicle and aspirin-treated groups; and

Figure 7 shows the effect of NAC (30 mM; 2h incubation; 37 ⁰ C) on inhibition of collagen-induced platelet aggregation in whole blood in vitro in the presence of increasing concentrations of aspirin (1 -100 mM). The additional effect of NAC was statistically significant (P, 0.01 ; ANOVA, n=9).

NAC

The therapeutic potential of N-acetyl-L-cysteine (NAC) was first recognised in the treatment of paracetamol overdose. Patients who have taken a paracetomol overdose exhibit severe depletion of hepatic glutathione (GSH), a crucial endogenous thiol that is central to a range of detoxification processes, including that of paracetamol. Glutamate- cysteine ligase (GCL), catalyses the initial and rate-limiting step of GSH synthesis and the rate of activity of GCL is heavily dependent on the availability of cysteine. In the case of a paracetamol overdose there is severe GSH depletion as the supply of cysteine is outstripped by demand. NAC acts by diffusion across cell membranes, a capability that is facilitated by the relative non-polarity imparted by the acetylation at the amine of cysteine, prior to de-acetylation by intracellular acetylases, resulting in intracellular accumulation of cysteine. NAC administration therefore increases the amount of cysteine which in turn allows for increased GSH synthesis.

Alongside its antioxidant activity, GSH also regulates various intracellular redox-dependent signalling cascades and is involved in the regulation of inflammatory processes. NAC might, therefore, be expected to enhance intracellular antioxidant defences where the intracellular machinery necessary for its conversion to GSH is present and active. This is an important point because, whilst NAC is often referred to as an "antioxidant", it might only impart an antioxidant effect through increasing intracellular GSH. Alternatively, NAC might have its own antioxidant

effects mediated by the thiol entity, but it is not necessarily the case that possession of a sulphydryl moiety will impart antioxidant properties; indeed homocysteine is seen as a pro-oxidant in some circles.

NAC as an antioxidant

NAC in oral form has previously been shown to improve endothelial function in a rat model of diabetes (Pieper GM, Siebeneich W. Oral administration of the antioxidant, N-acetylcysteine, abrogates diabetes- induced endothelial dysfunction. J Cardiovasc Pharmacol 1998;32(1 ):101 - 5), primarily via an antioxidant effect. Additionally, high levels of NAC are associated with enhanced effects of endothelium and drug-derived NO in platelets (Loscalzo J. N-Acetylcysteine potentiates inhibition of platelet aggregation by nitroglycerin. J Clin Invest 1985 Aug;76(2):703-8 and Stamler J, Mendelsohn ME, Amarante P, Smick D, Andon N, Davies PF, et al. N-acetylcysteine potentiates platelet inhibition by endothelium- derived relaxing factor. Circ Res 1989;65(3):789-95). However, the specific effect of NAC at therapeutically relevant concentrations on GSH metabolism, platelet function and fibrinolytic potential, particularly in patients with type 2 diabetes, has not previously been investigated.

The inventors have carried out work in isolated rat arteries which indicates that NAC can cause a rapid partial reversal of oxidative stress-mediated endothelial dysfunction in vitro. Figure 1 is a histogram showing the impact of NAC (1 mM) on endothelial function in an in vitro model of oxidative stress. Oxidative stress was induced by chemical inactivation of endogenous Cu/Zn superoxide dismutase (SOD) by the Cu-chelator, diethydithiocarbamic acid (DETCA; 10 mM 10 min incubation), reflected in a reduced maximal vasodilator response to ACh (10 μM). Full responses to ACh were restored by subsequent incubation with NAC or Cu/Zn SOD (150 U/ml).

Electron paramagnetic resonance (EPR) studies have also been carried out in washed human platelets from patients with type 2 diabetes following in vitro addition of NAC. These studies suggest an increased antioxidant capacity in this setting. Figure 2 shows the impact of NAC (10, 30 and 100 μM) on antioxidant activity in washed human platelets in Tyrode's buffer (37 ⁰ C; n=6). 30 μM NAC is shown to depress formation of the oxidised spin-adduct (CP ^" ) from the reduced counterpart, indicative of antioxidant activity.

However, parallel EPR studies to investigate the direct antioxidant potential of NAC indicate that it is a relatively poor antioxidant per se, whilst the endogenous intracellular antioxidant, GSH, for which NAC is a precursor, has powerful antioxidant properties (Figure 3).

Fig 3 shows two graphs to indicate the ability of NAC and GSH to reduce the spin adduct, 4-oxo-tempo to the spin-silent tempone-H, as a measure of reducing power and antioxidant capability (37 ⁰ C; n=6). These graphs show that GSH is a much more powerful antioxidant than NAC.

Experiments have also shown that in vitro treatment of platelets from patients with type 2 diabetes with therapeutically relevant NAC (10, 30 100 μM) is a credible antiplatelet agent, capable of reducing thrombin-induced aggregation by up to 50% (n=14; Fig 4A). Its actions are primarily mediated via an antioxidant effect (Fig 4B); the effect is associated with prior conversion of NAC to the critical intracellular antioxidant, GSH (Fig 4C). There was also concomitant protection of NOS-derived NO at the 30 μM concentration (Fig 5).

Taken together, these results indicate that conversion of NAC to GSH is a pre-requisite for powerful antioxidant activity and that in vitro delivery of NAC at concentrations relevant to oral therapy impacts on platelet activation through an antioxidant effect mediated via prior conversion to GSH.

Details of EPR studies

Patients with type 2 diabetes were recruited from the local population and subjects were requested not to take non-steroidal anti-inflammatory drugs or other agents that might interfere with platelet activity for 10 days prior to sampling. 250 ml of venous blood was drawn from the antecubital fossa using an intravenous cannula (BD Venflon 18G, BD Medical, Oxford, UK) and transferred into tubes containing 3.8% trisodium citrate or 7.5 g L ^"1 K3EDTA. Blood counts (Beckman Coulter ACT8; Beckman Coulter, High Wycombe, UK) were conducted on the trisodium citrate sample, in triplicate.

Sampling took place at approximately 9.30am to minimise the impact of diurnal variation of blood constituents, particularly platelets (Dalby et al., 2000) and antioxidant levels (Valencia et a!., 2001 )), following a 12 h fast.

The reducing power of NAC and GSH was determined by incubating each agent (10 μM - 1 mM) in phosphate buffer saline (PBS; Invitrogen, Paisley, UK) in the presence of the spin adduct, 4-oxo-tempo (100 μM final concentration; Axxora, Nottingham, UK). The ability of each agent to chemically reduce the spin adduct to spin-silent Tempone-H was monitored at intervals throughout a 90 min incubation period (37 ⁰ C) using EPR spectrometry (Miniscope MS200; Magnettech, Germany; parameter settings: BO-field, 3356 G; sweep width, 50 G; sweep time, 30 s; modulation amplitude, 1500 mG; microwave power, 20 mW; microwave

frequency, 9.3 GHz). Controls without antioxidant were run to confirm that 4-oxo-tempo in the absence of antioxidants maintained a constant spin signal.

A novel in vitro model of oxidative stress was used to assess the antioxidant effects of NAC and GSH in vitro. The spin trap (CP-H; 1 mM final concentration, Axxora, Nottingham, UK) was added to HEPES- containing Tyrode's buffer in order to trap radicals generated from the HEPES buffer itself (Grady et al., 1988; Simpson et al., 1988; Habib and Tabata, 2004). The resulting spin-adduct (CP ^" ) generates an EPR signal which is detectable by EPR spectrometry. The antioxidant effects were determined by incubating each antioxidant (10 μM - 10 mM - to reflect both plasma and cellular concentrations of these antioxidants) with Tyrode's buffer in the presence of the spin trap. Controls without antioxidants were run to determine the background auto-oxidation of the spin trap. Measurements were acquired every 30 min for 4 h. All EPR parameters were the same as for the experiment described above.

Blood samples were aliquoted into four groups: a control with no NAC, and three treatment groups (10, 30 and 100 μM NAC: concentrations which are relevant for oral dosing and enhancing GSH levels, while maintaining a low side-effect profile; Pendyala and Creaven, 1995).

Platelets were washed twice (Turnbull et al., 2006) and platelet pellets resuspended in Tyrode's buffer (composition in mM: 137 NaCI, 2.7 KCI, 1 .05 MgSO ₄ , 0.4 NaH ₂ PO ₄ , 12.5 NaHCO ₃ , 5.6 Glucose, 10 HEPES and 0.8 CaCI; pH 7.4). Platelet counts were determined (Beckman Coulter ACT8; Beckman Coulter, High Wycombe, UK) and standardised to 100 x 10 ⁹ platelets L ^"1 using Tyrode's buffer for use in all experiments requiring washed platelets.

Platelet aggregation was assessed using impedance (whole blood) aggregometry in a 4-channel aggregometer (Chrono-Log Model 700 Lumi- Aggregometer; LabMedics, Manchester, UK). 1 ml of diluted blood (500 μl whole blood and 500 μl of 0.9% saline) was pre-warmed at 37 ⁰ C for 5 min. Platelet aggregation was assessed by measuring the area under the curve in response to the agonists ADP (Labmedics, Manchester, UK; concentrations: 0.625, 1.25, 2.5, 5, 10 μM) and thrombin (0.125, 0.25, 0.5, 1 Unit ml ^"1 ) for 6 min. Experiments were conducted at 37 ⁰ C with a stirring speed of 1000 r.p.m. and impedance gain of 0.05.

GSH and GSSG levels were measured in platelet extracts by means of enzyme recycling (Rahman et al., 2006). GCL activity was determined in control (untreated) platelet extracts using a coupled assay with pyruvate kinase and lactate dehydrogenase (Seelig and Meister, 1985). Protein concentrations of the samples were also quantified (Bradford, 1976; Coomassie Protein Assay Kit, Perbio Science UK Ltd, Northumberland, UK).

Antioxidant effects of washed platelets were assessed similarly to the previous EPR experiment with Tyrode's buffer. The spin trap, CP-H (Axxora, Nottingham, UK), was added (1 mM final concentration) to each of the washed platelet samples and the formation of the spin adduct (CP ^" ) was detected using EPR under the same experimental conditions and parameters as described above. A control with Tyrode's buffer alone (no platelets) with CPH was run alongside the samples to allow assessment of the relative power of platelets themselves to depress the EPR signal generated.

ROS generation was determined using lucigenin-derived chemiluminescence (LDCL; Li et al., 1998) in whole blood and washed platelets. Whole blood was diluted using DMEM containing HEPES (100 nM) at a final dilution of 1 :10 (blood:DMEM) and then incubated for 10 min in a Lumi-Aggregometer (Chrono-Log Model 700 Lumi-Aggregometer, LabMedics, Manchester, United Kingdom; settings: temperature 37 ⁰ C; stirring speed 1000 r.p.m.; luminescence gain 2). 125 μM lucigenin (Chirkov et al., 1999), a luminescent probe for detecting ROS, was then added to each sample and incubated for a further 10 min - the luminescent signal (area under the curve) generated corresponds to the basal level of ROS in each sample. 10 U. ml ^"1 thrombin was added to samples to stimulate the release of ROS and the LDCL signal quantified over 40 mins. 300 U. ml ^"1 superoxide dismutase (SOD; Chirkov et al., 1999) was added to verify the sensitivity of the assay for detecting O ₂ ^'" , which should rapidly quench the LDCL signal. Platelet-derived ROS release was also determined by substituting whole blood for washed platelets.

The appearance of NO ₂ ^" in the supernatant from collagen-activated washed platelets provides an assessment of platelet NOS activity (Tsikas, 2004). Platelets were activated at 37 ⁰ C with 20 μM collagen (Labmedics, Manchester, UK) and incubated for 5 min before measuring the level of NO ₂ ^" in solution, using a chemiluminescence NO Analyser (Sievers 28Oi NO analyzer; GE Analytical Instruments, Colorado, USA). Parallel experiments were carried out and aliquots of platelets were incubated with the NOS inhibitor, L-NAME (200 μM), for 10 min prior to activation with collagen, in order to establish platelet NOS-derived NO ₂ ^" .

Results are expressed as the mean ± s.e.m., unless otherwise specified. Concentration-response curves were fitted using non-linear regression

(sigmoidal dose-response, variable slope), except for calibration curves which were fitted to linear regression. Curve fitting and statistical tests were performed using GraphPad Prism software (version 5.00). The distribution of data was assessed using the Kolmogorov-Smirnov test. In the event, all data sets followed Gaussian distribution and therefore parametric tests were appropriate. One-way ANOVA was used to compare more than two groups and one factor. Concentration-response curves were analysed by two-way ANOVA. Dunnett or Bonferroni post hoc tests were conducted, where appropriate. P<0.05 was considered to be statistically significant.

The finding of the study by the inventors is that NAC per se is a weak antioxidant and is unlikely to have any direct antioxidant effect at concentrations that equate to those found in plasma after tolerable oral NAC dosing (10-100 μM). However, 2 hr incubations of equivalent concentrations of NAC with human platelets from patients with type 2 diabetes caused a concentration-dependent increase in intracellular GSH (reduced form), concomitant with a reduction in both basal and agonist- stimulated ROS detection, and had a significant impact on agonist-induced aggregation. There was also a significant effect of 30 μM NAC on platelet nitric oxide synthase (NOS)-derived nitrite accumulation, suggesting that this concentration of NAC at least might either enhance NOS activity or act to protect the endogenous anti-platelet agent nitric oxide (NO) in this setting.

The EPR assays indicate that high μM - low mM concentrations of NAC are required to show any significant impact on ROS-mediated oxidation of the spin-trap, CPH. In contrast, GSH was active at concentrations as low as 30 μM. This finding is important because it suggests that the antioxidant effects of NAC often found in vitro, typically in response to

>1 mM concentrations might be mediated by a direct antioxidant effect that is different from that seen with in vivo oral dosing, whereby prior conversion to GSH is likely to be a pre-requisite for antioxidant activity. This is of crucial importance because it implies that, for NAC to be effective, the rate limiting enzyme for GSH synthesis (GCL) must be fully functional.

These investigations were carried out in relation to therapeutically relevant NAC concentrations on platelet function and biochemistry, with a view to the possible use of this agent as an antithrombotic drug in conditions where oxidative stress might contribute to heightened risk of thrombosis. Previous data in this field indicated that intramuscular GSH injections have an anti-platelet and pro-fibrinolytic impact, whilst exceptionally high concentrations of NAC also have an antiplatelet effect in obese subjects. The results in the present work indicated that, in platelets from patients with type 2 diabetes, NAC significantly reduced aggregation in response to thrombin, an effect that was not agonist-specific, given that ADP-induced aggregation was similarly affected. There was no significant additional benefit of increasing the concentration of NAC from 30 to 100 μM, perhaps suggesting that 30 μM approximates the concentration required to generate a maximal effect. NAC also caused a concentration-dependent effect on intraplatelet GSH and both basal and thrombin-induced ROS detection by luminescence. In keeping with the functional data, however, 100 μM NAC had no further benefit over the 30 μM concentration with respect to the antioxidant capacity of platelets, as measured by EPR.

Oral NAC represents a suitable antiplatelet agent for pro-thrombotic conditions characterised by oxidative stress. In type 2 diabetes, for example, oxidative stress is implicated in both the disease aetiology and in diabetes-induced platelet dysfunction due, in part, to depleted intraplatelet

GSH. However, it is clear that prior conversion of NAC to GSH is essential for a significant antioxidant effect at realistic therapeutic levels; GCL enzyme activity assays in platelets from these patients with type 2 diabetes confirmed that the enzyme was present and active.

Experiments in platelets from healthy control subjects were similar to those reported above, although the impact was less dramatic.

In vitro work to determine the impact of NAC on the efficacy of aspirin is also reported below. NAC is provided at a typical plasma concentration that can be obtained by tolerable oral dosing prior to generating an aspirin concentration-response curve. This gives an indication as to whether the benefits are additive or synergistic and assist in determining the extent of aspirin dose reduction that will be possible with combined therapy.

The above indicates that NAC is a useful adjunctive therapy with aspirin, providing the opportunity of a "combination pill" with NAC and aspirin that could (a) reduce the aspirin dose, helping to limit the gastrotoxic side- effects whilst, at the same time (b) not increase the number of pills that have to be taken by the patient. Use of both agents at the lowest effective dose will reduce the risk of dose related side effects and the combination will be a particularly effective for patients at higher risk of thrombosis i.e. usually those with previous events or high /multiple risk factors. The combined therapy will also reduce the risk of non compliance which is higher with multiple therapies.

Further experiments were undertaken to determine the impact of NAC in blood from patients to which aspirin is added.

The protocol was designed to test the effect of an achievable plasma concentration of NAC (30 μM) on the in vitro concentration-response curve to aspirin in whole blood from patients with type 2 diabetes. Two different agonists were used to stimulate platelet activation: thrombin, to match our previous experiments (n=2 to date) and collagen, which acts almost exclusively via an aspirin-sensitive pathway (n=9).

Our results, shown in Figures 6 and 7, clearly indicate that thrombin- induced aggregation is substantially inhibited by prior incubation (2 h; 37 ⁰ C) with NAC (30 μM) in the absence of aspirin and that the effect is of roughly equal amplitude in equivalent samples that have been co-treated with aspirin (30 μM; Fig 1 ).

In order to explore the impact of NAC specifically on the cyclo-oxygenase (COX) - aspirin-sensitive pathway - specifically, we determined the effect of NAC on the log-concentration response effect of aspirin on collagen- induced aggregation. There were two specific findings from this study: a. aspirin failed to completely inhibit the aggregation of platelets in samples from patients with type 2 diabetes. This is not unexpected on account of the well-documented resistance to aspirin effects amongst this population, but it does serve to highlight the issue surrounding aspirin resistance. b. NAC caused a small but consistent and statistically significant additional inhibition of platelet aggregation over that seen with aspirin.

Interpretation

From these and our previous experiments with thrombin and ADP, we concluded that NAC, at concentrations that are easily achievable in the plasma with oral ingestion, inhibits platelet aggregation induced by

thrombin, ADP and collagen. The NAC effect on thrombin- and collagen- induced aggregation is seen both in the absence and presence of aspirin, although the amplitude of the effect is most pronounced using thrombin.

Conclusions

We conclude that NAC could represent a useful adjunct to aspirin therapy in that it provides additional anti-thrombotic effects over and above those of aspirin in samples from this population. The extent to which this might reduce the aspirin dose necessary for a combi-pill is difficult to gauge on account of the different effects of both aspirin and NAC on different agonists, but we expect this to become evident after a further scheduled clinical trial.

With reference to the additional effect of NAC over aspirin alone in collagen-treated platelets, although the effect is small, it is significant at the P<0.01 level using ANOVA (grouped analysis). The ANOVA test we have used is the correct test for these kind of data - it takes into account all of the data and also allows for pairing (given that the experiments were carried out in parallel on the same samples). The inference is that, although the effect is small, it is consistent both from sample to sample and across the concentration response curve.

Furthermore, it is important to note that the x-axis scale of Figure 7 is logarithmic. To determine the concentration of aspirin required to give the same size of response (equipotent molar ratio), one can draw a horizontal line across at a convenient level (say 75%) and perpendicular down at the line intercept. The anti-log of the values obtained gives you the concentration of aspirin required to give the selected size of response. In this case, the values are 5.3 μM for NAC treated vs 7.6 μM for aspirin alone, hence a -30% reduction in aspirin dose. This represents a

significant reduction in the amount of aspirin administered for a given effect. It can be postulated that, in reality, this is a modest forecast of the potential for NAC to reduce aspirin dose because of the point discussed below.

It is crucial to understand that platelet activation is a process that is controlled simultaneously by a range of processes that interlink in vivo (e.g. ADP, thrombin, collagen, platelet activating factor and others can all activate platelets). Most inhibitors of platelet function, including aspirin, only act on one pathway (hence aspirin having a reasonable impact on collagen that stimulates the arachidonic acid to TXA2 pathway mediated by COX, but a lesser effect in thrombin-mediated effects). This point might underlie aspirin's lack of bleeding complications, but could also underlie the growing evidence that it is less effective (particularly in diabetes) than was once thought. NAC has a modest effect on collagen-induced activation (and ADP for that matter) and a profound effect on thrombin. If one looks at platelet activation as the complex process that it clearly is, it seems likely that the combined effect of aspirin and NAC will have a broad effect across a number of pathways and might, therefore, have a more profound effect on outcome than either agent alone.