Antithrombotic dose: Some observations from published clinical trials

Abstract

The clinical doses of antithrombotics—antiplatelet and anticoagulant agents—need to balance efficacy and safety. It is not clear from the published literature how the doses currently used in clinical practice have been derived from preclinical and clinical data. There are few large randomised controlled trials (RCTs) that compare outcomes with different doses vs placebo. For newer antithrombotics, RCT doses appear to have been chosen to maximise the probability of demonstrating noninferiority when compared to established agents such as warfarin or clopidogrel. Data from RCTs show that aspirin is an effective antithrombotic at doses below 75 mg daily, and that direct oral anticoagulants reduce the risk of stroke in patients with coronary disease at doses 1/4 of those recommended in atrial fibrillation. Lower doses than those currently recommended are safer and still maintain substantial efficacy.

1. INTRODUCTION

Thromboembolism is a leading cause of morbidity and mortality1 and consequently pharmacological prevention, primary or secondary, with antiplatelet and/or anticoagulant therapy is widely advocated.2 Dose determines both therapeutic effect and the risk of adverse events, but whilst the former reaches a plateau, the latter may increase exponentially with higher doses.3, 4 This risk–benefit balance implies an optimal level of platelet or clotting factor inhibition which will depend on dose but can be problematic to monitor clinically or in the laboratory. Warfarin therapy is monitored because of its multiple effects on coagulation and high risk of drug interactions and because there is a validated assay. The precision of international normalised ratio (INR) is modest—aside from analytical variation, biological variation is such that the within‐individual confidence interval around any single estimate is about ±0.4.5 Cross‐sectional data show that in patients with nonvalvular atrial fibrillation (AF) treated with warfarin, the lowest total mortality, which usefully summarises efficacy and safety, is seen at an INR of 2.46 but INR targets remain controversial. The main utility of INR, as with most therapeutic drug monitoring, is probably low and high levels as a guide to compliance and toxicity, respectively.

Large RCTs that investigate the impact of different doses on hard endpoints provide the best guidance as to when antithrombotic benefits can outweigh harms. Interpretation of net efficacy is, however, difficult in published studies as few, understandably, include a placebo treatment group. Here we consider data on common contemporary antithrombotics.

2. ASPIRIN

Aspirin is a nonselective cyclooxygenase (COX) inhibitor that has been used at different doses depending on the indication (Table 1).7 Adverse drug reactions (ADRs) of COX inhibitors are well known (Table 2) and their risk and intensity are less at lower doses, at which aspirin is still remarkably effective as a platelet antiaggregant.7

Table 1.

Aspirin indications for chronic disease, at different doses7, 8, 9

Recommended daily dose (mg)	Indication	ED50 (mg)
50–150	Platelet thromboxane synthesis inhibition	25
600–1200	Analgesia	600
1000–6000	Anti‐inflammatory	? 2000

ED₅₀, mean population oral dose of aspirin required to produce 50% of the maximal effect.

Table 2.

Adverse drug reactions with chronic nonselective cyclooxygenase inhibitors such as aspirin10, 11, 12

Common (>10%)

Increased bleeding, from any site Reduced renal function Increased blood pressure Peptic ulceration

1–5%

Exacerbation of asthma, rhinitis

<1%

Liver dysfunction Rashes: urticaria, erythema multiforme, vasculitis, angioedema

The antithrombotic effect of aspirin appears to be predominantly related to its inhibition of platelet thromboxane synthesis, which is catalysed by COX.7 The mean population oral dose of aspirin required to produce 50% of the maximal antiplatelet effect (ED₅₀), is <30 mg8, 9 (Table 1). The optimal mean population dose of aspirin in secondary prevention of arterial thrombosis, in systematic reviews appears to be 75–150 mg,39 less favourable outcomes with high doses commonly attributed to the adverse consequences of inhibition of endothelial prostacyclin.7 Higher doses may be prescribed to provide some redundancy to allow for imperfect compliance and differences in body size13 and pharmacokinetics.

Even in secondary prevention, in the Dutch transient ischaemic attack study,14 30 mg of aspirin daily, with which there was less haemorrhage and fewer ADRs, appeared to be as effective as 283 mg, suggesting dose–response for net efficacy had plateaued. The efficacy of doses below 75 mg was confirmed in the European Stroke Prevention Study 2,15 in which aspirin 50 mg daily for 2 years, compared to placebo, reduced stroke by 18% (P = .01); the number needed to treat for 1 year for 1 patient to benefit (avoid a stroke: NNTB), was 75. Combination with dipyridamole (a phosphodiesterase inhibitor which potentiates the action of platelet nitric oxide) conferred nearly additive benefit.16 The benefits of dual antiplatelet therapy after acute myocardial ischaemia are now established, although with adenosine diphosphate (ADP)‐receptor antagonists rather than dipyridamole due to poor tolerability.2, 16

Meta‐analysis of aspirin treatment compared to placebo amalgamates potentially heterogeneous studies. In secondary prevention, mean reductions in stroke (which varies with subtype17) and coronary events by about 1/5 and total mortality by 1/10 were observed over 5 years.18 Total mortality in individual studies was only significantly lower in 1 study,19 where it was 21% lower in patients treated for 1 month after acute myocardial infarction with 160 mg daily, compared to placebo; a benefit which was sustained out to 10 years.20 The NNTB for 5 weeks to prevent 1 vascular event was 53 and 1 death was 39. The number needed to be treated for 1 year for 1 patient to sustain an episode of bleeding (number needed to treat for harm, NNTH), was around 171. Other studies failed to show a reduction in mortality, perhaps because event rates were too low in lower risk trial populations.

Treatments to achieve remission in life‐threatening illness such as haematological malignancies are usually successful, with very low NNTBs but with high rates of ADRs and so very low NNTHs. Dose can be titrated carefully against disease variables and ADRs may be viewed as acceptable. If the disease responds to treatment, patients are likely to endure ADRs but in prevention, any ADRs may be viewed as unacceptable. ADRs on aspirin therapy (Table 2) are the same whether used in patients at high or low cardiovascular risk and in lower risk settings usually outweigh the benefits.18, 21 Meta‐analysis shows the net benefits of primary prevention in asymptomatic patients to be small: NNTB was 265 and NNTH 210.22

In the recently reported ASPREE study,23 the second largest primary prevention study to date, compared to placebo, 100 mg of aspirin daily (>3 times the estimated ED₅₀), reduced cardiovascular events by 11% (not significant, the NNTB for 1 year appeared to be around 1000) but was associated with a significant increase in total mortality (NNTH to cause a death was 699). A lower dose of aspirin may reduce this risk, but benefit in primary prevention appears marginal.

3. ADP RECEPTOR ANTAGONISTS

Clopidogrel is the most widely prescribed of the ADP receptor antagonists, which are used exclusively in secondary prevention. The absence of large RCTs to compare outcomes in patients randomised to different clopidogrel doses has, unusually, led to a fixed, flat 75‐mg daily maintenance dose. The oral ED₅₀ of clopidogrel is about 40 mg in healthy subjects,24 but pharmacokinetic variations mediated by common genetic polymorphisms and drug–drug interactions have predicated a 1‐size fits all approach.25 Placebo comparisons demonstrate the NNTB to prevent 1 cardiovascular event is 121 in the 2 weeks post‐acute myocardial infarction,26 and 33 in the year after elective percutaneous coronary intervention.27 There may be no mortality benefit.28

Interestingly, the fixed‐dose paradigm has continued with ticagrelor, a reversible ADP receptor antagonist that is not subject to the same pharmacokinetic variability as clopidogrel.25 Fixed doses of ADP receptor antagonists are not questioned in guidelines or elsewhere in striking contrast to, for example, statins which are also prescribed for coronary prevention but over a more than 60‐fold dose range.29

4. DIRECT ORAL ANTICOAGULANTS

Direct oral anticoagulants (DOACs) are direct acting inhibitors of either thrombin or factor Xa, and are increasingly favoured over warfarin for anticoagulation. These drugs are licenced at set doses for specific indications and their effects on coagulation are not routinely monitored in the laboratory. There is acknowledgement that dose might be individualised, particularly in renal failure, but there is limited evidence as to how this may affect outcomes.30 Lower doses than those employed in acute treatment of venous thromboembolism have been shown to be sufficient in secondary prevention.31, 32

Clinical studies addressing net outcomes at different doses are limited. Most published studies have been noninferiority trials of a fixed‐dose of a DOAC vs an INR‐monitored variable‐dose warfarin in patients with AF. Meta‐analysis33 demonstrated that warfarin, compared to placebo, in patients with AF reduced stroke by 64% (NNTB for 1 year to prevent 1 stroke was 37 in primary and 12 in secondary prevention) and total mortality by 26% (absolute risk reduction 1.6%/annum). The NNTB to save 1 life was therefore 62, whilst the NNTH for 1 patient to sustain a significant extracranial haemorrhage was 333,33 a substantial net benefit that may be difficult to surpass safely. RE‐LY was the only clinical trial that randomised participants to 2 different doses, 110 and 150 mg twice daily of dabigatran, a direct thrombin inhibitor, compared to warfarin in patients with AF.34 The higher dose of dabigatran reduced ischaemic stroke but increased bleeding and did not reduce mortality34—age and renal function were important covariates.35 Meta‐analysis showed a significant reduction in stroke and total mortality compared to warfarin, with significantly less intracranial haemorrhage but a small excess of gastrointestinal bleeding.36 Lower doses of DOACs showed similar reductions in stroke and systemic emboli in meta‐analysis36 compared to warfarin, consistent with a plateau in efficacy.

Total and cardiovascular mortality were reduced by rivaroxaban compared to placebo in patients with coronary disease when employed at 2.5 mg twice daily (relative risk reduction, 32 and 34%, P = .005, .004, respectively after acute coronary syndrome37; 22 and 18%, P = .02, .01 in stable coronary disease38) but not significantly at 5 mg (relative risk reduction 5 and 6% respectively),37 possibly a consequence of increased bleeding and other adverse effects. The NNTB for 1 year with 2.5 mg twice daily after acute coronary syndrome, compared to placebo,37 to prevent a vascular event or death was 95, whilst the NNTH to cause bleeding requiring medical treatment was only 26, probably partly because patients were also on dual antiplatelet therapy.

This dose of rivaroxaban was 1/4 of that used in the AF clinical trials but as in AF, the main benefit was a reduction in stroke, by about 1/2.38 It may be that higher doses were selected in the comparative trials DOAC with warfarin in AF in the pursuit of superior efficacy, to facilitate marketing. Current DOAC doses may therefore be higher than necessary and increase risk without benefit.

5. CONCLUSION

It is desirable to avoid unnecessarily high doses of antithrombotics, which increase the risk and intensity of bleeding and other adverse effects. Clinical trial evidence shows similar efficacy of aspirin in the secondary prevention of thromboembolism at doses below the commonly recommended dose of 75 mg daily, consistent with known pharmacodynamics and with the potential benefit of greater safety. DOACs reduce stroke and total mortality in coronary patients at doses 1/4 of those recommended in atrial fibrillation. It is useful for clinicians to be aware of the substantial antithrombotic efficacy at lower doses, especially should patients sustain, or be at increased risk of, bleeding or other ADRs.

5.1. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY.40

COMPETING INTERESTS

There are no competing interests to declare.

Dimmitt SB, Floyd CN, Ferner RE. Antithrombotic dose: Some observations from published clinical trials. Br J Clin Pharmacol. 2019; 85: 2194–2197. 10.1111/bcp.13968