NPC Archive Item: Modelling study finds no net benefit from aspirin plus clopidogrel versus aspirin alone in people with AF unable to take warfarin

30 November 2011

This modelling study used data from the ACTIVE-W and ACTIVE-A studies to allow a comparison of dual therapy with aspirin plus clopidogrel versus aspirin alone in people with atrial fibrillation (AF) who were unable to take warfarin. It found that if there is any net clinical difference between the two treatments in terms of the risk of death or disability arising from strokes or bleeds, it is small and could either be in favour of the combination or, indeed, in favour of aspirin alone

Level of evidence:
Level 3 (other evidence) according to the SORT criteria.

Action
Health professionals should continue to follow NICE guidance on the management of AF and offer patients aspirin or warfarin therapy, based on an assessment of their risk of stroke (see below). Patients who would normally be offered warfarin but who are unable to take it should be offered aspirin prophylaxis. Health professionals should also note that clopidogrel is not licensed for use in combination with aspirin for the prophylaxis of stroke in people with AF.

A NICE technology appraisal relating to the use of dabigatran▼ for the prevention of stroke and systemic embolism in patients with AF who have certain risk factors is being developed (expected in December 2011). Other novel anticoagulants may also be licensed and marketed for this indication in the future (see the On The Horizon Rapid Review of the ARISTOTLE trial of apixaban▼).

What is the background to this?
NICE guidance on the management of AF recommends that patients should be considered for either aspirin or warfarin therapy, based on an assessment of their risk of stroke. Patients at high risk should usually be offered warfarin, those at low risk should be offered aspirin. Those at intermediate risk should be offered either warfarin or aspirin, with a decision made on an individual basis after considering the risks and benefits of the alternatives for that person. A shared decision aid produced by the NPC is available to support this decision. If warfarin is contraindicated, NICE recommends aspirin prophylaxis. Clopidogrel is not licensed for use in combination with aspirin in prophylaxis of stroke in people with AF.

Two randomised controlled trials (RCTs) have examined the effects of dual therapy with clopidogrel plus aspirin in people with AF. The ACTIVE W study compared warfarin with dual therapy in people with AF and at least one major cardiovascular (CV) risk factor. The study was stopped early after a median of 15 months when it was found that clopidogrel plus aspirin was associated with a significantly greater risk of major vascular events than warfarin (for more details see MeReC Extra 24, September 2006).

The ACTIVE A study compared clopidogrel plus aspirin with aspirin alone in similar patients to those recruited to ACTIVE W, but for whom warfarin was unsuitable, over a median of 3.6 years. The mean CHADS2 score was 2.0. Clopidogrel plus aspirin reduced the risk of a vascular composite end point compared with aspirin alone. However, this benefit has to be balanced against a significant increase in the risk of major bleeding. The number needed to treat (NNT) to prevent a major vascular event was similar to the number needed to harm (NNH) for major bleeds – around 42 over 3.6 years. ACTIVE A is discussed in more detail in On The Horizon Rapid Review 331.

The authors of the modelling study reviewed here state: ‘Comparing the number of events in trial groups without regard for event type would yield misleading results, because equal weight would be given to events of obviously different clinical significance’. They therefore aggregated data on events from the three antiplatelet therapy groups in the two ACTIVE trials to weight the risk of different outcomes (extra-cranial bleeds, haemorrhagic stroke, death, etc.) to take clinical significance into account. They used three weighting systems to adjust the risk of the different outcomes observed, relative to the impact of ischaemic stoke: weighting by risk of death (i.e. taking into account the risk of death after, for example, a major bleed compared to the risk of death after an ischaemic stroke); by risk of death or disability (EuroQol EQ-5D score 0.6 or less); or by a previously published system. They were then able to convert the rates of the different outcomes (bleeds, strokes, etc.) seen among patients taking either dual therapy or aspirin alone to a ‘common currency’ of the rate of ‘ischaemic stroke equivalents’. The difference in the rate of ischaemic stroke equivalents between the two groups would indicate whether there was a net benefit or net harm from using dual therapy compared to aspirin in the ACTIVE A study.

What does this study claim?
Adding clopidogrel to aspirin therapy prevented a net 0.57 ischemic stroke equivalents (95% CI −0.12 to 1.24) per 100 patient years of treatment when weighted by hazard for death after ischaemia or haemorrhage. When weighted by death or disability after ischaemia or haemorrhage, dual therapy prevented a net 0.67 ischemic stroke equivalents compared to aspirin alone (95%CI −0.03 to 1.18). Neither difference was statistically significant. Net benefit or harm was also not statistically significant in any subgroup (defined by age, sex, previous CV event, diabetes or CHADS2 score).

So what?
This study suggests that if there is any net clinical difference between aspirin monotherapy and aspirin plus clopidogrel dual therapy in patients similar to those recruited to the ACTIVE A study, it is small. Furthermore, the confidence intervals indicate that this could either be a net benefit in favour of the combination or, indeed, in favour of aspirin monotherapy. The study is helpful insofar as it enables some more nuanced comparison of outcomes from these two treatment options. However, it has several limitations. The authors highlight three of these: firstly, the definition of disability was applied retrospectively, and the absence of EQ-5D data for every patient may have skewed the results. Secondly, all deaths were assumed to have been related to bleeding or stroke, but it is possible that patients died from other causes. Thirdly, the results apply to the patients recruited to the ACTIVE A trial, but may not apply for patient groups who have different risk of major bleeds or risk of stroke (for example as indicated by a different CHADS2 score).

However, a further limitation is the absence of a prospective or retrospective estimation of the impact of outcomes by patients themselves. Although the EQ-5D is a patient-based measure of disability, it does not capture a patient’s view on the implications, for him or her, of either a stroke or a bleed. Presenting the data in a readily accessible format may be helpful. The NPC has prepared a shared patient decision aid (PDA) which shows the benefits and risks of warfarin and aspirin in AF, for different baseline risks of stroke. A shared PDA comparing aspirin and aspirin plus clopidogrel has not been prepared, since the latter is unlicensed and is not recommended by NICE. However, a Cates plot which provides a pictorial illustration of the results of ACTIVE A is included in part 3 of the NPC’s recorded e-learning event on AF (slide 26).

The problem of how to manage patients with AF at high risk of stroke for whom warfarin is unsuitable is likely to change in the future, with the advent of several novel anticoagulants. For example, the AVERROES study of apixaban versus aspirin in people with AF not suitable for warfarin was terminated early after a mean of 1.1 years because of a clear benefit in favour of apixaban. The rate of the primary outcome of stroke or systemic embolism was 1.6% per year among patients assigned to apixaban and 3.7% per year among those assigned to aspirin (hazard ratio [HR], 95%CI 0.32 to 0.62, p<0.001). The differences in risk of death and risk of major bleeding were not statistically significant but this is difficult to interpret because as the trial was terminated early, it may have been underpowered to detect a true difference in the risk of either of these outcomes. In addition, there are likely to remain some patients who are at high risk of stroke but for whom none of the novel anticoagulants or warfarin is suitable.

Dabigatran is the only one of the novel oral anticoagulants which is currently licensed for the prevention of stroke in patients with AF. A NICE technology appraisal relating to the use of dabigatran for this indication is being developed (expected December 2011). Rivaroxaban has received a positive opinion for this indication (with slight differences in the specified risk factors). A NICE technology appraisal relating to rivaroxaban for this indication is in progress (expected May 2012).

Study details
Connolly SJ, et al. Net clinical benefit of adding clopidogrel to aspirin therapy in patients with atrial fibrillation for whom vitamin K antagonists are unsuitable. Ann Intern Med 2011;155:579–86

Design Modelling study, based on the results of the ACTIVE A and ACTIVE W RCTs.

Patients 10,041 patients with AF, 7554 of whom were not candidates for warfarin therapy.

Intervention and comparison Clopidogrel 75 mg/day plus aspirin 75 to 100 mg/day, versus aspirin 75 to 100 mg/day.

Outcomes and results Data on events from the three antiplatelet therapy groups in the two ACTIVE trials were aggregated and used to weight the risk of different outcomes relative to the outcomes associated with ischaemic stoke, by three weighting systems: by risk of death; by risk of death or disability (EuroQol EQ-5D score 0.6 or less); or by a previously published system. The net clinical benefit of dual antiplatelet therapy was defined as the sum of weighted event incidence with dual antiplatelet therapy subtracted from the sum of weighted event incidence on control treatment, expressed as ischemic stroke equivalents prevented per 100 patients years. See Table below.

Table. Net clinical benefit of adding clopidogrel to aspirin therapy in the ACTIVE A trial, according to three event-weighting systems and no weighting

*positive indicates a net benefit, negative indicates a net harm
**all major events weighted as 1.0

Study sponsorship: Bristol Meyers Squibb and Sanofi Aventis.
Further information on atrial fibrillation can be found on NHS Evidence and in the e-learning section of the NPC website.

Please comment on this MeReC Rapid Review using our feedback form.

Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you