Antiplatelet agents and anticoagulants for hypertension

Abstract

Background

Elevated systemic blood pressure results in high intravascular pressure but the main complications, coronary heart disease (CHD), ischaemic strokes and peripheral vascular disease (PVD), are related to thrombosis rather than haemorrhage. Some complications related to elevated blood pressure, heart failure or atrial fibrillation, are themselves associated with stroke and thromboembolism. Therefore it is important to investigate if antithrombotic therapy may be useful in preventing thrombosis‐related complications in patients with elevated blood pressure.

Objectives

To conduct a systematic review of the role of antiplatelet therapy and anticoagulation in patients with high blood pressure, including those with elevations in both systolic and diastolic blood pressure, isolated elevations of either systolic or diastolic blood pressure, to address the following hypotheses: (i) antiplatelet agents reduce total deaths and/or major thrombotic events when compared to placebo or other active treatment; and (ii) oral anticoagulants reduce total deaths and/or major thromboembolic events when compared to placebo or other active treatment.

Search methods

Electronic databases (MEDLINE, EMBASE, DARE, CENTRAL, Hypertension Group specialised register) were searched up to January 2011. The reference lists of papers resulting from the electronic searches and abstracts from national and international cardiovascular meetings were hand‐searched to identify missed or unpublished studies. Relevant authors of studies were contacted to obtain further data.

Selection criteria

Randomised controlled trials (RCTs) in patients with elevated blood pressure were included if they were of at least 3 months in duration and compared antithrombotic therapy with control or other active treatment.

Data collection and analysis

Data were independently collected and verified by two reviewers. Data from different trials were pooled where appropriate.

Main results

Four trials with a combined total of 44,012 patients met the inclusion criteria and are included in this review. Acetylsalicylic acid (ASA) did not reduce stroke or 'all cardiovascular events' compared to placebo in primary prevention patients with elevated blood pressure and no prior cardiovascular disease. In one large trial ASA taken for 5 years reduced myocardial infarction (ARR 0.5%, NNT 200), increased major haemorrhage (ARI 0.7%, NNT 154), and did not reduce all cause mortality or cardiovascular mortality. In one trial there was no significant difference between ASA and clopidogrel for the composite endpoint of stroke, myocardial infarction or vascular death. 
 In two small trials warfarin alone or in combination with ASA did not reduce stroke or coronary events. 
 The ATC meta‐analysis of antiplatelet therapy for secondary prevention in patients with elevated blood pressure reported an absolute reduction in vascular events of 4.1% as compared to placebo. Data on the 10,600 patients with elevated blood pressure from the 29 individual trials included in the ATC meta‐analysis was requested but could not be obtained.

Authors' conclusions

Antiplatelet therapy with ASA for primary prevention in patients with elevated blood pressure provides a benefit, reduction in myocardial infarction, which is negated by a harm of similar magnitude, increase in major haemorrhage.

The benefit of antiplatelet therapy for secondary prevention in patients with elevated blood pressure is many times greater than the harm.

Benefit has not been demonstrated for warfarin therapy alone or in combination with aspirin in patients with elevated blood pressure. Ticlopidine, clopidogrel and newer antiplatelet agents such as prasugrel and ticagrelor have not been sufficiently evaluated in patients with high blood pressure. Newer antithrombotic oral drugs such as dabigatran, rivaroxaban, apixaban and endosaban are yet to be tested in patients with high blood pressure.

Further trials of antithrombotic therapy including with newer agents and complete documentation of all benefits and harms are required in patients with elevated blood pressure.

Plain language summary

Antiplatelet agents and anticoagulants for hypertension

Daily aspirin reduces the incidence of heart attacks to a small degree, but increases the incidence of major bleeding events to a similar degree in patients treated for high blood pressure who have not had a prior stroke or heart attack. In patients with high blood pressure who have had a stroke or heart attack, the benefits of daily low‐dose aspirin outweigh the harms. There is no evidence of benefit for antithrombotic therapy with warfarin alone or in combination with aspirin in patients with high blood pressure. The benefits and harms of the newer drugs glycoprotein IIb/IIIa inhibitors, clopidogrel, prasugrel, ticagrelor and oral antithrombotic agents such as dabigatran and rivaroxaban for patients with high blood pressure have not been studied in clinical trials.

Background

Although systemic (arterial) elevations in blood pressure result in high intravascular pressure, the main complications of elevated blood pressure, coronary heart disease (CHD) events, ischaemic stroke and peripheral vascular disease (PVD), are related to thrombosis. The association between elevated blood pressure and risk for stroke and CHD has a linear relationship, with increasing risk for higher blood pressures (Collins 1994, Collins 1990a, Collins 1990b). Indeed an increase in the diastolic blood pressure by 10 mm Hg is associated with a 33% increase in CHD events and a 50% increase in stroke events (Collins 1994, Collins 1990a). In middle and old age, there appears to be a direct relationship between the level of blood pressure and the risk of cardiovascular death without any evidence of a threshold down to a blood pressure as low as 115/75 mmHg (PSC 2002). In accordance with the observational data, blood pressure reduction trials have shown a 16% reduction in CHD and 38% reduction in stroke (Collins 1990b).

In addition it should be recognised that some of the complications related to elevated blood pressure, heart failure or atrial fibrillation, are themselves associated with thromboembolism (Lip 2001a). Increasing evidence also points towards a prothrombotic or hypercoagulable state conferred by the presence of elevated blood pressure, as evident by abnormalities of coagulation (Lip 1994, Lee 1997), platelets (Lip 1994, Lee 1997) and endothelial function (Vanhoutte 1996, Lip 1997, Lip 2000) in such patients.

It therefore seems plausible that use of antithrombotic therapy may be of particular benefit in preventing the thrombosis‐related complications of elevated blood pressure (Lip 2001a). The antithrombotic agent, acetylsalicylic acid (ASA), is established as an effective agent for secondary prevention in patients with proven occlusive vascular disease. However, it is not recommended for primary prevention and it is unclear whether it has a role in patients with an increased risk of thrombotic complications such as those with elevated blood pressure. Warfarin has also been found to be useful as thromboprophylaxis in patients with elevated blood pressure and atrial fibrillation (Aguilar 2009, SPAF‐II 1994) but if blood pressures remain uncontrolled, such therapy carries significant risk, especially from intracranial haemorrhage.

Objectives

To conduct a systematic review of the role of antiplatelet therapy and anticoagulation in patients with elevated blood pressure, including elevations in systolic or diastolic alone or together. To determine whether antiplatelet agents reduce total deaths and/or major thrombotic events in these patients when compared to placebo or other active treatment; and to determine whether oral anticoagulants reduce total deaths and/or major thromboembolic events in these patients when compared to placebo or other active treatment.

It should be noted that patients with atrial fibrillation (Aguilar 2009), heart failure (Lip 2001) and pre‐eclampsia (Duley 2004) are excluded, as they are the subjects of separate Cochrane reviews.

Methods

Criteria for considering studies for this review

Types of studies

Single or double blind randomised controlled trials comparing antiplatelet drugs or oral anticoagulation with placebo or active treatment were included. Cohort‐studies, non‐randomised controlled studies, and open‐label studies were excluded.

Types of participants

Patients with at least mild increases in blood pressure or isolated systolic or diastolic increases in blood pressure as defined (for practical reasons) by the WHO‐ISH Guidelines for Management of Hypertension 1999 and/or British Society of Hypertension Guidelines for Management of Hypertension 2004 (WHO‐ISHGMH 1999, Williams 2004) were included. In trials prior to 1999, WHO‐ISH guidelines valid at the time of the study/publication were used (WHO‐ISH 1993) (and noted in the review). Generally a systolic blood pressure of >=140 mmHg and/or a diastolic blood pressure of >=90 mmHg was considered to be elevated and fit the criteria.

Patients with atrial fibrillation, congestive heart failure, pre‐eclampsia, eclampsia or pulmonary hypertension were excluded.

The presence or absence of elevated blood pressure related target organ damage at baseline, was analysed separately if possible. For example participants who have had a stroke and have elevated blood pressure represent secondary prevention as compared to studies in individuals who have elevated blood pressure but have no prior vascular disease.

The reader should note that all definitions of elevated blood pressure, including that used in the review, are arbitrary. The crucial question concerns the question "who benefits from this treatment?". Whilst the absolute benefit may be expected to increase with the level of blood pressure, since blood pressure is a risk factor for the events to be prevented, the iatrogenic risk (haemorrhage) is also expected to increase with increasing blood pressure.

Types of interventions

Treatment duration of at least 3 months with antiplatelet agents (aspirin, NSAIDs, dipyridamole, clopidogrel, ticlopidine, prasugrel, ticagrelor ) or oral anticoagulants (warfarin, dabigatran, rivaroxaban, apixaban and endosaban and other vitamin K antagonist anticoagulants) were included. Available data on concomitant treatment were collected, where available.

Types of outcome measures

Primary outcome measure(s): 
 All cause mortality and cardiovascular mortality (stroke, myocardial infarction, sudden death, thromboembolic events).

Secondary outcome measure(s): 
 All non‐fatal cardiovascular events (stroke, myocardial infarction, thromboembolic events such as acute coronary syndrome, acute limb ischaemia, pulmonary embolism, deep vein thrombosis), as a composite endpoint

All major bleeding events (fatal, non‐fatal) as a composite endpoint. A major bleed was defined as haemorrhagic stroke, or major blood loss defined as a drop in haemoglobin of > 2 g/dl with adequate hydration, or urgent transfusion with final haemoglobin after equilibration of less than pre‐bleed level, or orthostatic hypotension, or supine BP < 90/60 mmHg.

Tertiary outcome measure(s): 
 All cardiovascular events·(sudden death, fatal, non‐fatal: stroke, myocardial infarction, thromboembolic events, coronary revascularization) as a composite endpoint.

Any interaction between risk factors for cardiovascular disease or bleeding, and concomitant treatment were analysed if appropriate data were available. Further details will be obtained from trial authors, if possible.

Search methods for identification of studies

The Database of Abstracts of Reviews of Effectiveness (DARE) and the Cochrane Database of Systematic Reviews were searched for related reviews.

The following electronic databases were searched for primary studies:

a)    The Cochrane Central Register of Controlled Trials (2010, Issue 4)

b)    Bibliographic databases, including MEDLINE (2005‐January 2011), EMBASE (2010‐January 2011), and the Hypertension Group Specialised Register (all years). The Hypertension Group Specialised Register includes controlled trials from searches of AGRICOLA, Allied and Complementary Medicine (AMED), BIOSIS, CAB Abstracts, CINAHL, Cochrane Central Register of Controlled Trials, EMBASE, Food Science and Technology Abstracts (FSTA), Global Health, International Pharmaceutical Abstracts (IPA), LILACS, MEDLINE, ProQuest Dissertations & Theses, PsycINFO, SCIRUS, and Web of Science.

c)    International Clinical Trials Registry Platform (WHO‐ICTRP)

Electronic databases were searched using a strategy combining the Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE: sensitivity‐maximizing version (2008 revision) with selected MeSH terms and free text terms relating to antiplatelets, anticoagulants and hypertension.  No language restrictions were used.  The MEDLINE search strategy (Appendix 1) was translated into EMBASE (Appendix 2), Cochrane CCTR (Appendix 3), and the Hypertension Group Specialised Register (Appendix 4) using the appropriate controlled vocabulary as applicable. The search strategies for the previously‐published (2004) version of this review are in Appendix 5.

Other sources:

d)    Abstracts from national and international hypertension meetings were studied to identify unpublished studies and relevant authors of these studies were contacted to obtain further details.

e)    Reference lists of all papers and relevant reviews identified

f)    Authors of relevant papers were contacted regarding any further published or unpublished work

Data collection and analysis

a) Three reviewers (GYHL, DCF and GD), independently, selected trials that met the inclusion criteria.

b) The data extracted includes information such as patient characteristics, concomitant treatments, study eligibility, quality, and outcomes.

c) Trial quality criteria

Assessment of quality for each trial was made in accordance with guidelines in the Cochrane Handbook (Higgins 2011) based on adequacy of randomization, degree of blinding, and loss to follow up.

d) Contacting trialists

For unpublished studies or where data are incomplete in published papers, attempts were made to contact authors or researchers to obtain further details. Where relevant the drug company was contacted to attempt to obtain unpublished trial data on newer antiplatelet drugs that may have been used in patients with hypertension.

e) Resolution of differences

In the rare instances where the three reviewers (GYHL, DCF and GD) disagreed over the grading and inclusion of the studies, a majority decision was used to resolve any differences.

f) Appropriate statistical analyses were used (e.g. fixed or random effects model, testing for heterogeneity, etc). The outcome measures are reported as relative risk, absolute risk reduction (ARR) with number needed to treat (NNT) or relative risk, absolute risk increase (ARI) with number needed to harm (NNH). Sub‐group analyses based on the presence or absence of target organ damage (primary versus secondary prevention); the dose of aspirin (< 325 mg vs > 325 mg); for oral anticoagulants based on INR 2.0 to 3.0 vs INR > 3.0 were planned. Due to lack of sufficient data, funnel‐plot analyses could not be done for the correction for publication bias. At present, we do not intend to use individual patient data.

Results

Description of studies

3044 potentially relevant publications were initially identified and screened for retrieval: 2088 papers were excluded because of the coexistence of atrial fibrillation or generally lack of suitability of the study [see Figure 1]. Finally, 947 studies were considered to be eligible for inclusion in this review, and in total 888 non randomised studies were excluded.

1.

Study flow diagram.

Among the studies evaluating non‐steroidal antiinflammatory drugs (including aspirin), clopidogrel and ticlopidine, 33 studies were not included for further analysis for various reasons: short duration of the studies and inability to obtain data for the hypertensive subgroup (see Characteristics of excluded studies). Nine potentially appropriate randomised controlled trials remained. One study (TPT 1998) had two publications, with two different definitions for elevated blood pressure used: patients with systolic blood pressure >145 mm Hg (2000 publication), and patients treated with antihypertensive drugs at entry or during the trial (1998 publication). Both sets of data reported on different outcomes in hypertensive subgroups. These data were included as much as possible without duplication.

Three studies evaluating warfarin were excluded, since detailed data on the hypertensive patients was not available (CARS 2001, SPIRIT 1997, WARS 2001). One study remained for further analysis (Huynh 2001). Five studies investigating glycoprotein IIb/IIIa inhibitors or ticlopidine and their impact on cardiovascular and haemorrhagic events lacked sufficient follow up data (EPILOG 1998, EPIC 1999, Finelli 1991, Novo 1996, Schuhlen 2001). A further publication was excluded due to the inability to acquire the relevant data (TASS 1998).

The ATC meta‐analysis (APTC 1994) included mostly secondary prevention trials (142 out of 145) and compared the effects of antiplatelet therapy (at least 1 month duration) to placebo on a composite outcome measure ‐ major vascular events (defined as non‐fatal myocardial infarction, non‐fatal stroke or vascular deaths). Individual patient data for patients with baseline diastolic blood pressure >90 mmHg was reported from 29 of the trials. These trials must have been mostly secondary prevention trials and probably would have met the inclusion criteria of this review. However, several attempts were made to obtain more information from the ATC authors as to the identity of and data from these 29 trials, but no replies were received.

Most of the excluded trials were less than one month duration, in particular the 5 GPIIb/IIIa inhibitor and ticlopidine trials were only observing the effects over a few days. We did not think that this observational period would be sufficient to make any meaningful statements about the prevention of thromboembolic events in hypertensive patients. Furthermore the NSAID trials assessed the effects on blood pressure itself rather than thromboembolic complications.

Risk of bias in included studies

The only randomised placebo‐controlled trial designed to investigate the effects of antithrombotic therapy (aspirin) on cardiovascular events and haemorrhagic complication in treated hypertensive patients was the large Hypertension Optimal Treatment (HOT 1998) trial. This trial was a primary prevention trial in patients with elevated blood pressure (DBP between 100‐115 mmHg). The ASA component of this trial was a double‐blind placebo controlled design comparing ASA 75 mg once daily with identical placebo tablets. Randomization was computer‐generated and patients were stratified for a number of geographic variables. A clinical event committee evaluated events masked to treatment allocation. Major cardiovascular events were defined as fatal and non‐fatal myocardial infarction, fatal and non‐fatal stroke and other cardiovascular deaths. Silent myocardial infarction was defined as new Q or QS waves without clinical signs of myocardial infarction.

Among the other included studies, we tried to obtain further information from the authors. In the majority of the studies we were unable to retrieve the data, which was largely due to the length of time since publication of the original trials. However, details were successfully retrieved from two studies (CAPRIE 1996, Huynh 2001). One study (TPT 1998) had a duplicate publication including a subgroup analysis with a different definition of hypertension, systolic blood pressure >145 mm Hg (2000 publication), than in the data set provided from the original trial, patients on antihypertensive medication at entry or during study (1998 publication). Both sets of data have been considered separately.

CAPRIE study (CAPRIE 1996): A secondary prevention trial in 19,185 patients with recent ischaemic stroke, recent MI or symptomatic peripheral arterial disease. Randomisation: computer‐generated balanced blocks of four treatments with random sequential allocation of drug supplies packaged in predetermined order, each carton containing tablets for four patients; degree of blinding: a central validation committee validated the reported outcome events unaware of randomisation code; independent statistical centre performed analysis before randomisation scheme was provided; losses to follow up: 22 in clopidogrel group, 20 in ASA group; withdrawn nil. Definition of endpoints: all‐cause mortality, vascular death, myocardial infarction, ischaemic or haemorrhagic stroke, all‐cause rehospitalization, hospitalization for ischaemic events (unstable angina, TIA, limb ischaemia) or for bleeding events.

Huynh et al (Huynh 2001): A secondary prevention trial in patients with unstable angina or non‐ST elevation MI and prior CABG. Patients with uncontrolled systemic hypertension (>180/95 mmHg) were excluded. Randomisation: warfarin (INR 2.0‐2.5) + placebo; aspirin 80 mg OD + placebo; or warfarin + aspirin; double‐blind; degree of blinding: independent pharmacist adjusted warfarin dosage, placebo‐warfarin had regular blood tests and mock; placebo‐warfarin adjustments, endpoint evaluation before unblinding of treatment arm; losses to follow up: 3 patients (2.2% of 135 patients). Definition of endpoints: composite endpoint of all cause mortality, myocardial infarction or unstable angina requiring hospital admission. Other endpoints were coronary revascularization. This study was under powered by a low event rate to detect a difference between treatment groups.

Thrombosis Prevention Trial (TPT 1998): Primary prevention trial in patients with high risk of ischaemic heart disease. Patients with current or recent history of possible peptic ulceration or history of possible or definite MI or stroke were excluded. Randomisation: double blind, placebo controlled; degree of blinding: independent reviewer unaware of treatment group; losses to follow up: 58 (1.1% of 5499 patients); data for subgroup analysis of patients treated with antihypertensive medication at entry or during trial was reported in a subsequent publication. Definition of endpoints: fatal and non‐fatal events (i.e. coronary death and fatal and non‐fatal myocardial infarction), Stroke was a secondary end‐point, with results for thrombotic and haemorrhagic events to be distinguished as far as possible. Definition of all events was based on WHO criteria.

Effects of interventions

Aspirin versus placebo (HOT 1998 and TPT 1998)

All cause mortality: All‐cause mortality was reported in one trial (HOT 1998). We were unable to obtain data relating to all cause mortality from TPT 1998. In the HOT 1998 study, mortality was 3.0% in the aspirin treated patients compared with 3.2% in the placebo control group and was not different (p=0.36). The cardiovascular and non‐cardiovascular mortality, were also not different.

Cardiovascular events: Cardiovascular events were evaluated in two studies. In the HOT 1998 trial aspirin reduced myocardial infarction (ARR 0.5%, NNT 200) and major cardiovascular events (ARR 0.57%, NNT 176) as compared to placebo. However, when all cardiovascular events including silent MI were considered, the reduction failed to reach statistical significance. Similarly in the TPT 1998 no significant difference was observed in the cardiovascular events in the subgroup with elevated blood pressure: ASA 8.5% versus placebo 7.5% (TPT 1998). When the results of both studies were pooled for all cardiovascular events there was no significant difference with ASA as compared to placebo (OR 0.92; 95% CI: 0.81 to 1.05, see Analysis 1.2).

1.2. Analysis.

Comparison 1 Aspirin vs placebo, Outcome 2 All Cardiovascular events.

Stroke: Ischaemic strokes were reported in two trials (HOT 1998, TPT 1998). In the pooled results there was no significant difference between ASA and placebo (OR 0.94, 95% CI: 0.76 to 1.17, see Analysis 1.1).

1.1. Analysis.

Comparison 1 Aspirin vs placebo, Outcome 1 Stroke.

Haemorrhage: The haemorrhagic effects of ASA compared to placebo were evaluated in one study (HOT 1998). There was a significant increase in major bleeds (ARI 0.65%, NNH 154) and minor bleeds (ARI 0.73%, NNH 137) with ASA as compared to placebo. These increases are mainly explained by increased gastrointestinal bleeds (72 versus 34 for major bleeds and 30 versus 18 for minor bleeds) and nasal bleeds (22 versus 12 for major bleeds and 66 versus 24 for minor bleeds). There was no significant difference in fatal haemorrhagic events (ASA 7 versus placebo 8).

ASA versus clopidogrel (CAPRIE 1996) 
 
 The composite endpoint of stroke, myocardial infarction or vascular death has been evaluated in one study in patients with elevated blood pressure and was not different in patients taking ASA, 12.1%, as compared to clopidogrel, 11.0% (CAPRIE 1996). Data on all‐cause mortality, cardiovascular mortality, all cardiovascular events, stroke, myocardial infarction, or haemorrhagic events were not reported.

Aspirin and/or warfarin (Huynh 2001)

This study with a low event rate was underpowered to detect a difference in the hypertensive subgroup of patients. Data on all‐cause mortality, cardiovascular mortality, or all cardiovascular events were not reported.

This study evaluated the effects of ASA and warfarin alone or in combination on the composite endpoint of death and cardiovascular events (MI, stroke, angina, cardiac arrest and revascularization procedures).

Revascularization procedures were included in cardiovascular events as an endpoint, but no events occurred. There were no significant differences in this composite end point in the hypertensive subgroup (ASA 2/14, warfarin 5/15, ASA+warfarin 5/14).

Haemorrhage: There was no significant difference in the haemorrhagic complications (ASA 0/14, warfarin 0/15, ASA+warfarin 1/14) (Huynh 2001). The single haemorrhagic event occurred in the group treated with ASA and warfarin.

Warfarin versus no warfarin (TPT 1998)

Data on all‐cause mortality, cardiovascular mortality, or all cardiovascular events were not reported.

Coronary events: Coronary events (coronary death, fatal and non‐fatal myocardial infarction) were evaluated in one study comparing warfarin and warfarin plus aspirin (warfarin group) versus aspirin and placebo (no‐warfarin group) in patients with elevated blood pressure (TPT 1998). Coronary events were not different in the warfarin group, 7.4%, as compared to the no‐warfarin group, 8.6%. In the same study the hypertensive subgroup was analysed according to the occurrence of strokes (excluding haemorrhagic strokes). Strokes were not different with warfarin (aspirin plus warfarin or warfarin alone), 2.8%, compared to no‐warfarin group (aspirin or placebo), 3.2%. Combined cardiovascular events (stroke and coronary events) were not different between warfarin and no‐warfarin groups.

Haemorrhagic strokes: Data for haemorrhagic strokes were only available in pooled form combining warfarin and aspirin compared to placebo. (TPT 1998). No significant difference between the two arms was found. No haemorrhagic strokes occurred in the placebo group (N=342) but 4 such events were observed in the group taking warfarin and aspirin (N=338). Due to the low event rate, no meaningful statistical analysis was possible.

Discussion

Elevated blood pressure potentially confers a prothrombotic state (Lip 1994, Lip 1997, Lip 2000). Antithrombotic drugs such as aspirin, dipyridamole, clopidogrel, ticlopidine, ticagrelor, prasugrel, GPIIb/IIIa inhibitors and oral anticoagulants may have a particular role in the treatment of patients with elevated blood pressure. This antithrombotic action might prevent death and/or cardiovascular thrombotic events. However, data from previous studies suggests that antithrombotic therapy is associated with an increase in haemorrhagic complications. This risk might be amplified in hypertensive patients, potentially reversing the benefit/harm ratio for this therapy.

The primary analysis in the present review is based on 4 high quality prospective randomised double‐blind controlled trials. Two of these trials evaluated the effects of ASA as compared to placebo (HOT 1998, TPT 1998), whilst one trial compared the effects of aspirin, warfarin and aspirin plus warfarin (Huynh 2001). A further trial investigated the outcome of aspirin versus clopidogrel (CAPRIE 1996).

Currently there is only one large RCT with antithrombotic therapy that solely recruited patients with elevated blood pressure [HOT 1998]. Several of the potentially suitable trials enrolled up to 66% of patients with elevated blood pressure, but despite considerable effort, we were unable to obtain the data for the subgroup of patients with elevated blood pressure. One large meta‐analysis of prolonged antiplatelet therapy (APTC 1994) assessed outcome data in 142 trials in 'high risk' patients (some vascular disease or other conditions implying an increased risk of occlusive vascular disease) and 3 trials in 'low risk' patients. These authors were able to get individual patient data on the subgroup of patients with diastolic blood pressure of >90 mm Hg from 29 of these trials. They have reported this data on the composite outcome measure major vascular events (defined as non‐fatal myocardial infarction, non‐fatal stroke or vascular deaths). We were unable to get further information from the authors as to which trials were involved or the details of the data from the individual trials. Thus this data could not be included in this systematic review.

The ATC meta‐analysis of prolonged antiplatelet therapy compared to placebo in 100,000 patients with or without elevated blood pressure reduced all cause mortality (ARR 1.1%, NNT 91) which was mostly due to vascular deaths (ARR 1.0%, NNT 100); nonfatal myocardial infarction (ARR 1.1%, NNT 91); non‐fatal stroke (ARR 0.5%, NNT 200); and the composite outcome measure major vascular events (ARR 2.4%, NNT 42). Haemorrhagic events were not reported. Thus the benefit of antiplatelet therapy for secondary prevention in patients with or without elevated blood pressure has been established by the ATC meta‐analysis.

In the subgroup of patients with elevated blood pressure (29 trials, 10,600 patients), antiplatelet therapy significantly reduced major vascular events to a greater extent than the whole analysis (ARR 4.1%, NNT 25). All cause mortality, cardiovascular mortality, non fatal MI, nonfatal stroke and haemorrhagic events were not reported in the subgroup of patients with elevated blood pressure. This data suggests that the absolute benefit for antiplatelet therapy for secondary prevention is greater in patients with elevated blood pressure than in patients with normal blood pressure.

The HOT trial demonstrated a significant reduction by aspirin as compared to placebo in major cardiovascular events (ARR 0.6%, NNT 176) and myocardial infarction (ARR 0.5%, NNT 200). This absolute benefit for primary prevention in patients with elevated blood pressure is substantially lower than that seen in the ATC analysis for secondary prevention. Furthermore, strokes, cardiovascular and all cause mortality were not significantly decreased by ASA in the HOT trial. In particular, the stroke rate was not affected by ASA, and as not all patients had CT scans, the proportion of ischaemic and haemorrhagic strokes in HOT is uncertain. ASA was also associated with a significant increase in major bleeds, (ARI 0.7%, NNH 200) and minor bleeds, (ARI 0.7%, NNH 137). The differences in these events are mainly explained by increased gastrointestinal bleeds with ASA as compared to placebo (72 versus 34 for major bleeds and 30 versus 18 for minor bleeds) and nasal bleeds (22 versus 12 for major bleeds and 66 versus 24 for minor bleeds). There was no significant difference in fatal haemorrhagic events (ASA 7 versus placebo 8). Since the magnitude of the harm is similar to the benefit, ASA cannot be recommended for primary prevention in patients with elevated blood pressure.

The ATC meta‐analysis of the 3 primary prevention trials based on 28,000 patients also showed no significant difference in all cause mortality, nonfatal stroke or cardiovascular events. The significant decrease in nonfatal MI (ARR 0.5%, NNT 200) in the antiplatelet group compared to placebo is similar to that seen in the HOT trial. Haemorrhagic events were not reported in the primary prevention population.

There haven't been any head to head comparisons of ASA and clopidogrel in patients with elevated blood pressure. The subgroup analysis of patients with elevated blood pressure (CAPRIE 1996) demonstrated a non‐significant trend towards a decreased composite endpoint of stroke, myocardial infarction or vascular death with clopidogrel as compared to ASA. This would suggest that clopidogrel is a reasonable alternative for patients with elevated blood pressure and cardiovascular disease who are intolerant to ASA. However, further studies are required comparing other antiplatelet drugs with ASA in patients with elevated blood pressure.

Adequate blood pressure control is advisable when using antithrombotic therapy in patients with elevated blood pressure, as suggested by the Thrombosis Prevention Trial (TPT 1998). A retrospective observational analysis from this trial (2000 publication) showed that treatment with ASA compared to placebo reduces major cardiovascular events particularly strokes in people with SBP 130‐145 mm Hg, lower BP (SBP<130 mm Hg), whereas risk reduction was not seen in patients with systolic blood pressure >145 mm Hg. This suggests that the benefit of aspirin therapy may occur primarily in patients with well controlled blood pressure.

There are only limited data available on warfarin as an antithrombotic agent in patients with elevated blood pressure. There was no significant difference in cardiovascular events or haemorrhagic complications when comparing warfarin with placebo or ASA or the combination of ASA and warfarin. However, there was a non‐significant trend suggesting that warfarin is inferior to aspirin in preventing thromboembolic events in this group. Warfarin also seems to be associated with a higher rate of haemorrhagic complications. Based on currently available data, warfarin cannot be recommended for primary prevention in patients with elevated blood pressure.

There were no data available from subgroup analyses of patients with elevated blood pressure treated with glycoprotein IIb/IIIa inhibitors. However, oral glycoprotein IIb/IIIa inhibitors have failed to demonstrate an advantage over ASA after percutaneous coronary revascularization (EXCITE 2000). These patients are at high risk of thromboembolic events.

There are no data available with newer antiplatelets (i.e. prasugrel, ticagrelor) or oral anticoagulants (i.e. dabigatran, rivaroxaban, apixaban, endosaban) as these have not been tested as yet in patients with elevated blood pressure.

Authors' conclusions

Implications for practice.

Antiplatelet therapy with ASA for primary prevention in patients with elevated blood pressure provides a benefit, reduction in myocardial infarction, which is negated by a harm of similar magnitude, increase in major haemorrhage.

For secondary prevention in patients with elevated blood pressure the reduction of major vascular events with antiplatelet therapy is many times greater than the increase in major haemorrhagic events and appears to be greater than for secondary prevention in patients with normal blood pressure.

Antithrombotic therapy with warfarin alone or in combination with ASA in patients with elevated blood pressure has not been proven to reduce cardiovascular events and has a trend towards increased haemorrhagic events.

Ticlopidine, clopidogrel and newer antiplatelet agents such as prasugrel and ticagrelor have not been sufficiently evaluated in patients with high blood pressure. Newer antithrombotic oral drugs such as dabigatran, rivaroxaban, apixaban and endosaban are yet to be tested in patients with high blood pressure.

Implications for research.

Further trials of antiplatelet therapies in patients with elevated blood pressure are required to identify subgroups in whom benefits outweigh the harms and to determine whether there are differences in the effectiveness of different drugs. Specific attention to the standardized documenting of haemorrhagic events is essential. All strokes should be investigated with computerized tomography to differentiate between those of ischaemic and haemorrhagic origin.

Feedback

Hypertension and the ATC meta‐analysis, 31 October 2007

Summary

This review concludes that ASA is efficacious for secondary prevention of cardiovascular events in people with elevated blood pressure based on the ATC meta‐analysis (29 trials, 10,600 patients). However, the Cochrane authors were unable to obtain the data for their own analysis of the subgroup of high risk patients with hypertension. Additionally, haemorrhagic events were not reported and may have altered the risk/benefit ratio. Cleland challenged the validity of the ATC meta‐analysis, making the following points: 
 ‐ The Antiplatelet Trialists' Collaboration meta‐analysis was state‐of‐the‐art for its time but not robust by modern standards. 
 ‐ The findings of the meta‐analysis were driven largely by data from small positive trials, with a dearth of small negative trials. 
 ‐ Most clinicians agree that meta‐analysis is useful for confirming robust clinical trial results ‐ which are lacking for long‐term aspirin therapy ‐ or calculating the size of trial required to provide a robust result; however, proof depending on meta‐analysis alone is now generally considered weak.[1] 
 The Cochrane review authors argue that, since (1) the ATC meta‐analysis included 100,000 high risk patients with or without elevated blood pressure and (2) there was an overall reduction in mortality demonstrated, consequently, the benefit of antiplatelet therapy for secondary prevention extends to patients both with or without elevated blood pressure. This is a logical fallacy. All of the benefit might have occurred in the subgroup of non hypertensive patients. 
 For all these reasons, the benefit of antiplatelet therapy for secondary prevention in patients with elevated blood pressure should not be considered established. 
 The primary endpoint, which determines the main conclusions of the review, should have included safety endpoints rather than just efficacy endpoints. Fatal bleeding and intracranial bleeding are particularly important and should be separately included in the primary or secondary endpoints rather than using all major bleeding events (fatal, disabling, and other non‐fatal) as a composite endpoint. 
 Because of excluding observational and population based studies from consideration in the safety analysis of the small warfarin trials (Huynh and Throm Prev Trial '98), the bleeding complications were very likely understated. Adding the safety data concerning vitamin k inhibitors from previous RCTs, observational studies, and population based studies to the safety and efficacy results of the two warfarin RCTs reviewed here, the implications for practice should state, "Vitamin k inhibitors should be contraindicated as an antithrombotic therapy approach to improving clinical outcomes in patients with elevated blood pressure." 
 Likewise, the implications for research section should state, "Because of the bleeding risk and lack of efficacy in published RCTs, further trials with vitamin k inhibitors as antithrombotic treatment to improve clinical outcome in patients with hypertension would be unethical."

1. Cleland JG. Chronic Aspirin Therapy for the Prevention of Cardiovascular Events: A Waste of Time, or Worse? Nature Clinical Practice Cardiovascular Medicine. May 30, 2006; 3(5):234‐235.

Reply

We thank Dr Cundiff for his comments. Many patients with hypertension have associated vascular disease or comorbidities, e.g. previous myocardial infarction (MI) or stroke, and thus, many are currently treated with aspirin as secondary prevention. He claims that the benefit of antiplatelet therapy for 'secondary prevention' in patients with elevated blood pressure should not be considered established. Given that 'secondary prevention' implies those patients who are post‐event, this point is invalid. 
 Dr Cundiff also has concerns over the general interpretation of the Antiplatelet (now Antithrombotic) Trialists' Collaboration meta‐analysis. All meta‐analyses are limited by the studies included (or not, as the case may be), but the counter‐arguments to John Cleland's stance have been repeatedly debated in detail, with many concerns raised regarding his views and interpretation of the Antithrombotic Trialists' Collaboration meta‐analysis. [1,2] Indeed, many early trials of antiplatelet treatment were rather small to detect moderate benefits reliably, which is why meta‐analyses were needed. Dr Cundiff's suggestion that 'all of the benefit might have occurred in the subgroup of non hypertensive patients' is not supported by the detailed subgroup data provided in the earlier Antiplatelet Trialists' Collaboration meta‐analysis manuscript. [3] 
 We agree that bleeding is an important consideration in antithrombotic therapy, and have already included discussion of this aspect in our review. However, the objective was to assess the impact of antithrombotic therapy on major adverse events in hypertension (i.e. mortality, vascular events, etc), as reported as primary endpoints in various trials, and this was agreed when we submitted our Cochrane review protocol. Our objective was not to redefine primary endpoints in the various trials, especially bleeding definitions were variable between the different trials. 
 Dr Cundiff comments on our inclusion of the small warfarin trials (Huynh 2001 and TPT 1998), and the bleeding complications were very likely understated. We already recognise the limitations of these small studies in our review, and by its nature, many randomized trials only include a small selection of those screened. There is inadequate data on warfarin in hypertension per se. 
 Although one may presume that warfarin is associated with more bleeding, this was not seen in the recent BAFTA clinical trial amongst elderly atrial fibrillation patients in the primary care setting, where warfarin (INR2‐3) was superior to aspirin 75 mg for stroke prevention but the rates of major bleeding were no different between warfarin and aspirin. [4]

1. Baigent C, Collins R, Peto R. Article makes simple errors and could cause unnecessary deaths. BMJ. 2002 Jan 19; 324(7330):167. 
 2. Sudlow C, Sandercock P, Warlow C. Antiplatelet therapy and atherosclerotic events. Commentary is inaccurate. BMJ. 2002 Apr 13; 324(7342):917. 
 3. Antiplatelet Trialists' Collaboration. Collaborative overview of randomised trials of antiplatelet therapy. I. Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. BMJ 1994; 308:81‐106. 
 4. Mant J, Hobbs FD, Fletcher K, Roalfe A, Fitzmaurice D, Lip GY, Murray E; BAFTA investigators; Midland Research Practices Network (MidReC). Warfarin versus aspirin for stroke prevention in an elderly community population with atrial fibrillation (the Birmingham Atrial Fibrillation Treatment of the Aged Study, BAFTA): a randomised controlled trial. Lancet. 2007 Aug 11; 370(9586):493‐503.

Contributors

David K. Cundiff, MD, Occupation Physician 
 Submitter agrees with default conflict of interest statement: "I certify that I have no affiliations with or involvement in any organization or entity with a financial interest in the subject matter of my feedback."

What's new

Date	Event	Description
11 October 2011	New citation required but conclusions have not changed	one new author added: Girish Dwivedi
11 October 2011	New search has been performed	updated search (no new trials found), included Risk of Bias Tables, edited Abstract and Plane Language Summary to improve readability

History

Protocol first published: Issue 3, 2001
 Review first published: Issue 3, 2004

Date	Event	Description
13 August 2008	Amended	Converted to new review format.
14 November 2007	New search has been performed	Minor update.
31 October 2007	Feedback has been incorporated	Response to feedback added.
15 October 2007	Feedback has been incorporated	Feedback added.
25 May 2004	New citation required and conclusions have changed	Substantive amendment.

Appendices

Appendix 1. MEDLINE search strategy

Database: Ovid MEDLINE(R) 1948 to Present with Daily Update 
 Search Date: 13 January 2011 
 ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ 
 1 exp Anticoagulants/ (46017) 
 2 (anticoagulant$ or anti‐coagulant$).tw. (33373) 
 3 exp Antithrombins/ (3501) 
 4 (antithromb$ or anti‐thromb$).tw. (23370) 
 5 (apixaban or dabigatran or endosaban or rivaroxaban).tw. (319) 
 6 exp Coumarins/ (33584) 
 7 (coumarin$ or chromonar or coumestrol or esculin or isocoumarin$ or psoralens or pyranocoumarins or umbelliferones).tw. (7383) 
 8 Warfarin/ (11622) 
 9 (warfarin or coumadine or warfant or coumadin or marevan or aldocumar or tedicumar).tw. (11905) 
 10 exp Platelet Aggregation Inhibitors/ (20698) 
 11 (alprostadil or dipyridamol$ or disintegrin or epoprostenol or iloprost or ketanserin or milrinone or pentoxifylline or s‐nitrosoglutathione or nitrosothiols or ticlopidine or clopidogrel or trapidil).mp. (47776) 
 12 (antiplatelet adj1 (therap$ or regime$ or treatment$ or intervention$ or agent$ or drug$)).tw. (6985) 
 13 (platelet adj1 (antiaggreg$ or anti‐aggreg$ or inhibitor$ or antagonist$)).tw. (1599) 
 14 exp Anti‐Inflammatory Agents, Non‐Steroidal/ (55659) 
 15 nsaid$.tw. (13529) 
 16 ((non‐steroid$ or nonsteroid$) adj2 (antiinflammat$ or anti‐inflammat$)).tw. (23120) 
 17 (ampyrone or antipyrine or apazone or bufexamac or clofazimine or clonixin or curcumin or diclofenac or diflunisal or dipyrone or epirizole or etodolac or fenoprofen or flurbiprofen or ibuprofen or indomethacin or etoprofen or ketorolac or meclofenamic acid or mefenamic acid or mesalamine or naproxen or niflumic acid or oxyphenbutazone or phenylbutazone or piroxicam or prenazone or sulfasalazine or sulindac or suprofen or tolmetin).mp. (81663) 
 18 (aspirin$ or acetylsalicylic acid or salicylat$ or salicylic$).mp. (62706) 
 19 or/1‐18 (316746) 
 20 exp hypertension/ (183481) 
 21 hypertens$.tw. (253176) 
 22 exp blood pressure/ (219902) 
 23 (blood pressure or bloodpressure).tw. (176314) 
 24 or/20‐23 (501994) 
 25 randomized controlled trial.pt. (293970) 
 26 controlled clinical trial.pt. (80614) 
 27 randomized.tw. (216908) 
 28 placebo.tw. (123069) 
 29 drug therapy/ (28697) 
 30 randomly.tw. (148068) 
 31 trial.tw. (252500) 
 32 groups.tw. (1007126) 
 33 or/25‐32 (1517223) 
 34 animals/ not (humans/ and animals/) (3395814) 
 35 (preeclamp$ or eclamp$ or congestive heart failure or chf).ti. (17653) 
 36 33 not (34 or 35) (1233294) 
 37 19 and 24 and 36 (3308) 
 38 limit 37 to ed=20050101‐20110112 (1047)

Appendix 2. EMBASE search strategy

Database: EMBASE <1980 to 2011 Week 01> 
 Search Date: 13 January 2011 
 ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ 
 1 exp Anticoagulant Agent/ (369435) 
 2 (anticoagulant$ or anti‐coagulant$).tw. (40470) 
 3 Antithrombin/ (5562) 
 4 (antithromb$ or anti‐thromb$).tw. (28635) 
 5 (apixaban or dabigatran or endosaban or rivaroxaban).mp. (1528) 
 6 Coumarin/ (4175) 
 7 (coumarin$ or chromonar or coumestrol or esculin or isocoumarin$ or psoralens or pyranocoumarins or umbelliferones).tw. (9558) 
 8 Warfarin/ (44863) 
 9 (warfarin or coumadine or warfant or coumadin or marevan or aldocumar or tedicumar).tw. (16807) 
 10 exp Antithrombocytic Agent/ (201982) 
 11 (alprostadil or dipyridamol$ or disintegrin or epoprostenol or iloprost or ketanserin or milrinone or pentoxifylline or s‐nitrosoglutathione or nitrosothiols or ticlopidine or clopidogrel or trapidil).mp. (76463) 
 12 (antiplatelet adj1 (therap$ or regime$ or treatment$ or intervention$ or agent$ or drug$)).tw. (9762) 
 13 (platelet adj1 (antiaggreg$ or anti‐aggreg$ or inhibitor$ or antagonist$)).tw. (2070) 
 14 exp Nonsteroid AntiInflammatory Agent/ (338110) 
 15 nsaid$.tw. (19154) 
 16 ((non‐steroid$ or nonsteroid$) adj2 (antiinflammat$ or anti‐inflammat$)).tw. (29181) 
 17 (ampyrone or antipyrine or apazone or bufexamac or clofazimine or clonixin or curcumin or diclofenac or diflunisal or dipyrone or epirizole or etodolac or fenoprofen or flurbiprofen or ibuprofen or indomethacin or etoprofen or ketorolac or meclofenamic acid or mefenamic acid or mesalamine or naproxen or niflumic acid or oxyphenbutazone or phenylbutazone or piroxicam or prenazone or sulfasalazine or sulindac or suprofen or tolmetin).mp. (131632) 
 18 (aspirin$ or acetylsalicylic acid or salicylat$ or salicylic$).mp. (154051) 
 19 or/1‐18 (653132) 
 20 exp hypertension/ (346505) 
 21 hypertens$.tw. (312291) 
 22 (blood pressure or bloodpressure).mp. (332104) 
 23 or/20‐22 (642678) 
 24 randomized controlled trial/ (287318) 
 25 crossover procedure/ (29926) 
 26 double‐blind procedure/ (100818) 
 27 random$.tw. (617758) 
 28 (crossover$ or cross‐over$).tw. (54599) 
 29 placebo$.tw. (153949) 
 30 (doubl$ adj blind$).tw. (115336) 
 31 assign$.tw. (175885) 
 32 allocat$.tw. (57607) 
 33 or/24‐32 (910325) 
 34 (animal$ not (human$ and animal$)).mp. (3242311) 
 35 (preeclamp$ or eclamp$ or congestive heart failure or chf).ti. (20631) 
 36 33 not (34 or 35) (816407) 
 37 19 and 23 and 36 (5733) 
 38 2010$.em. (1172596) 
 39 37 and 38 (754)

Appendix 3. Cochrane CCTR search strategy

Database: Wiley Cochrane Central Register of Controlled Trials (2010 Issue 4) 
 Search Date: 13 January 2011 
 ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ 
 #1MeSH descriptor Anticoagulants explode all trees (7293) 
 #2(anticoagulant* or anti‐coagulant* or apixaban or dabigatran or endosaban or rivaroxaban):ti,ab in Clinical Trials (1885) 
 #3 MeSH descriptor Antithrombins explode all trees (722) 
 #4 (antithromb* or anti‐thromb*):ti,ab in Clinical Trials (2010) 
 #5MeSH descriptor Coumarins explode all trees (1398) 
 #6 (coumarin* or chromonar or coumestrol or esculin or isocoumarin* or psoralens or pyranocoumarins or umbelliferones):ti,ab (172) 
 #7 MeSH descriptor Warfarin explode all trees (957) 
 #8 (warfarin or coumadine or warfant or coumadin or marevan or aldocumar or tedicumar):ti,ab (1290) 
 #9 MeSH descriptor Platelet Aggregation Inhibitors explode all trees (7354) 
 #10 (alprostadil or dipyridamol* or disintegrin or epoprostenol or iloprost or ketanserin or milrinone or pentoxifylline or s‐nitrosoglutathione or nitrosothiols or ticlopidine or clopidogrel or trapidil):ti,ab,kw (4771) 
 #11antiplatelet near/1 (therap* or regime* or treatment* or intervention* or agent* or drug*):ti,ab (812) 
 #12platelet near/1 (antiaggreg* or anti‐aggreg* or inhibitor* or antagonist*):ti,ab (174) 
 #13MeSH descriptor Anti‐Inflammatory Agents, Non‐Steroidal explode all trees (12806) 
 #14nsaid*:ti,ab in Clinical Trials (1778) 
 #15((non‐steroid* or nonsteroid*) near/2 (antiinflammat* or anti‐inflammat$)):ti,ab (658) 
 #16(ampyrone or antipyrine or apazone or bufexamac or clofazimine or clonixin or curcumin or diclofenac or diflunisal or dipyrone or epirizole or etodolac or fenoprofen or flurbiprofen or ibuprofen or indomethacin or etoprofen 
 or ketorolac or meclofenamic acid or mefenamic acid or mesalamine or naproxen or niflumic acid or oxyphenbutazone or phenylbutazone or piroxicam or prenazone or sulfasalazine or sulindac or suprofen or tolmetin):ti,ab,kw (11189) 
 #17(aspirin* or acetylsalicylic acid or salicylat* or salicylic*):ti,ab,kw (7906) 
 #18#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 (31954) 
 #19MeSH descriptor Hypertension, this term only (12192) 
 #20hypertens*:ti,ab in Clinical Trials (21167) 
 #21MeSH descriptor Blood Pressure explode all trees (20526) 
 #22(blood pressure or bloodpressure):ti,ab in Clinical Trials (31705) 
 #23#19 or #20 or #21 or #22 (47715) 
 #24#18 and #23 (2371) 
 #25(preeclamp* or eclamp*):ti in Clinical Trials (540) 
 #26 #24 AND NOT #25 (2331)

Appendix 4. Hypertension Group Specialised Register search strategy

Database: Hypertension Group Specialised Register 
 Search Date: 13 January 2011 
 ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ 
 1All fields (anticoag* or anti‐coag* or antithromb* or anti‐throm* or coumarin* or chromonar or coumestrol or esculin or isocoumarin or psoralens or pyranocoumarin* or umbelliferones or warfarin or coumadin* or warfant or marevan or aldocumar or tecicumar) (67) 
 2All fields (platelet aggregation inhibit* or alprostadil or dipyridamol* or disintegrin or epoprostenol or iloprost or ketanserin or milrinone or pentoxifylline or s‐nitrosoglutathione (234) 
 3All fields (nitrosothiols or ticlopidine or clopidogrel or trapidil or anti‐platelet therap* or antiplatelet therap* or anti‐platelet treatment* or antiplatelet treatment* or anti‐platelet intervention* or antiplatelet intervention* or anti‐platelet regime* or 
 antiplatelet regime* or anti‐platelet agent* or antiplatelet agent*) (31) 
 4All fields (anti‐platelet drug* or antiplatelet drug* or platelet antiaggreg* or platelet anti‐aggreg* or platelet inhibitor* or platelet antagonist* or non‐steroidal anti‐inflammator* or non‐steroidal antiinflammator* or nonsteroidal anti‐inflammator* or 
 nonsteroidal antiinflammator*) (75) 
 5All fields (ampyrone or antipyrine or apazone or bufexamac or clofazimine or clonixin or curcumin or diclofenac or diflunisal or dipyrone or epirizole or etodolac or fenoprofen or flurbiprofen or ibuprofen or indomethacin) (94) 
 6All fields (etoprofen or ketorolac or meclofenamic acid or mefenamic or mesalamine or naproxen or niflumic or oxyphenbutazone or phenylbutazone or piroxicam or prenazone or sulfasalazine or sulindac or suprofen or tolmetin) (49) 
 7 All fields (apixaban or dabigatran or endosaban or rivaroxaban) (0) 
 8 1 or 2 or 3 or 4 or 5 or 6 or 7 (440)

Appendix 5. Search strategies from 2008 version of review

The Cochrane Controlled Trials Register (CENTRAL) was searched using the strategy with the keywords outlined below. This was updated by searching MEDLINE 2000‐2001 on Ovid using a standard RCT filter (Dickersin 1996) and EMBASE 1998 to 2001 using an EMBASE RCT filter (Lefebvre 1996).  The NHS Database of Abstracts of Reviews of Effectiveness and other relevant databases were searched to identify eligible studies and review articles. Relevant foreign papers were translated.  Abstracts from national and international hypertension meetings were studied to identify unpublished studies and relevant authors of these studies was contacted to obtain further details.  Searches of reference lists of papers were made.

Central search strategy (2008 review): 
 1 HYPERTENSION  
 2 HIGH BLOOD PRESSURE  
 3 BLOOD PRESSURE  
 4 ((#1 or #2) or #3)  
 5 ANTICOAGULANTS*:ME  
 6 ANTICOAGULANT*  
 7 ANTI‐COAGULANT*  
 8 ANTITHROMBINS*:ME  
 9 ANTITHROMB*  
 10 ANTI‐THROM*  
 11 COUMARINS*:ME  
 12 COUMARIN*  
 13 WARFARIN  
 14 WARFARIN*:ME  
 15 (((((((((#5 or #6) or #7) or #8) or #9) or #10) or #11) or #12) or #13) or #14) 
 16 PLATELET‐AGGREGATION‐INHIBITORS*:ME  
 17 PLATELET*  
 18 ANTI‐PLATELET*  
 19 ASPIRIN*:ME  
 20 ASPIRIN  
 21 ANTI‐INFLAMMATORY‐AGENTS‐NON‐STEROIDAL*:ME  
 22 (NONSTEROID* near ANTIINFLAMM*)  
 23 (NON‐STEROID* near ANTIINFLAM*)  
 24 (NONSTEROID* near ANTIINFLAM*)  
 25 (NON‐STEROID* near ANTI‐INFLAM*)  
 26 (NONSTEROID* near ANTI‐INFLAM*) 
 27 NSAID*  
 28 TICLOPIDINE  
 29 CLOPIDOGREL  
 30 DIPYRIDAMOL  
 31 ((((((((((((((#16 or #17) or #18) or #19) or #20) or #21) or #22) or #23) or #24) or #25) or #26) or #27) or #28) or #29) or#30)  
 32 (#4 and #15)  
 33 (#31 and #32)

Data and analyses

Comparison 1. Aspirin vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Stroke	2	20128	Odds Ratio (M‐H, Fixed, 95% CI)	0.94 [0.76, 1.17]
2 All Cardiovascular events	2	20128	Odds Ratio (M‐H, Fixed, 95% CI)	0.92 [0.81, 1.05]
2.1 All	2	20128	Odds Ratio (M‐H, Fixed, 95% CI)	0.92 [0.81, 1.05]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

CAPRIE 1996.

Methods	randomised, blinded, international
Participants	19,185 patients with previous stroke, MI or symptomatic PVD
Interventions	ASA 325 mg vs. clopidogrel 75 mg
Outcomes	stroke, MI, vascular death
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Comment: Probably done
Allocation concealment (selection bias)	Low risk	Comment: Probably done
Blinding (performance bias and detection bias) All outcomes	Low risk	Comment: Probably done
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Comment: Probably done
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Obtained from medical records; review authors do not believe this will introduce bias.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: Only around 20 patients lost to follow up in each group
Selective reporting (reporting bias)	Low risk	Comment: Reported all predetermined endpoints

HOT 1998.

Methods	PROBE, placebo double‐blind for ASA, DBP<90mmHg DBP<85mmHg DBP<80mmHg
Participants	19,193 pt with DBP 105‐115mmHg age 50‐80
Interventions	ACE or beta‐blocker Felodipine Diuretic Aspirin 75 mg or placebo
Outcomes	Major CV events, + silent MI, All MI, All MI + silent MI, All CVA, CV mortality, total mortality; Fatal bleed, non‐fatal bleed, minor bleed
Notes	multi center/national
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Comment: Probably done
Allocation concealment (selection bias)	Low risk	Comment: Probably done
Blinding (performance bias and detection bias) All outcomes	Low risk	Comment: Probably done
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Comment: Probably done
Blinding of outcome assessment (detection bias) All outcomes	Low risk	A clinical event committee evaluated events masked to treatment allocation: review authors do not believe this will introduce bias.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Review authors do not believe that this was large enough to introduce bias.
Selective reporting (reporting bias)	Unclear risk	Comment: Reported all predetermined endpoints

Huynh 2001.

Methods	double‐blind randomized trial
Participants	135 patients with prior CABG with UA or non‐ST elevation MI
Interventions	ASA+placebo or Warfarin+placebo or ASA+Warfarin
Outcomes	all cause mortality, MI, UA, reperfusion procedure
Notes	INR 2‐2.5
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Comment: Probably done
Allocation concealment (selection bias)	Low risk	Comment: Probably done
Blinding (performance bias and detection bias) All outcomes	Low risk	Comment: Probably done
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Comment: Unblinded pharmacists or physicians, not otherwise involved in the study and patient care, adjusted patient care
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Obtained from medical records: review authors do not believe this will introduce bias
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Comment: The study was terminated prematurely due to difficulty in recruiting because of high rate conventional and investigative procedures although only 3 patients (2.2% of 135 patients) were lost to follow up
Selective reporting (reporting bias)	Unclear risk	Comment: As above
Other bias	Unclear risk	Comment: Underpowered by a low event rate to detect a difference between treatment groups.

TPT 1998.

Methods	double‐blind placebo controlled
Participants	5499 men age 45‐69 increased risk of CHD
Interventions	ASA 75mg vs placebo + Warfarin vs placebo
Outcomes	1. fatal and non‐fatal CHD 2. CVA
Notes	1998 publication: on antihypertensive medications during trial 2000 publication: subgroup analysis of ASA group, according to BP: <130, 130‐145, >145
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Comment: Probably done
Allocation concealment (selection bias)	Low risk	Comment: Probably done
Blinding (performance bias and detection bias) All outcomes	Low risk	Comment: Probably done
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Comment: Probably done
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Obtained from medical records: review authors do not believe this will introduce bias
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: Lost to follow up: only 1.1%
Selective reporting (reporting bias)	High risk	Comment:Beneficial effects of low dose aspirin demonstrated following subgroup analyses

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Ajani 2000	unable to obtain data from hypertensive group.
APTC 1994	unable to obtain data from hypertensive subgroup (paper contains on one figure [Figure 7] of a subgroup of 'high risk' patients from 29 trials who had recorded diastolic blood pressure >90mmHg only, with no information on the individual trials)
Bhatt 2000	unable to obtain data from hypertensive subgroup (subgroup analysis of CAPRIE)
Bhatt 2001	unable to obtain data from hypertensive subgroup (subgroup analysis of CAPRIE)
Califf 2000	observational study
Caravaca 1995	retrospective study
CARS 2001	unable to obtain data of hypertensive subgroup
Cook 1997	self‐selected post trial ASA use
Cusson 1992	treatment duration only 7 days
Drapkin 1972	unable to obtain data of hypertensive subgroup
EPIC 1999	unable to obtain data of hypertensive subgroup
EPILOG 1998	ABCIXIMAB treatment only 12 hour duration
ESPS‐2 1998	unable to obtain data from hypertensive subgroup
Ferri 1993	treatment duration of only 5 days
Finelli 1991	treatment duration of only 60 days
Fisher 2001	calculation of the effect of "Clopidogrel vs placebo"
Forbes 1999	duration of treatment only 7 days
Furey 1993	duration of treatment only 7 days
Gurwitz 1996	duration of treatment only 4 weeks
Harker 1999	Subgroup analysis of CAPRIE study, no data available of hypertensive subgroup
Harpaz 2000	ASA not randomised, retrospective analysis of Bezafibrate Infarction Prevention study of women who were not randomised to Bezafibrate
Hartmann 1995	duration of treatment only 15 days
Harvey 1995	duration of treatment only 4 weeks
Hermida 1997	duration of treatment only 7 days
Herskovits 1985	unable to obtain data from hypertensive subgroup
Hess 1985	unable to obtain data of hypertensive subgroup
Houston 1995	duration of treatment only 3 weeks
Houtsmuller 1984	Cardiovascular events and CVA were not evaluated
Johnson 1996	duration of treatment only 1 month
Klassen 1993	duration of 4 weeks
Klassen 1995	duration of 4 weeks
Lee 1999	prospective cohort study
Lemak 1986	unable to obtain required data
Magagna 1991	Duration of NSAID therapy only 1 month
Magagna 1994	Duration of NSAID therapy only 1 month
Maggioni 1992	ASA not randomised
Mehta 2001	duration of treatment only 4 weeks
Mills 1982	duration of Aspirin treatment only 7 days duration of Indomethacin treatment only 7 days
Minuz 1990	duration of NSAID therapy only 2 weeks
Minuz 1995	duration of NSAID therapy only 3 days
Morgan 1993	Duration of NSAID therapy too short
Morgan 2000	treatment duration 3 weeks
Nawarskas 1999	treatment duration 2 weeks
Novo 1996	No suitable endpoints reported
Oosterga 1998	ASA not randomised
Pancera 1996	treatment duration 3 days
Peto 1998	not placebo controlled, not blinded (avoid ASA)
PHS 1989	unable to obtain data from hypertensive subgroup
Polonia 1995	treatment duration : ASA 2 weeks Indomethacin 1 week
Polonia 1996	treatment duration 1 week
PPP 2001	unable to obtain data from hypertensive subgroup
Pritchard 1997	treatment duration 3 weeks
Proudman 2000	control group treated with corticosteroids, methotrexate and cyclosporin ‐ unsuitable for this analysis as effecting BP and other variables
Puranen 1998	unable to obtain relevant endpoint data of hypertensive subgroup
PURSUIT 1999	Duration of treatment only up to 96 hours
Saloheimo 2001	not randomised, only observational study
Schreiber 1992	open label ASA, treatment duration, unable to retrieve data
Schuhlen 2001	not randomised, duration of treatment only 4 weeks
Sivenius 1993	unable to obtain data of hypertensive subgroup
Smith 1993	duration of Aspirin treatment only 2 weeks
SPIRIT 1997	unable to obtain data of hypertensive subgroup
Sturrock 1994	treatment duration only 4 days
Takeuchi 1991	treatment duration only 1 week
TASS 1998	unable to obtain data of hypertensive subgroup
Thakur 1999	treatment duration only 1 month
Tison 1994	treatment duration 8 weeks
Tsuji 2000	single injection of ASA
UK‐TIA 1991	unable to obtain data of hypertensive subgroup
Walter 1981	patients on antihypertensives excluded, remaining patients SBP 119.4 (3.9) mmHg and DBP 73.5 (2.4)mmHg, ASA (n=17), Warf (n=14)
WARS 2001	unable to obtain data
Whelton 2001	treatment duration 6 weeks

Contributions of authors

Prof Lip and Dr Felmeden contributed to the protocol design and jointly undertook the first version of this review, published in 2004. Dr. Dwivedi joined them as an author for the 2011 update of this review.

Sources of support

Internal sources

• University of Birmingham Centre for Cardiovascular Sciences, City Hospital, Birmingham, UK.

External sources

• No sources of support supplied

Declarations of interest

Dirk Felmeden and Girish Dwivedi: no potential conflicts of interest.

Professor Lip has served as a consultant for Bayer, Astellas, Merck, AstraZeneca, Sanofi‐Aventis, BMS/Pfizer, Biotronik, Portola and Boehringer Ingelheim and has been on the speakers bureau for Bayer, BMS/Pfizer, Boehringer Ingelheim, and Sanofi‐Aventis.

The Haemostasis, Thrombosis and Vascular Biology Unit with which all three authors are affiliated undertakes clinical trials of antithrombotic therapy in cardiovascular disease and stroke, and has received research funding from various pharmaceutical companies involved in thrombosis and antithrombotic therapy.

New search for studies and content updated (no change to conclusions)