sign in sign up
<<
Home , Antithrombotic, Thrombosis prevention

Journal of Human Hypertension (2005) 19, 185–196 & 2005 Nature Publishing Group All rights reserved 0950-9240/05 $30.00 www.nature.com/jhh ORIGINAL ARTICLE in hypertension: a Cochrane Systematic review

DC Felmeden and GYH Lip Haemostasis, , and Vascular Biology Unit, University Department of Medicine, City , Birmingham, UK

Although elevated systemic pressure (BP) results on one large trial, ASA taken for 5 years reduced in high intravascular pressure, the main complications (ARR, 0.5%, NNT 200 for 5 years), of hypertension are related to thrombosis rather than increased major haemorrhage (ARI, 0.7%, NNT 154), and haemorrhage. It therefore seemed plausible that use of did not reduce all cause mortality or cardiovascular antithrombotic therapy may be useful in preventing mortality. In two small trials, alone or in thrombosis-related complications of elevated BP. The combination with ASA did not reduce or coronary objectives were to conduct a systematic review of the events. Glycoprotein IIb/IIIa inhibitors as well as ticlopi- role of antiplatelet therapy and anticoagulation in dine and have not been sufficiently eval- patients with BP, to address the following hypotheses: uated in patients with elevated BP. To conclude for (i) antiplatelet agents reduce total deaths and/or major primary prevention in patients with elevated BP, anti- thrombotic events when compared to placebo or other therapy with ASA cannot be recommended active treatment; and (ii) oral reduce total since the magnitude of benefit, a reduction in myocar- deaths and/or major thromboembolic events when dial infarction, is negated by a harm of similar magni- compared to placebo or other active treatment. A tude, an increase in major haemorrhage. For secondary systematic review of randomised studies in patients prevention in patients with elevated BP, antiplatelet with elevated BP was performed. Studies were included therapy is recommended because the magnitude of the if they were 43 months in duration and compared absolute benefit is many times greater. Warfarin therapy antithrombotic therapy with control or other active alone or in combination with in patients with treatment. One meta-analysis of antiplatelet therapy for elevated BP cannot be recommended because of lack of secondary prevention in patients with elevated BP demonstrated benefit. Further trials of antithrombotic reported an absolute reduction in vascular events of therapy are required in patients with elevated BP. 4.1% as compared to placebo. Acetylsalicylic acid (ASA) Journal of Human Hypertension (2005) 19, 185–196. did not reduce stroke or ‘all cardiovascular events’ doi:10.1038/sj.jhh.1001807 compared to placebo in primary prevention patients with Published online 27 January 2005 elevated BP and no prior cardiovascular disease. Based

Keywords: aspirin; warfarin; antithrombotic therapy

Introduction and stroke by more than half.1–3 In middle and old age, there appears to be a direct relationship Although systemic (arterial) hypertension results in between the level of BP and the risk of cardiovas- high intravascular pressure, the main complications cular death without any evidence of a threshold of hypertension, that is, coronary heart disease down to a BP of at least 115/75 mmHg.4 Correspond- (CHD), ischaemic and peripheral vascular ingly, BP reduction trials have shown a reduction in disease (PVD), are related to thrombosis rather than CHD by 16% and stroke by 38%. It should not be haemorrhage. The association between hypertension forgotten that hypertension is a for both and risk for stroke and CHD almost appears to have a haemorrhagic as well as ischaemic strokes. dose–response relationship, with increasing risk for 1–3 It should be recognised that some of the complica- higher blood pressures (BP). Indeed an increase in tions related to hypertension, such as or the diastolic blood pressure (DBP) by 10 mmHg is atrial fibrillation, are themselves associated with associated with an increase in CHD risk by one-third stroke and thromboembolism.5 Increasing evidence also points towards a prothrombotic or hypercoagul- Correspondence: Professor GYH Lip, Haemostasis, Thrombosis, able state conferred by the presence of hypertension, and Vascular Biology Unit, University Department of Medicine, as evident by abnormalities of ,6,7 plate- City Hospital, Birmingham, UK. 6,7 8–10 E-mail: [email protected] lets and endothelial function in such patients. Received 18 October 2004; accepted 18 October 2004; published It therefore seems plausible that use of antithrom- online 27 January 2005 botic therapy may help to prevent thrombosis-related Antithrombotic therapy in hypertension DC Felmeden and GYH Lip 186 complications of hypertension.5 The antithrombotic congestive heart failure, pre-eclampsia, eclampsia or agent, aspirin, has long been used in the treatment pulmonary hypertension were excluded. and secondary prevention of many of the complica- tions of hypertension, but until recently little (c) Types of interventions information has been available on its role in the Treatment duration of at least 3 months with management of asymptomatic, hypertensive indivi- antiplatelet agents (aspirin, nonsteroidal anti-in- dual. Warfarin has also been found to be useful as flammatory (NSAIDs), , clopido- thromboprophylaxis in hypertensive patients with grel, glycoprotein IIb/IIIa inhibitors (GPIIb/IIIa atrial fibrillation,11,12 but if blood pressures remain inhibitors) (ie )) or oral anticoagulants uncontrolled, such therapy carries significant risks, (warfarin and other K antagonist antic- especially from intracranial haemorrhage. oagulants) were included. Available data on con- comitant treatment were collected, where available.

Objectives Types of outcome measures A systematic review of the role of antiplatelet therapy and anticoagulation was conducted in (a) Primary outcome measure(s) patients with systemic (arterial) hypertension, in- All cause mortality and cardiovascular mortality cluding those with systodiastolic hypertension, (stroke, myocardial infarction, sudden death, throm- isolated systolic hypertension and diastolic hyper- boembolic events). tension, to address the following hypotheses: (i) antiplatelet agents reduce total deaths and/or major (b) Secondary outcome measure(s) thrombotic events in patients when compared to All nonfatal cardiovascular events (stroke, myocar- placebo or other active treatment; and (ii) oral dial infarction, thromboembolic events such as anticoagulants reduce total deaths and/or major acute coronary syndrome, acute limb ischaemia, thromboembolic events in patients with hyperten- pulmonary , ), as a sion when compared to placebo or other active composite end point. All major bleeding events treatment. (fatal, nonfatal) as a composite end point. A major It should be noted that antithrombotic therapy in bleed was defined as haemorrhagic stroke, or major atrial fibrillation11 and heart failure13 per se are blood loss defined as a drop in haemoglobin of considered in detail in separate Cochrane reviews. 42 g/dl with adequate hydration, or urgent transfu- In atrial fibrillation, for example, the presence of sion with final haemoglobin after equilibration of hypertension increases the risk of stroke and less than prebleed level, or orthostatic hypotension, thromboembolism, and such patients are at moder- or supine blood pressure under 90/60 mmHg. ate-to-high risk; hypertension and atrial fibrillation also commonly coexist. (c) Tertiary outcome measure(s) All cardiovascular events (sudden death, fatal, nonfatal: stroke, myocardial infarction, thromboem- bolic events, coronary revascularisation) as a com- Methods posite end point. Any interaction between risk Criteria for considering studies for this review factors for cardiovascular disease or bleeding, and concomitant treatment were analysed if appropriate (a) Types of studies data were available. Further details will be obtained Single- or double-blind randomised controlled trials from trial authors, if possible. comparing antiplatelet drugs or oral anticoagulation The search strategy for identification of studies is with control or active treatment were included. summarised in Table 1 (also see Collaborative Cohort-studies, nonrandomised controlled studies, Review Group Research Strategy17). and open-label studies were excluded.

Methods of the review (b) Types of participants Patients with at least mild hypertension or isolated Two reviewers (GYHL and DCF), independently, systolic or diastolic hypertension as defined (for selected suitable trials for the inclusion of the practical reasons) by the WHO–ISH Guidelines for review. There was a review of the inclusion criteria Management of Hypertension 1999 and/or British for each question, the search strategies, the metho- Society of Hypertension Guidelines for Management dology criteria, and methods for pooling the data. of Hypertension 199914,15 were included. In trials Nonrandomised studies would not usually be help- prior to 1999, WHO–ISH guidelines valid at the time ful in a review of the effects of these drugs as any of the study/publication were used.16 Generally, a estimates of treatment effects are likely to be biased, systolic blood pressure (SBP) of X160 mmHg and/or but they have been included as they may help in a DBP of X100 mmHg was considered to be elevated assessing side effects in the case of anticoagulants. and fit the criteria. Patients with atrial fibrillation, Assessment of trial quality will be made in accor-

Journal of Human Hypertension Antithrombotic therapy in hypertension DC Felmeden and GYH Lip 187 Table 1 (a) Cochrane search strategy: (a) I and (b) II

(a) Cochrane Search Strategy I 1. 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 filter35 and EMBASE 1998–2001 using an EMBASE RCT filter.36 2. The NHS Database of Abstracts of Reviews of Effectiveness and any other relevant database were searched to identify eligible studies and review articles. Relevant foreign papers will be translated. 3. 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. 4. Searches of reference lists of papers were made (see Table 2).

(b) Cochrane Search Strategy II 1. HYPERTENSION 2. HIGH BLOOD PRESSURE 3. BLOOD PRESSURE 4. ((#1 or #2) or #3) 5. ANTICOAGULANTS*:ME 6. * 7. ANTI-COAGULANT* 8. *:ME 9. ANTITHROMB* 10. ANTI-THROM* 11. *:ME 12. * 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)

dance with guidelines in the Cochrane Handbook,18 1527 papers were excluded because of the coex- with consideration of adequacy of randomisation, istence of atrial fibrillation or generally lack of degree of blinding, and losses to follow-up for each suitability of the study (see Quorum diagram, trial. Figure 1). Finally, 946 studies were considered to In the rare instances where the two reviewers be eligible for inclusion in this review, and in total (GYHL and DCF) disagreed over the grading and 887 nonrandomised studies were excluded. inclusion of the studies, recourse to a third reviewer Among the studies evaluating NSAIDs (including was made. Appropriate statistical analyses were aspirin), clopidogrel and ticlopidine, 33 studies applied (eg type of therapeutic effect measure, were not included for further analysis because of choice of fixed or random effects model, testing of the short duration of the studies and in 13 publica- heterogeneity, etc). The outcome measure was tions, we were unable to obtain data of the reported as relative hazard rate (RR) with 95% hypertensive subgroup; thus, nine potentially ap- confidence interval (CI). propriate randomised controlled trials remained evaluating the effects of these drugs in hypertensive patients. One study was double reported, but two Results different definitions for hypertension were used: 19 Description of studies patients with SBP 4145 mmHg, or patients treated with antihypertensive drugs at entry or during A total of 2482 potentially relevant publications the trial.20 Both sets of data reported on different were initially identified and screened for retrieval: aspects of antithrombotic therapy in hypertensive

Journal of Human Hypertension Antithrombotic therapy in hypertension DC Felmeden and GYH Lip 188 Potentially relevant publications identified and screened for retrieval: 2130 (NS) 284 (WA) Paper excluded on the basis of the title and abstract 68 (GP) (generally due to lack of suitability of study design or intention)

1423 (NS) 1 102 (WA) Papers retrieved for more detailed 22(GP) evaluation:

707 (NS) Reasons for exclusion: 182 (WA) NS WA GP 66 (GP) Non-randomised 649 178 60 Lack of 3 month follow up 33 0 5 Multifactorial intervention 1 0 0 unable to obtain data of hypertensive subgroup 13 3 1 other 2 0 0 ______Papers included: Total: 698 181 66

9 (NS) 1 (WA) 0 (GP) Papers coalesced into RCTs (further publications of single studies grouped):

Potentially appropriate RCTs (with or without appropriate outcome data):

8+1 (NS) 1 (WA) 0 (GP) RCTs excluded in systematic review, due to unusable outcome data: 4

RCTs with useful outcome data included in the review

4+1 (NS) Abbrevation 1 (WA) NS: Non-steroidal anti-inflammatory drugs 0 (GP) including aspirin, clopidogrel and ticlopidine WA: Warfarin GP: Glycoprotein IIb/IIIa inhibitor

Figure 1 Quorum diagram.

subgroups. These data were considered separately trials were only observing the effects over a few and duplicated data were not considered twice. days. We did not consider that this observational In three studies evaluating warfarin, detailed data period would be sufficient to make any meaningful on the hypertensive patients were not made avail- statements about the prevention of thromboembolic able to us and had to be excluded.21–23 One study events in hypertensive patients. Furthermore, the remained for further analysis.24 Five studies inves- NSAID trials assessed the effects on BP itself rather tigating ticlopidine and glycoprotein IIb/IIIa inhibi- than thromboembolic complication. tors and their impact on cardiovascular and haemorrhagic events lacked sufficient follow-up 25–29 data. A further publication was excluded due Methodological quality of included studies to the inability to acquire the relevant data.30 Most of the excluded trials were less than 1-month The only randomised placebo-controlled trial de- duration; in particular, the five GPIIb/IIIa inhibitor signed to investigate the effects of antithrombotic

Journal of Human Hypertension Antithrombotic therapy in hypertension DC Felmeden and GYH Lip 189 therapy (aspirin) on cardiovascular events and Aspirin vs placebo haemorrhagic complication in treated hypertensive patients was the large HOT study (Hypertension (a) All-cause mortality 31 All-cause mortality was reported in one trial Optimal Treatment randomised trial ). This trial 31 was a primary prevention trial in patients with (HOT). In the HOT study, mortality was 3.0% in elevated BP (DBP between 100 and 115 mmHg). the aspirin-treated patients compared with 3.2% in The aspirin component of this trial was a the placebo-control group and was not different double-blind placebo-controlled design. Major (P ¼ 0.36). The cardiovascular and noncardiovascu- cardiovascular events were defined as fatal and lar mortality, were also not different. Data relating to all cause mortality from nonfatal myocardial infarction, fatal and nonfatal 20 stroke and other cardiovascular deaths. Silent Trial was not available. myocardial infarction was defined as new Q or QS waves without clinical signs of myocardial (b) Cardiovascular events infarction. Cardiovascular events were evaluated in two stu- Among the other included studies, we tried to dies.20,31 In the HOT trial, aspirin reduced MI obtain further information from the authors. In the (ARR ¼ 0.5%, NNT ¼ 200) and major cardiovascular majority of the studies, we were unable to retrieve events (ARR ¼ 0.57%, NNT ¼ 176) as compared to the data, which was largely due to the fact of placebo. However, when all cardiovascular events the length of time since publication of the original including silent MI were considered, the reduction trials. However, details were successfully requested 24,32 failed to reach statistical significance. Similarly, in from two studies and one study having pub- the Thrombosis Prevention Trial, no significant lished a subgroup analysis with a different defini- difference was observed in the cardiovascular tion hypertension (SBP4145 mmHg19) than in the 20 events in the hypertensive subgroup between the data set provided from the original trial. Both aspirin, 8.5% and placebo, 7.5%.20 When the results sets of data have been considered separately. These of both studies were pooled for all cardiovascular trials are summarised in Table 2. For the character- events, there was no significant difference with istics of the excluded studies of the Cochrane aspirin as compared to placebo (OR 0.93; 95% CI: review, see Table 3. 0.82–1.06). CAPRIE (Clopidogrel vs Aspirin in Patients at Risk of Ischaemic Events32) was a secondary pre- vention trial in patients with recent ischaemic (c) Stroke stroke, recent myocardial infarction (MI) or sympto- Ischaemic strokes were reported in two trials.20,31 In matic peripheral arterial disease. End points were the pooled results, there was no significant differ- all-cause mortality, vascular death, myocardial ence between the aspirin and the control group (OR infarction, ischaemic or haemorrhagic stroke, all- 0.94; 95% CI: 0.76–1.17). cause rehospitalisation, hospitalisation for ischae- mic events (unstable , TIA, limb ischaemic) or for bleeding events. (d) Haemorrhage Huynh et al24 reported a secondary prevention The haemorrhagic effects of aspirin alone when trial in patients with unstable angina or non-ST compared to placebo were evaluated in the HOT elevation MI and prior CABG. Patients with uncon- study.31 There was a significant increase in major trolled systemic hypertension (4180/95 mmHg) bleeds (ARI ¼ 0.65%, NNH ¼ 154) and minor bleeds were excluded. Randomisation: warfarin (INR (ARI ¼ 0.73%, NNH ¼ 137) with aspirin as compared 2.0–2.5) þ placebo; aspirin 80 mg OD þ placebo; or to placebo. The differences in these events were warfarin þ aspirin; double-blind. Definition of mainly explained by increased gastrointestinal (72 end points: composite end point of all-cause vs 34 for major bleeds and 30 vs 18 for minor bleeds) mortality, MI or unstable angina requiring hospital and nasal bleeds (22 vs 12 for major bleeds and 66 vs admission. This study was underpowered by a low 24 for minor bleeds) in the aspirin group. However, event rate to detect a difference between treatment there was no significant difference in fatal haemor- groups. rhagic events (seven (aspirin) vs eight (placebo)). The Thrombosis Prevention Trial19,20 was a pri- mary prevention trial in patients with high risk of ischaemic heart disease. Data for subgroup analysis Aspirin vs clopidogrel of patients treated with antihypertensive at entry or during trial were reported in a subsequent The composite end point of stroke, MI or vascular publication. Definition of end points: fatal and death has been evaluated in one study in hyperten- nonfatal events (ie, coronary death and fatal and sive patients was not different in patients taking nonfatal myocardial infarction). Stroke was a sec- aspirin, 12.1% as compared to clopidogrel, 11.0%.32 ondary end point, with results for thrombotic and Data on all-cause mortality, cardiovascular mortal- haemorrhagic events to be distinguished as far as ity, all cardiovascular events, stroke, myocardial possible. infarction or haemorrhagic events were not reported.

Journal of Human Hypertension ora fHmnHypertension Human of Journal 190

Table 2 Characteristics of included studies

Study Methods Participants Interventions Outcomes Notes Allocation

CAPRIE 199632 Randomised, blinded, Previous stroke, MI or Aspirin 325 mg vs Stroke, MI, vascular A international symptomatic PVD clopidogrel 75 mg death HOT hypertension in therapy Antithrombotic (Hanson et al31 ) PROBE, placebo 19 193 pt with DBP ACE or BB felodipine Major CV events, Multi centre/national A double-blind for 105–115 mmHg age ; aspirin 75 mg +silent MI, All MI, All aspirin, 50–80 years or placebo MI+silent MI, All CFlee n Y Lip GYH and Felmeden DC DBPo90 mmHg CVA, CV mortality, tot DBPo85 mmHg mortality; fatal bleed, DBPo80 mmHg nonfatal bleed, minor bleed Huynh et al24 135 patients with prior Aspirin +placebo or All-cause mortality, INR 2–2.5 A CABG with UA or non- warfarin+placebo or MI, UA, reperfusion ST, elevation MI, 5499 aspirin +warfarin procedure men aged 45–69 years increased risk of CHD Meade et al19 Double-blind, placebo- Aspirin 75 mg vs 1. Fatal and nonfatal Subgroup analysis of A controlled placebo (+warfarin vs CHD aspirin group, placebo) 2. CVA according to SBP: o130, 130–145, 4145 mmHg on antihypertensive meds during trial Thrombosis Double-blind, placebo- 5499 men aged 45–69 Aspirin 75 mg vs 1. Fatal and non-fatal A Prevention Trial controlled years increased risk of placebo+warfarin vs CHD 199820 CHD placebo 2. CVA

CAPRIE: Clopidogrel vs aspirin in patients at risk of ischaemic events; HOT: Hypertension Optimal Treatment randomised trial; MI: myocardial infarction; PROBE: prospective, randomised, open, blinded end point, PVD: peripheral vascular disease; SBP: systolic blood pressure; DBP: diastolic blood pressure; CHD coronary heart disease; CVA: cerebrovascular accident; CV: cardiovascular; UA: unstable angina; INR: International Normalised Ratio; CABG: coronary bypass graft. Antithrombotic therapy in hypertension DC Felmeden and GYH Lip 191 Table 3 Characteristics of excluded studies of the Cochrane review

Reasons for exclusion Study references

Unable to obtain data for hypertensive group 21, 22, 23, 26, 30, 33, 37, 38, 39, 44, 45, 51, 56, 57, 64, 78, 81, 84, 88, 95 Treatment duration o3 months 27, 29, 43, 46, 48, 49, 50, 53, 54, 55, 58, 60, 61, 62, 65, 66, 68, 69, 70, 71, 72, 73, 74, 76, 79, 80, 82, 85, 89, 90, 91, 92, 93, 94, 97 Open label, not randomised, observational or retrospective study 25, 40, 41, 42, 52, 63, 67, 75, 77, 86, 87 Unsuitable end points 28, 59,

Other Calculation of the effect of ‘Clopidogrel vs placebo’ 47 Control group treated with , methotrexate and 83 cyclosporin—unsuitable for this analysis as effecting BP and other variables Patient on antihypertensives excluded, remaining patients SBP 96 119.4(3.9) mmHg and DBP 73.5(2.4) mmHg

BP, blood pressure; SBP, systolic BP; DBP, diastolic BP.

Aspirin and/or warfarin were not different between warfarin and no-warfarin groups. This study by Huynh et al24 with a low event rate was underpowered to detect a difference in the (b) Haemorrhagic strokes hypertensive subgroup of patients. Data on all-cause Data for haemorrhagic strokes were only available in mortality, cardiovascular mortality or all cardiovas- pooled form combining warfarin and aspirin com- cular events were not reported. This study evaluated pared to placebo.20 No significant difference be- the effects of aspirin and warfarin alone or in tween the two arms was found. No haemorrhagic combination on the composite end point of death strokes occurred in the placebo group (N ¼ 342), but and cardiovascular events (MI, stroke, angina, four such events were observed in the group taking cardiac arrest and revascularisation procedures). warfarin and aspirin (N ¼ 338). Owing to the low There were no significant differences in this event rate, no meaningful statistical analysis was composite end point in the hypertensive subgroup possible. (aspirin 2/14, warfarin 5/15, aspirin þ warfarin 5/14). There was no significant difference in the haemorrhagic complications (ASA, 0/14, warfarin, Safety data 0/15, aspirin þ warfarin, 1/14).24 The single haemor- rhagic event occurred in the group treated with In patients of 60 years or older with isolated systolic aspirin and warfarin. hypertension (SBP4160–219 mmHg and DBPo95 mmHg, patients never taking NSAIDs (n ¼ 2882) were compared with patients on chronic Warfarin vs no warfarin NSAID intake (n ¼ 861), defined as reporting NSAID intake on at least 50% of the patient forms.98 There In the Thrombosis Prevention Trial,20 data on all- was a tendency towards lower mortality (RR (95% cause mortality, cardiovascular mortality or all CI), 0.77 (0.56–1.06)) and higher incidence of cardiovascular events were not reported. bleeding 1.13 (0.63–2.05) with chronic NSAID intake. (a) Coronary events Coronary events (coronary death, fatal and nonfatal Discussion myocardial infarction) were evaluated in one study comparing warfarin and warfarin plus aspirin Hypertension potentially confers a prothrombotic (warfarin group) vs aspirin and placebo (no-warfarin state.6,7,10 Antithrombotic drugs such as aspirin, group) in hypertensive patients.20 Coronary events dipyridamole, clopidogrel, ticlopidine, GPIIb/IIIa were not different in the warfarin group, 7.4 vs 8.6% inhibitors and oral anticoagulants may have a in no-warfarin group. In the same study, the particular role in the treatment of hypertensive hypertensive subgroup was analysed according to patients. This antithrombotic action might prevent the occurrence of strokes (excluding haemorrhagic death and/or cardiovascular thrombotic events. strokes). Strokes were not different with warfarin However, data from previous studies suggest that (aspirin plus warfarin or warfarin alone), 2.8%, antithrombotic therapy is associated with an in- compared to no-warfarin group, 3.2%. Combined crease in haemorrhagic complications. This risk cardiovascular events (stroke and coronary events) might be amplified in hypertensive patients,

Journal of Human Hypertension Antithrombotic therapy in hypertension DC Felmeden and GYH Lip 192 potentially reversing the benefit/harm ratio for this patients is substantially lower than that seen in the therapy. Antiplatelet Trialists’ Collaboration analysis for The primary analysis in the present review is secondary prevention. Furthermore, strokes, cardio- based on four high-quality prospective randomised vascular and all-cause mortality were not signifi- double-blind controlled trials. Two of these trials cantly different in the HOT trial. In particular, evaluated the effects of aspirin as compared to aspirin did not affect the stroke rate, and as not all placebo,20,31 while one trial compared the effects of patients had CT scans, the proportion of ischaemic aspirin, warfarin and aspirin plus warfarin.24 A and haemorrhagic strokes in HOT is uncertain. further trial investigated the outcome of aspirin vs There was also a significant increase in major clopidogrel.32 bleeds, ARI ¼ 0.7%, NNH ¼ 200) and minor bleeds, Currently, there is only one large RCT with ARI ¼ 0.7%, NNH ¼ 137). The differences in these antithrombotic therapy that recruited solely hyper- events were mainly explained by increased gastro- tensive patients.31 Several of the potentially suitable intestinal (72 vs 34 for major bleeds and 30 vs 18 for trials enrolled up to 66% of patients with hyperten- minor bleeds) and nasal bleeds (22 vs 12 for major sion, but despite relentless efforts, we were unable bleeds and 66 vs 24 for minor bleeds) in the aspirin to obtain the data for the subgroup of hypertensive vs placebo group. However, there was no significant patients. One large meta-analysis of prolonged difference in fatal haemorrhagic events (seven in antiplatelet therapy assessed outcome data in 142 the aspirin vs eight in placebo group). Since the trials in ‘high-risk’ patients (some vascular disease magnitude of the harm is similar to the benefit, or other conditions implying an increased risk of aspirin cannot be recommended for primary pre- occlusive vascular disease) and three trials in ‘low- vention in patients with hypertension. risk’ patients.33 I was able to get data on the The Antiplatelet Trialists’ Collaboration meta- subgroup of hypertensive patients with DBP of analysis of the three primary prevention trials based 490 mmHg from 29 of these trials. They have on 28 000 patients also showed no significant reported these data on the composite outcome difference in all-cause mortality, nonfatal stroke or measure major vascular events (defined as nonfatal cardiovascular events. The significant decrease in myocardial infarction, nonfatal stroke or vascular nonfatal MI (ARR ¼ 0.5%, NNT ¼ 200) in antiplatelet deaths). Unfortunately, it was not possible to get group compared to placebo is similar to that seen in further information from the authors as to which the HOT trial. Haemorrhagic events were not trials were involved or the details of the data from reported in the low-risk patient population. the individual trials. Thus, these data could not Clopidogrel might be an alternative antithrombo- be included in the present analysis. The Antiplate- tic therapy to aspirin, although there has been no let Trialists’ Collaboration meta-analysis of pro- head-to-head of aspirin vs clopidogrel in hyperten- longed antiplatelet therapy compared to placebo sive patients. The subgroup analysis of hypertensive in 100 000 patients with or without hyperten- patients only demonstrated a nonsignificant trend sion reduced all-cause mortality (ARR ¼ 1.1%, towards a decreased composite end point of stroke, NNT ¼ 91) which was mostly due to vascular deaths MI or vascular death.32 This would suggest that in (ARR ¼ 1.0%, NNT ¼ 100); nonfatal myocardial in- hypertensive patients who are intolerant of aspirin farction (ARR ¼ 1.1%, NNT ¼ 91); nonfatal stroke that clopidogrel may be an effective alternative. (ARR ¼ 0.5%, NNT ¼ 200); and the composite out- However, further studies are required comparing come measure major vascular events (ARR ¼ 2.4%, clopidogrel and aspirin in hypertensive patients. NNT ¼ 42). Haemorrhagic events were not reported. Rigorous and aggressive BP control is advisable The benefit of antiplatelet therapy in secondary when using antithrombotic therapy in hypertension, prevention in patients with or without hypertension as suggested by the Thrombosis Prevention Trial.20 has been established by the Antiplatelet Trialists’ A subgroup analysis from this trial showed that Collaboration meta-analysis. treatment with aspirin compared to placebo reduces In the hypertensive subgroup of 29 trials (10 600 major cardiovascular events particularly strokes in patients), antiplatelet therapy significantly reduced people with normal (SBP 130–145 mmHg) or even major vascular events (ARR ¼ 4.1%, NNT ¼ 25). All- low BP (SBPo130 mmHg), whereas this risk re- cause mortality, cardiovascular mortality, nonfatal duction was not seen in patients with SBP MI, nonfatal stroke and haemorrhagic events were 4145 mmHg.19 This suggests that the benefit of not reported in the hypertensive subgroup. These aspirin therapy may occur primarily in patients with data suggest that the absolute benefit for antiplatelet well-controlled BP. therapy for secondary prevention is greater in There are only limited data available on warfarin patients with hypertension than in patients without as an antithrombotic agent in hypertension. Owing hypertension. to small numbers, there was no significant differ- The HOT trial demonstrated a significant reduc- ence in hypertensives when comparing warfarin tion by aspirin as compared to placebo in major with placebo or aspirin or the combination of cardiovascular events (ARR ¼ 0.6%, NNT ¼ 176) and aspirin and warfarin with regard to cardiovascular myocardial infarction (ARR ¼ 0.5%, NNT ¼ 200).31 events or haemorrhagic complication. However, This benefit for primary prevention in hypertensive there is a nonsignificant trend suggesting that

Journal of Human Hypertension Antithrombotic therapy in hypertension DC Felmeden and GYH Lip 193 warfarin is inferior to aspirin in preventing Duality of interest thromboembolic events in this group. Warfarin seems to be also associated with a higher rate of Our unit undertakes clinical trials of antithrombotic haemorrhagic complications. Based on currently therapy in cardiovascular disease and stroke, and available data, warfarin therapy in hypertension has received research funding from various pharma- cannot be recommended. There were no data ceutical companies involved in thrombosis and available from subgroup analyses of hyperten- antithrombotic therapy. sive patients treated with glycoprotein IIb/IIIa inhibitors. However, oral glycoprotein IIb/IIIa in- hibitors have failed to demonstrate an advantage References over aspirin after percutaneous coronary revascular- isation.34 These patients are at high risk of throm- 1 Collins R, MacMahon S. Blood pressure, antihyperten- boembolic events. sive treatment and the risks of stroke and of coronary heart disease. Br Med Bull 1994; 50: 272–298. 2 Collins R et al. Blood pressure, stroke, and coronary heart disease. Part 2, Short-term reductions in blood Conclusion pressure: overview of randomised drug trials in their epidemiological context. Lancet 1990; 335: 827–838. For primary prevention in hypertensive patients, 3 Collins R et al. Blood pressure, stroke, and coronary antiplatelet therapy with aspirin cannot be recom- heart disease. Part 1, Prolonged differences in blood mended since the magnitude of benefit of a reduc- pressure: prospective observational studies corrected tion in myocardial infarction is similar to the for the regression dilution bias. Lancet 1990; 335: magnitude of harm of an increased risk of major 765–774. haemorrhagic events. For secondary prevention in 4 Prospective Studies Collaboration. Age-specific rele- hypertensive patients, antiplatelet therapy is recom- vance of usual blood pressure to vascular mortality: mended, as the magnitude of the benefit is greater a meta-analysis of individual data for one million than that in a nonhypertensive population. adults in 61 prospective studies. Lancet 2002; 360: 1903–1913. Antithrombotic therapy with warfarin alone or in 5 Lip GYH, Edmunds E, Beevers DG. Should patients combination with aspirin in patients with hyperten- with hypertension receive antithrombotic therapy? sion cannot be recommended because of lack of J Intern Med 2001; 249: 205–214. effectiveness in reducing cardiovascular events as 6 Lip GYH, Beevers DG. Abnormalities of rheology and well as a trend towards increased haemorrhagic coagulation in hypertension. J Hum Hypertens 1994; 8: events. Glycoprotein IIb/IIIa inhibitors as well as 693–702. ticlopidine and clopidogrel have not been suffi- 7 Lee AJ. The role of rheological and haemostatic factors ciently evaluated in order to recommend their use in in hypertension. J Hum Hypertens 1997; 11: 767–776. hypertensive patients. 8 Vanhoutte PM. Endothelial dysfunction in hyperten- As we have excluded patients with atrial fibrilla- sion. J Hypertens Suppl 1996; 14: S83–S93. 9 Lip GY, Blann AD. Does hypertension confer a tion from this analyses, the recommendations do not prothrombotic state? Virchow’s triad revisited. Circula- apply to patients in atrial fibrillation with or tion 2000; 101: 218–220. without hypertension. This review adds to the 10 Lip GYH et al. Relationship of endothelium, thrombo- Antiplatelet Trialists’ Collaboration meta-analysis genesis, and haemorrheology in systemic hypertension as we include the specific designed prospective to ethnicity and left ventricular hypertrophy. Am J placebo-controlled HOT trial with information on Cardiol 1997; 80: 1566–1571. bleeding complications and correction of the esti- 11 Segal JB et al. Anticoagulants or antiplatelet therapy mated benefit on cardiovascular events of aspirin for non-rheumatic atrial fibrillation and flutter (Co- use in hypertension. chrane Review). In: Cochrane Library, Issue 2. Update This review is also published in the Cochrane Software: Oxford, 2001 CD001938. 12 SPAF II Study. Warfarin versus aspirin for prevention Database of Systematic Reviews, as follows: Lip of thromboembolism in atrial fibrillation: Stroke GYH, Felmeden DC. Antiplatelet agents and anti- Prevention in Atrial Fibrillation II Study. Lancet coagulants for hypertension. Cochrane Database 1994; 343: 687–691. Syst Rev. 2004;(3):CD003186. 13 Lip GYH, Gibbs CR. Antiplatelet agents versus control or anticoagulation for heart failure in sinus rhythm (Cochrane Review). In: The Cochrane Library, Issue 4. John Wiley & Sons, Ltd: Chichester, UK, 2001 Acknowledgements CD003333. We acknowledge the support of the Sandwell & West 14 WHO–ISHGMH. World Health Organization–Interna- Birmingham NHS Trust Research and tional Society of Hypertension Guidelines for the management of hypertension. Guidelines Subcommit- Development Program for the Haemostasis, Throm- tee. J Hypertens 1999; 17: 151–183. bosis and Vascular Biology Unit. We thank our 15 Ramsay L et al. Guidelines for management of colleagues in the Cochrane Hypertension Group for hypertension: report of the third working party of the helpful comments and advice, as well as assisting British Hypertension Society. J Hum Hypertens 1999; with the detailed Cochrane peer review process. 13: 569–592.

Journal of Human Hypertension Antithrombotic therapy in hypertension DC Felmeden and GYH Lip 194 16 WHO Guidelines Sub-Committee. Guidelines for the Hypertension Optimal Treatment randomised trial. the management of mild hypertension: memorandum Lancet 1998; 351: 1755–1762. from a World Health Organization/International So- 32 CAPRIE Steering Committee. A randomised, blinded, ciety of Hypertension meeting. J Hypertens 1993; 11: trial of clopidogrel versus aspirin in patients at risk 905–918. of ischaemic events (CAPRIE). Lancet 1996; 348: 17 Clarke M, Oxman AD (eds). Cochrane Reviewers’ 1329–1339. Handbook 4.2.0 [updated March 2003]. In: The Co- 33 Antiplatelet Trialists’ Collaboration. Collaborative chrane Library, Issue 2, 2003. Oxford: Update Software. overview of randomised trials of antiplatelet ther- 18 Mulrow CD, Cook DJ, Davidoff F. Systematic reviews: apy—I: Prevention of death, myocardial infarction, critical links in the great chain of evidence. Ann Intern and stroke by prolonged antiplatelet therapy in various Med 1997; 126: 389–391. categories of patients. Antiplatelet Trialists’ Collabora- 19 Meade TW, Brennan PJ. Determination of who may tion. BMJ 1994; 308: 81–106. derive most benefit from aspirin in primary preven- 34 O’Neill WW et al. Long-term treatment with a platelet tion: subgroup results from a randomised controlled glycoprotein- antagonist after percutaneous trial. BMJ 2000; 321: 13–17. coronary revascularization. EXCITE Trial Investigators. 20 The Medical Research Council’s General Practice Evaluation of oral xemilofiban in controlling thrombo- Research Framework. Thrombosis prevention trial: tic events. N Engl J Med 2000; 342: 1316–1324. randomised trial of low-intensity oral anticoagulation 35 Dickersin K, Herxheimer A. Introduction: the quality with warfarin and low-dose aspirin in the primary of the medical evidence: is it good enough? Int prevention of ischaemic heart disease in men at J Technol Assess Health Care 1996; 12: 187–189. increased risk. Lancet 1998; 351: 233–241. 36 Lefebvre C, McDonald S. Development of a sensitive 21 O’Connor CM et al. Comparison of two aspirin doses search strategy for reports of randomized controlled on ischemic stroke in post-myocardial infarction trials in EMBASE. Paper presented at the Fourth patients in the warfarin (Coumadin) Aspirin Reinfarc- International Cochrane Colloquium, Adelaide, Austra- tion Study (CARS). Am J Cardiol 2001; 88: 541–546. lia, 20–24 October 1996. 22 Mohr JP et al. A comparison of warfarin and aspirin for 37 Ajani UA et al. Alcohol consumption and risk of the prevention of recurrent ischemic stroke. N Engl J coronary heart disease by status. Circulation Med 2001; 345: 1444–1451. 2000; 102: 500–505. 23 The Stroke Prevention in Reversible Ischemia 38 Bhatt DL et al. Reduction in the need for hospitaliza- Trial (SPIRIT) Study Group. A randomized trial of tion for recurrent ischemic events and bleeding with anticoagulants versus aspirin after cerebral ischemia clopidogrel instead of aspirin. CAPRIE investigators. of presumed arterial origin. Ann Neurol 1997; 42: Am Heart J 2000; 140: 67–73. 857–865. 39 Bhatt DL et al. Superiority of clopidogrel versus 24 Huynh T et al. Aspirin, warfarin or the combination aspirin in patients with prior cardiac . Circula- for secondary prevention of coronary events in tion 2001; 103: 363–368. patients with acute coronary syndromes and prior 40 Califf RM et al. Prediction of 1-year survival after coronary artery bypass surgery. Circulation 2001; 103: for acute myocardial infarction in the 3069–3074. global utilization of and TPA for oc- 25 Schuhlen H et al. Incidence of thrombotic occlusion cluded coronary trial. Circulation 2000; 101: and major adverse cardiac events between two and 2231–2238. four weeks after coronary stent placement: analysis of 41 Caravaca F et al. Haemodynamic changes induced by 5,678 patients with a four-week ticlopidine regimen. J the correction of anaemia by erythropoietin: role of Am Coll Cardiol 2001; 37: 2066–2073. antiplatelet therapy. Nephrol Dial Transplant 1995; 10: 26 Thel MC et al. Clinical risk factors for ischemic 1720–1724. complications after percutaneous coronary interven- 42 Cook ME et al. Comparative effects of , tions: results from the EPIC trial. The EPIC Investiga- , and on renal function. J Rheuma- tors. Am Heart J 1999; 137: 264–273. tol 1997; 24: 1137–1144. 27 Kleiman NS et al. Diabetes mellitus, glycoprotein IIb/ 43 Cusson JR et al. Effect of on blood pressure, IIIa blockade, and : evidence for a complex endocrine and renal responses to chronic dosing with interaction in a multicenter trial. EPILOG Investiga- captopril in patients with essential hypertension. tors. Circulation 1998; 97: 1912–1920. Blood Press 1992; 1: 162–167. 28 Novo S et al. Effects of captopril and ticlopidine, 44 Drapkin A, Merskey C. Anticoagulant therapy after alone or in combination, in hypertensive patients acute myocardial infarction. Relation of therapeutic with intermittent claudication. Int Angiol 1996; 15: benefit to patient’s age, sex, and severity of infarction. 169–174. JAMA 1972; 222: 541–548. 29 Finelli C et al. Ticlopidine lowers plasma fibrinogen in 45 Forbes CD, for the ESPS Investigators. Secondary patients with polycythaemia rubra vera and additional stroke prevention with low-dose aspirin, sustained thrombotic risk factors. A double-blind controlled release dipyridamol alone and in combination. study. Acta Haematol 1991; 85: 113–118. Thromb Res 1998; 92: S1–S6. 30 Hass WK, et al., for the Ticlopidine Aspirin Stroke 46 Ferri C et al. Enhanced blood pressure response to Study Group. A randomised trial comparing ticlopi- cyclooxygenase inhibition in salt-sensitive human dine hydrochloride with aspirin for the prevention of essential hypertension. Hypertension 1993; 21(6 Part stroke in high-risk patients. N Engl J Med 1989; 321: 1): 875–881. 501–507. 47 Fisher LD, Gent M, Buller HR. Active-control trials: 31 Hansson L, et al., for the HOT Study Group. Effects of how would a new agent compare with placebo? A intensive blood-pressure lowering and low-dose aspir- method illustrated with clopidogrel, aspirin, and in in patients with hypertension: principle results of placebo. Am Heart J 2001; 141: 26–32.

Journal of Human Hypertension Antithrombotic therapy in hypertension DC Felmeden and GYH Lip 195 48 Forbes CD et al. Clopidogrel compatibility with 66 Magagna A et al. Hemodynamic and humoral effects of concomitant cardiac co-: a study of its low-dose aspirin in treated and untreated essential interactions with a beta-blocker and a calcium uptake hypertensive patients. Blood Press 1994; 3: 236–241. antagonist. Semin Thromb Hemost 1999; 25: 55–60. 67 Maggioni AP et al. The risk of stroke in patients with 49 Furey SA, Vargas R, McMahon FG. Renovascular acute myocardial infarction after thrombolytic and effects of nonprescription ibuprofen in elderly hyper- antithrombotic treatment. Gruppo Italiano per lo tensive patients with mild renal impairment. Pharma- Studio della Sopravvivenza nell’Infarto Miocardico II cotherapy 1993; 13: 143–148. (GISSI-2), and The International Study Group. N Engl J 50 Gurwitz JH et al. The impact of ibuprofen on the Med 1992; 327: 1–6. efficacy of antihypertensive treatment with hydro- 68 Mehta SR et al. Effects of pretreatment with clopido- chlorothiazide in elderly persons. J Gerontol A Biol grel and aspirin followed by long-term therapy in Sci Med Sci 1996; 51: M74–M79. patients undergoing percutaneous coronary inter- 51 Harker LA, Boissel JP, Pilgrim AJ, Gent M. Comparative vention: the PCI-CURE study. Lancet 2001; 358: safety and tolerability of clopidogrel and aspirin: 527–533. results from CAPRIE. CAPRIE Steering Committee 69 Mills EH, Whitworth JA, Andrews J, Kincaid-Smith P. and Investigators. Clopidogrel versus aspirin in pa- Non-steroidal anti-inflammatory drugs and blood tients at risk of ischaemic events. Drug Saf 1999; 21: pressure. Aust N Z J Med 1982; 12: 478–482. 325–335. 70 Minuz P, Barrow SE, Cockcroft JR, Ritter JM. Effects of 52 Harpaz D et al. Effect of aspirin on the mortality in non-steroidal anti-inflammatory drugs on women with symptomatic or silent myocardial infarc- and biosynthesis in patients with mild tion. Am J Cardiol 1996; 78: 1215–1219. essential hypertension. Br J Clin Pharmacol 1990; 30: 53 Hartmann D et al. Study on the possible interaction 519–526. between and atenolol in hypertensive 71 Minuz P et al. Amlodipine and haemodynamic effects patients. Arzneimittelforschung 1995; 45: 494–498. of cyclo-oxygenase inhibition. Br J Clin Pharmacol 54 Harvey PJ et al. Effect of indomethacin on blood 1995; 39: 45–50. pressure control during treatment with nitrendipine. 72 Morgan T, Anderson A. Interaction of indomethacin Blood Press 1995; 4: 307–312. with felodipine and enalapril. J Hypertens Suppl 1993; 55 Hermida RC et al. Influence of aspirin usage on blood 11(Suppl 5): S338–S339. pressure: dose and administration-time dependencies. 73 Morgan TO, Anderson A, Bertram D. Effect of Chronobiol Int 1997; 14: 619–637. indomethacin on blood pressure in elderly people 56 Herskovits E et al. Preventive treatment of cerebral with essential hypertension well controlled on transient ischemia: comparative randomized trial of amlodipine or enalapril. Am J Hypertens 2000; 13: pentoxifylline versus conventional antiaggregants. Eur 1161–1167. Neurol 1985; 24: 73–81. 74 Nawarskas JJ, Townsend RR, Cirigliano MD, Spinler 57 Hess H, Mietaschk A, Deichsel G. Drug-induced SA. Effect of aspirin on blood pressure in hypertensive inhibition of platelet function delays progression of patients taking enalapril or losartan. Am J Hypertens peripheral occlusive arterial disease. A prospective 1999; 12(8 Part 1): 784–789. double-blind arteriographically controlled trial. Lancet 75 Oosterga M et al. Effects of aspirin on angiotensin- 1985; 1: 415–419. converting inhibition and left ventricular 58 Houston MC et al. The effects of nonsteroidal anti- dilation one year after acute myocardial infarction. inflammatory drugs on blood pressures of patients Am J Cardiol 1998; 81: 1178–1181. with hypertension controlled by verapamil. Arch 76 Pancera P et al. Changes in peripheral hemodynamics Intern Med 1995; 155: 1049–1054. and vasodilating prostaglandins after high-dose short- 59 Houtsmuller AJ et al. The influence of teiclopidine on term ibuprofen in chronically treated hypertensive the natural course of retinal vein occlusion. Agents patients. Prostaglandins Leukot Essent Fatty Acids Actions Suppl 1984; 15: 219–229. 1996; 54: 217–222. 60 Johnson AG et al. Potential mechanisms by which 77 Peto R et al. Randomised trial of prophylactic nonsteroidal anti-inflammatory drugs elevate blood daily aspirin in British male doctors. BMJ 1988; 296: pressure: the role of endothelin-1. J Hum Hypertens 313–316. 1996; 10: 257–261. 78 Steering Committee of the Physicians’ Health Study 61 Klassen D et al. Assessment of blood pressure during Research Group. Final report on the aspirin compo- treatment with or ibuprofen in hypertensive nent of the ongoing Physicians’ Health Study. N Engl patients treated with hydrochlorothiazide. J Clin J Med 1989; 321: 129–135. Pharmacol 1993; 33: 971–978. 79 Polonia J et al. Influence of non-steroidal anti- 62 Klassen DK, Jane LH, Young DY, Peterson CA. inflammatory drugs on renal function and 24 h Assessment of blood pressure during naproxen ther- ambulatory blood pressure-reducing effects of apy in hypertensive patients treated with nicardipine. enalapril and nifedipine gastrointestinal therapeutic Am J Hypertens 1995; 8: 146–153. system in hypertensive patients. J Hypertens 1995; 13: 63 Lee IM et al. and risk of stroke in male 925–931. physicians. Stroke 1999; 30: 1–6. 80 Polonia J, Gama G, Santos A. [Reduction of the 64 Lemak NA, Fields WS, Gary Jr HE. Controlled trial of antihypertensive effects of enalapril by indomethacin. aspirin in cerebral ischemia: an addendum. Neurology Its independence from renal sodium retention]. Rev 1986; 36: 705–710. Port Cardiol 1996; 15: 460, 485–492. 65 Magagna A et al. Indomethacin raises blood pressure 81 Collaborative Group of the Primary Prevention Project. in untreated essential hypertensives: a double-blind Low-dose aspirin and vitamin E in people at cardio- randomly allocated study versus placebo. J Hypertens vascular risk: a randomised trial in general practice. Suppl 1991; 9: S242–S243. Lancet 2001; 357: 89–95.

Journal of Human Hypertension Antithrombotic therapy in hypertension DC Felmeden and GYH Lip 196 82 Pritchard G et al. Do trandolapril and indomethacin 90 Sturrock ND, Lang CC, Struthers AD. Indomethacin influence renal function and renal functional reserve and cyclosporin together produce marked renal in hypertensive patients? Br J Clin Pharmacol 1997; 44: in humans. J Hypertens 1994; 12: 145–149. 919–924. 83 Proudman SM et al. Treatment of poor-prognosis early 91 Takeuchi K et al. No adverse effect of non-steroidal rheumatoid arthritis. A randomized study of treatment anti-inflammatory drugs, sulindac and with methotrexate, cyclosporin A, and intra-articular sodium, on blood pressure control with a calcium corticosteroids compared with sulfasalazine alone. antagonist, nifedipine, in elderly hypertensive pa- Arthritis Rheum 2000; 43: 1809–1819. tients. Tohoku J Exp Med 1991; 165: 201–208. 84 Puranen J, Laakso M, Riekkinen Sr PJ, Sivenius J. 92 Thakur V, Cook ME, Wallin JD. Antihypertensive effect Efficacy of antiplatelet treatment in hypertensive of the combination of fosinopril and HCTZ is resistant patients with TIA or stroke. J Cardiovasc Pharmacol to interference by nonsteroidal anti-inflammatory 1998; 32: 291–294. drugs. Am J Hypertens 1999; 12(9 Part 1): 925–928. 85 Mahaffey KW et al. Stroke in patients with acute 93 Tison P et al. Treatment of hypertension with dihy- coronary syndromes: incidence and outcomes in the dropyridine calcium antagonists and aspirin. Blood platelet glycoprotein IIb/IIIa in unstable angina. Re- Press Suppl 1994; 1: 57–60. ceptor suppression using integrilin therapy (PURSUIT) 94 Tsuji T, Yasu T, Katayama Y, Imanishi M. Different trial. The PURSUIT Investigators. Circulation 1999; 99: patterns of renal prostaglandins I2 and E2 in patients 2371–2377. with essential hypertension with low to normal or high 86 Saloheimo P, Juvela S, Hillbom M. Use of aspirin, renin activity. J Hypertens 2000; 18: 1091–1096. epistaxis and untreated hypertension as risk factors for 95 UK-TIA Study Group. The United Kingdom transient primary intracerebral hemorrhage in middle-aged and ischaemic attack (UK-TIA) aspirin trial: final results. J elderly people. Stroke 2001; 32: 399–404. Neurol Neurosurg Psychiatry 1991; 54: 1044–1054. 87 Schreiber TL et al. Randomized trial of thrombolysis 96 Walter E, Kaufmann W, Oster P. Does chronic aspirin versus heparin in unstable angina. Circulation 1992; treatment increase blood pressure in man? Klin 86: 1407–1414. Wochenschr 1981; 59: 297–299. 88 Sivenius J et al. Antiplatelet therapy is effective in 97 Whelton A et al. Cyclooxygenase-2—specific inhibitors primary prevention of myocardial infarction in pa- and cardiorenal function: a randomized, controlled tients with a previous cerebrovascular ischemic event. trial of and in older hypertensive Arch Neurol 1993; 50: 710–713. osteoarthritis patients. Am J Ther 2001; 8: 85–95. 89 Smith SR, Coffman TM, Svetkey LP. Effect of low-dose 98 Celis H et al. Interaction between nonsteroidal anti- aspirin on thromboxane production and the antihy- inflammatory drug intake and calcium-channel block- pertensive effect of captopril. J Am Soc Nephrol 1993; er-based antihypertensive treatment in the Syst-Eur 4: 1133–1139. trial. J Hum Hypertens 2001; 15: 613–618.

Journal of Human Hypertension