Aspirin in the Primary Prevention of Angina Pectoris in a Randomized Trial of United States Physicians JOANNE. MANSON,M.D., Boston, Massachusetts, DIEDERICKE. GROBBEE,M.D., Rotterdam, The Netherlands, MEIRJ. STAMPFER,M.D., JAMESO. TAYLOR,M.D., SAMUELZ. GOLDHABER,M.D., J. MICHAELGAZlANO,M.D., PAULM. RIDKER,M.D., JULIEE. BURING,D.Sc.,CHARLESH. HENNEKENS,M.D., Boston, Massachusetts
PURPOSE: The objective of this study was to exAmine the effect of low-dose aspirin (325 mg on alternate days) on the primary prevention of angina pectoris in the United States Physician~' Health Study. Despite a postulated role of platelets in atherogenesis and myocardial ischemi% the effect of chronic platelet inhibition on the development of clinical angina pectoris is nnlmown. SUBJECTS AND METHODS:The Physic~ns' Health Study is a randomiT~d, double-blind, placebo-controlled trial among 22,071 male physiciRn~ aged 40 to 84 years, free from previous myocardial infarction, stroke, and transient cerebral ischemia at entry, and followed for an average of 60,2 months. The 21,738 physicla~s who were also free from angina pectoris at baseline constituted the study population for the present nn~lyl~S. RESULTS: During 106,652 person-years of follow-up, 331 patients with confirmed angina poctoris were diagnosed, 194 of whom underwent a coronary revascnl~rization procedure (coronary artery bypass graft surgery or coronary angioplasty). As compared to participants assigned placebo, the relative risk of confirmed angina pectoris in the aspirin group was 1.!0 (95% confidence interval [CI], 0.88 to 1.38). For coronary revascularization, the relative risk was 1.19 (95% CI, 0.88 to 1.59). After simultaneous control for other coronary risk factors in a proportionalh~7~rds model, these relative risks remained near unity at 1.07 (95% CI, 0~4 to 1.36) and 1.11 (95% CL 0.81 to 1.52), respectively. When the
risks of angina pectoris were exRmined according to year of randomiT~tion in the trial, there was no pattern of increasing benefit with longer duration of treatment. CONCLUSION:These randomiT~d trial data indicate that chronic platelet inhibition with lowdose aspirin for an average duration of 60~2 months does not reduce the incidence of angina pectoris.
ngina pectoris generally reflects advanced ath-
A erosclerotic narrowing of one or more coronary arteries [1]. The presence of platelets in atherosclerotic plaques [2], the observation that platelet adhesion to damaged endothelium may promote plaque formation [2-4], and the demonstration that aspirin may inhibit atheroma formation in monkeys and rabbits fed atherogenic diets [5,6] suggest that platelet activity is associated with atherogenesis. Furthermore, platelet aggregation at the site of atherosclerotic narrowing may precipitate acute anginal episodes in patients with both fixed and variable lesions in a coronary artery [7]. Although low-dose aspirin is effective in primary [8,9] and secondary [10-13] prevention of myocardial infarction, presumably by inhibition of thrombosis [14,15], there are no data concerning this agent and the development of angina pectoris. We examined the effect of low-dose aspirin on the primary prevention of angina pectoris in the United States Physicians' Health Study, a randomized trial of 22,071 men aged 40 to 84 years followed for an average period of 60.2 months.
SUBJECTSAND METHODS From the Channing Laboratory, Department of Medicine, Harvard Medical School and Brigham and Women's Hospital (JEM, MJS, JOT, SZG, JMG, PMR, CHH), and the Department of Preventive Medicine, Harvard Medical School (JEB, CHH), Boston, Massachusetts; and the Department of Epidemiologyand Biostatistics, ErasmusUniversity Medical School (DEG), Rotterdam, The Netherlands. This work was supported by Grants HL-26490, HL-34595, CA-34944, and CA 40360 from the National Institutes of Health. Requestsfor reprints should be addressed to JoAnn E. Manson, M.D., 55 Pond Avenue, Brookline, Massachusetts02146. Manuscript submitted June 5, ]990, and accepted in revised form August 28, 1990.
772
December 1990
The Physicians' Health Study [8,9] is a randomized, double-blind, placebo-controlled trial designed to test the effect of low-dose aspirin on risks of cardiovascular disease and the effect of beta carotene on the incidence of cancer. A detailed description of the subjects and methods has been previously p u b l i s h e d [8]. Briefly, 22,071 U.S. male physicians aged 40 to 84 years at entry and free from prior myocardial infarction, stroke, and tran-
The American Journal of Medicine Volume 89
ASPIRINAND ANGINAPECTORIS/ MANSONET AL
sient cerebral ischemia were assigned at random, using a two-by-two factorial design, to one of four treatment groups: aspirin and beta carotene, aspirin and beta carotene placebo, aspirin placebo and beta carotene, or aspirin placebo and beta carotene placebo. A total of 11,037 physicians were randomized to aspirin, 325 mg on alternate days (Bufferin®, Bristol-Myers, New York, New York), and 11,034 to aspirin placebo. Every 6 months for the first year and annually thereafter, the participants were mailed a supply of monthly calendar packs containing study medication and brief questionnaires asking about their compliance with the assigned treatment and the occurrence of any relevant events. The reported consumption of aspirin or other platelet-active drugs was 85.7% in the aspirin group and 14.2% in the placebo group. The questionnaires also included items about cigarette smoking, systolic and diastolic blood pressure, serum cholesterol, height and weight, diagnosis of diabetes mellitus, parental history of myocardial infarction, alcohol use, and frequency of vigorous exercise. The aspirin component was terminated early (January 25, 1988) due to a statistically extreme 44% reduction in the risk of a first myocardial infarction among those in the aspirin group [8,9]. By this date, participants had been followed for an average of 60.2 months (range, 45.8 to 77.0 months). Vital status was known for all physicians, and follow-up data on morbidity were 99.7% complete.
Endpoints The primary endpoint in the present analyses is the incidence of confirmed angina pectoris. Of the 22,071 randomized physicians, 333 reported angina and/or coronary revascularization at baseline and were excluded. The present analyses include the 21,738 physicians free from myocardial infarction, angina pectoris, coronary revascularization, stroke, and transient cerebral ischemia at baseline. When participants reported a first diagnosis of angina pectoris after randomization, we sent supplementary questionnaires to elicit details about diagnostic tests and treatment. A report of angina pectoris was classified as confirmed by a committee of physicians if, on the supplementary questionnaire, a positive diagnostic test (a confirmatory exercise tolerance test, exercise thallium, exercise radionuclide ventriculogram, or narrowing of 50% or greater of a coronary artery on angiogram) was reported, or if the participant reported a coronary revascularization procedure (coronary artery bypass graft surgery or percutaneous transluminal coronary angioplasty). The physicians reviewing the confirmatory evidence for angina were unaware of the aspirin treatment assignment of the participants. Analyses
are presented separately for total confirmed angina pectoris (n = 331) and for the subgroup of coronary revascularization (n = 194). The available information did not permit a distinction between stable and unstable categories of angina pectoris. Randomized physicians reporting a first myocardial infarction prior to diagnosis of angina or coronary revascularization procedure were, however, excluded.
Endpoint Validation A validation study of angina pectoris confirmed by supplementary questionnaire was conducted among 100 randomly selected participants who reported this diagnosis. Consent for release of medical records regarding pertinent diagnostic tests was requested. Consent was obtained from 99 of the 100 subjects, among whom medical records were available for review in 90. Medical record review by a physician masked to the aspirin treatment assignment corroborated at least one positive diagnostic test result for angina pectoris in 88 of 90 participants {97.8%). Records were not requested to confirm self-reports of coronary revascularization procedures.
Statistical Analyses Incidence rates were calculated by dividing the numbers of events by person-time of follow-up for the aspirin and placebo groups, after adjustment for age (5-year categories) and beta carotene treatment assignment. Relative risks were calculated as the ratio of the incidence rate in the aspirin group to that in the placebo group. Multivariate relative risks were derived from proportional-hazards models, which controlled for multiple coronary risk factors simultaneously. For each relative risk, we calculated the 95% confidence interval (CI) [16]. To assess possible modifying effects of other coronary risk factors, analyses were performed within strata of these variables. To evaluate the effect of duration of aspirin therapy on the risk of confirmed angina, relative risks were examined separately according to year of randomization in the trial.
RESULTS The baseline characteristics of the analyzed participants were virtually identical in the aspirin and placebo groups (Table I). During an average of 60.2 months of follow-up (106,652 person-years), there were 331 subjects with confirmed angina pectoris, 194 of whom underwent coronary artery bypass graft surgery or percutaneous transluminal coronary angioplasty. When compared to those assigned placebo, the relative risk of confirmed angina pectoris among participants randomized to receive aspirin was 1.10 (95%
December 1990 The American Journal of Medicine Volume 89
773
ASPIRIN AND ANGINA PECTORIS / MANSON ET AL
TABLE I BaselineCharacteristicsof ParticipantsAccordingto Treatment Assignment,as Reportedon the Initial Questionnaire*
Mean age (years) (SEM) Mean body massindex (kg/m2) (SEM) Mean systolic BP (mm Hg) ( S E M ) Mean diastolic BP (mm Hg) (SEM) Hypertension (%)t Mean serum cholesterol (SEM):~ Diabetes mellitus (%) Parental history of MI (%)§ Cigarette smoking Current (%) Past (%) Never (%) Alcohol use Daily (%) Weekly (%) Rarely (%)
Placebo Group
Aspirin Group
(n = 10,879)
(n = 10,859)
53.1 (0.09) 24.9(0.03) 126.0(0.12) 78.8 (0.08) 13.5 211.8 (0.73) 2.3 13.0
53.1 (0.09) 24.9(0.03) 126.0(0.12) 78.9 (0.08) 13.6 211.7 (0.71) 2.3 13.0
11.1 38.9 50.0
11.1 39.5 49.4
24.9 48.6 26.5
24.8 49.7 25.5
* Excludesphysicianswith anginapectorisor coronaryrevascularizationat baseline.Risk
factors are self-reported. SEM denotesstandarderror of the mean. BP denotesblood
pressure. * Definedas self-reportedsystolicBP > 160 mm Hgand/or diastolicBP >- 95 mm Hg, or takingantihypertensivemedication. * Units are mg/dL. To calculatemmol/L, divideby 38.7. Reportedby onlyone third of participants. § MI denotesmyocardialinfarction, occurringin either parentbeforeage60.
CI, 0.88 to 1.38) (Table II). For coronary revascularization procedures, the relative risk was 1.19 (95% CI, 0.88 to 1.59). After simultaneous control for other coronary risk factors in a proportionalhazards model, these relative risks were not materially altered, being 1.07 (95% CI, 0.84 to 1.36) and 1.11 (95% CI, 0.81 to 1.52), respectively. There was no evidence for any modifying effect of age, cigarette smoking, blood pressure, diabetes mellitus, or other coronary risk factors on the influence of aspirin. To assess the possibility that a benefit of aspirin on the risk of angina pectoris is delayed due to the need for prolonged platelet inhibition, the relative risk of confirmed angina was examined according to year of randomization in the trial (Figure 1). Relative risks were not significantly different from unity at any time point in the trial, and there was no pattern of an increasing benefit from aspirin with longer duration of treatment. Specifically, in the fifth and final year of randomized treatment, the relative risk in the aspirin group was 1.18 (95% CI, 0.71 to 1.98).
increasing benefit from aspirin with prolonged treatment was evident. Low-dose aspirin irreversibly inhibits platelet cyclooxygenase activity, preventing thromboxane A2 production and inhibiting platelet aggregation [17,18]. The effect of aspirin in preventing myocardial infarction likely results from prevention of thrombotic occlusion of coronary arteries that are severely affected by atherosclerotic disease [14,15]. Platelets have also been implicated, however, in the development of atherosclerotic plaque [2-4], suggesting a possible role in atherogenesis. The hypothesis that platelet-arterial wall interactions may contribute to the development of atherosclerosis is supported by the finding that a platelet-derived factor appears to stimulate proliferation of arterial smooth muscle cells [19,20]. Several animal studies have suggested that aspirin may have an inhibitory effect on atheroma formation; aspirin has been demonstrated to inhibit the progression of coronary atherosclerosis in monkeys fed an atherogenic diet [5] and, in combination with dipyridamole, to inhibit plaque formation in hypercholesterolic rabbits [6]. Furthermore, animal studies have suggested that transient decreases in coronary blood flow may be related to acute platelet aggregation that can be prevented by aspirin therapy [21,22]. Despite plausible mechanisms for chronic platelet inhibition in the primary prevention of angina pectoris, the present data indicate that the risk of angina and the rate of coronary revascularization were not influenced by low-dose aspirin. These observations
TABLEII Relative Risk (RR) of Confirmed AnginaPectorisand Coronary RevascularizationProceduresAccordingTo AspirinTreatment Assignment
Number of person-years Confirmed angina pectoris Cases
RR* (95% CI)t Multivariate RR* (95% CI) Coronary revascularization§ Cases
COMMENTS
These data from a randomized trial of U.S. physicians demonstrate no apparent benefit of low-dose aspirin on the incidence of confirmed angina pectoris or the utilization of coronary revascularization procedures during an average treatment period of 60.2 months. In analyses performed separately by year of randomization in the trial, no pattern of an 774
PlaceboGroup (n = 10,879)
Aspirin Group (n = 10,859)
53,298
53,354
158
173
1.0
1.10
Referent 1.0 Referent
(0.88, 1.38) 1.07 (0.84, 1.36)
89
105
RR* (95% CI)
1.0 Referent
1.19 (0.88, 1.59)
Multivariate RR~ (95% Cl)
1.0 Referent
1.11 (0.81, 1.52)
* Adjustedby age(5-yearcategories)and betacaroteneassignment. t CI denotesconfidenceinterval. t MultivariateRR includessimultaneouscontrol for age (5-year categories), cigarette smoking(never, past, current), bodymassindex,systolicbloodpressure(BP), diastolic BP, history of hypertension (yes, no), history of high serum cholesterol (yes, no), diabetes mellitus (yes, no), parental myocardialinfarction before age 60 (yes, no), alcohol(>-daily, weekly,rarely), vigorousexercise(~once/week, < once/week), and beta caroteneassignment. § Coronaryrevascularizationproceduresinclude angina-relatedcoronaryartery bypassgraft surgeryand percutaneoustransluminalcoronaryangioplasty.
December 1990 The American Journal of Medicine Volume 89
ASPIRIN AND ANGINA PECTORIS / MANSON ET AL
suggest that the role of platelets in the occlusion of coronary arteries leading to infarction may be different from their role in the development of atherosclerotic stenosi8. Although platelet inhibition with low-dose aspirin may prevent a sudden thrombosis of a stenotic coronary artery, it does not appear to prevent or retard the process of coronary stenosis or transient ischemia. Some potential limitations of the present study should be considered. The average duration of the randomized aspirin component of the trial was only 60.2 months. Atherosclerosis is a multistage and gradual process of alterations of the vessel wall that takes many years to lead to symptoms of myocardial ischemia [23-25]. However, when analyses were performed separately according to year of treatment and follow-up in the trial, no trend towards a beneficial effect of aspirin was apparent. A small to moderate effect of low-dose aspirin after a longer period of intervention cannot be excluded. It is unlikely, however, that a randomized study of sufficient size and duration to evaluate this possibility will ever be conducted. A second potential limitation is that the dose of aspirin used in this study may have been insufficient to inhibit platelet-arterial wall interactions and platelet-related atherogenesis. Although low-dose aspirin inhibits platelet cyclooxygenase activity and associated platelet aggregation [15], the efficacy of this agent in suppressing platelet-derived growth factors or other factors stimulating atheroma formation is unknown. A further possibility that deserves consideration is that the number of angina cases in the aspirin group may have been increased due to the prevention of myocardial infarction by aspirin and the shifting of some individuals from the category of myocardial infarction to that of angina. Although this is a theoretical possibility, we have no direct evidence that this did in fact occur. Participants with angina were included in the analyses regardless of their subsequent status concerning the development of myocardial infarction. Although it is unlikely that such a shift would have been of large enough magnitude to obscure a true protective effect of aspirin on the risk of angina, it is possible that it could account for the slight excess in the number of cases of angina and coronary revascularization observed in the aspirin group. In conclusion, these data from a randomized trial of U.S. physicians indicate that low-dose aspirin for an average duration of 60.2 months does not reduce the incidence of angina pectoris or the utilization of angina-related coronary revascularization procedures among men without a previous diagnosis of coronary heart disease. No evidence of an emerging benefit with longer duration of treatment and fol-
2.4 2.2 o')
==
2.0 t.8 1.6
t.2 t.0 abJ >
-J W n.
0.8 0.6 0.4 0.2 0 I
2
3
4
5+
YEAR OF TREATMENT AND FOLLOW-UP Figure 1. Relative risks (adjusted for age and beta carotene a s s i g n m e n t ) of c o n f i r m e d angina p e c t o r i s in t h e aspirin group, as c o m p a r e d with the placebo group, according to year of randomization in the trial. Vertical lines represent 95% confidence intervals.
low-up was observed. These data suggest that, although low-dose aspirin is effective in the prevention of coronary occlusion leading to myocardial infarction, it does not appear to prevent or delay the onset of angina pectoris.
ACKNOWLEDGMENT We are indebted to the Steering Committee and staff of the Physicians' Health Study, and, in particular, to the 22,071 dedicated and conscientious physicians who are participating in this ongoing trial.
REFERENCES 1. Roberts WC. The coronary artery and left ventricle in clinically isolated angina pectoris: a necropsy study. Circulation 1976; 54: 388-90. 2. Packham MA, Mustard JF. The role of platelets in the development and complications of atherosclerosis. Semin Hematol 1986; 23: 8-26. 3. CazenaveJP, Packham MA, Guccione MA, Mustard JF. Inhibition of platelet adherence to damaged surface of rabbit aorta. J Lab Clin Med 1975; 86: 55153. 4. Buchanan MR, Dejana E, Gent M, Mustard JF, Hirsch J. Enhanced platelet accumulation onto injured carotid arteries in rabbits after aspirin treatment. J Clin Invest 1981; 67: 503-8. 5. Pick R, Chediak J, Glick G. Aspirin inhibits the development of coronary atherosclerosis in cynomolgus monkeys (macaca ascicularis) fed an atherogenic diet. J Clin Invest 1979; 63: 158-62. 6. Koster JK, Tryka AF, H'Doubler P, Collins JJ. The effect of low-dose aspirin and dipyridamole upon atherosclerosis in the rabbit. Artery 1981; 9: 405-13. 7. Epstein SE, Palmeri ST. Mechanisms contributing to precipitation of unstable angina and acute myocardial infarction: implications regarding therapy. Am J Cardiol 1984; 54: 1245-52. 8. The Steering Committee of the Physicians Health Study Research Group. Preliminary report: findings from the aspirin component of the ongoing Physicians' Health Study. N Engl J Med 1988; 318: 262-4. 9. The Steering Committee of the Physicians Health Study Research Group. Final report on the aspirin component of the ongoing Physicians' Health Study. N Engl J Med 1989; 321: 129-35.
December 1990
The American Journal of Medicine
Volume 89
775
ASPIRIN AND ANGINA PECTORIS / MANSON ET AL 10. Resnekov L, Chediak J, Hirsh J, Levis HD. Antithrombotic agents in coronary artery disease. Chest 1989; 95: 52S-72S. 11. Canner PL. Aspirin in coronary heart disease. Comparison of six clinical trials. Isr J Med Sci 1983; 19: 413-23. 12, Antiplatelet Trialists' Collaboration. Secondary prevention of vascular disease by prolonged antiplatelet treatment. Br Med J 1988; 296: 320-31. 13, Hennekens CH, Buring JE, Sandercock P, Collins R, Peto R. Aspirin and other antiplatelet agents in the secondary and primary prevention of cardiovascular disease. Circulation 1989; 80: 749-56. 14. Fox KA, Bergmann SR, Mathias CJ, eta/. Scintigraphic detection of coronary artery thrombi in patients with acute myocardial infarction. Am J Cardio11984; 4: 975-86. 15. DeWood MA, Spores J, Notske R, eta/. Prevalence of total coronary occlusion during the early hours of transmural myocardial infarction. N Engl J Med 1980; 303: 897-902. 16. Miettinen O. Estimability and estimation in case-referent studies. Am J Epidemiol 1976; 103: 226-35. 17. Moncada S, Vane JR. Arachidonic acid metabolites and the interactions between platelets and blood-vessel walls. N En81J Med 1979; 300: 1142-7.
776
18. Kessler CM. Anticoagulation and thrombolytic therapy: practical considerations. Chest 1989; 95: 245S-56S. 19, Ross R, Glomset J, Kariya B, Harker L. A platelet-dependent serum factor that stimulates the proliferation of arterial smooth muscle cells in vitro. Proc Natl Acad Sci 1974; 71: 1207-10. 20. Mehta J. Platelets and prostaglandins in coronary artery disease: rationale for use of platelet-suppressive drugs. JAMA 1983; 249: 2818-23. 21. Folts JD, Crowell EB, Rowe G. Platelet aggregation in partially obstructed vessels and its elimination with aspirin. Circulation 1976; 54: 365-70. 22,Aiken JW, Gorman RR, Shebuski RJ. Prevention of blockage of partially obstructed coronary arteries with prostacyclin correlates with inhibition of platelet aggregation. Prostaglandins 1979; 17: 483-94. 23. Osborn GR. The incubation period of coronary thrombosis. London: Butterworths, 1963. 24. Ross R. The pathogenesis of atherosclerosis--an update. N Engl J Med 1986; 314: 488-500. 25. Newman WP III, Freedman D, Voors AW, et aL Relation of serum lipoprotein levels and blood pressure to early atherosclerosis: the Bogalusa Heart Study. N Engl J Med 1986; 314: 138-44.
December1990 The American Journalof Medicine Volume89
PURPOSE: The objective of this study was to exAmine the effect of low-dose aspirin (325 mg on alternate days) on the primary prevention of angina pectoris in the United States Physician~' Health Study. Despite a postulated role of platelets in atherogenesis and myocardial ischemi% the effect of chronic platelet inhibition on the development of clinical angina pectoris is nnlmown. SUBJECTS AND METHODS:The Physic~ns' Health Study is a randomiT~d, double-blind, placebo-controlled trial among 22,071 male physiciRn~ aged 40 to 84 years, free from previous myocardial infarction, stroke, and transient cerebral ischemia at entry, and followed for an average of 60,2 months. The 21,738 physicla~s who were also free from angina pectoris at baseline constituted the study population for the present nn~lyl~S. RESULTS: During 106,652 person-years of follow-up, 331 patients with confirmed angina poctoris were diagnosed, 194 of whom underwent a coronary revascnl~rization procedure (coronary artery bypass graft surgery or coronary angioplasty). As compared to participants assigned placebo, the relative risk of confirmed angina pectoris in the aspirin group was 1.!0 (95% confidence interval [CI], 0.88 to 1.38). For coronary revascularization, the relative risk was 1.19 (95% CI, 0.88 to 1.59). After simultaneous control for other coronary risk factors in a proportionalh~7~rds model, these relative risks remained near unity at 1.07 (95% CI, 0~4 to 1.36) and 1.11 (95% CL 0.81 to 1.52), respectively. When the
risks of angina pectoris were exRmined according to year of randomiT~tion in the trial, there was no pattern of increasing benefit with longer duration of treatment. CONCLUSION:These randomiT~d trial data indicate that chronic platelet inhibition with lowdose aspirin for an average duration of 60~2 months does not reduce the incidence of angina pectoris.
ngina pectoris generally reflects advanced ath-
A erosclerotic narrowing of one or more coronary arteries [1]. The presence of platelets in atherosclerotic plaques [2], the observation that platelet adhesion to damaged endothelium may promote plaque formation [2-4], and the demonstration that aspirin may inhibit atheroma formation in monkeys and rabbits fed atherogenic diets [5,6] suggest that platelet activity is associated with atherogenesis. Furthermore, platelet aggregation at the site of atherosclerotic narrowing may precipitate acute anginal episodes in patients with both fixed and variable lesions in a coronary artery [7]. Although low-dose aspirin is effective in primary [8,9] and secondary [10-13] prevention of myocardial infarction, presumably by inhibition of thrombosis [14,15], there are no data concerning this agent and the development of angina pectoris. We examined the effect of low-dose aspirin on the primary prevention of angina pectoris in the United States Physicians' Health Study, a randomized trial of 22,071 men aged 40 to 84 years followed for an average period of 60.2 months.
SUBJECTSAND METHODS From the Channing Laboratory, Department of Medicine, Harvard Medical School and Brigham and Women's Hospital (JEM, MJS, JOT, SZG, JMG, PMR, CHH), and the Department of Preventive Medicine, Harvard Medical School (JEB, CHH), Boston, Massachusetts; and the Department of Epidemiologyand Biostatistics, ErasmusUniversity Medical School (DEG), Rotterdam, The Netherlands. This work was supported by Grants HL-26490, HL-34595, CA-34944, and CA 40360 from the National Institutes of Health. Requestsfor reprints should be addressed to JoAnn E. Manson, M.D., 55 Pond Avenue, Brookline, Massachusetts02146. Manuscript submitted June 5, ]990, and accepted in revised form August 28, 1990.
772
December 1990
The Physicians' Health Study [8,9] is a randomized, double-blind, placebo-controlled trial designed to test the effect of low-dose aspirin on risks of cardiovascular disease and the effect of beta carotene on the incidence of cancer. A detailed description of the subjects and methods has been previously p u b l i s h e d [8]. Briefly, 22,071 U.S. male physicians aged 40 to 84 years at entry and free from prior myocardial infarction, stroke, and tran-
The American Journal of Medicine Volume 89
ASPIRINAND ANGINAPECTORIS/ MANSONET AL
sient cerebral ischemia were assigned at random, using a two-by-two factorial design, to one of four treatment groups: aspirin and beta carotene, aspirin and beta carotene placebo, aspirin placebo and beta carotene, or aspirin placebo and beta carotene placebo. A total of 11,037 physicians were randomized to aspirin, 325 mg on alternate days (Bufferin®, Bristol-Myers, New York, New York), and 11,034 to aspirin placebo. Every 6 months for the first year and annually thereafter, the participants were mailed a supply of monthly calendar packs containing study medication and brief questionnaires asking about their compliance with the assigned treatment and the occurrence of any relevant events. The reported consumption of aspirin or other platelet-active drugs was 85.7% in the aspirin group and 14.2% in the placebo group. The questionnaires also included items about cigarette smoking, systolic and diastolic blood pressure, serum cholesterol, height and weight, diagnosis of diabetes mellitus, parental history of myocardial infarction, alcohol use, and frequency of vigorous exercise. The aspirin component was terminated early (January 25, 1988) due to a statistically extreme 44% reduction in the risk of a first myocardial infarction among those in the aspirin group [8,9]. By this date, participants had been followed for an average of 60.2 months (range, 45.8 to 77.0 months). Vital status was known for all physicians, and follow-up data on morbidity were 99.7% complete.
Endpoints The primary endpoint in the present analyses is the incidence of confirmed angina pectoris. Of the 22,071 randomized physicians, 333 reported angina and/or coronary revascularization at baseline and were excluded. The present analyses include the 21,738 physicians free from myocardial infarction, angina pectoris, coronary revascularization, stroke, and transient cerebral ischemia at baseline. When participants reported a first diagnosis of angina pectoris after randomization, we sent supplementary questionnaires to elicit details about diagnostic tests and treatment. A report of angina pectoris was classified as confirmed by a committee of physicians if, on the supplementary questionnaire, a positive diagnostic test (a confirmatory exercise tolerance test, exercise thallium, exercise radionuclide ventriculogram, or narrowing of 50% or greater of a coronary artery on angiogram) was reported, or if the participant reported a coronary revascularization procedure (coronary artery bypass graft surgery or percutaneous transluminal coronary angioplasty). The physicians reviewing the confirmatory evidence for angina were unaware of the aspirin treatment assignment of the participants. Analyses
are presented separately for total confirmed angina pectoris (n = 331) and for the subgroup of coronary revascularization (n = 194). The available information did not permit a distinction between stable and unstable categories of angina pectoris. Randomized physicians reporting a first myocardial infarction prior to diagnosis of angina or coronary revascularization procedure were, however, excluded.
Endpoint Validation A validation study of angina pectoris confirmed by supplementary questionnaire was conducted among 100 randomly selected participants who reported this diagnosis. Consent for release of medical records regarding pertinent diagnostic tests was requested. Consent was obtained from 99 of the 100 subjects, among whom medical records were available for review in 90. Medical record review by a physician masked to the aspirin treatment assignment corroborated at least one positive diagnostic test result for angina pectoris in 88 of 90 participants {97.8%). Records were not requested to confirm self-reports of coronary revascularization procedures.
Statistical Analyses Incidence rates were calculated by dividing the numbers of events by person-time of follow-up for the aspirin and placebo groups, after adjustment for age (5-year categories) and beta carotene treatment assignment. Relative risks were calculated as the ratio of the incidence rate in the aspirin group to that in the placebo group. Multivariate relative risks were derived from proportional-hazards models, which controlled for multiple coronary risk factors simultaneously. For each relative risk, we calculated the 95% confidence interval (CI) [16]. To assess possible modifying effects of other coronary risk factors, analyses were performed within strata of these variables. To evaluate the effect of duration of aspirin therapy on the risk of confirmed angina, relative risks were examined separately according to year of randomization in the trial.
RESULTS The baseline characteristics of the analyzed participants were virtually identical in the aspirin and placebo groups (Table I). During an average of 60.2 months of follow-up (106,652 person-years), there were 331 subjects with confirmed angina pectoris, 194 of whom underwent coronary artery bypass graft surgery or percutaneous transluminal coronary angioplasty. When compared to those assigned placebo, the relative risk of confirmed angina pectoris among participants randomized to receive aspirin was 1.10 (95%
December 1990 The American Journal of Medicine Volume 89
773
ASPIRIN AND ANGINA PECTORIS / MANSON ET AL
TABLE I BaselineCharacteristicsof ParticipantsAccordingto Treatment Assignment,as Reportedon the Initial Questionnaire*
Mean age (years) (SEM) Mean body massindex (kg/m2) (SEM) Mean systolic BP (mm Hg) ( S E M ) Mean diastolic BP (mm Hg) (SEM) Hypertension (%)t Mean serum cholesterol (SEM):~ Diabetes mellitus (%) Parental history of MI (%)§ Cigarette smoking Current (%) Past (%) Never (%) Alcohol use Daily (%) Weekly (%) Rarely (%)
Placebo Group
Aspirin Group
(n = 10,879)
(n = 10,859)
53.1 (0.09) 24.9(0.03) 126.0(0.12) 78.8 (0.08) 13.5 211.8 (0.73) 2.3 13.0
53.1 (0.09) 24.9(0.03) 126.0(0.12) 78.9 (0.08) 13.6 211.7 (0.71) 2.3 13.0
11.1 38.9 50.0
11.1 39.5 49.4
24.9 48.6 26.5
24.8 49.7 25.5
* Excludesphysicianswith anginapectorisor coronaryrevascularizationat baseline.Risk
factors are self-reported. SEM denotesstandarderror of the mean. BP denotesblood
pressure. * Definedas self-reportedsystolicBP > 160 mm Hgand/or diastolicBP >- 95 mm Hg, or takingantihypertensivemedication. * Units are mg/dL. To calculatemmol/L, divideby 38.7. Reportedby onlyone third of participants. § MI denotesmyocardialinfarction, occurringin either parentbeforeage60.
CI, 0.88 to 1.38) (Table II). For coronary revascularization procedures, the relative risk was 1.19 (95% CI, 0.88 to 1.59). After simultaneous control for other coronary risk factors in a proportionalhazards model, these relative risks were not materially altered, being 1.07 (95% CI, 0.84 to 1.36) and 1.11 (95% CI, 0.81 to 1.52), respectively. There was no evidence for any modifying effect of age, cigarette smoking, blood pressure, diabetes mellitus, or other coronary risk factors on the influence of aspirin. To assess the possibility that a benefit of aspirin on the risk of angina pectoris is delayed due to the need for prolonged platelet inhibition, the relative risk of confirmed angina was examined according to year of randomization in the trial (Figure 1). Relative risks were not significantly different from unity at any time point in the trial, and there was no pattern of an increasing benefit from aspirin with longer duration of treatment. Specifically, in the fifth and final year of randomized treatment, the relative risk in the aspirin group was 1.18 (95% CI, 0.71 to 1.98).
increasing benefit from aspirin with prolonged treatment was evident. Low-dose aspirin irreversibly inhibits platelet cyclooxygenase activity, preventing thromboxane A2 production and inhibiting platelet aggregation [17,18]. The effect of aspirin in preventing myocardial infarction likely results from prevention of thrombotic occlusion of coronary arteries that are severely affected by atherosclerotic disease [14,15]. Platelets have also been implicated, however, in the development of atherosclerotic plaque [2-4], suggesting a possible role in atherogenesis. The hypothesis that platelet-arterial wall interactions may contribute to the development of atherosclerosis is supported by the finding that a platelet-derived factor appears to stimulate proliferation of arterial smooth muscle cells [19,20]. Several animal studies have suggested that aspirin may have an inhibitory effect on atheroma formation; aspirin has been demonstrated to inhibit the progression of coronary atherosclerosis in monkeys fed an atherogenic diet [5] and, in combination with dipyridamole, to inhibit plaque formation in hypercholesterolic rabbits [6]. Furthermore, animal studies have suggested that transient decreases in coronary blood flow may be related to acute platelet aggregation that can be prevented by aspirin therapy [21,22]. Despite plausible mechanisms for chronic platelet inhibition in the primary prevention of angina pectoris, the present data indicate that the risk of angina and the rate of coronary revascularization were not influenced by low-dose aspirin. These observations
TABLEII Relative Risk (RR) of Confirmed AnginaPectorisand Coronary RevascularizationProceduresAccordingTo AspirinTreatment Assignment
Number of person-years Confirmed angina pectoris Cases
RR* (95% CI)t Multivariate RR* (95% CI) Coronary revascularization§ Cases
COMMENTS
These data from a randomized trial of U.S. physicians demonstrate no apparent benefit of low-dose aspirin on the incidence of confirmed angina pectoris or the utilization of coronary revascularization procedures during an average treatment period of 60.2 months. In analyses performed separately by year of randomization in the trial, no pattern of an 774
PlaceboGroup (n = 10,879)
Aspirin Group (n = 10,859)
53,298
53,354
158
173
1.0
1.10
Referent 1.0 Referent
(0.88, 1.38) 1.07 (0.84, 1.36)
89
105
RR* (95% CI)
1.0 Referent
1.19 (0.88, 1.59)
Multivariate RR~ (95% Cl)
1.0 Referent
1.11 (0.81, 1.52)
* Adjustedby age(5-yearcategories)and betacaroteneassignment. t CI denotesconfidenceinterval. t MultivariateRR includessimultaneouscontrol for age (5-year categories), cigarette smoking(never, past, current), bodymassindex,systolicbloodpressure(BP), diastolic BP, history of hypertension (yes, no), history of high serum cholesterol (yes, no), diabetes mellitus (yes, no), parental myocardialinfarction before age 60 (yes, no), alcohol(>-daily, weekly,rarely), vigorousexercise(~once/week, < once/week), and beta caroteneassignment. § Coronaryrevascularizationproceduresinclude angina-relatedcoronaryartery bypassgraft surgeryand percutaneoustransluminalcoronaryangioplasty.
December 1990 The American Journal of Medicine Volume 89
ASPIRIN AND ANGINA PECTORIS / MANSON ET AL
suggest that the role of platelets in the occlusion of coronary arteries leading to infarction may be different from their role in the development of atherosclerotic stenosi8. Although platelet inhibition with low-dose aspirin may prevent a sudden thrombosis of a stenotic coronary artery, it does not appear to prevent or retard the process of coronary stenosis or transient ischemia. Some potential limitations of the present study should be considered. The average duration of the randomized aspirin component of the trial was only 60.2 months. Atherosclerosis is a multistage and gradual process of alterations of the vessel wall that takes many years to lead to symptoms of myocardial ischemia [23-25]. However, when analyses were performed separately according to year of treatment and follow-up in the trial, no trend towards a beneficial effect of aspirin was apparent. A small to moderate effect of low-dose aspirin after a longer period of intervention cannot be excluded. It is unlikely, however, that a randomized study of sufficient size and duration to evaluate this possibility will ever be conducted. A second potential limitation is that the dose of aspirin used in this study may have been insufficient to inhibit platelet-arterial wall interactions and platelet-related atherogenesis. Although low-dose aspirin inhibits platelet cyclooxygenase activity and associated platelet aggregation [15], the efficacy of this agent in suppressing platelet-derived growth factors or other factors stimulating atheroma formation is unknown. A further possibility that deserves consideration is that the number of angina cases in the aspirin group may have been increased due to the prevention of myocardial infarction by aspirin and the shifting of some individuals from the category of myocardial infarction to that of angina. Although this is a theoretical possibility, we have no direct evidence that this did in fact occur. Participants with angina were included in the analyses regardless of their subsequent status concerning the development of myocardial infarction. Although it is unlikely that such a shift would have been of large enough magnitude to obscure a true protective effect of aspirin on the risk of angina, it is possible that it could account for the slight excess in the number of cases of angina and coronary revascularization observed in the aspirin group. In conclusion, these data from a randomized trial of U.S. physicians indicate that low-dose aspirin for an average duration of 60.2 months does not reduce the incidence of angina pectoris or the utilization of angina-related coronary revascularization procedures among men without a previous diagnosis of coronary heart disease. No evidence of an emerging benefit with longer duration of treatment and fol-
2.4 2.2 o')
==
2.0 t.8 1.6
t.2 t.0 abJ >
-J W n.
0.8 0.6 0.4 0.2 0 I
2
3
4
5+
YEAR OF TREATMENT AND FOLLOW-UP Figure 1. Relative risks (adjusted for age and beta carotene a s s i g n m e n t ) of c o n f i r m e d angina p e c t o r i s in t h e aspirin group, as c o m p a r e d with the placebo group, according to year of randomization in the trial. Vertical lines represent 95% confidence intervals.
low-up was observed. These data suggest that, although low-dose aspirin is effective in the prevention of coronary occlusion leading to myocardial infarction, it does not appear to prevent or delay the onset of angina pectoris.
ACKNOWLEDGMENT We are indebted to the Steering Committee and staff of the Physicians' Health Study, and, in particular, to the 22,071 dedicated and conscientious physicians who are participating in this ongoing trial.
REFERENCES 1. Roberts WC. The coronary artery and left ventricle in clinically isolated angina pectoris: a necropsy study. Circulation 1976; 54: 388-90. 2. Packham MA, Mustard JF. The role of platelets in the development and complications of atherosclerosis. Semin Hematol 1986; 23: 8-26. 3. CazenaveJP, Packham MA, Guccione MA, Mustard JF. Inhibition of platelet adherence to damaged surface of rabbit aorta. J Lab Clin Med 1975; 86: 55153. 4. Buchanan MR, Dejana E, Gent M, Mustard JF, Hirsch J. Enhanced platelet accumulation onto injured carotid arteries in rabbits after aspirin treatment. J Clin Invest 1981; 67: 503-8. 5. Pick R, Chediak J, Glick G. Aspirin inhibits the development of coronary atherosclerosis in cynomolgus monkeys (macaca ascicularis) fed an atherogenic diet. J Clin Invest 1979; 63: 158-62. 6. Koster JK, Tryka AF, H'Doubler P, Collins JJ. The effect of low-dose aspirin and dipyridamole upon atherosclerosis in the rabbit. Artery 1981; 9: 405-13. 7. Epstein SE, Palmeri ST. Mechanisms contributing to precipitation of unstable angina and acute myocardial infarction: implications regarding therapy. Am J Cardiol 1984; 54: 1245-52. 8. The Steering Committee of the Physicians Health Study Research Group. Preliminary report: findings from the aspirin component of the ongoing Physicians' Health Study. N Engl J Med 1988; 318: 262-4. 9. The Steering Committee of the Physicians Health Study Research Group. Final report on the aspirin component of the ongoing Physicians' Health Study. N Engl J Med 1989; 321: 129-35.
December 1990
The American Journal of Medicine
Volume 89
775
ASPIRIN AND ANGINA PECTORIS / MANSON ET AL 10. Resnekov L, Chediak J, Hirsh J, Levis HD. Antithrombotic agents in coronary artery disease. Chest 1989; 95: 52S-72S. 11. Canner PL. Aspirin in coronary heart disease. Comparison of six clinical trials. Isr J Med Sci 1983; 19: 413-23. 12, Antiplatelet Trialists' Collaboration. Secondary prevention of vascular disease by prolonged antiplatelet treatment. Br Med J 1988; 296: 320-31. 13, Hennekens CH, Buring JE, Sandercock P, Collins R, Peto R. Aspirin and other antiplatelet agents in the secondary and primary prevention of cardiovascular disease. Circulation 1989; 80: 749-56. 14. Fox KA, Bergmann SR, Mathias CJ, eta/. Scintigraphic detection of coronary artery thrombi in patients with acute myocardial infarction. Am J Cardio11984; 4: 975-86. 15. DeWood MA, Spores J, Notske R, eta/. Prevalence of total coronary occlusion during the early hours of transmural myocardial infarction. N Engl J Med 1980; 303: 897-902. 16. Miettinen O. Estimability and estimation in case-referent studies. Am J Epidemiol 1976; 103: 226-35. 17. Moncada S, Vane JR. Arachidonic acid metabolites and the interactions between platelets and blood-vessel walls. N En81J Med 1979; 300: 1142-7.
776
18. Kessler CM. Anticoagulation and thrombolytic therapy: practical considerations. Chest 1989; 95: 245S-56S. 19, Ross R, Glomset J, Kariya B, Harker L. A platelet-dependent serum factor that stimulates the proliferation of arterial smooth muscle cells in vitro. Proc Natl Acad Sci 1974; 71: 1207-10. 20. Mehta J. Platelets and prostaglandins in coronary artery disease: rationale for use of platelet-suppressive drugs. JAMA 1983; 249: 2818-23. 21. Folts JD, Crowell EB, Rowe G. Platelet aggregation in partially obstructed vessels and its elimination with aspirin. Circulation 1976; 54: 365-70. 22,Aiken JW, Gorman RR, Shebuski RJ. Prevention of blockage of partially obstructed coronary arteries with prostacyclin correlates with inhibition of platelet aggregation. Prostaglandins 1979; 17: 483-94. 23. Osborn GR. The incubation period of coronary thrombosis. London: Butterworths, 1963. 24. Ross R. The pathogenesis of atherosclerosis--an update. N Engl J Med 1986; 314: 488-500. 25. Newman WP III, Freedman D, Voors AW, et aL Relation of serum lipoprotein levels and blood pressure to early atherosclerosis: the Bogalusa Heart Study. N Engl J Med 1986; 314: 138-44.
December1990 The American Journalof Medicine Volume89
