Saturday
12 December 1992
No 8833
ORIGINAL ARTICLES Double-blind trial of aspirin in primary prevention of myocardial infarction in patients with stable chronic
angina pectoris
Clinical trials have demonstrated a prophylactic role for aspirin in myocardial infarction and in unstable angina pectoris. The Swedish Angina Pectoris Aspirin Trial (SAPAT) is the first prospective study of aspirin in stable angina.
2035 patients treatment with
patients
were
were
mg daily or placebo. All treated with sotalol for control of
symptoms. The median duration of follow-up
was
50 months. Compared with the placebo + sotalol group, the aspirin + sotalol group had a 34% (81 vs 124 patients) reduction in primary outcome events (myocardial infarction and sudden death; 95% confidence interval 24-49%; p=0·003) and the reduction observed in secondary outcome events (vascular events, vascular death, all cause mortality, stroke) ranged from 22% to 32%. Treatment withdrawal caused by adverse events occurred in 109 patients in the aspirin + sotalol group and 100 in patients in the placebo + sotalol group; major bleeds, including haemorrhagic stroke, occurred in 20 and
13 patients, respectively (not significant). The addition of a low dose of aspirin
significant
to sotalol benefit in terms of
cardiovascular events, including a significant reduction in the incidence of first myocardial infarction in patients with symptoms of stable angina
pectoris. Lancet 1992; 340: 1421-25.
Introduction inhibits
symptoms sotalol6
Aspirin cyclo-oxygenase-dependent platelet aggregation. Complete inhibition was obtained with aspirin in doses as low as 20-100 mg per day, while doses above 80 mg resulted in substantial inhibition of endogenous prostacyclin biosynthesis.l Aspirin 75-160 mg daily reduces mortality in myocardial infarction (MI)2 and unstable angina and prevents stroke after cerebral transient ischaemic attacks.’ Furthermore, aspirin has been effective in preventing MI in healthy American doctors.5
was
the standard treatment for all
patients.
randomised double-blind to
aspirin 75
treatment showed
In SAPAT, the main purpose was to compare aspirin and placebo in patients with a history of chronic stable angina pectoris without a previous MI. For control of anginal
Patients and methods Patients 2035 patients with symptoms of chronic stable angina pectoris treated with increasing doses of sotalol until optimal symptom control was obtained. The median dose was 160 (range 40-480) mg daily (table i). After showing good tolerance of sotalol for at least three weeks the patients were randomised double blind to aspirin 75 mg daily (n = 1009) or placebo (n = 1026). The inclusion criterion was a history of exertional chest pain for at least one month in patients aged 30-80. The protocol did not require further investigations such as exercise tests, radionuclide scans, or coronary angiography. Patients already on treatment with, or requiring, aspirin, anticoagulants, verapamil, or non-steroid anti-inflammatory drugs were excluded; as were, to avoid the risk of hypokalaemia, patients needing more than 50 mg of hydrochlorthiazide, 5 mg bendroflumethiazide or 40 mg frusemide daily. Further exclusion criteria were a resting heart rate below 55/min, ongoing treatment with class i antiarrhythmic drugs, a history of myocardial infarcdon, atrioventricular block n/m, symptoms of obstructive lung disease; active peptic ulcer, hypersensitivity to aspirin, juvenile diabetes, and uncontrolled late-onset diabetes. Informed consent was obtained from all patients before randomisation. The study protocol was approved by the ethics committees concerned and the Swedish Medical Products Agency. The study was conducted according to the Helsinki Declaration. were
’
Design SAPAT was a multicentre trial, in primary care centres in Sweden. Between June, 1985, and November, 1989, 2035 patients were randomised at ninety-four centres by 108 investigators. The
ADDRESSES. Cardiology Section, Department of Medicine, General Hospital, S-21401 Malmö, Sweden (S Juul-Moller, MD);
Sahlgren’s Hospital, Gothenburg (N Edvardsson, MD); County Hospital, Västerås (S Sorensen, PhD); and Bristol-Myers Squibb, Stockholm (B. Jahnmatz, BSc, A Rosén, MD R Omblus, BSc). Correspondence to Dr S Juul-Möller
1422
TABLE I-DEMOGRAPHIC DATA AT RANDOMISATION
Months0 KX)9
1026
12
993 988
24 974 956
Number of patients
Cumulative plot of
Mean
(SD) unless otherwise indicated; for sotalol
dose
interquartile range is given
median number of enrolled patients per centre was 19 (range 1-66). Twenty-five centres contributed half the patients. Follow-up stopped on Nov 1, 1991. 2 patients were lost to follow-up in respect of mortality and another 9 in respect of other endpoints. Of these 11patients, 6 were in the aspirin group and 5 in the placebo group. During a median follow-up of 50 (range 23-76) months, the patients were seen every 3-6 months. Every visit included a medical history and clinical examination. An electrocardiogram was recorded at randomisation and at the end of the study. Any spontaneously reported new symptoms were noted. An adverse event questionnaire was not used. Besides monitoring of drug prescription and of the blinded trial medication there was no procedure to control compliance. Every 3-6 months each centre was monitored by Bristol-Myers Squibb staff to control patient enrolment, case record forms, and the status of study drug supply. Twice a year an investigators’ meeting with an integrated educational session was held. During the last year of the follow-up period all patient files were examined at on-site visits by the two cardiologists in the executive committee, both of whom are independent of Bristol-Myers Squibb. Throughout the study and at the end of the study all endpoints and all events causing cessation of treatment were examined and classified by one of the cardiologists and then confirmed by the other, both cardiologists being blind to the trial medication.
Definition of terms The primary endpoint was the first occurrence of non-fatal or fatal MI (during hospitalisation) or sudden death. MI was defined on World Health Organization criteria’ by the presence of two or more of: (a) central typical chest pain, pulmonary oedema, syncope
shock; (b) a pathological Q wave and/or appearance or disappearance of localised ST-elevation followed by T-wave inversion in two or more of twelve standard leads; (c) increase in concentration of serum enzymes consistent with MI; and (d) necropsy fmdings of MI of an age corresponding to time of onset of symptoms. Sudden death was defined as witnessed, instantaneous or
death or death within one hour of onset of symptoms, as well as unwitnessed death that, depending on the circumstances, was regarded to have been sudden. About 80% of sudden deaths are caused by ischaemic heart disease.8 Secondary endpoints were vascular events (first occurrence of MI, stroke, or vascular death), vascular death (ie, fatal vascular events), all-cause mortality, and stroke. Cessation of study treatment was defined as the defrnite withdrawal of either sotalol or the aspirin/placebo component and occurred in 779 patients, including cessation caused by an endpoint. A temporary interruption of any study medication for up to 4 weeks was allowed and occurred in 149 patients (69 aspirin + sotalol
treated, 80 placebo + sotalol).
at
nsk
in
60 210 215
36 741 722
48 505 497
each of the
two treatment arms
72 20 16
primary endpoints (ASA=aspirin).
Bleeds were divided into major and minor. Major bleeds required transfusion, caused death, or had serious implications for the patient.
Statistical analysis The calculation of sample size was based on the expected incidence of MI. With an accumulated incidence of 10%, a presumed 35% risk reduction, a type I error of 0-05 (two-tailed), and a type II error of 0-20, the number of randomised patients was calculated to be 2000. Time to an endpoint was analysed by the Kaplan-Meier method and the logrank test, on the intention-to-treat principle. The chi-square test was used to compare frequencies of adverse effects. Prognostic factors and demographic data were analysed with respect to equal distribution between the two treatment regimens. The effect of co-variates was analysed according to the Cox proportional hazards model (Systat statistical package, survival
module). All statistical analyses were performed by independent of Bristol-Myers Squibb.
a
statistician
Results at randomisation for the two shown in table 1. Sex ratio, age, blood pressure, body weight, duration of angina pectoris, diabetes, smoking habits, treatment for hypertension, serum potassium, total cholesterol, and triglyceride levels, and concomitant use of calcium channel blocking drugs were virtually identical in the two groups.
The
groups
demographic data
are
Primary outcome event At the end of the follow-up period there were 205 patients with a primary endpoint, 81 in the aspirin + sotalol group and 124 in the placebo + sotalol group (p = 0-003, fig 1). Within the primary endpoint there was a significant decrease in non-fatal MI and
a
beneficial trend in sudden
TABLE II-ENDPOINTS
1423
TABLE III-REASONS FOR STOPPING TREATMENT PREMATURELY
TABLE V-BLEEDS I
*Requnng aspinn
*Figures m parentheses indicate fatal events Patients are listed by first bleeding event.
treatment
Comment. Patients are only mentioned by first event However, primary endpoints other reasons for cessation of treatment. CABG coronary artery bypass grafting, NSAID = non-steroidal anti-inflammatory
censor
drug
death in the aspirin + sotalol group, while the numbers of fatal MIs during hospitalisation were identical in the two groups
(table n).
Secondary outcome events The incidences of all secondary endpoints were lower in the aspirin + sotalol group (table II). Despite reductions of 22-32% only the 32% reduction in vascular events was statistically significant. Stroke developed in 66 patients (28 aspirin + sotalol, 38 placebo + sotalol). There were 59 nonhaemorrhagic strokes, 38 of which were verified by CT scan (23 aspirin + sotalol, 36 placebo + sotalol). The 7 haemorrhagic strokes were verified by CT scan or necropsy (5 aspirin + sotalol, 2 placebo + sotalol). 26 patients experienced transient ischaemic attacks (9 aspirin + sotalol, 17 placebo + sotalol; p 0-170). There was no difference in the number of coronary artery bypass procedures between the two groups (39 vs 40). =
Withdrawal of treatment, adverse events
patients stopped trial medication prematurely (362 aspirin + sotalol, 417 placebo + sotalol) (table III). The 6 protocol violations were caused by non-trial doctors prescribing medication not allowed in the study. There were 342 non-haemorrhagic adverse events, with an equal distribution between the two groups (table iv). 45 patients experienced two adverse events and 7 patients three. 43 patients had bleeds, 33 of which were major (table v). Of these, 14 were fatal (9 aspirin + sotalol, 5 placebo + sotalol). More than half of the major bleeds were in the gastrointestinal tract. 779
Analysis of risk factors In analysis of risk factors for primary endpoint event factors initially included in the model were treatment, sex, age (cut-off 60 years), treatment for hypertension, smoking, serum cholesterol (cut-off 6-5 mmol/1), serum triglycerides (2-0 mmol/1), heredity for MI, diabetes, duration of angina TABLE IV-NON-HAEMORRHAGIC ADVERSE EVENTS
*In 283 patients
I
pectoris (cut-off 2 years), heart rate (70/min) and dose of sotalol (159 mg per day). A variable was excluded if the relative risk was not significant at the 5% level. The following were found to be independent risk factors: treatment, sex, diabetes type n and serum cholesterol. The relative risks (RR) and 95 % confidence intervals were: F vs M, 0-44 (0-33-0-61); no diabetes vs diabetes 0-45 (0-270.72); aspirin + sotalol vs placebo + sotalol 0-68 (049-093); and serum cholesterol < 6.5 vs 6,5 mmol/l 0.70 > The RRs were (051-096). basically identical on trial stratification by medication-ie, the relative risks were the same in the two treatment groups.
Discussion In the prevention of cardiovascular events aspirin has been shown to be beneficial in a variety of primary and secondary prevention trials. The primary prevention of MI was demonstrated in the US Physicians Health Study. A significant 44% reduction in the risk of first infarction was observed in the group treated with aspirin 325 mg every other day.5 However, a trial in British male doctors of 500 mg aspirin daily showed no beneficial effect.9 A retrospective subgroup analysis in the US Physicians Health Study in 333 physicians with angina pectoris at the time of randomisation indicated greater than 50% reduction of risk of first MI."* Among the secondary prevention trials the second International Study of Infarct Survival showed a significant 45% reduction in 5 weeks reinfarction in the aspirin group (160 mg daily).z In unstable angina pectoris/ non-Q MI the RISC trial of 75 mg daily showed a 61 % reduction in MI and death at 90 days follow-up.3 In ISIS-3 a similar efficacy was also demonstrated in patients with developing MI treated with thrombolytic agents and aspirin." Thus, there appears to be a more favourable outcome in the primary as well as the secondary prevention trials with low doses of aspirin (75-160 mg daily). In SAPAT the primary endpoint, non-fatal MI and sudden death showed reductions of similar magnitude, while there was no difference in the rate of fatal MI during hospitalisation. Although the substantial reduction in sudden death was not significant, the results imply that aspirin may prevent sudden occlusion through inhibition of the proaggregatory and vasoconstrictive thromboxane Az. Other factors besides thromboxane Az, such as changes in sympathetic vascular tone and activation of platelet receptors, may also promote platelet aggregation.1z During the change from chronic stable angina to unstable angina and subsequent minor MI an increase in myocardial ischaemia may be initiated by platelet aggregation and vasoconstriction induced by local accumulation of thromboxane Az and by reductions in endothelially derived vasodilators and in inhibitors of platelet aggregation. 13
1424
Aspirin could thus have diminished the risk of developing unstable angina and subsequent MI by its effect on thromboxane Az. This would primarily inhibit the development of minor non-fatal MI capable of triggering electrical disorder, ventricular fibrillation, and sudden death. Major fatal infarctions are probably more often caused by subintimal plaque bleeding with subsequent rupture. 14,15 Platelet-active drugs cannot be expected to have an inhibitory effect on this plaque bleeding and rupture. When SAPAT was designed data on the incidence of MI in patients with chronic stable angina pectoris were based on patients not receiving beta-adrenoceptor blocking drugs, and those data suggested an incidence of 5 % or so per year. 16 The annual incidence of primary endpoints in our controls was lower at (3%). This discrepancy is compatible with the trend in Sweden towards a lower incidence of MI over the past decadel’ and may be related to a better diet and/or to the new drugs introduced for the treatment of angina.18 The standard treatment with sotalol in both groups may thus have contributed to the low incidence of primary endpoint. The selection criteria for the angina pectoris patients may also have had an important role. In 2035 patients followed up for 8485 patient-years in SAPAT the addition of aspirin 75 mg daily to standard sotalol reduced the incidence in the first MI or sudden death by 34% (p=0-003). The reduction of vascular events (including non-fatal MI, non-fatal stroke and vascular death) was 32% (p
12 December 1992
No 8833
ORIGINAL ARTICLES Double-blind trial of aspirin in primary prevention of myocardial infarction in patients with stable chronic
angina pectoris
Clinical trials have demonstrated a prophylactic role for aspirin in myocardial infarction and in unstable angina pectoris. The Swedish Angina Pectoris Aspirin Trial (SAPAT) is the first prospective study of aspirin in stable angina.
2035 patients treatment with
patients
were
were
mg daily or placebo. All treated with sotalol for control of
symptoms. The median duration of follow-up
was
50 months. Compared with the placebo + sotalol group, the aspirin + sotalol group had a 34% (81 vs 124 patients) reduction in primary outcome events (myocardial infarction and sudden death; 95% confidence interval 24-49%; p=0·003) and the reduction observed in secondary outcome events (vascular events, vascular death, all cause mortality, stroke) ranged from 22% to 32%. Treatment withdrawal caused by adverse events occurred in 109 patients in the aspirin + sotalol group and 100 in patients in the placebo + sotalol group; major bleeds, including haemorrhagic stroke, occurred in 20 and
13 patients, respectively (not significant). The addition of a low dose of aspirin
significant
to sotalol benefit in terms of
cardiovascular events, including a significant reduction in the incidence of first myocardial infarction in patients with symptoms of stable angina
pectoris. Lancet 1992; 340: 1421-25.
Introduction inhibits
symptoms sotalol6
Aspirin cyclo-oxygenase-dependent platelet aggregation. Complete inhibition was obtained with aspirin in doses as low as 20-100 mg per day, while doses above 80 mg resulted in substantial inhibition of endogenous prostacyclin biosynthesis.l Aspirin 75-160 mg daily reduces mortality in myocardial infarction (MI)2 and unstable angina and prevents stroke after cerebral transient ischaemic attacks.’ Furthermore, aspirin has been effective in preventing MI in healthy American doctors.5
was
the standard treatment for all
patients.
randomised double-blind to
aspirin 75
treatment showed
In SAPAT, the main purpose was to compare aspirin and placebo in patients with a history of chronic stable angina pectoris without a previous MI. For control of anginal
Patients and methods Patients 2035 patients with symptoms of chronic stable angina pectoris treated with increasing doses of sotalol until optimal symptom control was obtained. The median dose was 160 (range 40-480) mg daily (table i). After showing good tolerance of sotalol for at least three weeks the patients were randomised double blind to aspirin 75 mg daily (n = 1009) or placebo (n = 1026). The inclusion criterion was a history of exertional chest pain for at least one month in patients aged 30-80. The protocol did not require further investigations such as exercise tests, radionuclide scans, or coronary angiography. Patients already on treatment with, or requiring, aspirin, anticoagulants, verapamil, or non-steroid anti-inflammatory drugs were excluded; as were, to avoid the risk of hypokalaemia, patients needing more than 50 mg of hydrochlorthiazide, 5 mg bendroflumethiazide or 40 mg frusemide daily. Further exclusion criteria were a resting heart rate below 55/min, ongoing treatment with class i antiarrhythmic drugs, a history of myocardial infarcdon, atrioventricular block n/m, symptoms of obstructive lung disease; active peptic ulcer, hypersensitivity to aspirin, juvenile diabetes, and uncontrolled late-onset diabetes. Informed consent was obtained from all patients before randomisation. The study protocol was approved by the ethics committees concerned and the Swedish Medical Products Agency. The study was conducted according to the Helsinki Declaration. were
’
Design SAPAT was a multicentre trial, in primary care centres in Sweden. Between June, 1985, and November, 1989, 2035 patients were randomised at ninety-four centres by 108 investigators. The
ADDRESSES. Cardiology Section, Department of Medicine, General Hospital, S-21401 Malmö, Sweden (S Juul-Moller, MD);
Sahlgren’s Hospital, Gothenburg (N Edvardsson, MD); County Hospital, Västerås (S Sorensen, PhD); and Bristol-Myers Squibb, Stockholm (B. Jahnmatz, BSc, A Rosén, MD R Omblus, BSc). Correspondence to Dr S Juul-Möller
1422
TABLE I-DEMOGRAPHIC DATA AT RANDOMISATION
Months0 KX)9
1026
12
993 988
24 974 956
Number of patients
Cumulative plot of
Mean
(SD) unless otherwise indicated; for sotalol
dose
interquartile range is given
median number of enrolled patients per centre was 19 (range 1-66). Twenty-five centres contributed half the patients. Follow-up stopped on Nov 1, 1991. 2 patients were lost to follow-up in respect of mortality and another 9 in respect of other endpoints. Of these 11patients, 6 were in the aspirin group and 5 in the placebo group. During a median follow-up of 50 (range 23-76) months, the patients were seen every 3-6 months. Every visit included a medical history and clinical examination. An electrocardiogram was recorded at randomisation and at the end of the study. Any spontaneously reported new symptoms were noted. An adverse event questionnaire was not used. Besides monitoring of drug prescription and of the blinded trial medication there was no procedure to control compliance. Every 3-6 months each centre was monitored by Bristol-Myers Squibb staff to control patient enrolment, case record forms, and the status of study drug supply. Twice a year an investigators’ meeting with an integrated educational session was held. During the last year of the follow-up period all patient files were examined at on-site visits by the two cardiologists in the executive committee, both of whom are independent of Bristol-Myers Squibb. Throughout the study and at the end of the study all endpoints and all events causing cessation of treatment were examined and classified by one of the cardiologists and then confirmed by the other, both cardiologists being blind to the trial medication.
Definition of terms The primary endpoint was the first occurrence of non-fatal or fatal MI (during hospitalisation) or sudden death. MI was defined on World Health Organization criteria’ by the presence of two or more of: (a) central typical chest pain, pulmonary oedema, syncope
shock; (b) a pathological Q wave and/or appearance or disappearance of localised ST-elevation followed by T-wave inversion in two or more of twelve standard leads; (c) increase in concentration of serum enzymes consistent with MI; and (d) necropsy fmdings of MI of an age corresponding to time of onset of symptoms. Sudden death was defined as witnessed, instantaneous or
death or death within one hour of onset of symptoms, as well as unwitnessed death that, depending on the circumstances, was regarded to have been sudden. About 80% of sudden deaths are caused by ischaemic heart disease.8 Secondary endpoints were vascular events (first occurrence of MI, stroke, or vascular death), vascular death (ie, fatal vascular events), all-cause mortality, and stroke. Cessation of study treatment was defined as the defrnite withdrawal of either sotalol or the aspirin/placebo component and occurred in 779 patients, including cessation caused by an endpoint. A temporary interruption of any study medication for up to 4 weeks was allowed and occurred in 149 patients (69 aspirin + sotalol
treated, 80 placebo + sotalol).
at
nsk
in
60 210 215
36 741 722
48 505 497
each of the
two treatment arms
72 20 16
primary endpoints (ASA=aspirin).
Bleeds were divided into major and minor. Major bleeds required transfusion, caused death, or had serious implications for the patient.
Statistical analysis The calculation of sample size was based on the expected incidence of MI. With an accumulated incidence of 10%, a presumed 35% risk reduction, a type I error of 0-05 (two-tailed), and a type II error of 0-20, the number of randomised patients was calculated to be 2000. Time to an endpoint was analysed by the Kaplan-Meier method and the logrank test, on the intention-to-treat principle. The chi-square test was used to compare frequencies of adverse effects. Prognostic factors and demographic data were analysed with respect to equal distribution between the two treatment regimens. The effect of co-variates was analysed according to the Cox proportional hazards model (Systat statistical package, survival
module). All statistical analyses were performed by independent of Bristol-Myers Squibb.
a
statistician
Results at randomisation for the two shown in table 1. Sex ratio, age, blood pressure, body weight, duration of angina pectoris, diabetes, smoking habits, treatment for hypertension, serum potassium, total cholesterol, and triglyceride levels, and concomitant use of calcium channel blocking drugs were virtually identical in the two groups.
The
groups
demographic data
are
Primary outcome event At the end of the follow-up period there were 205 patients with a primary endpoint, 81 in the aspirin + sotalol group and 124 in the placebo + sotalol group (p = 0-003, fig 1). Within the primary endpoint there was a significant decrease in non-fatal MI and
a
beneficial trend in sudden
TABLE II-ENDPOINTS
1423
TABLE III-REASONS FOR STOPPING TREATMENT PREMATURELY
TABLE V-BLEEDS I
*Requnng aspinn
*Figures m parentheses indicate fatal events Patients are listed by first bleeding event.
treatment
Comment. Patients are only mentioned by first event However, primary endpoints other reasons for cessation of treatment. CABG coronary artery bypass grafting, NSAID = non-steroidal anti-inflammatory
censor
drug
death in the aspirin + sotalol group, while the numbers of fatal MIs during hospitalisation were identical in the two groups
(table n).
Secondary outcome events The incidences of all secondary endpoints were lower in the aspirin + sotalol group (table II). Despite reductions of 22-32% only the 32% reduction in vascular events was statistically significant. Stroke developed in 66 patients (28 aspirin + sotalol, 38 placebo + sotalol). There were 59 nonhaemorrhagic strokes, 38 of which were verified by CT scan (23 aspirin + sotalol, 36 placebo + sotalol). The 7 haemorrhagic strokes were verified by CT scan or necropsy (5 aspirin + sotalol, 2 placebo + sotalol). 26 patients experienced transient ischaemic attacks (9 aspirin + sotalol, 17 placebo + sotalol; p 0-170). There was no difference in the number of coronary artery bypass procedures between the two groups (39 vs 40). =
Withdrawal of treatment, adverse events
patients stopped trial medication prematurely (362 aspirin + sotalol, 417 placebo + sotalol) (table III). The 6 protocol violations were caused by non-trial doctors prescribing medication not allowed in the study. There were 342 non-haemorrhagic adverse events, with an equal distribution between the two groups (table iv). 45 patients experienced two adverse events and 7 patients three. 43 patients had bleeds, 33 of which were major (table v). Of these, 14 were fatal (9 aspirin + sotalol, 5 placebo + sotalol). More than half of the major bleeds were in the gastrointestinal tract. 779
Analysis of risk factors In analysis of risk factors for primary endpoint event factors initially included in the model were treatment, sex, age (cut-off 60 years), treatment for hypertension, smoking, serum cholesterol (cut-off 6-5 mmol/1), serum triglycerides (2-0 mmol/1), heredity for MI, diabetes, duration of angina TABLE IV-NON-HAEMORRHAGIC ADVERSE EVENTS
*In 283 patients
I
pectoris (cut-off 2 years), heart rate (70/min) and dose of sotalol (159 mg per day). A variable was excluded if the relative risk was not significant at the 5% level. The following were found to be independent risk factors: treatment, sex, diabetes type n and serum cholesterol. The relative risks (RR) and 95 % confidence intervals were: F vs M, 0-44 (0-33-0-61); no diabetes vs diabetes 0-45 (0-270.72); aspirin + sotalol vs placebo + sotalol 0-68 (049-093); and serum cholesterol < 6.5 vs 6,5 mmol/l 0.70 > The RRs were (051-096). basically identical on trial stratification by medication-ie, the relative risks were the same in the two treatment groups.
Discussion In the prevention of cardiovascular events aspirin has been shown to be beneficial in a variety of primary and secondary prevention trials. The primary prevention of MI was demonstrated in the US Physicians Health Study. A significant 44% reduction in the risk of first infarction was observed in the group treated with aspirin 325 mg every other day.5 However, a trial in British male doctors of 500 mg aspirin daily showed no beneficial effect.9 A retrospective subgroup analysis in the US Physicians Health Study in 333 physicians with angina pectoris at the time of randomisation indicated greater than 50% reduction of risk of first MI."* Among the secondary prevention trials the second International Study of Infarct Survival showed a significant 45% reduction in 5 weeks reinfarction in the aspirin group (160 mg daily).z In unstable angina pectoris/ non-Q MI the RISC trial of 75 mg daily showed a 61 % reduction in MI and death at 90 days follow-up.3 In ISIS-3 a similar efficacy was also demonstrated in patients with developing MI treated with thrombolytic agents and aspirin." Thus, there appears to be a more favourable outcome in the primary as well as the secondary prevention trials with low doses of aspirin (75-160 mg daily). In SAPAT the primary endpoint, non-fatal MI and sudden death showed reductions of similar magnitude, while there was no difference in the rate of fatal MI during hospitalisation. Although the substantial reduction in sudden death was not significant, the results imply that aspirin may prevent sudden occlusion through inhibition of the proaggregatory and vasoconstrictive thromboxane Az. Other factors besides thromboxane Az, such as changes in sympathetic vascular tone and activation of platelet receptors, may also promote platelet aggregation.1z During the change from chronic stable angina to unstable angina and subsequent minor MI an increase in myocardial ischaemia may be initiated by platelet aggregation and vasoconstriction induced by local accumulation of thromboxane Az and by reductions in endothelially derived vasodilators and in inhibitors of platelet aggregation. 13
1424
Aspirin could thus have diminished the risk of developing unstable angina and subsequent MI by its effect on thromboxane Az. This would primarily inhibit the development of minor non-fatal MI capable of triggering electrical disorder, ventricular fibrillation, and sudden death. Major fatal infarctions are probably more often caused by subintimal plaque bleeding with subsequent rupture. 14,15 Platelet-active drugs cannot be expected to have an inhibitory effect on this plaque bleeding and rupture. When SAPAT was designed data on the incidence of MI in patients with chronic stable angina pectoris were based on patients not receiving beta-adrenoceptor blocking drugs, and those data suggested an incidence of 5 % or so per year. 16 The annual incidence of primary endpoints in our controls was lower at (3%). This discrepancy is compatible with the trend in Sweden towards a lower incidence of MI over the past decadel’ and may be related to a better diet and/or to the new drugs introduced for the treatment of angina.18 The standard treatment with sotalol in both groups may thus have contributed to the low incidence of primary endpoint. The selection criteria for the angina pectoris patients may also have had an important role. In 2035 patients followed up for 8485 patient-years in SAPAT the addition of aspirin 75 mg daily to standard sotalol reduced the incidence in the first MI or sudden death by 34% (p=0-003). The reduction of vascular events (including non-fatal MI, non-fatal stroke and vascular death) was 32% (p
