Comparison of Clinical Outcomes for Women and Men after Acute Myocardial Infarction

Methods

Patient Selection

Details of the TIMI-II protocol have been reported before [17, 18]. Briefly, women and men younger than 76 years who had ischemic chest pain lasting 30 minutes or more, in whom treatment with recombinant tissue plasminogen activator (rt-PA) (alteplase; Activase, Genentech, South San Francisco, California) within 4 hours of symptom onset was feasible, were considered for the study. Exclusion criteria included a history of cerebrovascular disease; blood pressure greater than 180 mm Hg systolic or 110 mm Hg diastolic; systemic bleeding disorders; major surgery within the previous 2 weeks; recent prolonged cardiopulmonary resuscitation; percutaneous transluminal coronary angioplasty or severe trauma within the previous 6 months; previous coronary artery bypass grafting or prosthetic heart valve replacement; left bundle-branch block; dilated cardiomyopathy; and other serious illness.

Thrombolytic Therapy

The rt-PA used in the TIMI-II study was produced by the suspension culture method (G11044, supplied by Genentech). The total dosage of rt-PA used in the first 520 patients was 150 mg administered intravenously during a period of 6 hours. Because of an unacceptably high rate of intracranial hemorrhage [19], however, the dosage was reduced to 100 mg given during 6 hours in the remaining patients.

Assigned Strategies and Beta-Blocker Treatment

Patients were assigned randomly to receive one of two treatment regimens: 1) routine coronary angiography done 18 to 48 hours after study entry and percutaneous transluminal coronary angioplasty or coronary artery bypass grafting if angiography showed that the patient's anatomy was suitable [invasive strategy]; or 2) conventional care without coronary angiography and percutaneous transluminal coronary angioplasty, unless evidence showed either spontaneous or exercise-induced myocardial ischemia (conservative strategy) [17]. Coronary artery bypass grafting was done in patients assigned to either group for appropriate clinical indications.

In the TIMI-IIA substudy [18], 195 patients were assigned to receive an immediate invasive strategy and were excluded from the analysis. Enrollment of patients in TIMI-IIB (the β-blocker substudy) [20] at the seven TIMI-IIA clinical sites did not begin until enrollment in TIMI-IIA was completed.

In TIMI-IIB, patients assigned to receive immediate intravenous β-blocker therapy were given 15 mg of metoprolol as three 5-mg intravenous injections at 2-minute intervals, followed by oral metoprolol. Those assigned to receive deferred β-blocker therapy received 50 mg of oral metoprolol twice on day 6 and 100 mg twice each day thereafter.

Concomitant Care

Patients received intravenous lidocaine for 24 hours and intravenous heparin for 5 days. Aspirin (80 mg) was administered on the day of study entry according to the protocol in the first 488 patients and on the next day in the remaining patients; it was increased to 325 mg per day on day 6, when intravenous heparin was replaced by subcutaneous heparin.

End Points

The primary end point in TIMI-II was survival with no recurrent myocardial infarction at 42 days. Secondary end points included the ejection fraction at rest and during exercise at hospital discharge and at 6 weeks. Complications of therapy were also assessed. Follow-up examinations were conducted 6 weeks and 1 year after study entry. The vital status of all patients not individually examined was determined by telephone interview.

Statistical Analysis

Probability values and confidence intervals for percentages and proportions were calculated using standard methods to test differences between two independent proportions [21, 22]. Because of multiple comparisons in the TIMI-II secondary analyses, probability values between 0.01 and 0.001 for two-sided tests were considered to provide some evidence of differences, and values less than 0.001 were thought to provide strong evidence of differences. Comparisons between mortality and either mortality or myocardial infarction rates are based on Cox regression analysis [23, 24]. The Cox proportional-hazards model was examined for validity of assumptions using the Kaplan-Meier [23] estimates of survival (S[t]) for each covariate in graphs plotting the log (−log[S(t)]) against time to events (death and death or myocardial infarction), and the assumptions were tested for the comparison of men with women by introducing time-dependent variables for interaction [24]. Adjustment, when appropriate, was made for assignment to conservative or invasive strategy, rt-PA dose, and the TIMI-II baseline patient characteristics (age, previous myocardial infarction, anterior myocardial infarction, history of diabetes, history of hypertension, time from onset of symptoms to study entry, ongoing chest pain when rt-PA infusion was begun, history of congestive heart failure, race, and history of angina). Crude event rates and Kaplan-Meier event rates for mortality and for death or myocardial infarction were almost identical because of the completeness of follow-up information. Tests of interaction were done using crude event rates and the Breslow-Day statistic [25]. This report is based on an analysis file prepared at the TIMI Coordinating Center in January 1991.

Results

Patient Characteristics

Women were older than men, with mean ages of 62.2 and 56.6 years (P < 0.001), respectively, and more commonly had past medical histories that included congestive heart failure (P < 0.001), systemic hypertension (P < 0.001), or diabetes mellitus (P < 0.001). As a group, women were less likely than men to be classified as low risk (P < 0.001). The time from symptom onset to study entry was delayed in women compared with men (2.8 hours compared with 2.6 hours; P < 0.001). Fewer women than men were considered eligible for the β-blocker study (P < 0.001) (Table 1).

Invasive Procedures and Medical Treatments

The frequency of invasive procedures done within 6 weeks of study entry is summarized in Table 2. No differences between men and women were found in either the invasive or conservative treatment groups. Throughout this period, women were less likely than men to receive β-blockers, but calcium channel blockers were prescribed more frequently for women. Aspirin was used with similar frequency.

Complications

Coronary Angioplasty

Although coronary arterial dissection was observed more frequently in women than in men (17.2% compared with 10.0%; P = 0.002), within 24 hours after percutaneous transluminal coronary angioplasty, the frequencies of recurrent myocardial infarction, death, or the need to do emergent coronary bypass surgery did not differ statistically between women and men assigned to receive invasive procedures. Sex was not associated with the frequency of successful percutaneous transluminal coronary angioplasty [26].

Left Ventricular Function

More of the 597 women in the TIMI-II study had resting left ventricular ejection fractions greater than 55% on radionuclide ventriculography than did the 2742 men before hospital discharge (33.5% compared with 29.0%; P = 0.03) and at 6-week follow-up (32.8% compared with 27.1%; P = 0.005). However, a greater proportion of these women than men did not have radionuclide studies analyzed because of death or inability to measure ejection fraction at hospital discharge (24.0% compared with 16.6%; P < 0.001) and 6-week follow-up (34.2% compared with 25.5%; P < 0.001).

Patient Outcome: Mortality and Morbidity

One-year follow-up data were available for 3316 (99.3%) patients [27]. Event rates according to sex and treatment strategy are summarized in Table 3. The cumulative 6-week mortality rate was higher for women than for men (9% compared with 4%; P < 0.001). Combined reinfarction and death also was more common among women than men (15.9% compared with 9.5%; P < 0.001). These differences (for both the invasive and conservative treatments) persisted at 1-year follow-up (Figures 1 and 2).

The occurrence of myocardial infarction or death was predictably lower for low-risk women and men. The mortality rate was 1.9% for women compared with 1.6% for men (P = 0.78) 6 weeks after study entry, and it was 3.1% compared with 2.5% (P = 0.66) 1 year after study entry. Among high-risk patients, the mortality rate was higher for women at 6 weeks than for men (11.7% compared with 5.4%; P < 0.001) and 1 year after study entry (15.6% compared with 7.9%; P < 0.001). Although the differences between men and women who were at low risk for an unfavorable outcome appear to be smaller than the differences among those who were not at low risk, no evidence was found for a relation between sex and risk stratum (for 1-year mortality, interaction P = 0.30, and for myocardial infarction or death by 1 year, interaction P = 0.62). Combined myocardial infarction and death occurred 59% more often among low-risk women than among low-risk men (10.0% compared with 6.3%; P = 0.09) 6 weeks from study entry and 44% more often (14.4% compared with 10.0%; P < 0.001) 1 year after study entry. For high-risk patients, the risk was 18.1% compared with 11.2% (P < 0.001) at 6 weeks and 24.0% compared with 15.4% 1 year after study entry (P < 0.001).

Among patients assigned to receive 100 mg of rt-PA and to the conservative strategy, the 6-week mortality rate was 7.5% for women and 3.8% for men (P = 0.01), and the 6-week incidence of reinfarction or death was 14.2% for women and 8.9% for men (P = 0.01). One year after study entry, the mortality rate was 11.1% among women and 6.5% among men (P = 0.01), and death or myocardial infarction occurred among 20.6% of women and 13.5% of men (P = 0.004).

In every decade of age, the frequency of death and combined reinfarction and death at 1-year follow-up was greater for women than men. On testing Cox model assumptions of proportional hazards, no significant interactions were found for time to event and sex for mortality (P = 0.26) and for the combined end point of myocardial infarction and death (P = 0.29). In a Cox regression model adjusted for baseline characteristics (see Methods and Table 1), the relative risk for death by 1 year among women was 1.39 (P = 0.03) compared with a relative risk for reinfarction or death of 1.28 (P = 0.02) for men (Table 4). The adjusted relative risks for death at 6 weeks and 1 year for women (compared with men) assigned to the conservative regimen were 1.51 (99% CI, 0.72 to 3.16) and 1.22 (CI, 0.67 to 2.22), respectively; the relative risks for reinfarction or death were 1.30 (99% CI, 0.77 to 2.18) and 1.23 (99% CI, 0.80 to 1.90), respectively. Cardiovascular disease accounted for similarly large percentages of the deaths occurring among men (129 of 163 or 79.1%) and women (53 of 72 or 73.6%).

Discussion

In the Framingham Heart Study [7], a longitudinal investigation involving 5127 women and men initially free of coronary heart disease, the prognosis after myocardial infarction for women was distinctly worse than for men. The Multicenter Investigation for the Limitation of Infarct Size investigators examined the influence of sex and race on prognosis after myocardial infarction [9]. The cumulative mortality rate at 48 months was 36% for women and 21% for men. Our unadjusted findings are consistent with those in the Framingham Study and in the Multicenter Investigation for the Limitation of Infarct Size.

The poor prognosis for women with coronary heart disease after an acute myocardial infarction has sometimes been attributed to advanced age or concomitant medical illnesses [8, 10, 11, 13, 28-31]. Our findings also suggest that women having myocardial infarction are typically older and have a higher prevalence of previous congestive heart failure, systemic hypertension, and diabetes mellitus than do their male counterparts. Thus, these differences account for much but probably not all of the observed sex-related differences in clinical outcome.

Influence of Sex on Prognosis

Angiographic observations have shown that coronary artery reperfusion and subsequent reocclusion rates are similar for women and men (TIMI-I and TIMI-IIA, unpublished data). Similarly, women have an overall reduction in early mortality, as do men, after thrombolytic therapy [32, 33]. Mueller and colleagues [34] reported no differences in the outcomes for men or women treated with an invasive strategy compared with a conservative strategy. In the GISSI-I (Gruppo Italiano per lo Studio della Streptochinasi nell'infarto Miocardico) trial [32, 33], 2313 women were included. Women receiving intravenous streptokinase (1.5 MU) (n = 1157) had a lower 21-day mortality rate compared with women given placebo (18.5% compared with 22.6%). The mortality rate among women treated with streptokinase, however, was twice that for men (18.5% compared with 8.8%, respectively). In the Anglo-Scandinavian Study of Early Thrombolysis [35], women treated with rt-PA and heparin (n = 577) had a 21% relative reduction in mortality rate compared with a 28% relative reduction for men. Similarly, whereas both women and men included in the Second International Study of Infarct Survival [36] benefited from thrombolysis and platelet inhibition, the relative reductions in mortality rates after treatment with streptokinase, aspirin, or their combination were greater for men than for women (27% compared with 15%, 23% compared with 17%, and 44% compared with 30%, respectively).

Investigators in the International Study of rt-PA and streptokinase [37] did not observe differences in morbidity for women or men relative to thrombolytic agent used but did find differences in mortality rates between women and men. Women treated with rt-PA had an in-hospital mortality rate of 13.6% compared with 7.6% for men. Similarly, the early mortality rate for women receiving streptokinase was more than twice that for men—15.4% and 6.7%, respectively.

The differences we found in the medical management of women compared with men, that is, the use of fewer β-blocker agents and more calcium channel antagonists (see Table 1), warrant further investigation. Perhaps women are more vulnerable to the adverse effects of β-blocker agents, or perhaps this useful class of medications [38] is not used enough in women. Although they probably did not occur by chance, these differences are too small to account for the differences in mortality rate observed after 1 year of follow-up.

Among patients with presumed coronary heart disease, previous studies found that women with abnormal results on exercise tolerance tests are less likely to be referred for diagnostic coronary angiography than are men [39]. Further, they are typically referred for revascularization at a more advanced stage of disease than are men, which contributes directly to their higher operative and periprocedural rates of mortality and morbidity [40]. Two studies have brought additional attention to sex differences in case management [41, 42]. Whether the observed practice and referral patterns represent overuse of invasive procedures in men or underuse in women is not clear [43, 44].

Because of potential referral bias for women and men with coronary artery disease [39-45], TIMI clinical centers may have identified men with a wide range of illness severity but only those women who were more seriously ill. Because study eligibility criteria would tend to make recruited women and men more similar to each other, the influences of patient selection cannot be directly evaluated.

Women enrolled in TIMI-II were older and had a history of diabetes more frequently than did men, and women were at higher risk for an unfavorable outcome than were the men in the study. Most of the deaths in TIMI-II occurred among high-risk patients.

Left Ventricular Function: Prognosis

Previous studies have shown that women who have myocardial infarction are at increased risk for congestive heart failure, despite preserved systolic function as assessed by radionuclide angiography [46]. As with men, women with congestive heart failure have poor short-and long-term prognoses [47]. Women have a greater ejection fraction at rest than do men, yet they have a decreased ejection fraction in response to exercise that is unexplained by differences in exercise capacity [48]. In TIMI-II, women were more likely than men to have resting ejection fractions greater than 55%, but the test was done on fewer women. Further research on effects of sex on ventricular function after infarction is needed.

Coronary heart disease is the leading cause of death among women and men in the United States and other industrialized nations. Although mortality rates have decreased progressively in the past 30 years, studies have suggested that women still have a worse prognosis than men after myocardial infarction [49]. This difference can be explained, in large part, by their older age and the presence of underlying medical conditions such as hypertension or diabetes mellitus [50]. To determine the contribution, if any, of sex-specific differences requires further studies of clinical outcome after myocardial infarction in patients for whom referral bias is excluded.