OUP user menu

Do men benefit more than women from an interventional strategy in patients with unstable angina or non-ST-elevation myocardial infarction? The impact of gender in the RITA 3 trial

T.C. Clayton, S.J. Pocock, R.A. Henderson, P.A. Poole-Wilson, T.R.D. Shaw, R. Knight, K.A.A. Fox
DOI: http://dx.doi.org/10.1016/j.ehj.2004.07.032 1641-1650 First published online: 2 September 2004

Abstract

Aims The RITA 3 trial randomized patients with non-ST-elevation myocardial infarction or unstable angina to strategies of early intervention (angiography followed by revascularization) or conservative care (ischaemia or symptom driven angiography). The aim of this analysis was to investigate the impact of gender on the effect of these two strategies.

Methods and results In total, 1810 patients (682 women and 1128 men) were randomized. The risk factor profile of women at presentation was markedly different to men. There was evidence that men benefited more from an early intervention strategy for death or non-fatal myocardial infarction at 1 year (adjusted odds ratios 0.63, 95% confidence interval 0.41–0.98 for men and 1.79, 95% confidence interval 0.95–3.35 for women; interaction p-value=0.007). Men who underwent the assigned angiogram were more likely to be put forward for coronary artery bypass surgery, even after allowing for differences in disease severity.

Conclusion An early intervention strategy resulted in a beneficial effect in men which was not seen in women although caution is needed in interpretation. Further research is needed to evaluate why women do not appear to benefit from early intervention and to identify treatments that improve the prognosis of women.

See page 1559 for the editorial comment on this article (doi:10.1016/j.ehj.2004.07.002).

Keywords Gender; Unstable angina; Non-ST-elevation myocardial infarction; Interventional strategy; Conservative strategy; Coronary revascularization

Introduction

Controversy exists over differences in revascularization treatment and outcomes of men and women with acute coronary syndromes, and there are conflicting data from randomized trials. Several studies have assessed the outcomes of men and women with coronary artery disease following coronary artery bypass grafting surgery (CABG)1–6 or percutaneous coronary intervention (PCI)3,7–14 with many of these reporting worse outcomes in women. This could partially explain differences in the treatment of men and women with acute coronary syndromes that have been reported.15–19 However, after adjustment for other factors, similar outcomes in men and women have been reported.20–23 In particular, in establishing predictors of death, myocardial infarction or severe recurrent ischaemia for the TIMI risk model20 and of in-hospital mortality for the GRACE model,21 gender did not emerge as an independent risk factor.

In assessing the impact of an early intervention strategy in men and women with unstable coronary disease, the results from two recent clinical trials appear to lead to conflicting conclusions. The FRISC II trial (Fragmin and Revascularisation during Instability in Coronary artery disease II) trial indicated a benefit of early intervention for death or myocardial infarction in men which was not seen in women.24 The TACTICS-TIMI 18 trial (Treat angina with Aggrastat and determine Cost of Therapy with an Invasive or Conservative Strategy – Thrombolysis In Myocardial Infarction 18) reported a benefit of early intervention for death, myocardial infarction or re-hospitalisation that was seen in both men and women.25 In addition, a recent observational study in patients with non-ST-elevation acute coronary syndromes assessing the impact of very early intervention reported a better long-term outcome of death or myocardial infarction in women compared to men.26 The impact of gender on the treatment and outcome of patients with non-ST-elevation myocardial infarction or unstable angina therefore remains uncertain and RITA 3 (Randomized Intervention Trial of unstable Angina) provides the opportunity for further insight as to whether such gender differences exist.

The results from the RITA 3 trial indicated that patients benefited from an early intervention strategy in comparison to a conservative strategy largely due to a reduction in the incidence of refractory angina.27 A subgroup analysis suggested that men benefited from a strategy of early intervention in terms of the primary outcomes of (i) death, myocardial infarction or refractory angina at 4 months, and (ii) death or myocardial infarction at 1 year, whereas no such benefit was seen in women. In this paper we test the hypothesis that the differences in the benefit of early intervention between women and men are due to differences in the baseline clinical characteristics and disease severity. In addition, differences in the management of men and women randomized to early intervention will be explored.

Methods

The design of RITA 3 has been reported previously.27 Briefly, between Nov 1997 and Oct 2001, 1810 patients were recruited from 45 hospitals in England and Scotland. Multi-centre national ethics committee approval and local ethics committee approval were obtained. Patients were considered for the trial if they had suspected cardiac chest pain at rest and documented evidence of coronary artery disease, and if the participating cardiologist deemed either interventional or conservative strategies acceptable treatment options. The exclusion criteria are detailed in the main clinical report. All eligible patients who consented to participate were randomized within 48 h of the index episode of suspected cardiac chest pain to initial interventional or conservative treatment strategies.

It was intended that patients randomized to the intervention group underwent coronary angiography within 72 h of randomization. Patients randomized in centres without on-site angiographic facilities were transferred to an intervention centre. The need for revascularization in those undergoing early angiography was determined by the clinician responsible. The protocol recommended revascularization in major epicardial vessels with ⩾70% diameter stenosis (⩾50% for left main stem).

Patients randomized to the conservative treatment strategy were managed with antianginal and antithrombotic medication with the objective of providing the best medical treatment available at the time of randomization. Coronary angiography could be undertaken for failure of the conservative treatment strategy. During the index admission this constituted at least one of (i) recurrence of ischaemic pain at rest or on minimal exertion, with transient or persistent electrographic evidence of ischaemia despite full antianginal medication, or (ii) the inability to withdraw intravenous antianginal or antithrombotic medication without recurrence of ischaemic chest pain. Following discharge from the index hospital admission, failure of the conservative treatment strategy included exertional angina despite optimal antianginal medication.

Patients were seen at 4 months and 1 year, and follow-up is to be maintained annually to 5 years following randomization.

Q-wave myocardial infarction was diagnosed when new Q-waves were detected on a follow-up electrocardiogram when compared with the baseline electrocardiogram. Other myocardial infarctions were diagnosed when a clinical event was accompanied by electrographic evidence of acute myocardial infarction associated with a rise in at least one cardiac enzyme to twice the upper limit of normal. In patients for whom baseline levels of enzymes were already raised, a further twofold increase beyond the first 24 h following randomization was required.

Refractory angina was diagnosed during the index admission if recurrence of chest pain was associated with electrographic evidence of myocardial ischaemia, and led to a revascularization procedure within 24 h of the onset of pain. Following discharge from the index hospital admission, refractory angina was diagnosed if the patient was readmitted with an episode of cardiac chest pain associated with electrographic evidence of myocardial ischaemia. Episodes of refractory angina occurring up to 1 year after randomization were reported.

All patients provided written informed consent and RITA 3 complies with the Declaration of Helsinki.

Statistical methods

All analyses were performed by intention to treat except for analyses based on patients who underwent randomized coronary arteriography. Differences in baseline characteristics and disease severity between women and men were compared with t-tests and χ2 tests. The main outcome considered was death or non-fatal myocardial infarction at 1 year, with further analyses investigating death or non-fatal myocardial infarction over all available follow-up at the date the last randomized patient had been in the trial for 1 year. Logistic regression was used to examine gender differences in outcomes at 1 year (including interaction tests) and results are presented as odds ratios together with 95% confidence intervals (CI). Adjustment was made for baseline characteristics that were independently associated with the outcome (from a forward stepwise procedure). Cox regression was used for evaluating effects of gender and treatment on the outcome of death or non-fatal myocardial infarction over follow-up beyond 1 year and Kaplan–Meier graphs are presented for women and men separately.

In addition the outcomes of death alone and death, myocardial infarction or refractory angina at 1 year were also considered. In assessing the impact of disease severity on planned revascularization procedures among patients randomized to the intervention arm, cross-tabulations were made of the proposed management plan by disease severity in men and women separately. Men and women were also stratified according to risk on the basis of available factors included in the TIMI risk score20 to assess whether these characteristics explained any gender differences. Finally, the benefits and hazards of intervention were investigated in men and women separately according to their body mass index levels.

Results

Baseline characteristics

Of 1810 subjects randomized into RITA 3, 682 (38%) were women. There were some clear differences in baseline characteristics between women and men at presentation (Table 1). Women were, on average, slightly older than men, with more severe angina and higher cholesterol levels. Although men had higher diastolic blood pressure levels, women were more commonly treated for hypertension. More men had a previous myocardial infarction and were smokers and men more frequently had raised serum creatinine kinase or troponins. There were no differences in body mass index, prevalence of diabetes or ST depression on the baseline electrocardiogram.

View this table:

Table 1. Gender distribution of baseline characteristics

Men (n=1128)Women (n=682)p-valuea
ap-value from comparison of proportions or t-tests for continuous variables unless otherwise stated.
b Trend test.
c Treated with drugs.
Age (years)Mean (SD)61.8(10.5)63.3(9.9)0.002
Body mass index (kg/m2)Mean (SD)27.7(4.6)28.1(5.4)0.14
Blood pressure (mmHg)
SystolicMean (SD)132(22)131(23)0.38
DiastolicMean (SD)75(14)71(13)<0.001
Heart rate (beats/min)Mean (SD)68(13)70(14)<0.001
Total cholesterol (mmol/l)Mean (SD)5.5(1.1)5.8(1.3)<0.001
>5.5(%)475(43)378(57)<0.001
Smoking status (%)Never197(17)220(32)<0.001b
Ex-smoker540(48)266(39)
Current smoker390(35)196(29)


Medical history (%)Hypertensionc337(30)295(43)<0.001
Previous myocardial infarction352(31)149(22)<0.001
Diabetes147(13)97(14)0.47
Hypercholesterolaemiac345(31)234(34)0.095


Medication (%)Beta-blocker832(74)467(68)0.016
Calcium antagonist384(34)245(36)0.42
Nitrate (oral or intravenous)554(49)362(47)0.1
Potassium channel activator90(8)59(9)0.61
Aspirin1031(91)631(93)0.4
ACE-inhibitor197(17)130(19)0.39
Statin496(44)314(46)0.39
Angina grade in 4 weeks prior to interview (%)Any angina pre-admission750(67)485(71)0.043
Grade 3/4 pre-admission360(32)289(42)<0.001
Electrocardiographic evidence of myocardial ischaemia (%)ST depression (⩾0.1 mV)406(36)254(37)0.59
ST elevation (⩾0.1 mV)98(9)41(6)0.038
Ischaemic T-wave abnormalities787(70)511(75)0.02
Left bundle branch block35(3)29(4)0.2
Q-waves suggesting previous myocardial infarction (%)247(22)106(16)0.001
Evidence of CAD on previous arteriogram (%)131(12)49(7)0.002
Creatinine kinase (%)Total(n=1026)(n=618)
Raised levels222(22)77(12)<0.001
Troponins (%)Total(n=148)(n=63)
Raised levels117(79)41(65)0.032

Revascularization procedures

Of the 895 patients randomized to intervention, 865 (97%) underwent the assigned coronary angiogram within a median of 2 days. Significant disease was taken as ⩾50% diameter stenosis in the left main stem or ⩾70% diameter stenosis in any major epicardial vessel. Men tended to have more severe coronary artery disease (Table 2). Among the women, 37% had no significant coronary disease compared to only 12% of men. Over a quarter of men had significant stenosis in three vessels compared to 16% of women. Overall, 32% of men had significant disease in the proximal left anterior descending coronary artery compared to 21% of women.

View this table:

Table 2. Gender distribution of disease severity and management plan in patients in the Intervention arm

Men (n=545)Women (n=350)p-value
a LMS is taken as affecting the LAD and the CFX.
b Trend test.
c Revascularization procedure vs no revascularization procedure.
Patients undergoing randomized angiography522343
Significant disease (%)
LMS36(7)12(4)0.032
LAD327(63)156(46)<0.001
Proximal LAD165(32)73(21)0.001
CFX253(49)99(29)<0.001
RCA266(51)119(35)<0.001


Number of vessels with significant disease (%)a
060(12)126(37)<0.001b
1177(34)103(30)
2148(29)60(18)
3134(26)53(16)


Management plan (%) (among those with at least 1 diseased vessel)
CABG153(33)40(19)0.063c
PCI200(44)111(52)
Medical106(23)64(30)

LMS, left main stem; LAD, left anterior descending; CFX, circumflex; RCA, right coronary artery.

Cross-tabulations of the number of diseased vessels by the proposed management in men and women separately indicate gender differences in the intended treatment plan (Table 3). A PCI was planned in a similar proportion of men (39%) and women (33%) in the intervention group, but CABG was planned for 30% of men compared to 12% of women. If differences in disease severity are ignored, women were less likely to be put forward for a revascularization procedure with an unadjusted odds ratio (women vs. men) of 0.37 (95% CI 0.28–0.49). This compares to an odds ratio of 0.69 (95% CI 0.48–1.00) after adjusting for differences in the number of diseased vessels. For those patients with one-vessel disease, there was little difference in the proportion of men and women with a planned revascularization procedure (almost entirely PCI). In addition, for those with two and three-vessel disease, the proportion of men and women where the planned management was PCI was also similar. However, women with two and three-vessel disease were less likely than men to be put forward for CABG. In patients with two-vessel disease, 22% of women compared to 34% of men had a planned management of CABG, and in patients with three-vessel disease these percentages were 51% and 72% for women and men, respectively. After adjustment for the disease severity (including the number of diseased vessels, and disease of the left main stem and proximal left anterior descending, which were all independent predictors of undergoing CABG), the odds ratio of being put forward for a CABG procedure (women vs. men) was 0.43 (95% CI 0.26–0.72) (compared to the unadjusted odds ratio of 0.32 [95% CI 0.22–0.46]). The above results were very similar after further adjustment for body mass index. Further, left ventricular function using a wall motion score to group patients did not explain these differences with substantially more men than women put forward for CABG in each left ventricular score grouping.

View this table:

Table 3. Proposed management plan by number of significantly diseased vessels and gender

Proposed planNumber of significantly diseased vesselsTotal
0123
a All three patients with no significantly diseased vessel put forward of PCI had at least one vessel with ⩾50% stenosis.
b Of the six patients put forward for CABG with one significantly diseased vessel, three had a diseased left main stem of 40%, one had extensive disease throughout the left anterior descending (LAD), and two had a stenosis of ⩾90% in the LAD (including one with a 50% stenosis in the right coronary artery).
MenCABG0(0)6b(3)50(34)97(72)153(29)
PCI2a(3)121(68)66(45)13(10)202(39)
Medical58(97)50(28)32(22)24(18)164(32)
Total patients60177148134519


WomenCABG0(0)0(0)13(22)27(51)40(12)
PCI1a(1)73(71)31(53)7(13)112(33)
Medical125(99)30(29)15(25)19(36)189(55)
Total patients1261035953341

Similar patterns were seen when the proposed management plan was tabulated by the site of significantly diseased vessels (not shown). For example, approximately half of the men with a significantly diseased proximal left anterior descending were put forward for CABG compared to a quarter of women.

The impact of gender on the primary outcome

For the primary outcome of death or non-fatal myocardial infarction at 1 year there was evidence that the effect of an early intervention strategy compared to a conservative strategy differed between men and women (Table 4). There were 38 (7.0%) men in the intervention arm experiencing death or myocardial infarction by 1 year compared to 59 (10.1%) men in the conservative arm, whereas among women 30 (8.6%) in the intervention arm compared to 17 (5.1%) in the conservative arm experienced this outcome (unadjusted interaction p-value=0.011). In order to consider the impact of the baseline characteristics of patients, a multivariable analysis was conducted. Factors independently associated with an increased risk of the outcome were age, ST depression (⩾0.1 mV), heart rate, previous myocardial infarction, diabetes and severe angina. After adjustment for these factors, the odds ratio for men (intervention vs. conservative) was 0.63 (95% CI 0.41–0.98) and for women 1.79 (95% CI 0.95–3.35) (interaction p-value=0.007). In the conservative arm, women had a decreased risk compared to men (adjusted odds ratio 0.39, 95% CI 0.22–0.70) whereas in the intervention arm the risks were similar (adjusted odds ratio 1.11, 95% CI 0.66–1.85). These results were very similar after further adjustment for body mass index.

View this table:

Table 4. Outcomes of death, myocardial infarction or refractory angina within 1 year of randomization

OutcomeIntervention (n=895a)Conservative (n=915a)Adjusted odds ratio (95% CI)Interaction p-valueb
a Intervention group: men n=545, women n=350. Conservative group: men n=583, women n=332.
b The p-values are from an interaction test assessing whether the effect of treatment differs according to gender.
Deaths (%)
Men23(4.2)28(4.8)0.78(0.44–1.41)0.031
Women18(5.1)8(2.4)2.43(1.01–5.84)


Myocardial infarctions
Men1934
Women1510


Refractory angina
Men3371
Women2635


Death or myocardial infarction (%)
Men38(7.0)59(10.1)0.63(0.41–0.98)0.007
Women30(8.6)17(5.1)1.79(0.95–3.35)


Death, myocardial infarction or refractory angina (%)
Men67(12.3)125(21.4)0.48(0.34–0.67)0.002
Women55(15.7)46(13.9)1.14(0.74–1.76)

Differences were also apparent with the addition of refractory angina to the endpoint of death or myocardial infarction at 1 year with an estimated halving of the odds in men whereas there was little evidence of an impact of intervention in women. After adjustment, the odds ratio for men (intervention vs. conservative) was 0.48 (95% CI 0.34–0.67), and for women 1.14 (95% CI 0.74–1.76) (interaction p-value=0.002).

There was also an indication of differences in death or myocardial infarction between men and women over the longer term (Fig. 1) with an adjusted hazard ratio for men of 0.61 (95% CI 0.44–0.85) and for women 1.09 (95% CI 0.70–1.71) (interaction p-value=0.042). The Kaplan–Meier curves display the observed treatment differences for death or myocardial infarction for all available follow-up at the date the last patient randomized had been in the trial for 1 year, and show the apparent benefit of intervention in men which is not seen in women.

Fig. 1

Cumulative risk of death or myocardial infarction by gender.

Risk stratification and outcome

Although the TIMI risk score was developed for the outcome of death, myocardial infarction or recurrent ischaemia within 14 days, it was also shown to be predictive for each of the components of the combined outcome, and for death or myocardial infarction.20 The majority of risk factors used in constructing the TIMI risk score are available for RITA 3 patients including age, risk factors for coronary artery disease (except family history), ST-deviation, and raised cardiac markers. In addition aspirin at randomization and severity of angina prior to the index event were also available. Details of the extent of the coronary disease were only available for those patients randomized to the intervention arm and so are not included in this analysis. Patients were divided into lower, moderate and higher risk categories with approximately a third in each according to their derived risk score.

The effect of treatment was examined in men and women separately for each category to explore whether this accounted for the observed gender differences. For both men and women there was an increase in the event rates for each increase in the risk score (from 5.6% to 14.4% in men, and from 3.0% to 10.0% in women). In those with lower risk the event rates were very similar in the intervention and conservative arms for both men (6.1% vs. 5.1%) and women (2.3% vs. 3.8%). However in men with moderate and higher risk scores, those in the intervention arm had a lower event rate than those in the conservative arm (5.4% vs. 9.5% and 10.3% vs. 17.9% in the moderate and higher risk groups respectively), whereas women in the intervention arm had higher rates (13.4% vs. 3.4% and 11.7% vs. 8.2% in the two risk groups respectively). The results indicate that in men the benefit of intervention was in patients at moderate or higher risk, whereas this pattern was not observed in women where no benefit was seen even in the higher risk patients.

The effect of the randomized intervention was examined by body mass index group (with cut-offs at 25 and 30 kg/m2) in men and women separately. The results suggested any benefit of intervention in men was consistent across all body mass index groups with no benefit of intervention seen in any body mass index group for women.

The timing of outcome events

Among men, the event rate of death or myocardial infarction in the first 7 days following randomization was similar in the 2 treatment groups (3.44 per 1000 days in the intervention arm compared to 3.73 per 1000 days in the conservative arm), whereas after the first 7 days the event rate is consistently lower in the intervention arm. Eight of the 13 events in the first 7 days occurring among men in the intervention arm were related to the randomized angiogram or assigned PCI. Two of the 15 events by 7 days occurring among men in the conservative arm were related to a PCI. Among women the event rate in the first 7 days was increased in the intervention arm (5.85 per 1000 days) compared to the conservative arm (1.30 per 1000 days), whereas after the first 7 days the rates in the two arms were remarkably similar. Nine of the 14 events in the first 7 days among women in the intervention arm were related to the randomized angiogram or assigned PCI. None of the three events among women in the conservative arm were related to a procedure. There were seven CABG-related deaths in the intervention arm (five men, two women) in the first year of follow-up compared to three CABG-related deaths (all men) and one CABG-related myocardial infarction (in a male) in the conservative arm (Table 5).

View this table:

Table 5. Procedure-related events within 1 year of randomization

Procedure-related eventMenWomen
Intervention (n=545)Conservative (n=583)Intervention (n=350)Conservative (n=332)
Deaths
Randomized angiogram02
PCI2010
CABG5320
None1625138


Total2328188


Myocardial infarctions
Angiogram3010
PCI5260
CABG0100
None1131810


Total19341510

Outcome by treatment received

In order to assess the impact of revascularization on event rates, an exploratory analysis was undertaken excluding those events that occurred within 3 days of a revascularization procedure. Although the numbers are small, the event rates among men and women in the intervention arm within the first 7 days were reduced by approximately a third (2.13 per 1000 days in men and 3.78 per 1000 days in women). Excluding events within 3 days of a revascularization procedure had little impact on rates in the conservative arm.

In patients randomized to the intervention group, 19 (36%) women with three-vessel disease were treated medically compared to 24 men (18%). Among these, patients five women and five men experienced an outcome of death or myocardial infarction by 1 year (nine of these 10 patients died within a year of randomization). None of these 10 patients underwent a revascularization procedure.

Thus the difference in the effects of the randomized treatment strategies between men and women reflects, in part, a low early event rate in women in the conservative arm, and a high early event rate in the interventional arm that cannot fully be explained by events related to revascularization procedures. Further, the differences in the event rates between men and women in the intervention arm were enhanced after excluding those with no significant disease. At 1 year in patients with one, two and three significantly diseased vessels there were 8 (4.5%), 13 (8.8%), and 15 (11.2%) men respectively who died or had a myocardial infarction compared to 7 (6.8%), 9 (15%), and 10 (18.9%) women.

Discussion

In RITA 3, women and men differed with respect to a number of important characteristics at admission for the index event. Furthermore, men who were randomized to intervention had more severe coronary artery disease than women (based on percentage diameter stenosis) both in terms of the location of the disease and in the number of diseased vessels. These findings are consistent with those reported in FRISC II and TACTICS-TIMI 18.24,25 The benefit of an early interventional strategy in RITA 3, in terms of the primary outcome of death or myocardial infarction at 1 year, was observed in men but not in women. There are several reasons why such gender differences may have arisen and these are detailed below.

Disease severity

It has been suggested that the benefit of early intervention in patients with unstable angina or non-ST-elevation myocardial infarction is greatest in those with more severe coronary disease,25 whereas women in RITA 3 may represent a particularly low risk group. A strategy of early intervention will identify those with more severe disease earlier than a conservative approach, and these patients may benefit from early revascularization. In RITA 3, a patient could be randomized if the participating cardiologist was satisfied that either strategy was an acceptable option and as a result the patients in RITA 3 had a lower risk profile than in earlier trials. In RITA 3 women presented with worse symptoms than men but after adjustment for baseline risk factors, gender differences in outcomes remained. In addition, there were no significant differences between men and women in the TIMI risk score based on the risk factors available. This did not change after the inclusion of ⩾50% diameter stenosis as an additional component among patients in the intervention arm. In contrast to men, women in the moderate and higher risk TIMI groups showed no benefit of an early intervention.

It has been suggested recently that men and women with unstable angina or non-ST-elevation myocardial infarction have different levels of biomarkers at presentation, and that marker-positive women had improved outcomes with early intervention, whereas marker-negative women had improved outcomes with a conservative strategy.28 In RITA 3, men presented with raised levels of creatinine kinase more often than women although the overall number of patients with raised levels was low.

Complications from revascularization procedures

Women may experience more complications from invasive treatment. Previous studies have reported that target vessel size (or measures of body size used as a surrogate) help to explain the differences in outcomes between women and men following CABG or PCI,4,29 although a recent study combining data from five trials and one registry reported that a small target vessel did not account for the higher mortality seen in women following PCI.7 In RITA 3, body mass index (used as a surrogate for vessel size) did not explain the apparent gender differences with men showing a consistent benefit from intervention across all body mass index groups and no benefit was seen in women in any group. Although based on a fairly small number of events, there is a suggestion that some, but not all, of the excess risk in women randomized to early intervention relates to an increased procedure-related risk in the first week together with a low event rate in the conservative arm. However, after excluding events within 3 days of a revascularization procedure, an early excess risk among women in the intervention arm was still apparent.

Referral bias

In patients with comparable disease severity, there may be a gender bias in referral for angiography, PCI or CABG. Having adjusted for the number of significantly diseased vessels there was an indication that men with two- and three-vessel disease were more likely than women to be put forward for CABG (52% vs. 36%), whereas for PCI there was little evidence of a difference (28% vs. 34%). This may represent a less aggressive approach for treating women, particularly women with small coronary arteries. Women with two- or three-vessel disease may more commonly represent a particularly high risk group of patients with small coronary arteries. Among the patients with two- or three- vessel disease, 34 women (30%) were treated medically compared to 56 men (20%). Of these patients, seven women (21%) experienced the primary outcome compared to six men (11%).

Limitations and comparison with other trials

The results may have arisen by chance (or at least be an exaggeration of the true impact of gender on the effect of early intervention). Caution is needed in the interpretation of any differences arising from subgroup findings especially as trials are rarely powered for such analyses.30 In addition, detailed analyses of the possible explanations for the differences observed are post hoc and limited by the low numbers. One explanation of the lack of benefit of early intervention in women could be the very low event rate observed in the conservative group, possibly due to a high proportion of women in this group with normal or near normal coronaries. However, there are several factors which suggest that gender differences may exist in this population of patients. The interaction between gender and treatment was one of only five subgroup analyses undertaken for the primary endpoint of death or myocardial infarction within 1 year of randomization.27 Moreover, we found a gender difference for death alone, and also with the inclusion of refractory angina to the primary outcome. In addition there was some evidence that men benefited more from early intervention in terms of any symptoms of angina at 1 year (interaction p-value=0.027) but not in terms of severe symptoms (interaction p-value=0.23).

In addition, the results are consistent with those of FRISC II which also observed a benefit of early intervention on death or myocardial infarction in men alone. While TACTICS-TIMI-18 demonstrated a benefit of early intervention in both men and women, there are possible explanations for the differences seen with RITA 3 and FRISC II. In TACTICS-TIMI-18, by design, patients were routinely treated with a glycoprotein IIb/IIIa and this is suggested as a reason for the improved outcome among women receiving invasive therapy. 25 However, there is some uncertainty as to the benefit of routine treatment with glycoprotein IIb/IIIa in women with non-ST-elevation acute coronary syndromes (particularly in those without raised troponin concentrations)31 and alternative reasons for the differences between the TACTICS-TIMI-18 trial, and the RITA 3 and FRISC II trials have been suggested;32 for example, differences in the procedural risks and risk profiles of enrolled patients. There may also be other differences in the clinical practice of patients with unstable angina or non-ST-elevation myocardial infarction in the US compared to the UK and Scandinavia.

A further determinant of the impact of early intervention is in the definition of a procedure-related myocardial infarction. In RITA 3, a consistent definition for both procedure and non-procedure related infarctions was used, whereas both FRISC II and TACTICS-TIMI-18 used more stringent definitions of myocardial infarction following a revascularization procedure. Furthermore, a conservatively treated patient in TACTICS-TIMI-18 was defined as having sustained a myocardial infarction with any increase in enzyme or marker concentration above the normal diagnostic limit for normal controls, compared to FRISC II where usually a twofold increase was required. Thus, the beneficial effect of early intervention observed in women in TACTICS-TIMI-18 may, in part, be influenced by the definition of myocardial infarction used. An additional analysis presented in the main RITA 3 paper indicated that significant differences between the groups would be observed by using the ESC/ACC criteria of myocardial infarction largely as a result of the lower threshold for diagnosis of myocardial infarction among the conservatively treated group. We repeated the subgroup analysis of death or myocardial infarction using these criteria to assess whether this might explain differences seen between RITA 3 and TACTICS-TIMI-18. At 1 year, 73 (13.4%) men in the intervention group and 114 (19.6%) in the conservative group would have met the criteria for death or myocardial infarction using these criteria, and for women the numbers were 39 (11.1%) and 42 (12.7%), respectively (interaction p-value 0.24). Thus, these data support the hypothesis that some of the observed differences between the trials may be due to the definitions of myocardial infarction used.

Conclusions

In summary, an early intervention strategy in patients with unstable angina or non-ST-elevation myocardial infarction resulted in a beneficial effect in men which was not seen in women. Establishing why women may fair less well from an early intervention strategy is not straightforward and while there are several potential reasons for this, caution is needed in the interpretation of findings from such subgroups. Further research is needed to evaluate why women do not appear to benefit from such intervention and also to identify treatments to improve the prognosis of women. To this end, it is recommended that a collaborative effort with the investigators from the other major trials in this area is conducted and also that randomized trials for interventions in women with unstable angina or non-ST-elevation myocardial infarction are undertaken.

Acknowledgments

We thank all the investigators, co-ordinators and support staff of the RITA 3 trial together with all the medical and nursing staff in the recruitment and intervention centres who made the trial possible. In particular, we thank all the patients who participated in the trial. RITA 3 was funded by a competitive grant from the British Heart Foundation, and the British Heart Foundation received a donation from Aventis Pharma. Additional governmental support (Culyer) was obtained to reimburse interventional centres for part of the costs of PCI catheters and stents.

For a complete list of investigators, committee members and participating centres please refer to Lancet 2002;360:750.

References

  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
View Abstract