Beneficial effects of aggressive low-density lipoprotein cholesterol lowering in women with stable coronary heart disease in the Treating to New Targets (TNT) study

Under-representation of women in many major cardiovascular trials has continued to limit the evidence base needed to devise optimal management strategies for women with coronary heart disease (CHD), and has created a mandate for gender-specific reporting.1

Whereas clinical trial data have unequivocally established the beneficial effects of lipid-lowering treatment with statins in the prevention of future events among women at high risk for cardiovascular disease,2^–5 it is unknown whether reduction of low-density lipoprotein cholesterol (LDL-C) below the National Cholesterol Education Program (NCEP) Adult Treatment Panel (NCEP ATP) III target of <2.6 mmol/l (100 mg/dl) will provide added benefit.

In 2004, NCEP introduced a more aggressive, but optional, LDL-C goal of <1.8 mmol/l (70 mg/dl) for subjects at high risk of CHD.6 This lower LDL-C goal is consistent with the recent secondary prevention guidelines of the American Heart Association and American College of Cardiology, which state that it is reasonable to reduce LDL-C levels to <1.8 mmol/l (70 mg/dl) in any patient with established CHD.7 The results of the Treating to New Targets (TNT) study showed that reducing LDL-C to a mean level of 2.0 mmol/l (77 mg/dl) with atorvastatin 80 mg in patients with stable CHD provided significant clinical benefit beyond treatment with atorvastatin 10 mg to an LDL-C level of approximately 2.6 mmol/l (100 mg/dl).8 TNT therefore affords the opportunity to examine whether these benefits of intensive versus standard levels of lipid lowering are equally applicable to women. The analysis among women was a predefined secondary assessment of TNT; comparison with data for men is provided.

METHODS

Study design and efficacy analyses

The protocol and outcome measures for the TNT trial have been published previously.8 9 Drugs that were criteria for exclusion included lipid-regulating drugs not specified as study treatment in the protocol, probucol, fibrates and derivatives, bile-acid sequestering resins, other HMG-CoA reductase inhibitors, niacin, supplemental fish oils, and drugs known to be associated with an increased risk of rhabdomyolysis—for example, erythromycin, clarithromycin, mibefradil dihydrochloride, nefazodone and fluvoxamine.

In brief, patients with clinically evident CHD (defined as previous myocardial infarction (MI), previous or present angina with objective evidence of atherosclerotic CHD, and/or who had undergone a coronary revascularisation procedure) started 8 weeks of open-label treatment with atorvastatin 10 mg/day. After this run-in period, 10 001 patients with an LDL-C <3.4 mmol/l (130 mg/dl) were randomised to double-blind treatment with either atorvastatin 10 or 80 mg/day. Patients were followed up for a median of 4.9 years. The primary end point was the time to the first occurrence of a major cardiovascular event, defined as CHD death, non-fatal non-procedure-related MI, resuscitated cardiac arrest and fatal or non-fatal stroke.

Safety analyses

According to the protocol, deaths (including those due to non-cardiovascular causes) were adjudicated by the endpoint committee. Because they were not protocol defined, non-serious and serious adverse events were coded and reported according to industry standards by the investigator with responsibility for the patient. Information reported on all observed or volunteered adverse events included onset, duration, severity and the investigator’s opinion of the relationship to the study treatment. Adverse events included adverse drug reactions, illnesses with onset during the study and exacerbation of previous illnesses.

Statistical analysis

All analyses were performed on an intention-to-treat basis. Baseline differences between genders were performed using t-tests. Statistical significance of treatment effect on end points was tested using the log-rank test. Hazard ratios (HRs) and their confidence intervals (CIs) are presented when appropriate. Homogeneity tests for treatment interaction with various factors were performed under the proportional hazards assumption and model. The number needed to treat (NNT) calculations are based on the total number of cardiovascular events and reflect the number of patients that would have to be treated to prevent one major cardiovascular event over the period of the trial.

RESULTS

Patient population

Of the 18 469 patients screened, 666 women and 2215 men were deemed ineligible to enter the open-label run-in period (gender information was missing for 124 screen failures). A further 1127 women and 4332 men were excluded after the open-label phase (with gender information missing for two run-in failures). The randomised cohort of 10 001 patients with CHD included 1902 women (961 receiving atorvastatin 10 mg and 941 receiving atorvastatin 80 mg) and 8099 men (4045 receiving atorvastatin 10 mg and 4054 receiving atorvastatin 80 mg). Two patients who were resuscitated were not given drugs. Within both female and male subgroups, the atorvastatin 80 mg and 10 mg treatment groups were well matched at baseline.

Table 1 shows the baseline characteristics. Women were older than men, were less likely to be white, and had a higher body mass index and a lower estimated glomerular filtration rate (calculated from serum creatinine using the Modification of Diet in Renal Disease equation10). In addition, women more often had hypertension and diabetes, were more likely to have angina and less likely to have a history of MI and to have undergone coronary artery bypass graft surgery. The proportion of patients having had a prior coronary angioplasty and having peripheral arterial disease and congestive heart failure was also higher in the female than in the male subgroup.

At baseline, 1.4% of women were using some form of birth control and 28.5% (equally distributed between treatment groups) were receiving menopausal hormone therapy (these values increased to 2.5% and 36.0%, respectively, during the course of the study). At baseline, women were also more likely than men to be receiving β-blockers and angiotensin converting enzyme inhibitor or angiotensin II receptor blocker treatment.

Lipid values

Women had higher total cholesterol, triglycerides, and high-density lipoprotein cholesterol (HDL-C) levels, but similar LDL-C at baseline (following the 8-week run-in with atorvastatin 10 mg/day) compared with men (table 1). Reductions in LDL-C, total cholesterol, and triglyceride levels from baseline to week 12 in patients given atorvastatin 80 mg were similar in women and men and were maintained for the duration of the study. Changes from baseline in HDL-C were not significant for either women or men.

Cardiovascular outcomes

In women and men, intensive treatment with atorvastatin 80 mg significantly decreased the rate of major cardiovascular events compared with atorvastatin 10 mg (fig 1). In both gender subgroups there was a gradually increasing difference in the cumulative event rate over time. In women, the relative and absolute reductions in the rate of the primary end point were 27% and 2.7%, respectively. In men, the corresponding rate reductions were 21% and 2.2%. The effect of atorvastatin 80 mg versus 10 mg on the primary outcome was not significantly different by gender. If these results were extrapolated to clinical practice, and we account for all major cardiovascular events (including multiple events per patient) the NNT to prevent one cardiovascular event during 4.9 years of treatment with atorvastatin 80 mg versus atorvastatin 10 mg would be 29 for women and 30 for men.

The hazard ratio for each of the components of the primary outcome were less than or equal to unity for both women and men and not significantly different by gender. Figure 2 shows outcomes and hazard ratios for secondary end points. Tests for evidence of heterogeneity by gender were not significant for gender for any of the secondary cardiovascular efficacy outcomes.

Total, cardiovascular and non-cardiovascular mortality

In women, there was a non-significant excess of total deaths in the atorvastatin 80 mg group compared with the atorvastatin 10 mg group (58 vs 45; p = 0.13) (table 2). This excess in total mortality (13 deaths) was the result of a non-significant reduction in cardiovascular deaths (HR = 0.83, 95% CI 0.49 to 1.42; p = 0.50) and a significant excess of non-cardiovascular deaths (HR = 2.38, 95% CI 1.30 to 4.37; p = 0.004) in the atorvastatin 80 mg group compared with the atorvastatin 10 mg group. In women the 4.9 year NNT to prevent one cardiovascular death was 175, whereas the corresponding NNT to produce one excess non-cardiovascular death was 49. In contrast, in men there were no significant differences in total, cardiovascular, or non-cardiovascular mortality between the atorvastatin 80 mg and atorvastatin 10 mg groups.

Additional analyses of the non-cardiovascular mortality in women identified that the excess risk was limited to an excess in cancer mortality (2.0% vs 0.6%, p = 0.006). No specific body system and no single cancer type contributed disproportionately to this difference in non-cardiovascular mortality between treatment groups among women. Rates of cancer death among men were similar for atorvastatin 80 mg (1.6%) and atorvastatin 10 mg (1.7%). Small numeric differences between groups were observed for some types of cancer, but the excess was sometimes seen in the atorvastatin 80 mg group—for example, glioblastoma (five men (0.1%) receiving atorvastatin 80 mg versus one man (0.0%) receiving atorvastatin 10 mg), and sometimes in the atorvastatin 10 mg group—for example, colon cancer (one man (0.0%) receiving atorvastatin 80 mg versus five men (0.1%) receiving atorvastatin 10 mg). There was no imbalance between treatment groups in either gender in cancer adverse events.

Safety

Myalgia was the most common adverse event related to treatment. Rates of myalgia were slightly higher among women than men, but similar between treatment groups for both women and men. Discontinuations due to treatment-related adverse events in the atorvastatin 10 mg group versus the atorvastatin 80 mg group were 3.0% versus 6.5%, respectively, for women and 2.4% versus 3.7%, respectively, for men. Persistent liver function test abnormalities in the atorvastatin 10 mg group versus the atorvastatin 80 mg group were 0.2% versus 2.6%, respectively, for women and 0.2% versus 0.9%, respectively, for men. No woman or man experienced a persistent elevation in creatine kinase (defined as two consecutive measurements >10 times the upper limit of normal within 4–10 days).

DISCUSSION

Recent evidence has demonstrated that only one-third of women at high risk with documented CHD received the recommended lipid-modifying treatment, and as few as 12% attained optimal lipid levels in their managed care setting.11 Despite presenting with higher risk characteristics and having higher in-hospital risk, women with acute coronary syndromes have been shown to be less likely to receive statin treatment at discharge and are overall treated less aggressively than men.12 There is, therefore, substantial opportunity to improve lipid management and reduce cardiovascular morbidity and mortality among women.

Efficacy analyses

The current secondary analysis of the TNT study demonstrates that among women with clinically evident CHD, intensive lipid-lowering treatment with atorvastatin 80 mg to LDL-C values well below currently recommended levels was associated with a significant 27% reduction in the rate of major cardiovascular events compared with a more moderate regimen of atorvastatin 10 mg (p = 0.049). These findings are consistent with those seen among men and in the overall patient population.

This reduction in cardiovascular risk associated with atorvastatin 80 mg is above the clinical benefits provided by treatment with atorvastatin 10 mg (ie, this is a dose comparison, not a placebo-controlled trial), and the attainment of LDL-C levels currently regarded as optimal treatment targets. The degree of LDL-C lowering found with atorvastatin 10 mg during the open-label period was similar or greater than that reported among women in previous secondary prevention studies in which statin treatment provided highly significant reductions in the rate of major coronary events compared with placebo.2^–4

Hazard ratios and confidence intervals for secondary outcomes were consistent with the overall study population and there was no significant heterogeneity of treatment effect for gender, indicating that the significant results of the main study for major coronary events, any coronary event, cerebrovascular events, hospitalisation for congestive heart failure, and any cardiovascular event hold true for both women and men. As for the total study cohort, the effect of atorvastatin 80 mg on peripheral artery disease did not differ significantly from that of atorvastatin 10 mg in women or men.

Hormonal influences on lipoprotein levels in women are complex, change throughout the lifespan, and are influenced by the administration of oral contraceptives and hormone therapy (during the TNT study, 2.5% of women were using some form of birth control and 36.0% were receiving menopausal hormone therapy).13 Baseline lipid values among women and men in TNT showed some significant differences. Triglyceride levels were significantly higher among women than among men, and this may reflect the increased levels of remnant-like particle cholesterol and triglyceride found in postmenopausal compared with premenopausal women and in women with CHD compared with healthy women.14 15 The higher level of baseline HDL-C seen among women in TNT is typical of gender differences in HDL-C levels that emerge at puberty and are maintained throughout life.16

Baseline levels and reductions in LDL-C in response to treatment were similar in women and men. A small increase in absolute LDL-C level was observed for both treatment groups over the duration of the study.

Based on the results of the current analysis, the NNT to prevent one cardiovascular event (over a period of 4.9 years) among women is 29. This value is consistent with that observed for men (30) and that reported for the total patient cohort (30).8 These findings are consistent with a prior meta-analysis, which included primary and secondary prevention trials and assessed the effect of statins on CHD and reported an NNT of 31 to prevent one coronary event among women.17 It is important to note that the NNT values for these previous analyses are based on comparisons with placebo, while those for TNT are derived from the differential between atorvastatin 80 mg and the clinical benefits achievable with atorvastatin 10 mg.

There was no significant difference, for either women or men, between the treatments in their effects on all-cause mortality. Similar findings were reported in a recent meta-analysis of eight secondary prevention lipid-lowering studies, in which lipid lowering, compared with placebo, had no significant effect on death from any cause.18 Consistent with the current study, CHD-related events in this analysis were significantly reduced by 20% and the corresponding NNT value among women was estimated at 26 (although again this was compared with placebo and not standard lipid lowering treatment). A similar lack of effect on all-cause-mortality when comparing intensive versus moderate statin treatment in patients with stable CHD was also reported in a recent comparative meta-analysis.19 However, this analysis included the TNT trial as one of the largest sources of data, therefore, the fact that the results resemble those of the TNT trial is not surprising.

Safety analyses

The finding of excess non-cardiovascular mortality in women, and more specifically cancer mortality, raises an important question about the safety of high-dose atorvastatin. There are, however, a number of reasons that this finding should be viewed with care. Multiple secondary analyses comparing small numbers of events increase the likelihood of a false positive result due to chance. The fact that no single cause of death by body system and no single cancer type contributed disproportionately to the difference in non-cardiovascular mortality between treatment groups, and that there was no difference between groups in new cancer diagnoses or adverse events associated with cancer, lend further support to the possibility that this represents a chance finding.

Similar concerns about a potential statin-associated increase in risk for cancer have been raised in earlier placebo-controlled trials. For example, although the overall incidence of fatal or non-fatal cancer in the Cholesterol and Recurrent Events (CARE) trial was not significantly different between treatment groups, the incidence of breast cancer was higher in patients given pravastatin.20 However, in the subsequent Long-term Intervention with Pravastatin in Ischaemic Disease (LIPID) study, there was no increase in the number of newly diagnosed breast cancers among women assigned to receive pravastatin, leading the authors to suggest that the excess rate of breast cancer in the CARE study was a chance finding.21 Likewise, in the PROspective Study of Pravastatin in the Elderly at Risk (PROSPER), new cancer diagnoses were more frequent in pravastatin-treated patients than in those receiving placebo.22 In a subsequent meta-analysis of several large studies, no significant effect of pravastatin on cancer rates was demonstrated.23

Perhaps the most reassuring piece of evidence that the excess cancer mortality risk in women receiving intensive atorvastatin treatment may be a spurious finding is that a similar pattern to TNT was not observed in the Incremental Decrease in End Points through Aggressive Lipid Lowering (IDEAL) study, in which patients who had a previous MI were randomised to prescription treatment with atorvastatin 80 mg or simvastatin 20 mg (the dose was increased to 40 mg at week 24 in those patients whose plasma total cholesterol remained >4.9 mmol/l (190 mg/dl), or whose LDL-C remained >3.0 mmol/l (115 mg/dl)).24 In contrast to TNT, women receiving atorvastatin 80 mg in IDEAL experienced fewer non-cardiovascular deaths than women in the moderate lipid-lowering comparator group (25 deaths in the atorvastatin 80 mg group versus 36 deaths in the simvastatin group) (Pedersen T, unpublished data, 2006). Nevertheless, future studies should continue to monitor carefully gender differences in both cardiovascular and non-cardiovascular mortality in patients treated with statins.

The reduction in cardiovascular events in women and men was not associated with any persistent increases in creatine kinase. The occurrence of liver function abnormalities was low in both treatment groups and both gender groups, but higher among women than men receiving atorvastatin 80 mg. The higher rate of discontinuations in the female compared with the male subgroup is consistent with other analyses, which have reported a reduced likelihood to discontinue study medication among men.2 25 In TNT, the higher maximum concentration of atorvastatin in women compared with men (Cmax approximately 20% higher) may contribute to this gender difference.26 Plasma concentrations of atorvastatin are also higher (approximately 40% for Cmax) in elderly subjects (age ≥65 years) than in young adults. The fact that women were significantly older than men in TNT (mean age 60.4 years for men, 63.5 years for women) may also potentially affect both benefit and risk among the female subgroup.

CONCLUSIONS

In conclusion, among women with stable CHD, reduction of LDL-C levels to well below currently recommended targets with atorvastatin 80 mg produced significant reductions in risk for major cardiovascular events compared with atorvastatin 10 mg. More data are needed to clarify the effects of aggressive lipid-lowering treatment on cardiovascular versus non-cardiovascular mortality in these women. The data from the current study do not provide compelling evidence of a favourable or adverse effect of aggressive lipid lowering on total mortality in women.