In this report, we evaluated the efficacy and tolerability of lovastatin in women who participated in the Expanded Clinical Evaluation of Lovastatin (EXCEL) study [7]. Patients in the EXCEL study had moderately elevated baseline levels of plasma total and low-density lipoprotein (LDL) cholesterol, typical of those seen in the usual clinical practice. The study design provided for large treatment groups so that the incidence of abnormal laboratory safety tests and clinical adverse events (particularly involving liver and muscle) could be reliably determined among treatment groups even at low frequency. Although an analysis based on gender was not an a priori objective of the EXCEL study design, the large sample size provided an opportunity to investigate the effect of lovastatin in women, to assess the effect of concomitant administration of estrogen-replacement therapy on the efficacy of lovastatin, and to detect any clinically important gender differences in efficacy and tolerability.

Methods

Top Methods Results Discussion Author & Article Info References

The design of this study and the methods used have been described previously [7]. In summary, the EXCEL study evaluated the efficacy and tolerability of lovastatin using a diet- and placebo-controlled, randomized, double-blind design. The protocol excluded premenopausal women unless surgically sterilized; however, data classifying women as surgically sterilized or postmenopausal were not collected. After a minimum of 4 weeks on an American Heart Association Step I diet, 8245 patients (3390 women) with moderate primary hypercholesterolemia (total cholesterol, 6.21 to 7.76 mmol/L [240 to 300 mg/dL]; LDL cholesterol 4.14 mmol/L (160 mg/dL); and triglycerides < 3.95 mmol/L [350 mg/dL]) from 362 clinical centers were instructed to continue diet and were randomly assigned to 48 weeks of treatment with one of the following five regimens: placebo, or 20 or 40 mg lovastatin once daily, or 20 or 40 mg twice daily. The study was approved by institutional review boards and participants gave written informed consent. Laboratory tests were performed by SmithKline Beecham Laboratories (Clinical Trials Division, Van Nuys, California).

Efficacy Analyses

Total cholesterol, high-density lipoprotein (HDL) cholesterol, and triglycerides were measured using standard enzymatic techniques [7]. The LDL cholesterol was calculated using the Friedewald approximation [8]. Response to treatment was calculated by the formula percent change equals ([baseline level minus treatment level] divided by baseline level) x 100. A patient's baseline level was the mean of two measurements in 91% to 92% of patients in each treatment group. The treatment values represented the mean of up to four measurements taken every 12 weeks during treatment. In calculations of LDL cholesterol response, data from 606 to 650 (92.5% [lovastatin, 20 mg once daily] to 93.5% [lovastatin, 20 mg twice daily]) women and 897 to 941 (94.3% [lovastatin, 20 mg twice daily] to 95.3% [lovastatin, 20 mg once daily]) men in each treatment group were available for both baseline and treatment period measurements. For total LDL and HDL cholesterol, two-sided t-tests were used to assess differences in mean percentage change between placebo and each lovastatin treatment group. The Wilcoxon rank sum test was used for testing differences in triglyceride levels.

A less extensive approach, using linear statistical models, was used to assess the effects of concomitant use of estrogen. The covariates included in this modeling were baseline effects of age and lipid/lipoprotein levels.

Safety Analyses

Elevations in serum alanine aminotransferase and aspartate aminotransferase and muscle symptoms (complaints of pain, tenderness, or weakness) with increased levels of serum creatine kinase represent clinical and laboratory abnormalities associated with lovastatin therapy. Analyses presented here examine gender differences in both the incidence of elevations greater than the upper limit of normal and the mean levels of change from baseline to the end of treatment. Survival analysis methods specific for analyzing incidence have been described previously [9]. For serum transaminases, 95% confidence intervals for mean differences between women and men that did not overlap zero were considered statistically significant. The Cochran-Mantel-Haenszel test [10], after stratification by treatment group, was used to assess gender differences in muscle symptoms and in the incidence of elevated serum transaminases and creatine kinase. The Fisher exact test was used to evaluate treatment differences between placebo and lovastatin groups; this test was done separately for women and men.

The proportion of patients having a clinical adverse experience (any change in health status) that investigators rated blindly (without knowledge of treatment group assignment) as being possibly, probably, or definitely drug related was examined by gender and by estrogen use with lovastatin. These were categorized into serious drug-related adverse experiences (resulted in death; required or prolonged hospitalization; were cancers; or caused permanent disability), nonserious adverse experiences, and adverse experiences requiring discontinuation of therapy. The Cochran-Mantel-Haenszel test [10] was used to test for gender differences after stratification by treatment group.

Results

Top Methods Results Discussion Author & Article Info References

Characteristics of Women at Baseline

Of the total 8245 patients, 3390 (41%) were women (Table 1). The number of women randomized to each treatment group ranged from 655 to 695. The average age was 58.4 years, 89.8% were white, 18.3% reported that they currently smoked cigarettes, and 42.1% were hypertensive. The mean baseline total cholesterol level was 6.74 mmol/L (261 mg/dL), and the mean LDL cholesterol level was 4.60 mmol/L (178 mg/dL). Although gender differences were not evident in most baseline characteristics, the mean age of women was 4.5 years older. They were less likely to have a history of coronary heart disease (16.3% versus 37.3%), less likely to consume alcohol (43% versus 65%) and particularly seven or more drinks per week (5% versus 19%), had higher mean HDL cholesterol levels (1.32 mmol/L [51 mg/dL] versus 1.05 mmol/L [41 mg/dL]), and lower median triglyceride levels (1.65 mmol/L [147 mg/dL] versus 1.94 mmol/L [173 mg/dL]).

Lipid and Lipoprotein Response

Lovastatin produced dose-dependent (20 to 80 mg/d) reductions of 24.4% to 40.4% in LDL cholesterol and 9.4% to 17.9% in triglycerides, and increases of 6.7% to 8.6% in HDL cholesterol (Table 2). Lovastatin treatment groups differed (P < 0.001) from placebo in the mean percentage changes from baseline for all lipid and lipoprotein parameters. Overall, these results were similar to those found in men.

To assess the achievement of National Cholesterol Education Program (NCEP) [11] LDL cholesterol goals after 48 weeks of lovastatin treatment, patients were categorized into two groups: a lower-risk group without coronary heart disease and with fewer than two coronary heart disease risk factors, and a high-risk group with coronary heart disease or at least two coronary heart disease risk factors, or both. In the lower-risk women, the NCEP LDL cholesterol goal of less than 4.14 mmol/L (160 mg/dL) was reached by 26% given placebo and from 82% to 95% treated with lovastatin (20 to 80 mg/d) (Figure 1). In high-risk women, the NCEP LDL cholesterol goal of less than 3.36 mmol/L (130 mg/dL) was reached by 4% given placebo and from 40% to 87% treated with lovastatin (20 to 80 mg/d). Similar results were observed in men.

View larger version (17K): [in this window] [in a new window]   Figure 1. Percentage of women achieving National Cholesterol Education Program target levels for low-density cholesterol at week 48. The white bars at left indicate those women without coronary heart disease (CHD) and with fewer than two CHD risk factors (goal-low-density lipoprotein less than 4.14 mmol/L). The black bars at the right indicate those women with coronary heart disease or with two or more CHD risk factors (goal: low-density lipoprotein less than 3.36 mmol/L).

In both women and men, lovastatin-associated elevations in HDL cholesterol were dose dependent, and the percentage increases were larger when baseline levels of HDL cholesterol were low. When baseline HDL cholesterol levels were less than 0.91 mmol/L (35 mg/dL), lovastatin (20 to 80 mg/d) increased the placebo-corrected mean HDL cholesterol level in women by 1.6% to 11.5%. When baseline HDL cholesterol levels were 0.91 mmol/L or higher, the corresponding placebo-corrected increases were 4.7% to 5.9%. Fewer women than men (5% to 8%, versus 25% to 30%) had HDL cholesterol levels less than 0.91 mmol/L.

Safety Analyses

Successive elevations in alanine aminotransferase or aspartate aminotransferase greater than three times the upper limit of normal were found in one woman (0.2%) given placebo; two (0.4%) given 20 mg/d; three (0.5%) given 40 mg every evening; five (1.0%) given 20 mg twice daily, and eight (1.6%) given 80 mg/d. No statistically significant gender differences in incidence were observed. Mean alanine aminotransferase levels, however, showed less increase with lovastatin treatment in women (up to 0.05 µkatal/L [26.3%] at 80 mg/d) than in men (0.06 to 0.11 µkatal/L [19.6% to 30.1%] at 20 to 80 mg/d), and the 95% confidence intervals for the gender difference did not overlap zero in the lovastatin groups. Mean aspartate aminotransferase levels were also increased in women given lovastatin (0.07 to 0.09 µkatal/L [21.9 to 26.4%], 20 to 80 mg/d), but no statistically significant gender differences were observed.

Myopathy (defined in the study as muscle symptoms with creatine kinase greater than 10 times the upper limit of normal) was rare and too infrequent (4 women, 1 man) to assess treatment group differences or gender differences by statistical analyses. All five cases occurred within the first 6 months of the study. Two patients (one woman [on 80 mg/d] and one man [40 mg/d]) continued lovastatin therapy through the 48-week study, and in both, symptoms resolved and creatine kinase levels returned to normal. In the other three women, lovastatin therapy (80 mg/d) was discontinued, and within 30 days symptoms resolved and creatine kinase levels returned to normal. None of the five patients had myoglobinuria or acute renal failure; maximum creatine kinase elevations ranged from 33.34 to 171.70 µkatal/L (2000 to 10 300 U/L).

Muscle symptoms were reported frequently by women in the placebo group (6.8%) and by all lovastatin groups (range, 6.6% [20 mg twice daily] to 10.7% [40 mg twice daily]), as well as by men in each treatment group (P > 0.2 for gender difference). Women had a lower incidence (P < 0.01) of creatine kinase elevations above the upper limit of normal (21.4% in the placebo group and a range of 20.4% [40 mg once daily] to 26.9% (40 mg twice daily) among lovastatin groups; compared with 34.2% in the placebo group, and a range of 33.8% to 40.0% [20 to 80 mg/d] among lovastatin groups in men). In women given 80 mg/d lovastatin, the incidence of either muscle symptoms or any creatine kinase elevation above normal was slightly greater than the incidence in other lovastatin or placebo treatment groups and became significant (P < 0.002) for the combined findings of muscle symptoms and creatine kinase elevations above the upper limit of normal (4.2% versus 3.1% in men). Mean baseline creatine kinase levels were lower in women (1.67 µkat/L) than men in (2.45 µkatal/L). The changes in creatine kinase values during lovastatin treatment were highly variable, and no statistically significant gender differences were found.

Serious clinical adverse experiences attributable to drugs were rare, with no apparent gender differences (Table 3). In women, the incidence of drug-attributable nonserious clinical adverse experiences was 4% to 5.2% (20 to 80 mg/d) higher in the lovastatin treatment groups than in the placebo group, compared with a range of –0.3%(40 mg once daily) to 1.2% (40 mg twice daily) in men. Women had a higher incidence overall of adverse experiences than men (P < 0.01), and the level was similar for both younger and older women. The same gender pattern (P < 0.01 for difference between women and men) was seen for clinical adverse experiences that resulted in discontinuation of therapy; the difference in incidence between the lovastatin treatment and placebo groups ranged from 3.2% (20 mg twice daily) to 3.7% (40 mg twice daily) for women and from 0.1% (20 mg once daily) to 0.9% (40 mg once daily) for men. In women, the relative risk of discontinuing therapy for clinical adverse experiences was 2.5 (95% CI, 1.5 to 4.2) times greater in lovastatin treatment compared with placebo groups. Both the number of specific types of clinical adverse experiences and the occasions these experiences were reported was somewhat greater in the lovastatin treatment groups compared with the placebo group.

Effects of Concomitant Estrogen Therapy

Of the 3390 women, 788 (23.2%) reported taking concomitant estrogen therapy throughout the study, 2424 (71.5%) received no estrogen therapy, and 178 (5.3%) received estrogen therapy sporadically and were excluded from the comparison. The percentages of women receiving estrogen (21.7% [lovastatin, 20 mg once daily] to 25.6% [lovastatin, 40 mg once daily]) and receiving no estrogen (70.1% [lovastatin, 40 mg once daily] to 73.3% [lovastatin, 20 mg once daily]) were similar across treatment groups (see Table 1). Overall, women receiving estrogen therapy were slightly younger [56.4 versus 59.2 years], more likely to be white (95.7% versus 87.8%), and less likely to be hypertensive (35.7% versus 44.3%), or to have a history of coronary heart disease (9.4% versus 18.8%). In addition, the mean baseline LDL cholesterol level was lower for women receiving estrogen (4.41 mmol/L [171 mg/dL]) than for women receiving no estrogen (4.67 mmol/L [181 mg/dL]). The mean baseline HDL cholesterol level was higher (P = 0.004) for women receiving estrogen therapy (1.47 mmol/L [57 mg/dL]) than for women receiving no estrogen (1.27 mmol/L [49 mg/dL]). Baseline mean total cholesterol values were similar. Triglyceride values were variable and higher among women receiving estrogen (median, 1.72 mmol/L [154 mg/dL] versus 1.63 mmol/L [145 mg/dL]).

We explored the effect of concomitant estrogen therapy on response to lovastatin treatment in women. The LDL cholesterol reductions for estrogen users compared with nonusers in the lovastatin treatment groups were 23.7% versus 24.6% (20 mg once daily), 29.9% versus 31.6% (40 mg once daily), 32.9% versus 34.1% (20 mg twice daily), and 37.8% versus 41.1% (40 mg twice daily). After adjusting for baseline effects of age and lipid/lipoprotein levels, concomitant estrogen therapy in women had no statistically significant effect on lovastatin-induced changes in LDL cholesterol, HDL cholesterol, or triglyceride levels. Concomitant estrogen therapy had no statistically significant effect on the safety assessments monitored in this study. The incidence of drug-related adverse experience for estrogen users versus nonusers in the lovastatin treatment groups was 26.8% versus 29.2% (20 mg once daily), 29.5% versus 28.8% (40 mg once daily), 26.4% versus 28.2% (20 mg twice daily), and 32.9% versus 29.6% (40 mg twice daily).

Discussion

Top Methods Results Discussion Author & Article Info References

Elevated levels of total cholesterol and reduced levels of HDL cholesterol are related to the development of coronary heart disease in both women and men [12, 13]. Substantial evidence associating increased levels of LDL cholesterol with coronary heart disease forms the basis of the NCEP guidelines for the evaluation and treatment of hypercholesterolemia in adults [11], and an estimated 23 million U.S. women age 40 years or older are candidates for medical advice and intervention for high blood cholesterol levels [14]. Despite this, few clinical trials have included enough women to evaluate the effects of therapeutic agents on hypercholesterolemia in women. Although dietary intervention is the initial strategy of therapy for primary hypercholesterolemia, drug therapy may be necessary to reduce LDL cholesterol levels adequately. The objective of intervention is the attainment of desirable LDL cholesterol levels as recommended by the NCEP guidelines to reduce the risk for coronary heart disease [11]. Lovastatin, a potent inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A reductase, is one of the drugs available for the treatment of hypercholesterolemia. Reductions in total cholesterol and modest increases in HDL cholesterol have been documented in other clinical trials of lovastatin [15].

Therapy with lovastatin substantially lowered plasma LDL cholesterol (24.4% to 40.4%) in women, who were primarily in the postmenopausal age range. Significant, dose-dependent increases in HDL cholesterol (6.7% to 8.8%) and decreases in triglycerides (9.4% to 17.9%) were also observed. The effects of lovastatin on lipids and lipoproteins were similar in women and men. Other studies with smaller sample sizes have shown similar results for both genders [16, 17].

Substantial proportions of women achieved the appropriate NCEP LDL cholesterol goal when treated with diet and lovastatin. The importance of reaching the NCEP-recommended LDL cholesterol goal has been emphasized [11]. For women without coronary heart disease and less than two coronary heart disease risk factors, 82% treated with lovastatin (20 mg/d) and 95% treated with lovastatin (80 mg/d) achieved their NCEP LDL cholesterol goal of less than 4.14 mmol/L (160 mg/dL). By NCEP guidelines, some of these women qualified for diet-only intervention; they were treated with lovastatin in this study because the protocol was implemented before the publication of the NCEP guidelines. For women with coronary heart disease or two or more coronary heart disease risk factors, or both, 40% treated with 20 mg/d and 87% treated with 80 mg/d of lovastatin achieved their NCEP LDL cholesterol goal of less than 3.36 mmol/L (130 mg/dL). The dose-dependent increase in the number of women achieving NCEP LDL cholesterol goals suggests that titration would allow many of those inadequately managed at a lower dose to achieve their goal. In those women who did not fully meet this objective by taking lovastatin alone, the addition of a second agent (for example, a moderate dose of a bile acid sequestrant) has been shown to enhance LDL cholesterol lowering [15]. If the patient was initially taking a bile acid sequestrant and had not reached the recommended goal with this agent alone, the addition of a reductase inhibitor to the bile acid sequestrant represents an alternative strategy of combined drug treatment.

Previously published data from this study used statistical models to identify patient characteristics that modified the lipid/lipoprotein response to therapy with lovastatin [18]. Gender showed little effect; but gender interacted with age in women so that in older women (65 compared with 45 years) lovastatin produced a modestly enhanced reduction in LDL cholesterol level (42.8% versus 38.4% at 80 mg) [18]. This small age effect was independent of other patient characteristics and remains unexplained. Concomitant estrogen replacement therapy in women appeared to have no effect on the LDL cholesterol, HDL cholesterol, or triglyceride response to lovastatin.

As expected [12], estrogen therapy in women was associated with higher baseline HDL cholesterol levels. In addition, women overall had higher mean baseline HDL cholesterol levels (1.32 mmol/L [51 mg/dL]) than did men (1.05 mmol/L [41 mg/dL]), and only a small proportion of women had low HDL cholesterol (< 0.91 mmol/L [35 mg/dL]). In both women and men, increases in HDL cholesterol on lovastatin treatment were dose dependent and appeared to be enhanced when pretreatment levels were low. In patients with baseline levels less than 0.91 mmol/L, treatment with lovastatin 80 mg/d resulted in mean increases (corrected for placebo effects) of 11.6% in women and 10.3% in men.

Overall, the tolerability profile of lovastatin in women was very favorable and did not appear to be affected by concomitant estrogen replacement therapy. Marked elevations (more than three times the upper limit of normal) in serum transaminase levels were dose dependent above lovastatin, 20 mg/d and were infrequent in both women and men. Mean increases in alanine aminotransferase were lower in women than in men treated with lovastatin and may be related to increased alcohol consumption among men. Myopathy was rare, but four of the five reported cases occurred in women, all at the highest recommended dose of lovastatin (80 mg/d). Myopathy associated with lovastatin treatment has been reported in patients taking marketed drugs and has been attributed to a complex set of interactions involving increased circulating drug levels during combination therapies [19].

Women had a greater tendency to report nonserious drug-related clinical adverse experiences and more women discontinued lovastatin because of adverse experiences. Gender differences in the willingness to acknowledge and report symptoms may explain these differences [20]. It is also worth noting that the percentage of women who discontinued therapy because of drug-related clinical adverse experiences involved no more than 6% of any treatment group and thus does not detract appreciably from the overall evidence of the high tolerability of lovastatin in women.

Because gender-specific analyses and analyses based on estrogen use were not originally stated in the hypotheses to be tested in this study, the results of these retrospective subgroup analyses should be regarded as exploratory and interpreted with appropriate caution. Although patients were not initially stratified by subgroup, the randomization of large numbers of patients resulted in similarities across treatment groups in both the number of patients represented in each subgroup and their associated baseline characteristics. For the efficacy analyses, biases were minimized through stratification on variables identified by a linear regression model [18]. Gender differences in the safety analyses may be causally related to other variables in this study group.

A substantial portion of women who have high plasma LDL cholesterol levels may require pharmacologic intervention to achieve the LDL cholesterol goals recommended by the NCEP. The EXCEL Study included the largest number of women who have participated in a controlled drug evaluation trial. The demographic and clinical characteristics of these study participants are typical of a substantial portion of women who may be expected to be prescribed a lipoprotein-modifying drug. Findings from this study indicate that lovastatin, when used as an adjunct to diet, is a highly effective and generally well tolerated agent for controlling primary hypercholesterolemia in women as well as in men.