| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Obstetrics/Gynecology and Endocrinology, The Women’s Place and the Cancer Center, University of Virginia Health Sciences Center, Charlottesville, Virginia 22903-9301
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. Prevention of Cardiovascular...III. Prevention and Treatment...IV. Urogenital AtrophyV. Vasomotor InstabilityVI. Central Nervous System...VII. Safety and Side...VIII. ConclusionsReferences |
|---|
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. Prevention of Cardiovascular...III. Prevention and Treatment...IV. Urogenital AtrophyV. Vasomotor InstabilityVI. Central Nervous System...VII. Safety and Side...VIII. ConclusionsReferences |
|---|
Therapeutic goals to be achieved by estrogen alternatives might include the amelioration of symptoms from urogenital atrophy, vasomotor instability, and neurocognitive dysfunction; and the prevention of heart disease and osteoporosis (1, 2). Possible additional goals include a reduction in incidence of Alzheimer’s disease, macular degeneration, colon cancer, and mandibular bone loss (2). Rational use of alternatives for ERT requires knowledge of the relative efficacy of each treatment modality compared with that of estrogen itself. This review will examine data regarding the relative effectiveness of estrogen vs. nonestrogen alternatives such as 3-hydroxy-3-methyl-glutaryl-coenzyme A (HMG-CoA) reductase inhibitors or "statins," bisphosphonates, calcitonins, clonidine, and megestrol acetate, as well as the partial estrogen agonists/antagonists or SERMs (selective estrogen receptor modulators).
|
II. Prevention of Cardiovascular Disease |
|---|
|
TopAbstractI. IntroductionII. Prevention of Cardiovascular...III. Prevention and Treatment...IV. Urogenital AtrophyV. Vasomotor InstabilityVI. Central Nervous System...VII. Safety and Side...VIII. ConclusionsReferences |
|---|
A. Estrogens and primary prevention of cardiovascular disease
1. Treatment strategies. Two therapeutic goals,
primary and secondary prevention, have been
evaluated with respect to estrogens and cardiovascular disease. The
primary prevention strategy attempts to reduce the incidence
of initial cardiovascular events. Secondary prevention seeks
to diminish the rate of events in subjects who have evidence for
established disease whether or not they have had an event. The precise
efficacy of estrogens for primary prevention of heart
disease can only be estimated from observational studies since no
randomized, placebo-controlled, double-blind trials have been
completed. These observational data provide substantial support for the
cardioprotective effects of estrogen and the magnitude of effect.
Evidence from more than 30 epidemiological studies suggests a 35–50%
decrease in the incidence of initial coronary events with ERT (3, 4, 5, 6, 7, 8, 9).
2. Mechanisms of effect. Estrogen acts on the cardiovascular system by at least two different mechanisms: positive effects on the serum lipid profile and direct actions on blood vessels. It has been estimated that about 25% of the cardioprotective effect is due to lipid changes including increased high-density lipoprotein (HDL), decreased low-density lipoprotein (LDL), decreased lipoprotein (a) [Lp(a)], decreased LDL oxidation, and decreased vascular LDL uptake (10). The fact that estrogens improve lipid profiles has been used as evidence that estrogens should prevent heart disease. In the absence of prospective randomized trials, lipid levels are used as "indirect end points (surrogate end points)" to support the possibility that estrogens are cardioprotective. In the randomized, placebo-controlled, double blind Postmenopausal Estrogen/Progestin Intervention Trial (PEPI) (11), 875 women received various estrogen/progestin regimens. Estrogen alone decreased LDL cholesterol levels and increased HDL cholesterol. Concomitant use of medroxyprogesterone acetate (MPA) with estrogen blunted the increase in HDL cholesterol but did not affect the reduction in LDL cholesterol. Triglyceride levels increased with either therapy when compared with placebo. Micronized progesterone exerted less of a negative effect on HDL than MPA (11).
The remaining 65–75% of the cardioprotective effect of estrogen is believed to result from lipid-independent effects on the heart and blood vessels and on plaque formation possibly mediated by increased nitric oxide, decreased endothelin, increased PGI2, and decreased thromboxane A2 (12, 13). When delivered transdermally, estrogen is believed to exert its effects predominantly through nonlipid effects. This means of administration abrogates first-pass effects on the liver with resultant lipid alterations but allows direct blood vessel and vascular actions (14).
3. Evidence for efficacy. Data from the Nurse’s Health Study provide the strongest support for a cardioprotective effect of estrogen (15, 16). In this prospective observational study, 59,337 women were followed for a total of 662,891 patient-years. The authors reported that the relative risk (RR) of developing heart disease in current hormone replacement therapy (HRT) users was 0.60 for estrogen alone and 0.39 for estrogen plus a progestin (15). They also calculated the absolute numbers of patients benefited. After adjustment for multiple risk factors, it was estimated that 33 new cardiac events would be prevented per year per 100,000 women initiating estrogens usage at age 50 (15). For those starting at age 60, the number prevented would be 66. We considered that these data would be more clinically relevant if extrapolated to 10 yr of estrogen use. Assuming that the effects were linear with increasing duration of estrogen use, we calculated that by taking estrogen continuously for 10 yr, 330 new events would be prevented in 50-yr-old women and 660 events in 60-yr-old women (17).
Barrett-Connor and others advise caution in interpreting these observational data because of several confounding biases (18, 19, 20, 21). Women who take estrogen in the United States are self-selected for several factors that reduce the risk of cardiovascular disease—the "healthy woman" bias. There may be a "prevention" bias for women who spend a significant amount of time in doctor’s offices and become better informed about health issues or a "compliance" bias since these self-selected women continue to take estrogen long term. Compliance with taking a placebo has been shown to be associated with a decrease in incidence of heart disease. Data from the Coronary Drug Project showed that subjects who took 80% of their placebo had an RR of cardiovascular disease of 0.53, or a 50% reduction in heart disease (2). The Women’s Health Initiative, a large prospective randomized, controlled study of estrogen vs. placebo has been started in the United States to provide a definitive answer to several questions regarding the effects of ERT.
B. Comparison of statins with estrogens for primary prevention of
cardiovascular disease
1. Lipid levels. The ability of various agents to alter lipid
levels is considered to be an "indirect end point" to predict
efficacy in preventing heart disease. With this as a rationale, two
randomized studies have compared the effects of estrogen and the HMG
Co-A reductase inhibitors (statins) on lipid levels (22, 23) (Fig. 1
).
|
The clinical significance of these differences is as yet unknown. The atherogenic potential of estrogen-induced hypertriglyceridemia is unclear. Secondary-prevention studies have not shown that reducing LDL cholesterol in women is superior to increasing HDL cholesterol. HDL cholesterol level may be a better predictor of the risk of coronary heart disease in women than the total or LDL cholesterol level (24, 25, 26, 27). Lipoprotein (a) level may also be an independent predictor of the risk of coronary heart disease (28, 29, 30).
b. North American Study. In a second direct comparison,
Davidson et al.
(23) reported the effects of conjugated
estrogens alone, pravastatin alone, and the combination for management
of hypercholesterolemia in postmenopausal women. Trial design was
double blind, randomized, and placebo-controlled with four parallel
groups. Participants (n = 76) were randomly assigned to receive
conjugated estrogens, 0.625 mg/day; pravastatin sodium (Bristol
Myers Squibb, Princeton, NJ), 20 mg/day; conjugated
estrogens plus pravastatin; or a placebo for 16 weeks. We will present
only the comparative data between pravastatin and estrogen. Among
participants treated with conjugated estrogens, levels of calculated
LDL cholesterol (LDL-C) decreased by 13.5%, while increases in levels
of HDL-C (22.5%) and triglycerides (4.2%) were seen (Fig. 1
).
Participants in the pravastatin group achieved reductions of 25.4% in
calculated LDL-C levels. Levels of HDL-C increased slightly (3.7%),
and triglycerides decreased by 12.1%. The pravastatin group had a more
significant reduction in LDL-C. Estrogen improved HDL-C significantly
(22.5% vs. 3.7%) but also increased triglycerides.
c. Higher doses of statins.
Dose escalation may further
improve lipid levels as reported by Stein et al. (31) for
the Expanded Dose Simvastatin U.S. Study Group. This randomized,
multicenter, double-blind, parallel-group study evaluated the
lipid-altering efficacy and safety of simvastatin, 80 mg/day, a dose
twice the current maximum recommended dose. Male and female
hypercholesterolemic patients (n = 521) were randomly assigned in
a ratio of 2:3 to receive simvastatin 40 or 80 mg once daily,
respectively, for 24 weeks in conjunction with a lipid-lowering diet.
LDL decreased significantly (P < 0.001) from baseline
at weeks 18 and 24, with mean reduction of 38% and 46% for the 40-
and 80-mg groups. One third of patients on the 40- and 80-mg doses
achieved an LDL cholesterol reduction of 46% and 53%, respectively.
Decreases in apolipoprotein B, total cholesterol, and triglycerides
were also significantly greater among patients receiving 80 mg/day.
Simvastatin, 80 mg/day, provided substantial reductions in LDL
cholesterol, allowing most patients to reach target levels; it also had
an excellent safety and tolerability profile.
C. Effects of statins on new coronary events and survival
1. Primary prevention of heart disease. The AFCAPs/TexCPS
study examined the effect of 20 mg daily of lovastatin (or 40 mg/day if
LDL cholesterol remained above 110 mg/dl) vs. placebo as
primary prevention of acute coronary events (32). The study involved
5,608 men and 997 women without a prior history of cardiovascular
disease and with average LDL cholesterol and below-average HDL
cholesterol levels. Overall, the RR of first coronary event was 0.63
after an average follow-up of 5.2 yr in patients receiving lovastatin.
This group also had a decreased RR for myocardial infarction, unstable
angina, coronary revascularization procedures, coronary events, and
cardiovascular events. The effect of treatment with lovastatin on the
rate of first acute major coronary events (MCEs) was relatively greater
in women than in men (46% vs. 37% reduction); however, the
actual number of women who had a primary end point was small (20 of
997), and there were no statistically significant differences in
treatment effects between sexes.
D. Other agents with potential for prevention of cardiovascular
disease
1. Tamoxifen. Most of the SERMS exert both estrogen agonist
and estrogen antagonistic effects, depending upon the tissue examined.
Tamoxifen, used for prevention and treatment of breast cancer, exerts
estrogenic effects on the liver and thus lowers LDL cholesterol levels.
Lipid data are available from the studies of Love et al.
(33), which compared basal lipid levels with those obtained after 5 yr
of receiving either placebo or tamoxifen in 62 women (Fig. 1). After 5
yr of use, tamoxifen decreased LDL cholesterol concentrations by 30%.
HDL cholesterol also fell by 6.4% (7% for placebo) and
triglycerides increased 45.6% (23.8% for placebo). Lipoprotein
(a) was decreased with tamoxifen by 3.5% vs. a 1% decrease
with placebo. Data from the Royal Marsden Hospital tamoxifen
chemoprevention trial demonstrated similar effects on lipid levels (2, 34). Thus, tamoxifen appears to decrease LDL cholesterol, to reduce
lipoprotein (a), and to increase triglycerides (33, 34, 35).
The Tamoxifen Prevention Trial was originally powered to detect a cardiovascular protective effect. However, rather than the two-thirds projected, only one-third of accrued patients were postmenopausal. For this reason, the number of cardiovascular events was insufficient for adequate analysis. Nonetheless, the trend toward a similar number of events in tamoxifen- as in placebo-treated patients suggests that tamoxifen may not be cardioprotective (36).
2. Raloxifene. This SERM was recently approved by the Food and
Drug Administration (FDA) for the prevention of osteoporosis. In
preclinical studies in cholesterol-fed rabbits, raloxifene reduced
aortic accumulation of cholesterol (37), suggesting the potential of
inhibiting atherosclerosis. In randomized studies on lipid effects in
humans (38, 39), actions similar to those of estrogens on lipids but of
somewhat smaller magnitude are observed (Fig. 1). Reductions in total
cholesterol of 6.6% and LDL by 10.9% are seen with no change in HDL
or triglycerides. Both lipoprotein (a) and fibrinogen are reduced.
While not directly compared, this corresponds to an increase of 10.6%
in HDL and an increase in 10% in triglycerides with HRT.
Raloxifene exerts direct effects on the cardiovascular system as well as inducing lipid alterations. However, in ovariectomized placebo-treated cynomolgous monkeys (40), neither vasoactive effects nor protective effects against atherosclerosis were seen at either low or high doses of raloxifene. In comparison (according to blood levels), very high dose Premarin provided nearly complete protection. A prospective study is in progress (RUTH) to study cardiovascular endpoints during raloxifene treatment of postmenopausal women at risk for cardiovascular events.
3. Phytoestrogens. Phytoestrogens are plant-derived compounds that are isoflavones, bind to estrogen receptors, and have both estrogen agonist and antagonist properties. Soybeans are a rich source of the phytoestrogens, genistein, daidzein, and glytein. Populations with high levels of soy intake have lower rates of coronary heart disease and breast cancer (41, 42, 43, 44, 45). However, soy research is in its infancy. Animal studies and early human studies suggest that soy/isoflavones may inhibit bone resorption and/or stimulate bone formation (41, 46). Male cynomolgus monkeys fed high-soy diets showed a marked reduction in coronary artery disease, with an estrogen-like effect on coronary artery reactivity (47, 48, 49). Enhanced coronary vascular reactivity has been found in atherosclerotic female macaques (50). A recent meta-analysis (51) of the effect of soy on cholesterol in humans revealed that 47 g daily of soy was associated with a 12.9% decrease in LDL-cholesterol, 9.3% decrease in total cholesterol, and no change in HDL cholesterol. Interestingly, the greatest effect was seen in those with the highest pretreatment cholesterol. The amounts of total soy isoflavones that exert clinical effects approximate 40–80 mg/day (2, 41, 52).
E. Secondary prevention of cardiovascular disease
1. Estrogen. Nonrandomized secondary prevention studies
(53, 54, 55, 56) suggest that HRT use in women with established cardiovascular
disease reduces risk of death and future cardiovascular events.
Surprisingly, the randomized, prospective HERS trial (57) could not
confirm this beneficial effect. This study involved 2,763 women, mean
age 66.7 with severe coronary heart disease (CHD) who used continuous
combined estrogen (conjugated equine estrogens 0.625 mg) and
progesterone (MPA 2.5 mg) or placebo, with 4.1 yr of follow-up. The HRT
group showed an increase in CHD and mortality at year 1 when compared
with those using placebo. With continued use, it appeared that a
beneficial effect of HRT developed over time, with fewer deaths in
years 4 and 5. However, at the end of 5 yr on study, the number of
deaths, heart attacks, and CHD rates did not differ between the two
groups. Increases in thromboembolism of 4.1/1000 woman years (RR 2.89),
and gallstones (RR 1.38) were seen. Estrogen effects on lipid were
found with an 11% decrease in LDL cholesterol and a 10% increase in
HDL cholesterol. Many believe that the early death rate represented a
prothrombotic effect of estrogen and that the lack of differences
between groups was due to a detrimental effect of MPA, the progestin
used in the hormone group. The studies of Clarkson and associates in
monkeys (58) demonstrated that MPA could completely abrogate the
ameliorative effects of estrogens on the coronary arteries. The HERS
study has called into question the efficacy of estrogens for secondary
prevention. Definitive data regarding estrogens and secondary
prevention will probably become available with completion of the
Women’s Health Initiative.
Recent noncontrolled but prospective studies show different results from those reported in the HERS trial (59, 60, 61). Sullivan et al. (59) compared survival after coronary bypass surgery in women receiving and not receiving estrogen. The 10-yr survival rate in estrogen users was 81.4% compared with 65.1% in nonusers. Another study examined women after atherectomy in whom estrogen inhibited the rate of restenosis. The mean loss in minimal luminal diameter was -0.13 mm in estrogen users at 6 months after the procedure and -0.46 mm in nonusers, (P = 0.0006) (60). Clearly, longer term prospective randomized trials comparing estrogen/non-MPA progestin combinations with placebo are needed. In the meantime, the efficacy of estrogens for secondary prevention of heart disease must be held in question.
2. Statins. Randomized, prospective, controlled trials demonstrate that the statins reduce new cardiovascular events by approximately 30% and have favorable side effect profiles. Unfortunately, the statin studies have comprised a much larger proportion of men than women. In the Scandinavian simvastatin survival study, only 827 of 4,444 subjects, or 19%, were women (62, 63). Of the 52 deaths of women in the trial, only 30 were a result of CHD, 17/26 in the placebo group and 13/27 in the simvastatin group. According to Pederson et al., the probability that a woman would escape an MCE was 77.7% in the placebo group and 85.1% in the simvastatin group with RR of 0.65 (95% CI 0.47–0.90; P = 0.010). In the placebo group, mortality rate for women was less than half that for men. With only 52 deaths among women, demonstration of improved survival in women as a separate subgroup was unlikely. Nevertheless, simvastatin did reduce the risk of MCEs in women to about the same extent as it did in men. It also improved survival in patients aged 60 or more. This is the first trial to show that cholesterol lowering reduces MCEs in women and improves survival in older patients.
When results from men and women were combined (62, 63), it was found that simvastatin produced highly significant reductions in the risk of death and morbidity in patients with CHD followed for a median of 5.4 yr, relative to patients receiving standard care. Over the median follow-up period of 5.4 yr, one or more MCEs occurred in 622 (28%) of the 2,223 patients in the placebo group and 431 (19%) of the 2,221 patients in the simvastatin group for a highly significant 34% risk reduction with P < 0.00001. The results indicate that the addition of 20–40 mg simvastatin/day to the treatment regimens of CHD patients, with characteristics similar to those of postmenopausal women, should be beneficial.
The Cox proportional hazards model was used to assess the relationship between lipid values (baseline, year 1, and percent change from baseline at year 1) and MCEs. The reduction in MCEs within the simvastatin group was highly correlated with on-treatment levels and changes from baseline in total and LDL cholesterol, apolipoprotein B, and less so with HDL cholesterol, with no clear relationship with triglycerides. The authors estimate that each additional 1% reduction in LDL cholesterol reduces MCE risk by 1.7% (95% CI, 1.0% to 2.4%; P < 0.00001). Thus, the beneficial effect of simvastatin in individual patients in this study was determined mainly by the magnitude of the change in LDL cholesterol. These findings are consistent with current guidelines that emphasize aggressive reduction of this lipid in CHD patients.
Conclusions from these non "head to head" comparisons of the statins with HRT for primary and secondary prevention can only be tentative. Improvement of survival of 30% with the statins and 30–50% with estrogens are similar, and no data exist to prove either therapy superior to the other. In actuality, only the statins have been demonstrated by proper randomized trials to prevent MCEs. Based upon this analysis, it is reasonable to consider the statins as appropriate estrogen alternatives for the primary or secondary prevention of heart disease.
|
III. Prevention and Treatment of Osteoporosis |
|---|
|
TopAbstractI. IntroductionII. Prevention of Cardiovascular...III. Prevention and Treatment...IV. Urogenital AtrophyV. Vasomotor InstabilityVI. Central Nervous System...VII. Safety and Side...VIII. ConclusionsReferences |
|---|
A. Estrogen therapy for osteoporosis
Estrogen therapy reduces bone resorption and has been used
to both prevent and treat osteoporosis. The PEPI trial (69) revealed
that HRT improved bone mineral density by up to 4% in the lumbar spine
and 2% in the hip at 3 yr. To have the most protective effect on bone
density, estrogen therapy is best initiated at menopause. Nonetheless,
Ettinger and Grady (70) suggest that estrogen given to older
osteoporotic women can also increase bone density 5–10% and reduce
fracture risk. Based on epidemiological studies (71, 72, 73), HRT users
have a decreased risk of fracture (RR of 0.45). Studies suggest that
women may need only 0.3 mg of esterified estrogen (74, 75) plus 1,500
mg daily calcium to preserve bone, a regimen that causes little
mastalgia, bleeding, or endometrial hyperplasia, although there are no
fracture data. Whether bone loss is accelerated after discontinuation
of estrogen is controversial. However, 10 yr after discontinuation, the
bone mineral density and fracture risks were similar in women who had
used estrogen compared with those who had not (76).
1. Fracture data. Data regarding fracture risk reduction with estrogens derives almost exclusively from observational studies (77, 78, 79). One clinical trial of transdermal estrogen therapy in 75 postmenopausal women with osteoporosis observed for 1 yr revealed an RR of vertebral fractures of 0.39 compared with placebo (80). A series of 20 studies (79) revealed a risk reduction ranging from 80% to no effect in some studies. The reduction appears to average approximately 50%. The analysis of Grady et al. (8) estimated a lifetime reduction of vertebral fracture of 50% and hip fracture of 30% with long-term use of estrogen. A prospective cohort study (The Study of Osteoporotic Fractures) (81) among 9,704 women 65 and older, found the RR for nonspinal fractures for women receiving estrogen to be 0.66. Current users experienced a reduced hip fracture RR of 0.60. For women who started estrogen within 5 yr of menopause, the RR was 0.29 for hip fracture and 0.50 for all nonspinal fractures. More precise comparisons must await further studies.
B. Bisphosphonates for prevention of osteoporosis
Antiresorptive agents such as the bisphosphonates, calcitonin,
tamoxifen, and raloxifene are useful for the prevention or treatment of
osteoporosis. Bisphosphonates are stable analogs of pyrophosphate and
exert effects primarily on the osteoclast. Newer bisphosphonates such
as risendronate are being studied.
1. Comparison of alendronate and estrogen. Data regarding the
relative efficacy of the bisphosphonates vs. estrogen are
available for effects on bone density but not for fracture prevention.
The most commonly used bisphosphonate, alendronate, appears to exert an
antiresorptive potency similar to that of estrogen when used at low
dose (5 mg) (Fig. 2). The Early
Postmenopausal Intervention Cohort (EPIC) (82) study of recently
menopausal women included placebo, alendronate (2.5 or 5 mg/day), and
open-label estrogen/progesterone. Patients receiving placebo plus
calcium lost bone. Those receiving either 2.5 or 5 mg of
alendronate/day increased bone mass between 1 to 2% over baseline but
not to the same extent as the estrogen/progesterone group at 2% (Fig. 2). With measurement of total body bone density at 2 yr,
estrogen/progesterone improved bone density almost 2%, while 5 mg of
alendronate increased this parameter 1%, and 2.5 mg of alendronate
maintained baseline levels compared with a 2% loss with placebo. Thus
for prevention of bone loss, both estrogen/progesterone and 5 mg of
alendronate are effective with slightly better response with
estrogen/progesterone.
|
Preliminary data presented in abstract from the FACET trial (Fosamax Addition to Continuing Estrogen Therapy) (86) suggest that there may be added benefit in combining two antiresorptive agents. The combination of alendronate to an established estrogen and or estrogen/progestin regimen significantly increased BMD over the addition of placebo, with no increase in side effects.
2. Fracture and BMD data with alendronate. No direct comparative data to evaluate the reduction of fracture risk with estrogen vs. bisphosphonates are yet available. Consequently, it is necessary to compare data from other randomized or observational studies to draw tentative conclusions. Three multicenter, double-blind, placebo-controlled studies provide compelling evidence of the vertebral antifracture efficacy of alendronate. The first (85) involved 994 postmenopausal women (mean age, 64 yr) with osteoporosis who received either placebo for 3 yr, alendronate, 5 mg/day, for 3 yr, alendronate, 10 mg/day, for 3 yr, or alendronate, 20 mg/day, for 2 yr, and then 5 mg/day for a third year. All patients received 500 mg elemental calcium. BMD increased in patients receiving all alendronate dosage regimens and decreased in patients receiving placebo. Vertebral fractures occurred in 6.2% of patients receiving placebo and 3.2% of patients receiving alendronate; this represented a 48% reduction in numbers of women sustaining fractures (P < 0.04). Two or more new vertebral fractures occurred in 4.2% of patients receiving placebo (15 of 355) and 0.6% of patients receiving alendronate (3 of 526), a risk reduction of 87%. Patients in the placebo group who sustained new fractures lost 23.3 mm in height. Alendronate-treated patients who sustained one or more fractures lost only 5.9 mm in height, consistent with less severe fractures. Nonvertebral fractures occurred in 60 of 590 women receiving placebo and 73 of 1,012 receiving alendronate. The cumulative incidences (placebo vs. alendronate) were 12.6% and 9%, a 29% reduction in risk compared with placebo (P < 0.05).
In the Fracture Intervention Trial (FIT) (87), 2,027 women (mean age, 71 yr) with one or more vertebral fractures at baseline and reduced BMD were randomized at 11 centers in the United States to receive either placebo (n = 1,005) or alendronate (n = 1,022; 5 mg/day for 2 yr, 10 mg/day in year 3). Calcium and vitamin D were supplemented if the diet contained less than 1,000 mg/day calcium. After 3 yr, BMD increased significantly by 6.2% above placebo at the spine and by 4.7% above placebo at the total hip region. The rate of new clinically apparent vertebral fractures decreased by 47% compared with placebo. Similar decreases were seen in the frequency of hip and wrist fractures, but not for other types of fractures.
A follow-up to this study examined women without preexisting vertebral fractures. Women received 5 mg daily of alendronate for 2 yr followed by 10 mg daily thereafter (88). Only those with osteoporosis (i.e., bone density at the femoral neck of >2.5 SD below the normal adult mean) responded with a significant reduction of clinically evident fractures. The radiologically detected fractures were reduced by 44% in the total group of women taking alendronate [RR 0.56; 95% confidence limit (CL), 0.39–0.80]. Subgroup analysis demonstrated a significant reduction only in those with a baseline bone density T score of >2.5 SD and close to a significant reduction in those with scores of -2.5 to -2.0 (RR 0.54; 95% CL, 0.28–1.04). Too few episodes occurred in the remainder to be meaningful (i.e., 10 vs. 8 events).
These well designed trials suggest that alendronate is effective at preventing vertebral and nonvertebral fractures when compared with placebo, and 10 mg appears to be the most effective dose. Although hip fractures were reduced, the number of actual fractures was small.
C. Nasal calcitonin for osteoporosis
Nasal calcitonin is approved for treatment of osteoporosis in
women who are 5 or more years postmenopausal. It is not effective in
preventing bone loss in early postmenopausal women (89, 90). A
randomized double-blind trial of approximately 230 postmenopausal women
with osteoporosis in each group received either placebo or three doses
of nasal calcitonin (100, 200, and 400 IU/day) for 3 yr (89). An
increase of 1.0–1.5% over baseline in lumbar spine but not other
skeletal sites was seen at the end of year 1 (significant) with no dose
response apparent. Statistical significance over placebo was lost in
years 2 and 3.
1. Fracture data with nasal calcitonin. Overgaard et al. (91) studied the effects of intranasal salmon calcitonin in a 2-yr, double-blind, placebo-controlled trial of women aged 68–72 yr randomized to receive 50, 100, or 200 IU calcitonin or placebo daily plus 500-mg calcium supplement. Among 162 women completing the study, spinal BMD increased by 1% in the placebo group and by 3% in the group receiving 200 IU calcitonin. The numbers of fractures varied according to the method of assessment of vertebral morphometry. By one method, fractures occurred in seven patients receiving placebo and five receiving calcitonin; by the second method, fractures occurred in six patients receiving placebo and four receiving calcitonin. Peripheral fractures occurred in two placebo- and two calcitonin-treated patients. Although a decreased number of fractures was seen, pooling of all doses was not preplanned, and the small numbers of fractures make results uncertain.
Four-year interim results (92) from the 5-yr multicenter PROOF study (Prevent Recurrence of Osteoporotic Fractures) were presented in abstract form at the 1998 European Congress on Osteoporosis; 1,255 postmenopausal women with established osteoporosis were randomized to receive either placebo, or salmon calcitonin nasal spray (100, 200, or 400 IU/day) plus 1000 mg of calcium and 400 IU of vitamin D. At 4 yr, there was a 36% reduction in RR of new fracture with 200 IU compared with placebo (P = 0.020). A reduction of 18% and 23% in new fractures was seen among those treated with 100 IU or 400 IU compared with placebo. The mean increment in BMD over baseline for placebo, 100 IU, 200 IU, and 400 IU was 0.7, 1.2, 1.2 and 1.6%, respectively. The increases in lumbar spine BMD were statistically significantly increased in all treatment groups compared with placebo over baseline and at 2 yr compared with placebo, and up to 3 yr for the 400 IU. Despite modest increases in BMD, the analysis in the abstract suggests nasal calcitonin spray reduces the risk of new vertebral fractures in postmenopausal women with established osteoporosis. The data suggest a protective effect on hip fractures, but the study is not adequately powered to detect differences in hip fracture rates. At 4 yr, a 51% reduction in hip/femur fractures was noted in the 200-IU dose but the numbers are quite small: 1.3% (4/315) of patients in 200-IU group vs. 2.6% (8/305 of patients in the placebo group.
D. SERMs for prevention of osteoporosis
Studies in postmenopausal women using tamoxifen for prevention of
breast cancer have shown average increases in BMD of 1–2% per year
(93). Premenopausal women show a transient bone loss for the first 2 yr
with tamoxifen, felt to be due to the antiestrogenic effect of
tamoxifen. In the Tamoxifen Prevention Trial (36), a nonsignificant
reduction in fractures was observed.
Three large randomized placebo-controlled osteoporosis prevention trials (38, 39, 94, 96) with raloxifene demonstrated significant increases in BMD of 2% over placebo in hip, spine, and total body. Raloxifene acts as an antiresorptive agent, as does estrogen. Compared with placebo, there is between a 2–2.5% increase in spine, total hip, and total body (P < 0.001 for all comparisons).
Raloxifene increased total-body BMD from 1.8% to 2.5% over placebo after 24 months (38). This effect was lower than that seen with conjugated equine estrogens and MPA (5.2%) or 5 mg of alendronate/day (5.8%) after 24 months (82). In the total hip, the raloxifene group improved to a greater extent than those given placebo (1.8% to 2.3%). The effect was similar in magnitude to that observed in women receiving conjugated equine estrogens and MPA or 5 mg of alendronate (82). The effects were greater than effected by 200 IU of nasal calcitonin or tamoxifen.
1. Fracture data with raloxifene. The Multiple Outcomes of Raloxifene Evaluation (MORE) trial is a multicenter, double-blind, controlled study of 7,705 recently postmenopausal (average 4.5–5 yr) women with osteoporosis (mean age 66.5) treated with raloxifene (60 mg or 120 mg) or placebo. All received 500 mg of calcium and 400–600 IU vitamin D. Interim analysis at 2 yr (94) and at 3 yr (96), shows significant effects on vertebral fracture prevention of 38% (RR 0.62) in women with prevalent fractures and 52% (RR 0.48) without prior fractures.
E. Comparison among therapeutic agents for treatment of
osteoporosis (Fig. 2)
For women with documented osteoporosis, there are no direct
comparisons of effectiveness of different agents. Alendronate, 10 mg,
has been shown to increase density in the lumbar spine by 5–7% and
femoral neck by 4–6% (85). Estrogen alone or with progesterone
increased bone density by 2–5% in the lumbar spine and by 1–7% in
the hip in the PEPI trial (69). Nasal calcitonin at 200 IU improved
bone density 2–3% at 2 yr in lumbar spine (91). Raloxifene increased
BMD 2–3% per year (38), whereas tamoxifen increased BMD 1–2% (93).
While we have long-term epidemiological data concerning the effect of
estrogen on prevention of osteoporosis, we are lacking prospective
controlled data for estrogen and long-term data about alendronate,
nasal calcitonin, and raloxifene. Studies of efficacy of treatment of
osteoporosis with regards to prevention of fracture are limited by the
low annual rates of fractures. Uncertainty regarding efficacy and
safety may be related to design, execution, and interpretation of the
clinical studies rather than problems in the drugs themselves (95).
Head-to-head comparisons of various agents in preventing fracture are nonexistent. Efficacy is estimated by comparing results from different trials. Epidemiological studies suggest a relative reduction of fracture rate of 55% with estrogen use, even for women over 70 (8, 81). For alendronate, the decrease in lumbar spine and hip fractures is 50% (87); for nasal calcitonin, the decrease in lumbar fractures is 36% (92) and for raloxifene 38–52% (96). Based upon these data, one can conclude that alendronate, nasal calcitonin, and raloxifene could serve as alternatives for estrogen for prevention and treatment of osteoporosis.
|
IV. Urogenital Atrophy |
|---|
|
TopAbstractI. IntroductionII. Prevention of Cardiovascular...III. Prevention and Treatment...IV. Urogenital AtrophyV. Vasomotor InstabilityVI. Central Nervous System...VII. Safety and Side...VIII. ConclusionsReferences |
|---|
Another approach is to use sufficiently low doses of vaginal estrogen to achieve local effects without systemic estrogen absorption. In one study examining this possibility (99), plasma estrone and estradiol levels were measured in women receiving vaginal Premarin (Wyeth-Ayerst Laboratories, Philadelphia, PA) with gradual increases in doses from 0.3–2.5 g/day. Significant increases in systemic levels of estradiol were detected, and levels increased up to 60 pg/ml, a level comparable to that seen with oral Premarin. However, the lowest dose (0.3 g/day) produced complete maturation of vaginal mucosa with only minimal increases in plasma estrone and estradiol. Handa et al. (99) recently observed that patients on this regimen showed no increase in plasma estrone over baseline after 6 months of use. Disadvantages of the vaginal method involve irregular application intervals, bolus absorption, and low absorption capacity of a fat-based vehicle, necessitating use of emollient (the stickiness of which may lessen compliance).
Newer methods of delivering estrogen locally into the vagina without systemic absorption include a vaginal estrogen ring device (Estring, Pharmacia and Upjohn, Kalamazoo, MI) (100, 101) and low-dose vaginal creams currently being tested and developed. The vaginal ring provides nearly complete relief of symptoms. In open label studies (but with blinded review of vaginal cytology), similar efficacy was observed with the vaginal ring device as with conjugated systemic estrogens with respect to vaginal secretions, color, tissue integrity, urethral meatus integrity, and patient acceptance. Data reviewed indicate minimal systemic absorption from this device (2). Only 4% of women using the ring had endometrial thickness greater than 5 mm measured by ultrasound compared with 10% using the vaginal cream. Three percent experienced vaginal withdrawal bleeding from the ring compared with 21% using vaginal cream.
Our preliminary data support the possibility that very low doses of estradiol given vaginally can achieve local vaginal effects without systemic absorption of estradiol (102). Currently available RIAs for estradiol may not be sufficiently sensitive to detect minor increments in plasma estradiol with use of Estring or vaginal creams (102). Further studies are required to determine the safety of vaginal estradiol in women surviving breast cancer.
|
V. Vasomotor Instability |
|---|
|
TopAbstractI. IntroductionII. Prevention of Cardiovascular...III. Prevention and Treatment...IV. Urogenital AtrophyV. Vasomotor InstabilityVI. Central Nervous System...VII. Safety and Side...VIII. ConclusionsReferences |
|---|
). A major
side effect is weight gain. Long-term safety of Megestrol acetate in
patients surviving breast cancer is not known. The SSRI (Selective
Serotonin Reuptake Inhibitor) drugs may show a decrease in hot flashes
over placebo (107). Studies (41, 108) that have used phytoestrogens for
relief of hot flashes show conflicting results, varying from little
effect, to a significant reduction, to a reduction in the severity but
not frequency of hot flashes.
|
|
VI. Central Nervous System (CNS) Symptoms |
|---|
|
TopAbstractI. IntroductionII. Prevention of Cardiovascular...III. Prevention and Treatment...IV. Urogenital AtrophyV. Vasomotor InstabilityVI. Central Nervous System...VII. Safety and Side...VIII. ConclusionsReferences |
|---|
|
VII. Safety and Side Effects of Alternatives to Estrogen |
|---|
|
TopAbstractI. IntroductionII. Prevention of Cardiovascular...III. Prevention and Treatment...IV. Urogenital AtrophyV. Vasomotor InstabilityVI. Central Nervous System...VII. Safety and Side...VIII. ConclusionsReferences |
|---|
For women with breast cancer, bisphosphonates may be particularly attractive because of reduced risk of fractures from skeletal metastasis and the potential for slowing of progression of metastasis (120). By slowing remodeling, bone apparently becomes less hospitable for metastasis to grow. Reports of gastrointestinal symptoms including erosive esophagitis (121) have been reported with alendronate, although no significant differences have been found in controlled studies compared with placebo. Ettinger et al. (122) recently reported a retrospective data base analysis and telephone interview survey of 812 women with mean age of 68.7. They found that one of three women receiving alendronate for osteoporosis had upper gastrointestinal complaints in a large Kaiser HMO with one of eight women requiring medical treatment of acid-related disorders. Excluding women with prior history of acid-related disorders, the RR was 1.6 for alendronate users over nonusers. A significant trend was seen with increasing incidence as women age and with concurrent use of nonsteroidal antiinflammatory drugs. This study underscores the need to minimize the risk of esophagitis and to improve drug absorption by recommending that alendronate be taken first thing in the morning, on an empty stomach, with 8 ounces of water, and with the patient remaining upright for 30 min.
The nasal form of calcitonin has few side effects, nasal discomfort or ulceration, nausea, facial flushing, and diarrhea, and like subcutaneous calcitonin, may have an analgesic effect on bone pain (123).
With regard to raloxifene’s effect on breast, studies suggest a 50–70% decrease in risk of breast cancer at 3 yr. Preliminary data, presented in abstract form, from more than 7,000 women who participated in preliminary trials shows a 71% reduction of breast cancer and a decrease in endometrial cancer in the raloxifene group (124). No data are yet available on long-term rates of breast cancer with raloxifene, its use in women at high risk for breast cancer, or its use in woman with prior breast cancer. Uterine thickness was unchanged from placebo with no stimulation of atrophic endometrium. Side effects include leg cramps and a 2.5 increased RR of deep vein thrombosis. Raloxifene does not appear to be an effective agent to relieve the symptoms of vasomotor insufficiency. No effect was seen on vaginitis, migraine, headache, anxiety, or emotional lability. Very few data are available with respect to cognitive function, mood, or memory (38, 39).
Tamoxifen has been used primarily for prevention of recurrent breast cancer but is now approved for prevention of breast cancer in women at high risk. The Tamoxifen Prevention Trial (36), a prospective, randomized, blinded, controlled study of 13,000 high-risk women found a 45% decrease in development of breast cancer after 4.5 yr of treatment. Risks include increased risk of uterine cancer, cataracts, and blood clots (36). The absolute excess of deaths from endometrial cancer (125) during the whole decade after randomization was, in each of the three tamoxifen duration categories, about 1 or 2 per 1,000 (corresponding to an annual excess of about 0.2 per 1,000). There was a nonsignificant tendency for the excess of endometrial cancer deaths to be greater in the trials of longer duration of tamoxifen. Although this trend may well be real, the absolute excess was not large. Among 3,673 women allocated to 5 yr of tamoxifen in trials that provided cause-of-death information, there were seven endometrial cancer deaths and the cumulative risk during the whole of the first decade was about two deaths from endometrial cancer per 1,000 (81, 125).
Physicians should inquire routinely about the use of dietary supplements and other products and be alert to potential adverse effects with products, combinations of products, contaminants, or interactions with other drugs (126). It may be necessary to monitor with serial blood counts, chemistries, and liver function tests. Herbs with known harmful effects include those that contain carcinogenic pyrrolizidine alkaloids (borage, borage oils, coltsfoot, comfrey, and life root) or with hepatotoxic effects (life root, germander, chaparral, and some Chinese medicine combinations) (127). Eighteen cases of chaparral hepatotoxicity were reported to the FDA between 1992 and 1994 (128). Four developed cirrhosis and two had fulminant liver failure requiring liver transplantation. Kidney failure and kidney transplantation have also been reported. Licorice in high doses can cause pseudoaldosteronism. Other herbs reported harmful include pennyroyal, pokeroot (which can be fatal in children), sassafras (which contains the carcinogen safrole), and calamus (which contain the carcinogen cisisoasarone).
|
VIII. Conclusions |
|---|
|
TopAbstractI. IntroductionII. Prevention of Cardiovascular...III. Prevention and Treatment...IV. Urogenital AtrophyV. Vasomotor InstabilityVI. Central Nervous System...VII. Safety and Side...VIII. ConclusionsReferences |
|---|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1 Substantial data cited in this review are
published only in abstract and lack full peer review (86 92 94 96 102 124 ). In the opinion of the authors, these data generally
represent the results of large randomized trials in areas that are
rapidly advancing. They should provide information regarding clinical
decision making but should not yet be viewed as definitive. 
|
References |
|---|
|
TopAbstractI. IntroductionII. Prevention of Cardiovascular...III. Prevention and Treatment...IV. Urogenital AtrophyV. Vasomotor InstabilityVI. Central Nervous System...VII. Safety and Side...VIII. ConclusionsReferences |
|---|
