University of California Coordinating Center, Department of Medicine, Prevention Sciences Group, University of California, San Francisco, California 94105
Address all correspondence and requests for reprints to: Katharina Modelska, M.D., University of California, 74 New Montgomery Street, Suite 600, San Francisco, California 94105. E-mail: kmodelska{at}psg.ucsf.edu
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() |
---|
Tibolone itself has no biological activity; its effects are the results of the activity of its metabolites on various tissues (4). After administration, tibolone is quickly metabolized into 3-hydroxytibolone (3
-OH-tibolone) and 3ß-OH-tibolone compounds, which are also present in an inactive, sulfated form (4) (Fig. 1
). A third compound, the
4-isomer, is formed from tibolone directly or from the 3ß-OH-metabolites. The 3
- and 3ß-OH-metabolites bind solely to the ER, whereas the
4-isomer has affinity for PR and AR, but not ER.
|
Tibolone given orally is rapidly absorbed, appearing in the plasma within 30 min and peaking in 4 h. Tibolone is metabolized mainly in the liver and is excreted in the urine and feces. The elimination half-life is approximately 45 h.
In this review we will discuss the results from the randomized, controlled, and double blind trials (RCTs) of tibolone, with respect to its effects on climacteric symptoms, sexual function, endometrial and breast tissue, lipid metabolism, and bone mineral density (BMD).
![]() |
Materials and Methods |
---|
![]() ![]() ![]() ![]() ![]() |
---|
From over 390 clinical reports we have found 21 unique RCTs that assessed the clinical effects of tibolone (2.5 mg/d) in postmenopausal women. Summaries of the study design, duration of the trials, participant demographics, and different outcomes are presented in Tables 14. If a trial measured many different effects of tibolone in postmenopausal women, we presented the results of this 1 trial in more than 1 table. Selection criteria and characteristics of the study participants are described in the body of this review. Because the aim of this review is to summarize tibolones effects on postmenopausal women, we excluded trials involving premenopausal women. We also excluded retrospective analyses, nonrandomized and open label studies, and trials in which the use of a placebo was not specifically stated.
|
|
|
|
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() |
---|
Menopause-associated symptoms impair quality of life for many women. More than 75% of postmenopausal women experience hot flushes and sweating. Other symptoms, such as insomnia, headache, or fatigue, as well as changes in mood and libido may result directly from menopause or indirectly, such as effects of hot flushes on sleep. We found eight unique RCTs that assessed the effects of tibolone on climacteric symptoms (Table 1). In the first five RCTs (7, 8, 9, 10, 11) the effects of tibolone on climacteric symptoms were compared with those of placebo. In addition, one trial compared tibolone to E2 valerate (12), and two unique RCTs compared tibolone to the combination of E2 and noresthisterone acetate (E2/NETA) (13, 14). The results of the first combination trial with E2/NETA were presented in two different publications by Nathorst-Böös (13) and Hammar (15), and the results of the second trial were presented recently by Dören (14).
The selection criteria for all subjects involved in the RCTs were hot flushes, sweating, and other postmenopausal- related symptoms, such as fatigue, headaches, changes in mood, insomnia, and decrease in libido. All subjects selected for these RCTs had been postmenopausal for at least 12 months and free of any treatment for at least 3 months. Women after surgical menopause were included in two trials (12, 16). Excluded were patients who received any hormonal or psychotropic therapy in the 612 wk before the trials started and patients with a previous history of gynecological or malignant disease (7, 10, 11, 17). Also excluded were patients with hypertension (12, 14); as well as patients with any cardiovascular or cerebrovascular disease, thromboembolic disorder (18, 19, 20), history of renal or liver disease (14, 20); and patients who were treated with medication known to affect coagulation, fibrinolysis, or lipid or bone metabolism (21, 22). Women who smoked cigarettes were excluded in several trials (23, 24). In addition, tobacco use equivalent to more than five cigarettes per d was an exclusion criteria in one trial (25).
Effects of tibolone on hot flushes and sweating
Compared with placebo, most RCTs reported a significant reduction in hot flushes and sweating in women taking tibolone (Table 1). However, the magnitude of changes was not reported in all of the publications. When the effects of tibolone on hot flushes were compared with the effects of hormonal replacement therapy, a similar reduction of hot flushes was seen with both therapies. Because there was no placebo group in these trials, the magnitude of improvement of hot flushes and sweating could not be ascertained.
Effects of tibolone on other postmenopausal symptoms
Many trials reported a beneficial effect of tibolone on fatigability, frequency of headaches, psychological instability, and insomnia (7, 10, 11). Genazzani (9) found that the relief of hot flushes from the second month of treatment was associated with improvement of both mood and insomnia. The reported changes in postmenopausal symptoms varied among the trials, which may be due to variations in study duration, social differences in multicultural samples, and different scoring systems used. Two of the trials were very short (8, 9), and two trials did not describe clearly the methods by which the symptoms were measured (9, 12).
Effects of tibolone on mood
In one small trial of young women who had undergone oophorectomy and hysterectomy, Crona et al. (12) found that tibolone and E2 valerate reduced hot flushes and improved mood to a similar degree. In another trial, Genazzani (9) found that tibolone increased the concentration of ß-endorphins and proposed that this might contribute to the improved mood in postmenopausal women. However, mood was not directly assessed in this trial.
Effects of tibolone on sexual function
It is believed that tibolone may improve sexual function. This is plausible because T has been shown to increase libido and frequency of sexual activities (26, 27), and tibolone has androgenic activity (4). Specifically, as noted earlier, the 4-isomer of tibolone stimulates AR. In addition, tibolone may act indirectly to decrease SHBG concentrations and thereby increase the availability of T. Postmenopausal women have, on the average, 30% lower circulating T levels than those measured in premenopausal women (26, 27). In one recent trial, Dören and colleagues (14) found that women treated with tibolone had higher free T levels and lower SHBG levels than women treated with E2/NETA. However, sexual function was not assessed in this trial.
Two randomized and double-blind trials assessed the effects of tibolone on sexual function compared with placebo (10, 28). In the first trial by Nevinny-Stickel (10), performed almost 20 yr ago, there was no significant improvement in regard to libido in women taking tibolone. On the contrary, the recent double blind and placebo-controlled trial by Laan (28) has shown that treatment with tibolone significantly improved the physiological aspects of sexual function in postmenopausal women, such as vaginal blood flow and vaginal lubrication, and subjective measures, such as sexual desire and arousability, but there was no difference in the frequency of sexual intercourse, nonpenetrative sexual activity, or initiation and rejection of sexual activity between women who were taking tibolone vs. those receiving placebo (28). However, the small sample size in the subgroups may have diminished the power to detect statistical differences between these variables.
In two other trials, tibolone was compared with the combination of E2/NETA (13, 14). In the first trial by Nathorst-Böös (13), the improvement of sexual function with regard to frequency, satisfaction, and enjoyment was measured in women taking tibolone vs. those taking E2/NETA. However, almost 30% of the subjects dropped out, and the reasons for discontinuation were not reported. In the second trial by Dören (14), tibolone decreased SHBG and increased free T, as mentioned previously. The problem with the interpretation of both trials by Nathorst-Böös and Dören was that E2 increases SHBG levels and there were no placebo-control groups used.
Thus, although Laan and colleagues (28) have shown that tibolone increased sexual interest and desire in postmenopausal women compared with the effects of placebo, tibolones ability to increase sexual activity among postmenopausal women remains unclear and needs to be answered by another placebo-controlled trial with a larger sample size. If tibolone improves sexual function, it will be important to determine whether it does so by increasing free T in women with very low T concentrations or in all women regardless of androgen levels.
Tibolone and its effects on vaginal bleeding and endometrium
In the endometrium, tibolone is transformed into the 4-isomer by 3ß-hydroxysteroid dehydrogenase isomerase. This metabolite does not have estrogenic activity, but does have intrinsic progestagenic activity; therefore, it does not stimulate the endometrial tissue (4, 5). Thus, therapy with tibolone may not require the addition of progestagen to protect the endometrium (16).
Only 1 trial with a total of 94 women has assessed the effects of tibolone on vaginal bleeding compared with placebo (Table 2). In this trial, Berning (24) reported that 51% of women taking tibolone and 22% of women taking placebo had incidences of vaginal bleeding during 96 wk of treatment. He concluded that early postmenopausal women who took tibolone were 22.5 times more likely to sustain vaginal bleeding compared with women taking placebo. In this trial endometrial morphology assessed by curettage showed no evidence of endometrial stimulation in women with bleeding. However, hysteroscopy was not performed in the patients who bled, and therefore precise information about intrauterine abnormalities could not be collected. Berning also investigated the relationship among E2 levels, endometrial morphology, and vaginal bleeding among his study subjects. Interestingly, there was no relationship between E2 levels and the occurrence of bleeding in women who were taking tibolone.
Two other trials have compared the effects of tibolone with those of E2/NETA on vaginal bleeding (15, 29) (Table 2). The results showed that the combination of E2/NETA caused approximately twice as many episodes of bleeding as tibolone. Women who bled while taking tibolone were younger at menopause or were recently menopausal and therefore may have had higher remaining endogenous estrogen production than women who did not bleed. For this reason, tibolone is currently recommended only for women who are at least 1 yr postmenopausal.
In summary, treatment with tibolone caused significantly more vaginal bleeding than placebo, but about half as much vaginal bleeding as E2/NETA. The endometrial effects of tibolone need to be assessed in larger trials.
Tibolone and its effects on lipids
In postmenopausal women, there is typically an increase in total cholesterol and triglycerides, mostly due to an increase in low density lipoprotein cholesterol (LDL-C), whereas high density lipoprotein cholesterol (HDL-C) remains unchanged. Seven trials assessed the effects of tibolone on lipids and the clotting factors compared with those of placebo, and two other trials assessed these effects compared with those of E2/NETA (Table 3). Compared with placebo, tibolone reduced HDL-C by approximately 34% (11, 21, 22) and decreased triglycerides by approximately 25%, but had no effect on LDL-C and lipoprotein(a) (9, 17, 20, 21). These effects distinguish tibolone from estrogens, which decrease LDL-C and lipoprotein(a) and increase HDL-C. The reduction in HDL-C by tibolone may be due to an androgenic effect of the compound
4-isomer on hepatic lipase.
The androgenic effects of tibolone caused a decrease in SHBG concentrations and an increase in fibrinolytic activities (18, 19, 21, 29). Compared with placebo, tibolone caused increases in hemoglobin, antithrombin III, plasminogen, and platelet count (18, 19, 30). However, the number of serum parameters measured in these trials was limited. Therefore, the clinical implication of the effects on lipid metabolism and hemostasis, as reported in these trials, is not clear. Definite conclusions with regard to risks for cardiovascular disease or venous thromboembolism cannot be drawn from these trials.
Tibolone and osteoporosis
Postmenopausal bone loss results from an increase in bone resorption after menopause. Tibolones action on bone is mediated via stimulation of ER (4, 23, 25, 31, 32). Six trials have assessed the effects of tibolone on bone mass in postmenopausal women vs. placebo (23, 25, 31, 33, 34, 35), and one trial evaluated the effects of tibolone vs. estrogens (36) (Table 4). In early postmenopausal women who took tibolone vs. placebo, a 15% net increase in lumbar BMD was measured with computed tomography (23) and an approximately 3% net increase in lumbar BMD was measured with dual-energy x-ray absorptiometry (DXA) (35). In late postmenopausal women, tibolone suppressed bone turnover and prevented bone loss, mainly at the lumbar spine, in two trials compared with placebo (25, 34). Specifically, women with established osteoporosis who were treated with tibolone had a 12% net increase in lumbar BMD vs. placebo, measured with dual photon absorptiometry over 2 yr (34) and a 5% net increase in lumbar BMD measured with DXA over 2 yr (25).
In summary, tibolone decreases bone turnover and significantly improves BMD, especially trabecular BMD. However, these trials are small, relatively short in duration (2 yr), and provide no data about the effects of tibolone on fracture risk. Therefore, long-term trials assessing fracture reduction with tibolone vs. placebo are required to determine the role of tibolone in the management of postmenopausal osteoporosis.
Tibolone and breast cancer
The effect of tibolone on the breast cells has been studied extensively (4, 37, 38, 39, 40, 41). In particular, two studies have shown the effects of tibolone on the hormone-dependent human breast cancer cells (37, 39). Chetrite (41) found that the 3-OH-metabolites of tibolone are strong inhibitors of sulfatase activity and weak inhibitors of 17ß-hydroxysteroid dehydrogenase activity; therefore, they inhibit the formation of active estrogens in the breast. Gompel (37) found that tibolone has antiproliferative and proapoptotic activities in breast cancer cells. Thus, in breast cells, tibolone slows down the proliferation rate as well as increases differentiation and apoptosis. This has led to the view that tibolone might reduce breast cancer risk.
One small RCT evaluated the effect of tibolone on mammographic density (42), whereas another assessed the effect of tibolone on breast tenderness (15). Colacurci showed that 1 yr of tibolone treatment did not affect breast density in postmenopausal women with normal breast tissue compared with a control group and a hormonal replacement therapy group (42). However, this trial was very small (n = 44) and may have missed clinically important effects. Hammar (15) reported that postmenopausal women taking tibolone experienced breast tenderness significantly less frequently than those who were taking E2/NETA, but there was no placebo control. Tibolone should be tested in large trials for its effect in reducing the risk of breast cancer.
Adverse effects of tibolone
The adverse effects of tibolone have not been addressed in placebo-controlled trials. Drop-out rates due to adverse effects were also not reported in the RTCs we reviewed. We found only two trials of tibolone against E2/NETA that reported adverse effects. However, these trials used estrogen as a comparison, which made it impossible to determine the adverse effects due to tibolone. Further RCTs on tibolone against placebo should comprehensively report adverse events in tibolone and placebo groups.
Limitation of this review
We did not attempt a meta-analysis for several reasons. First, there were substantial methodological differences between these trials. The evidence in the investigated areas did not permit formal analysis. Most areas we reviewed had too few placebo-controlled trials to combine in a meta-analysis. For example, there are only four trials with menopausal symptoms as outcome (Table 1) and only one trial involving patients with vaginal bleeding (Table 2
). The effects of tibolone on hot flushes, sweating, and vaginal dryness were assessed by either using a three-point scoring system (7, 10, 11) or a five-point scoring system (15). The measurements of postmenopausal complaints (changes in mood, irritability, changes in sexual function) were made using different scores and different scales. The majority of investigators used the McCoy Sex Scale Questionnaire to assess sexual function, whereas the investigators in Sweden used the Swedish version of the McCoy Sex Scale Questionnaire (13), which contained more questions. Also, the trials differed with regard to patient age and demographics and inclusion and exclusion criteria.
Conclusion
A few RCTs have shown the estrogenic effects of tibolone in reducing hot flushes and sweating in postmenopausal women. It was confirmed that tibolone increases BMD, especially in women with established osteoporosis. Tibolone also may have beneficial, androgenic effects on sexual function. However, there have been no placebo-controlled trials conducted to determine whether tibolone would improve sexual function in postmenopausal women and, if so, whether the effect depends on baseline T levels.
Therefore, we can conclude that tibolone significantly reduces hot flushes and sweating and increases BMD in postmenopausal women. Other effects of tibolone in postmenopausal women, such as its influence on lipid metabolism, hemostasis, and sexual function, are less certain. In addition, the long-term effects of tibolone, particularly in reducing fractures, breast cancer, and cardiovascular disease are still unknown.
![]() |
Acknowledgments |
---|
![]() |
Footnotes |
---|
Received June 11, 2001.
Accepted September 24, 2001.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() |
---|