Raloxifene (Evista; Eli Lilly and Co., Indianapolis, IN) is approved for the prevention and treatment of postmenopausal osteoporosis.1,2 It is a selective estrogen receptor modulator with estrogen agonistic activity on a number of targets, including bone and lipid metabolism, and estrogen antagonistic activity on breast and uterine tissues.3
Clinical data indicate that raloxifene use by postmenopausal women is not associated with any adverse vaginal symptoms.4 These studies excluded women who complained of vaginal atrophy. Women who have specific complaints attributed to vaginal atrophy are frequently treated with vaginal estrogen such as low-dose conjugated estrogen cream (Premarin; Wyeth-Ayerst Laboratories, Philadelphia, PA).5 The nonhormonal moisturizer (Replens; Columbia Laboratories Inc., Aventura, FL) is another accepted treatment modality.6,7 Because raloxifene is a selective estrogen receptor modulator, it could theoretically block the effects of local estrogens on the vaginal mucosa. However, nonhormonal moisturizer should be equally effective with or without concomitant administration of raloxifene. This study examined whether raloxifene modified the response to a low-dose conjugated estrogen cream or nonhormonal moisturizer in postmenopausal women with preexisting vaginal atrophy.
MATERIALS AND METHODS
Healthy postmenopausal women between the ages of 42 and 80 who had their last menstrual period (natural, not surgical) at least 2 years before study entry and had at least two signs of vaginal atrophy (dryness, pallor, mucosal epithelial thinning, petechiae, loss of rugation of the vaginal lining, or labial atrophy) were included in the study.
Subjects were excluded from enrolling in the study for the following reasons: took estrogen within the previous 3 months or nonestrogen drugs or over the counter remedies for vaginal atrophy or hot flashes within the previous 2 months; had an abnormal cervical cytology screening smear at entry or within the previous 3 years; had a suspected or confirmed breast malignancy; or had a prior hysterectomy, bilateral oophorectomy, or endometrial ablation procedure. Subjects were also excluded if they had any disorder of the vulva that would preclude proper assessment of drug effects. Individual site ethical review board approval of the protocol and informed subject consent were required before study enrollment occurred.
This was a multicenter, parallel, placebo-controlled, randomized study. Eligible subjects were randomly assigned to one of four treatment groups: raloxifene plus nonhormonal moisturizer, raloxifene plus conjugated estrogen cream, oral placebo plus nonhormonal moisturizer; or oral placebo plus conjugated estrogen cream. Randomization codes were computer generated at a central coordinating center, with randomization stratified by study site. Subjects took one oral tablet of raloxifene (60 mg) or a matching placebo daily for 3 months. Conjugated estrogen cream (0.5 g) or one applicator full of nonhormonal moisturizer was inserted deep in the vagina each night before bedtime for the first 2 weeks, and then twice weekly thereafter for 3 months. Investigators and subjects were blinded to the oral study medication assignments, whereas the vaginal preparations were administered open label.
Efficacy measures included assessment of vaginal maturation, vaginal pH, urinary pH, examination (investigator visual assessment) of vaginal mucosa, subjective assessment (subject rating) of atrophy symptoms by daily diary, and overall satisfaction measured at baseline and after 3 months of treatment. Laboratory data, vital signs, and adverse events were recorded throughout the study period and analyzed to assess safety. Compliance with study medication was assessed by counting pills and inspecting remaining vaginal supplies at 3 months. Non-compliance was defined as an estimated use of less than 70% of study drugs.
The vaginal maturation value was obtained by cytologic examination of a smear obtained from the upper lateral vaginal wall. Samples were analyzed at a central pathology laboratory.8,9 Vaginal and urinary pH were measured at the individual study sites using hydrion pH paper strips (range 4–9).
Vaginal mucosa was evaluated by physical examination with five measures: rugal atrophy, pallor, petechiae, mucosal thinning (elasticity), and dryness. Each sign was graded using a descriptive assessment table (Table 1) with a 4-point ordinal scale (0 = none or normal to 3 = severe). The sum of these five scores provided a global score.
Subjects were provided with a diary to assess six symptoms including vaginal dryness, itching, painful urination, urinary urgency, vaginal mucosal bleeding, and painful intercourse. These six symptoms were also graded on a 4-point ordinal scale (none, mild, moderate, severe). A seventh score (subject global) was calculated by adding the six symptom scores. Subjects also kept a record of the number and severity of hot flashes. They assessed their overall satisfaction with treatment using a 7-point scale (1 = extremely dissatisfied to 7 = extremely satisfied).
An estimated sample size of 160 subjects would be necessary for the study to have 90% power to detect a difference between two treatment groups of approximately 1.0 standard deviation (SD) in the mean change from baseline to end point with a two-sided significance level of .05. The measures used to make this projection included vaginal maturation value, physician assessment, and subjective assessment.
The primary analysis measure was change from baseline to end point for objective and subjective parameters. Data from all subjects with at least one postbaseline measurement were included in the efficacy analysis, with the last observation carried forward if any data were missing. An analysis of variance model with treatment, investigator, and the treatment-by-investigator interaction as explanatory variables was used to test for significant overall and pairwise treatment effects between the four treatment groups. Since the terms for investigator and treatment by investigator were not significant, the P values for these terms are not shown. Subgroup analysis was also done by age; however, age was not a significant factor in any of the analyses, so those results are not shown. The 2 × 2 factorial design provides an estimate of interaction (“potentiation” or “antagonism”) between the two factors tested, in this case between “vaginal cream” and “raloxifene.” This design directly addresses whether the effects of these factors are additive and, when they are, provides efficient estimates of both cream and raloxifene main effects.
In the subjective assessment and the evaluation of vaginal mucosa, a treatment response was defined a priori as a rating shift of at least one ordinal category from baseline. A mean change of at least one unit in the direction of improvement was needed to show treatment efficacy.
Data from all subjects were included in all safety analyses. Tests for overall treatment effect with respect to rates of treatment-emergent adverse events were conducted using the χ2 test. All statistical significance tests were done at the 5% level.
For the study, 308 women were screened at 16 sites in the United States, and 187 were enrolled. The number of subjects enrolled at an individual site ranged from five to 28. The mean age was 59.4 years (range 42.6 to 80.9), 90.4% were white, the mean weight was 71 kg (range 45–120 kg), the mean height was 162 cm (range 140–183 cm), the mean body mass index was 27.2 kg/m2 (range 18–45), and the mean number of years postmenopausal was 9.5 (range 2–30). There were no significant differences at baseline between the treatment groups in demographic characteristics or signs or symptoms of vaginal atrophy (Table 2). Rate of compliance with the study medications was high (greater than 90% overall), and there were no significant differences between treatment groups.
A total of 157 subjects completed the study, whereas 30 subjects (16%) discontinued early. The reasons for discontinuation were personal (12 subjects), protocol violation (five subjects), and adverse events (13 subjects). One subject assigned to raloxifene plus conjugated estrogen cream treatment experienced a deep vein thrombophlebitis and discontinued. There were no significant differences between the four treatment groups in the incidence of discontinuations or adverse events.
There were no significant differences between the four treatment group means for vaginal maturation value, vaginal pH, or urinary pH at baseline (data not shown). At 3 months, all treatment groups except the placebo–nonhormonal moisturizer group demonstrated significant mean increases from baseline in vaginal maturation values (Table 3). Although vaginal maturation values increased significantly from baseline in the raloxifene–nonhormonal moisturizer group, this change was not significantly greater than the smaller increase seen in the placebo–nonhormonal moisturizer group. The 2 × 2 factorial analysis showed no statistical differences between raloxifene and the placebo when coadministered with either conjugated estrogen cream or nonhormonal moisturizer for mean changes in vaginal maturation value (Table 4). The conjugated estrogen cream was superior to nonhormonal moisturizer for the increase in vaginal maturation value (P < .001) in raloxifene- and placebo-treated groups.
In addition, all treatment groups demonstrated mean decreases in vaginal pH, with the conjugated estrogen cream groups showing significant decreases (P < .05) (Table 3). Conjugated estrogen cream was superior to nonhormonal moisturizer for reducing vaginal pH (P < .05) in raloxifene- and placebo-treated groups. There was little change in urinary pH relative to baseline, with no significant differences between treatment groups, although the raloxifene-conjugated estrogen cream group showed a significant increase from baseline (P < .05).
At baseline there were no significant differences between any of the treatment group means for the five signs of vaginal atrophy or for the global score (data not shown). Most signs were evaluated as mild to moderate in severity. Petechiae ranged from none to mild. After 3 months, the mean percentage changes from baseline in rugal atrophy, pallor, petechiae, mucosal thinning, dryness, and the calculated global score were negative in all treatment groups, which indicated improvement (Table 5). These changes from baseline were significant (P ≤ .015) in all treatment groups. There were no differences between raloxifene and placebo treatment pairs. The mean reductions in scores were significantly greater (P < .05) with conjugated estrogen cream treatment than with nonhormonal moisturizer treatment for the calculated global score as well as for all of the signs evaluated except vaginal dryness, which showed no difference between treatment groups.
There were no significant differences at baseline between any of the treatment groups for the six subjective assessment scores or for the global score (data not shown). Most symptoms were rated as none to mild across all groups. At 3 months, there were no significant differences between raloxifene and placebo treatment pairs in mean percentage changes for dryness, itching, dysuria and urgency, dyspareunia, and vaginal mucosal bleeding and no significant difference in global scores (Figure 1). For most parameters, the mean percentage change from baseline was negative, indicating symptom improvement. However, these mean percentage changes were small, and their SDs were relatively large. The change in global score, representing the sum of all of the six symptom scores, was the only parameter showing significantly greater improvement with conjugated estrogen cream treatment than with nonhormonal moisturizer or from baseline. Raloxifene did not modify the magnitude of improvement in the subject global scores when administered concomitantly with either vaginal preparation.
At baseline, there was no significant difference between any of the treatment groups in the mean daily number of hot flashes or hot flash severity score. The mean (± SD) across treatment groups in hot flash number was 1.24 (± 2.14), whereas the mean (± SD) severity score was 0.60 (± 0.85), indicating that hot flashes were generally considered mild. At 3 months, mean changes from baseline in hot flash number were statistically indistinguishable from zero within and between treatment groups. The mean changes in severity score were also not significant.
At baseline, there was no significant difference between any of the treatment groups in the overall satisfaction score. The median across treatment groups was 4.00, indicating neutral to mild satisfaction. After 3 months, median overall satisfaction values increased by 1.00 from baseline across treatment groups, indicating improvement (P < .03). There were no significant differences between raloxifene and placebo treatment pairs or between conjugated estrogen cream and nonhormonal moisturizer treatment pairs.
The objective of this study was to determine whether raloxifene therapy affected the efficacy of low-dose conjugated estrogen cream or nonhormonal moisturizer in treating the signs and symptoms of vaginal atrophy in postmenopausal women. All of the women enrolled in this study presented with objective signs of vaginal atrophy, rated as mild to moderate in severity. As has been shown previously, both low-dose conjugated estrogen cream and nonhormonal moisturizer improve signs and symptoms of vaginal atrophy.6,7 The study design offered a true blinded assessment of the effect of raloxifene. Nonhormonal moisturizer acted as the control for conjugated estrogen cream by providing the local application of a vaginal preparation without any known effect on the local estrogen receptor system. The use of a blinded placebo preparation for the conjugated estrogen cream would have been optimal to determine the true effect of vaginal estrogen, but because such a preparation was not available, this was not the goal of this study. Rather, the use of nonhormonal moisturizer allowed a comparison of the effectiveness of two accepted treatments for vaginal atrophy in the presence of raloxifene.
The present results support the overall neutral effect of raloxifene on the vaginal mucosa as well as on symptoms of vaginal atrophy, as previously reported.4 Both vaginal preparations were equally effective in the presence of either raloxifene or a placebo. This was true for both the objective evaluations and the subject self-assessments. Therefore, we conclude that raloxifene does not block the local effect of topical vaginal estrogens, even when a low-dose estrogen regimen is used. The current study does not address the effect of raloxifene alone in subjects reporting vaginal atrophy symptoms.
The study also compared the effects of low-dose conjugated estrogen cream and nonhormonal moisturizer on the signs and symptoms of vaginal atrophy. Nonhormonal moisturizer is a unique polyglycolic acid matrix product that has been studied extensively and must not be confused with inert type moisturizers. Bygdeman and Swahn showed that nonhormonal moisturizer and conjugated estrogen cream significantly reduced symptoms of vaginal itching, irritation, and dyspareunia with no efficacy differences between the two groups beyond the first week of treatment.7 Nachtigall found that both therapies significantly increased vaginal moisture, fluid volume, and elasticity and returned vaginal pH to a premenopausal level.6 In our study, conjugated estrogen cream significantly improved objective signs compared with nonhormonal moisturizer, whereas the subjective assessments of symptoms and overall satisfaction did not indicate a difference between conjugated estrogen cream and nonhormonal moisturizer. Our results are therefore compatible with those of prior studies.
The four treatment regimens were well tolerated. No difference in the number or severity of hot flashes was found between the raloxifene and placebo treatment pairs after 3 months. This is most likely a reflection of the study population, which had a low incidence of hot flashes at baseline.
In summary, the study shows that postmenopausal women with signs of vaginal atrophy receiving raloxifene may be effectively treated with either a low-dose conjugated estrogen cream or nonhormonal moisturizer for treating the atrophy.
1. Delmas PD, Bjarnason NH, Mitlak BH, Ravoux A-C, Shah AS, Huster WJ, et al. Effects of raloxifene on bone mineral density, serum cholesterol concentrations, and uterine endometrium in postmenopausal women. N Engl J Med 1997;337:1641–7.
2. Ettinger B, Black DM, Mitlak BH, Knickerbocker RK, Nickelsen T, Genant HK, et al. Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: Results from a 3-year randomized clinical trial. JAMA 1999;282:637–45.
3. Plouffe L Jr. Selective estrogen receptor modulators (SERMs) in clinical practice. J Soc Gynecol Investig 2000; 7(Suppl 1):S38–46.
4. Davies GC, Huster WJ, Lu Y, Plouffe L Jr, Lakshmanan M. Adverse events reported by postmenopausal women in controlled trials with raloxifene. Obstet Gynecol 1999;93:558–65.
5. Bachmann G. Urogenital ageing: An old problem newly recognized. Maturitas 1995;22 Suppl:S1–5.
6. Nachtigall LE. Comparative study: Replens versus local estrogen in menopausal women. Fertil Steril 1994;61:178–80.
7. Bygdeman M, Swahn ML. Replens versus dienoestrol cream in the symptomatic treatment of vaginal atrophy in postmenopausal women. Maturitas 1996;23:259–63.
8. Meisels A. The maturation value. Acta Cytol 1967;11:249.
© 2003 The American College of Obstetricians and Gynecologists
9. Wied GL. Hormonal cytology. In: Keebler CM, Somrak TM, eds. The manual of cytotechnology. Chicago: American Society of Clinical Pathologists Press, 1993:62–72.