Abstract
Context:
Women with primary ovarian insufficiency have significantly lower serum estradiol and T levels compared with regularly menstruating women. They also have significantly reduced bone mineral density (BMD).
Objective:
The objective of the study was to evaluate the efficacy of hormone replacement in maintaining BMD in these young women.
Design and Setting:
This was a randomized, double-blind, single-center, placebo-controlled clinical trial at the National Institutes of Health clinical center (Bethesda, Maryland).
Participants:
Young women with primary ovarian insufficiency participated in the study.
Interventions:
We compared the effect of estradiol and progestin replacement (n = 72) vs estradiol, progestin, and T replacement (n = 73) on BMD. We also compared findings with a contemporaneous control group of normal women (n = 70). All patients received transdermal estradiol (100 μg/d) plus oral medroxyprogesterone acetate 10 mg/d (12 d/mo) for a 3-month run-in period before being randomized in a double-blinded fashion to the addition of transdermal T (150 μg/d) or placebo.
Main Outcome Measure:
Change in BMD at the femoral neck was measured by dual-energy x-ray absorptiometry.
Results:
At screening, patients had significantly lower femoral neck BMD compared with control women (0.77 vs 0.81 g/cm2, P = .001) and did not differ in body mass index, age at menarche, or education level. Normal control women lost femoral neck BMD over the study period, whereas patients on estradiol and progestin therapy gained BMD; and at the end of the study period, femoral neck BMD of patients on estradiol and progestin therapy did not differ from that of control women (0.80 g/cm2 in both groups, P = .9). The addition of T showed no further benefit (percentage change in BMD 3.9 vs 2.4, respectively, P = .9). Nonetheless, using a repeated-measures model, the T group achieved a mean BMD in the femoral neck 0.015 g/cm2 higher than the placebo group at 3 years (95% confidence interval −0.005 to 0.034, P = .13). Similar findings were observed in the lumbar spine BMD as well.
Conclusion:
Long-term physiological transdermal estradiol replacement in combination with oral medroxyprogesterone acetate restores mean femoral neck BMD to normal in young women with spontaneous 46,XX primary ovarian insufficiency. However, the addition of physiological transdermal T replacement did not provide additional benefit.
Osteoporosis affects millions of postmenopausal women worldwide (1). Young women with hypogonadism are also at increased risk of reduced bone mineral density (BMD) (2). There is limited evidence upon which to guide management of reduced BMD in young women with estrogen deficiency. Estrogen increases BMD and prevents fractures in postmenopausal women (3), but its effect in young women with hypogonadism has not been well studied (4–9). In fact, it has been suggested that reduced BMD related to amenorrhea may not be completely reversible with estradiol therapy (10). Of concern, a recent report from an osteoporosis referral center found that a substantial percentage of estrogen-deficient young women with low BMD were treated with bisphosphonates (11), despite the fact that bisphosponates have long-term skeletal retention and could pose a teratogenic risk.
Spontaneous 46,XX primary ovarian insufficiency (POI) (12) (also known as premature ovarian failure and premature menopause) affects 1 in 100 women by age 40 years (13). POI is a state of hypogonadism, characterized by oligoamenorrhea, infertility, estrogen deficiency, and its associated symptoms. Women with POI are not only deficient in estrogen, but they also have significantly lower free and total T levels compared with regularly menstruating women of similar age (14). Deficiency in these two hormones increases the risk for osteoporosis (15–17). Young women with primary ovarian insufficiency have significantly reduced BMD compared with regularly menstruating women (2, 18).
Present hormone replacement therapy regimens were designed for women who experience normal menopause at approximately age 50 years. Surprisingly, there are few published data establishing optimal hormone replacement therapy regimens for younger women who experience primary ovarian insufficiency. T replacement in androgen-deficient men improves bone density, but little is known about the effect of T replacement in androgen-deficient women. Therefore, we sought to study the effects of estradiol and T replacement on bone health with a goal to identify optimal hormone replacement therapy for women with POI.
We chose to administer 100 μg/d estradiol by transdermal patch as a replacement dose of estrogen for young women (19) and to avoid the hepatic first-pass effect of oral estrogen on clotting factors. We chose the 150 μg/d dose of T because our goal was to mimic the production rate provided by normal ovarian function in young women. The total production rate of T in women having normal ovarian and adrenal function is approximately 300 μg/d (20). Approximately 50% of T is derived from the ovary and 50% from the adrenals (21). The regimen of 150 μg/d provides an approximation of the normal ovarian production rate of T, and in our experience, it increases the T levels to the normal range in this population of women.
This study addresses two important questions: 1) does transdermal replacement of physiological serum levels of T improve BMD as compared with estrogen/progestin replacement alone and 2) does transdermal physiological replacement of estradiol in combination with monthly oral medroxyprogesterone maintain BMD as effectively as normal ovarian function does in normal control women?
Materials and Methods
Study design
This was a 3-year prospective, double-blind, randomized, placebo-controlled clinical trial. From July 2000 to June 2006, we recruited 145 patients with primary ovarian insufficiency and 70 control women who had normal ovarian function. Patients were recruited by letters to physicians, notices in medical journals, and through the Internet. Concurrent controls were recruited by local advertisement. The Institutional Review Board of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health (Bethesda, Maryland), approved the study.
Women with spontaneous 46, XX primary ovarian insufficiency (22)
Inclusion criteria were as follows: 1) diagnosis of primary ovarian insufficiency (ie, at least 4 months of oligoamenorrhea and two FSH levels in the menopausal range, confirmed on two separate occasions at least 1 month apart before age 40 y); 2) no iatrogenic cause or known chromosomal abnormality; and 3) age between 18 and 42 years. Selecting women with spontaneous POI provided a well-characterized population with an estrogen-deficient condition relatively free of confounding factors (such as nutritional deficiency, chronic disease, chemotherapy, etc) that might affect bone health.
Exclusion criteria included factors affecting BMD. These criteria included smoking (more than two cigarettes per day); alcohol use (more than two drinks per day); body mass index (BMI) of 30 kg/m2 or greater and 19 kg/m2 or less; a previous history of hip fracture or other active hip pathology; major dermatological disorders or a history of skin sensitivity to adhesive bandages, tape, or transdermal matrix patches; a hirsutism (Ferriman Gallwey) score greater than 8; an acne score greater than 1; hysterectomy (so menstrual cycle patterns can be evaluated); baseline free T levels above the normal range and/or SHBG levels below the lower limit of normal; any prior treatment in the past 6 months known to affect bone other than estrogen (such as glucocorticoids); and a medical history of conditions affecting bone metabolism.
Control women
Controls were healthy, nonpregnant, regularly menstruating women (cycles between 21 and 35 d) who were nonsmokers, nonalcohol users, not taking medications, using nonhormonal contraceptive methods (ie, sterile or using barrier methods of contraception), and had no intention of conceiving within the 3 years of the study period. These women had a BMI of 19–30 kg/m2 and were between the ages 18 and 40 years.
Interventions
All patients were given the estradiol patch (100 μg/d) and progestin replacement therapy for a 3-month run-in period before randomization, such that all the patients began the T replacement portion of the study from a similar hormonal milieu. After confirmation of eligibility and signing the informed consent, patients were randomized in 1:1 fashion into two groups. One group received the estradiol patch (100 μg/d) and an investigational T patch (150 μg/d) (23), and the other group received the estradiol patch (100 μg/d) and the placebo patch. Both groups received cyclical 10 mg oral medroxyprogesterone acetate each day for 12 days of each calendar month. Calcium carbonate oral supplementation (two tablets of 0.650 g/d, 520 mg elemental calcium/d) was given to all participants, patients and controls. Vitamin D was not provided because at the time of trial design, its supplementation was recommended only for older individuals and in patients with vitamin D deficiency (24).
Study procedures
The evaluation included a complete medical history physical examination and laboratory screening. For control women, blood samples were drawn during the midfollicular phase of their menstrual cycle (d 5–12). At the midfollicular phase, the average serum estradiol is about 100 pg/mL in normal women. This is the level that the estradiol patch is designed to deliver on average. Admission staff recorded self-designated race/ethnicity. In addition, patients completed structured questionnaires to assess daily calcium intake (15), depression (16), and other lifestyle risk factors for osteoporosis such as exercise, use of contraceptives, hormone replacement therapy, smoking, alcohol intake, and reproductive history. Scores for hirsutism (25), clitoral index (26), acne (27), and frequency of depilation were assessed at each visit.
Bone mineral density
We measured areal BMD (grams per square centimeter) with a Hologic QDR4500 instrument (Hologic, Inc; the coefficient of variation is < 0.4%) (28), dual-energy x-ray absorptiometer at the posterior anterior lumbar spine (vertebrae L1-L4), femoral neck, and total hip as reported previously (2).
Hormonal assessment
Blood samples were drawn from fasted subjects at 8:00 am, separated within 1 hour, and frozen. We measured serum total T by RIA after extraction chromatography (Esoterix Endocrinology). We used an equilibrium dialysis to evaluate serum free T concentrations (29). Measurements of estradiol, FSH, and LH levels were performed at the National Institutes of Health Clinical Center. Estradiol was measured by competitive chemiluminescence immunoassay (Immulite 2000 analyzer). FSH and LH were measured by microparticle enzyme immunoassay (AxSYM System; Abbott Diagnostics).
Statistical analysis
We used Statistical Analysis Software (SAS Institute) to analyze the data. We compared quantitative clinical characteristics and hormones between groups of women using the Wilcoxon rank-sum test. To compare categorical variables, we used the Fisher's exact test. We tested correlation using Spearman rank correlation. The primary efficacy analysis of mean change from screening was assessed separately at 1, 2, and 3 years using available data from those visits and compared between treatment groups by the Wilcoxon rank-sum test. At a power level of 0.90 and an α-level of .05, the required sample size was calculated to be approximately 50 women in each group. Because of dropout and the study being stopped early, there were only 40 women with complete data on the change in the femoral neck BMD from screening to 3 years. Therefore, we used a repeated-measures model of BMD that incorporated all of the data at each visit on each woman in the trial. The model specified separate categorical effects for each visit and allowed treatment group and screening age to each affect BMD values at each postrandomization visit. The difference in model-predicted 3-year BMD means is reported as the treatment effect estimate for the trial.
The two primary outcome measures were: 1) change in BMD at the femoral neck measured by dual-energy x-ray absorptiometry compared between the women with POI treated with estradiol, progestin, and T (EPT) and the women treated with estradiol, progestin and placebo (EPP) alone; and 2) change in BMD at the femoral neck measured by dual-energy x-ray absorptiometry compared between the women with POI treated with EPP and the normal controls. All P values are two tailed. Results are presented as mean ± SD.
Results
We evaluated 395 women with POI (Figure 1). One hundred fourteen women did not meet eligibility criteria (having abnormal karyotype, smoking or alcohol use, pregnancy, and BMI > 30 kg/m2). Of the remaining subjects, 136 declined to participate due to a variety of reasons such as concerns over the long-term study, desire for fertility work-up or therapy, or did not want hormone replacement. A total of 145 subjects signed the informed consent and participated in the study. Before completion of the study, the data safety and monitoring board recommended stopping the trial early because it was determined that it would be futile to continue the study as the a priori goal (gain in femoral neck BMD related to T therapy) would not be achieved, even if the trial was allowed to complete. As a result, 45 women with POI and 33 controls completed the 3-year study. The mean length of follow-up was similar in both groups [EPP, 2.21 (1.01) y and EPT, 2.04 (1.04) y, P = .4]. There was no difference in the discontinuation rate by length of participation or reason for withdrawal between the EPT and EPP groups. Reasons for discontinuation included pregnancy, donor egg/other family-building efforts, inconvenience, and concerns over reported risks of hormone replacement.
Figure 1.
Clinical trial flow chart.
The original intent of the study was 2-fold. Accordingly, results are presented in two categories: comparisons between the patients using the EPP (n = 72) and the following: 1) patients using the EPT (n = 73), and 2) normal control women with regular menstrual cycles (n = 70).
Demographic characteristics at baseline
Table 1 shows demographic and reproductive characteristics of women with POI (EPT, n = 73, and EPP, n = 72) and concurrent control women (n = 70). At baseline, there was no difference between the T and placebo groups in age, marital status, BMI, height, age of onset of menstrual abnormality, age at diagnosis of POI, history of thyroid disease, or depression.
Table 1.
Baseline Demographic Characteristics
| EPT (n = 73) Mean (SD) | EPP (n = 72) Mean (SD) | Control (n = 70) Mean (SD) | P Value | |
|---|---|---|---|---|
| Age, y | 31.3 (5.8) | 33.0 (5.2) | 30.1 (7.3) | EPT vs EPP: .06 |
| EPP vs control: .013 | ||||
| Race n, % | ||||
| White | 66 (90.4%) | 56 (77.8%) | 51 (72.9%) | EPT vs EPP: .049 |
| African American | 6 (8.2%) | 7 (9.7%) | 11 (15.7%) | EPP vs control: .43 |
| Asian | 1 (1.4%) | 4 (5.6%) | 6 (8.6%) | |
| Other/unknown | 0 (0%) | 5 (6.9%) | 2 (2.9%) | |
| Marital status | ||||
| Married | 41 (56.2%) | 45 (62.5%) | 26 (37.7%) | EPT vs EPP 0.74 |
| Single | 31 (42.5%) | 26 (36.1%) | 43 (62.3%) | EPP vs control .003 |
| Divorced | 1 (1.4%) | 1 (1.4%) | 0 | |
| BMI, kg/m2 | 23.4 (3.2) | 23.8 (3.1) | 22.9 (2.8) | EPT vs EPP: .25 |
| EPP vs control: .16 | ||||
| Height, cm | 165.3 (6.4) | 164.6 (6.8) | 164.5 (6) | EPT vs EPP: .3 |
| EPP vs control: .8 | ||||
| Age at first menstrual abnormality, y | 24.4 (8.4) | 24.9 (7.9) | NA | EPT vs EPP: .76 |
| Age at diagnosis of POI, y | 28.5 (6.7) | 29.1 (6.8) | NA | EPT vs EPP: 0.49 |
| Hx of depression | 15 (20.6%) | 16 (22.2%) | 0 (0%) | EPT vs EPP: .84 |
| EPP vs control: <.001 | ||||
| Hx of thyroid disease | 7 (9.6%) | 9 (12.5%) | 0 (0%) | EPT vs EPP: .61 |
| EPP vs control: .003 |
Abbreviations: Hx, history; NA, not applicable.
Compared with normal controls, women with primary ovarian insufficiency on EPP were slightly older (EPP: 33.0 y vs control: 30.1 y), more likely to be married, more likely to be Caucasian (EPP: 78% vs control: 73%), and more likely to have a history of depression or thyroid disease (Table 1).
Transdermal estradiol therapy (100 μg/d) normalized serum estradiol levels
After the 3-month run-in period of therapy with the 100 μg estradiol patch and 10 mg medroxyprogesterone acetate for 12 days each month, serum estradiol, FSH, and LH levels were similar between the EPT and EPP groups (Table 2). Serum estradiol levels increased to normal control levels in both patient groups receiving the estradiol patch. Serum FSH levels after run-in were significantly lower in patients on hormone therapy compared with screening (P < .0001), but these levels were still significantly greater than control levels (P < .0001).
Table 2.
Serum Estradiol, FSH, and LH Levels Among Three Groups Over 3-Year Study Period
| EPT n = 73 Mean (SD) | EPP n = 72 Mean (SD) | Control n = 70 Mean (SD) | Comparison | Raw P Value | |
|---|---|---|---|---|---|
| Serum estradiol, pg/mL | |||||
| At screening | 40.5 (49.2) | 35.6 (32.4) | 102.7 (97.2) | EPT vs EPP | .87 |
| EPP vs control | <.0001 | ||||
| After run-in period | 94.9 (61.7) | 98.0 (102.1) | NA | EPT vs EPP | .56 |
| EPP vs control | .67 | ||||
| 12 mo | 88.2 (66.9) | 89.0 (54.7) | 70.6 (63.1) | EPT vs EPP | .5 |
| EPP vs control | .0066 | ||||
| 24 mo | 93.7 (61.0) | 81.6 (60.8) | NA | EPT vs EPP | .25 |
| 36 mo | 84.9 (63) | 75.1 (47.8) | NA | EPT vs EPP | .5 |
| Serum FSH, μIU/mL | |||||
| At screening | 93.0 (40.2) | 84.6 (42.0) | 6.3 (2.0) | EPT vs EPP | .16 |
| EPP vs control | <.0001 | ||||
| After run-in period | 24.8 (23.5) | 22.5 (23.3) | NA | EPT vs EPP | .30 |
| EPP vs control | <.0001 | ||||
| 12 mo | 20.5 (20.5) | 20.2 (16.4) | 7.2 (4.2) | EPT vs EPP | .6 |
| EPP vs control | <.0001 | ||||
| 24 mo | 20.7 (21.9) | 20.4 (18.7) | NA | EPT vs EPP | .6 |
| 36 mo | 19.5 (15.8) | 24.9 (22.7) | NA | EPT vs EPP | .4 |
| Serum LH, μIU/mL | |||||
| At screening | 55.6 (26.2) | 52.4 (30.3) | 7.9 (5.8) | EPT vs EPP | .27 |
| EPP vs control | <.0001 | ||||
| After run-in period | 15.4 (13.9) | 15.0 (17.4) | NA | EPT vs EPP | .25 |
| EPP vs control | .18 | ||||
| 12 mo | 13.8 (13.5) | 12.1 (13.1) | 6.5 (3.0) | EPT vs EPP | .4 |
| EPP vs Control | .25 | ||||
| 24 mo | 13.4 (14.1) | 14.1 (15.0) | NA | EPT vs EPP | .6 |
| 36 mo | 12.1 (11.9) | 14 (10.4) | NA | EPT vs EPP | .36 |
Abbreviation: NA, not applicable or available. Reference ranges for the follicular phase for estradiol (20–160 pg/mL), FSH (3–11 μIU/mL), and LH (3–14 μIU/mL) correspond to the National Institutes of Health Clinical Center laboratory database.
Transdermal T therapy (150 μg/d) normalized serum T levels
While off estrogen therapy, mean serum free T concentration was significantly lower in women with POI compared with controls [2.4 (1.1) vs 3.6 (1.6) pg/mL, P < .001]. After the run-in period (of mean 3.9 mo), on 100 μg estradiol and medroxyprogesterone acetate therapy, free T dropped significantly lower to 2.1 (1.3) pg/mL (P < .001). Treatment with 150 μg/d T patch significantly increased free T levels (measured at 1 y) within the normal range [4.7 (2.8) pg/mL], which was greater than free T levels in the placebo group [1.6 (0.7) pg/mL, P < .001]. Throughout 3 years, mean free and total T levels remained within the normal range (Figure 2).
Figure 2.
Mean (SEM) serum free and total T levels.
Effect of replacement of T on BMD
At screening, patient groups randomized to placebo patch or T patch had no significant difference in femoral neck BMD [0.77 (0.11) vs 0.77 (0.11) g/cm2, P = .9] or lumbar spine BMD [0.94 (0.12) vs 0.96 (0.15) g/cm2, P = .4]. During 3-year follow up, percentage change in BMD from screening was not statistically different between the EPT and EPP groups except for one instance of femoral neck BMD at 1 year (Figure 3A: 1 y, P = .045; other, P > .17). Similarly, the BMD at the lumbar spine was not statistically different at any time point between the EPT group and the EPP group (Figure 3B: all P > .56). Using all available data and a repeated-measures model to estimate a difference in the femoral neck BMD after 3 years gave an estimate that the T group would end with a mean BMD of 0.015 g/cm2 higher than the placebo group (suggestive but not statistically significant; 95% confidence interval −0.005–0.034).
Figure 3.
A, Mean (SEM) percentage change from screening in the femoral neck BMD. B, Mean (SEM) percentage change from screening in the lumbar spine BMD.
Estradiol and cyclic progestin therapy restored BMD to normal control levels
Women with POI gained BMD at both the femoral neck and lumbar spine while on estradiol and cyclic medroxyprogesterone acetate replacement therapy (Figure 3, A and B). At screening, the femoral neck BMD was significantly lower in women with POI as compared with control women [EPP, 0.76 (0.11); EPT, 0.77 (0.11) vs controls, 0.81 (0.09) g/cm2, P = .001]. The mean percentage gain in BMD at the femoral neck at the end of the 3-year study in the women with POI on estradiol and cyclic medroxyprogesterone acetate replacement therapy was 2.45% (4.71%), whereas the control women lost femoral neck BMD over the 3-year period by 2.61% (8.38%), P = .0002 (Figure 3A). This result remained highly significant after adjustment by the Bonferroni factor of 2. At the end of 3 years, these patients' femoral neck BMD did not significantly differ from controls [0.80 (0.12) vs 0.80 (0.11) g/cm2, P = .95].
At the initial evaluation, the mean lumbar spine BMD was significantly lower in women with POI as compared with control women [EPP, 0.94 (0.12); EPT, 0.96 (0.15) vs controls 1.00 (0.10) g/cm2, P = .01]. Women with POI gained BMD at the lumbar spine while on estradiol and cyclic medroxyprogesterone acetate replacement therapy (Figure 3B). Control women maintained their BMD at the lumbar spine over the 3 years. At the end of 3 years, the lumbar spine BMD of the women with POI on estradiol and cyclic medroxyprogesterone acetate replacement therapy did not differ significantly from normal control women [1.02 (0.11) vs 1.01 (0.11) g/cm2, P = .8]. The mean gain in lumbar spine BMD in the EPP patients at the end of the study was a surprising 7.7% (6.0%) compared with a change of 0.4% (3.3%) in controls. Also, during the brief run-in period of only a few months of estradiol/progestin therapy, the difference in the lumbar spine BMD between the patient and control groups became not significantly different [EPP, 0.97 (0.11) vs controls, 1.00 (0.11), P = .14]. During this brief run-in time, the patients on EPP had an unexpectedly high individual mean gain in lumbar spine of 3.3% (3.7%).
Compared with the controls, more women with POI had a femoral neck BMD Z score of −2 or less at the screening visit (19.2% EPT group, 15.3% for the EPP group, and 5.7% for the control group; EPT vs controls, P = .02, EPP vs controls, P = .1). At 36 months, controls and the women with POI had a similar percentages of women who had a femoral neck BMD Z score of −2 or less (0% EPT group, 4% for the EPP group, and 2.4% for the control group; EPT and EPP vs controls, P = 1.00). At the screening visit, the mean femoral neck BMD Z score was −0.7 (0.9) for the EPT group, −0.7 (1.0) for the EPP group, and −0.3 (0.8) for the control group (EPT vs controls and EPP vs controls, P = .04). At 36 months, the mean BMD Z score was −0.7 (0.8) for the EPT group, −0.3 (1.1) for the EPP group, and −0.3 (0.7) for the control group (EPT vs controls, P = .04, and EPP vs controls, P = .99). Similarly, at the screening visit, more women with POI had a lumbar spine BMD Z score of −2 or less: 24.7% women in the EPT group, 22.2% in the EPT group, and 7.1% in the control group (EPT vs controls, P = .006, and EPP vs controls, P = .02). At 36 months, controls and the women with POI had a similar percentage of women who had a lumbar spine BMD Z score of −2 or less (5% for the EPT group, 4% for the EPP group, and 4.9% for the control group; EPT and EPP vs. controls P = 1.0). At the screening visit, the mean lumbar spine BMD Z score was −0.7 (1.4) for the EPT group, −0.9 (1.1) for the EPP group, and −0.4 (0.9) for the control group (EPT vs controls, P = .25; EPP vs controls, P = .01). At 36 months, the mean BMD Z score was −0.3 (1.0) for the EPT group, −0.1 (1.1) for the EPP group, and −0.3 (0.9) for the control group (EPT vs controls, P = .7, and EPP vs controls, P = .5). The results of the change in the BMD Z score are presented in Supplemental Table 1.
Markers of bone turnover
Markers of bone turnover were measured only at the baseline (after 3 mo of estradiol and progestin therapy in the run-in period) and at 12 months. There was no difference in the markers of bone resorption or formation between the patients receiving T vs placebo at baseline or at 12 months (data not shown). As shown in Table 3, compared with normal controls, women with POI (treated with estrogen and progestin during 3 mo run-in period) had significantly higher serum alkaline phosphatase and bone-specific alkaline phosphates levels, markers of bone formation. There was no significant difference in the markers of bone resorption between the two groups.
Table 3.
Bone Turnover Markers in Normal Control Women and Women With POI on Estradiol and Progestin Replacement After 3 Month Run-In Period
| Control Women |
Women With POI Mean (SD, n) |
p Value | |||
|---|---|---|---|---|---|
| n | Mean (SD) | n | Mean (SD) | ||
| Urinary cross-collagen teleopeptide, nmol/L | 64 | 42.6 (19.4) | 137 | 42.8 (20.3) | .92 |
| Urine deoxypyridinolines, nmol/mmol creatinine | 69 | 6.5 (1.9) | 136 | 6.4 (2.1) | .40 |
| Serum bone-specific alkaline phosphates, ng/mL | 65 | 11.1 (4.4) | 137 | 12.6 (5.5) | .04 |
| Serum osteocalcin cis, ng/mL | 57 | 22.4 (7.8) | 134 | 23.8 (9.5) | .42 |
| Serum alkaline phosphatase, U/L | 16 | 37.1 (11.3) | 145 | 64.2 (19.6) | <.001 |
Safety and tolerability
Physiological replacement of T using the 150-μg patch was well tolerated. A total of five subjects (four in the T group and one in the placebo group) dropped out due to side effects including skin irritation, redness, rash, hirsutism, and oily skin. Four patients, two in each group, dropped out due to pregnancy. Patients randomized to T had minimal increases in acne score, hirsutism score, clitoral index, and facial depilation score after 36 months of treatment. None of these increases were significantly different from the findings in patients randomized to placebo treatment. In addition, there were no statistically significant differences between the two groups at baseline or at 36 months in terms of hemoglobin, liver function tests, lipid profile, chemistry panel, or local skin reactions.
Discussion
Here we provide evidence that physiological transdermal estradiol and oral progestin replacement therapy not only maintains BMD as well as normal ovarian function does, but it can actually restore BMD to normal levels in this population. We found the specific regimen of a 100-μg estradiol transdermal patch combined with cyclic medroxyprogesterone acetate was well tolerated and effective in improving bone mass. Bone resorption markers in these patients returned to normal and bone formation markers increased. However, the addition of T replacement did not significantly improve BMD. Clearly this study demonstrates that bones in young women can regain lost BMD due to a prior estrogen-deficient state.
We chose to administer estradiol by the transdermal patch for this study. Transdermal administration bypasses the liver and avoids the first-pass effect with the associated increase in clotting factors and increased risk of thromboembolic events (30–32). Daily serum measurements of estradiol in regularly menstruating women indicate that the mean estradiol level throughout the menstrual cycle is approximately 104 pg/mL (382 pmol/L) (33). The 100-μg estradiol patch provides serum levels in this range (19). Another advantage of administering estradiol by transdermal patch is that this method permits meaningful monitoring of therapy by measuring serum estradiol levels, should this become necessary. Evidence suggests that all forms of hormone replacement therapy may not be equal with regard to maintaining BMD in women with primary ovarian insufficiency. Recent findings demonstrate that using the oral contraceptive as hormone replacement therapy in these women may have inferior bone effects as compared with transdermal estradiol and cyclic progestin administration (34). The oral contraceptive provides an ovulation-suppressive dose of continuous combined estrogen and progestin, and this regimen did not have a positive effect on bone formation markers. In contrast, continuous physiological transdermal estradiol and cyclic monthly progestin in our study increased bone formation markers significantly. These findings are in accord with studies showing similar effect of estrogen treatment in young adults (both men and women) with hypogonadism such as aromatase deficiency (35) and in some studies for anorexia nervosa (36).
There was no statistically significant additional benefit of adding 150 μg/d T. We chose the 150-μg/d dose because our goal was to mimic the production rate provided by normal ovarian function in young women. The regimen of 150 μg/d provides an approximation of the normal ovarian production rate of T (the other half, ∼150 μg/d would be produced by the adrenals) and in our experience increased the T levels to the normal range in this population of women. The findings support a conclusion that the 150-μg/d dose of T was indeed a physiological replacement dose for these young women with POI.
Another approach to address the question of T replacement would be to exclusively enroll women with serum free T levels below the lower limit of normal. The goal of the present study was to work toward developing an optimal hormone replacement therapy regimen for women with POI in general, not a select group who must first be proven to have clinical androgen deficiency. In fact, presently there is no consensus as to how best to define and identify androgen deficiency in young women, and this would make a titration of dose a difficulty. It is possible that physiological T replacement might have a positive effect on BMD in a selected subset of women who have T levels below normal.
Our study has limitations. We had a high dropout rate over the 3-year period in both groups. Most women dropped out to pursue their plans for pregnancy. Because the current study was stopped early, it is difficult to determine the magnitude of a potential T benefit to BMD. Although the addition of T replacement did not show a statistically significant improvement in BMD, there is a suggestion of a positive effect (Figure 3A). A study that relies entirely on data from screening to 3 years would require 185 subjects per group to have an 80% power to detect a difference in the femoral neck BMD if the estimated difference of 0.015 g/cm2 is the true difference. This study was not powered to detect this amount of difference.
Strengths of this study are the long duration (3 y) and the large number of young estrogen-deficient women with a well-characterized estrogen-deficient condition that lacks confounding factors (such as nutritional deficiency, chronic disease, chemotherapy, etc). In conditions resulting in hypogonadism such as anorexia nervosa, the female athlete triad, and hypothalamic amenorrhea bone loss may be not be fully reversible with estrogen replacement therapy (37). However, spontaneous 46,XX primary ovarian insufficiency is hypergonadotropic hypogonadism without interplay of other confounding factors such as malnutrition, chemotherapy, or X chromosome haploinsufficiency. Women with spontaneous 46,XX primary ovarian insufficiency provide a unique opportunity to study the effects of estrogen deficiency on the bone health in young women without the effect of such confounding.
Our study demonstrates a potent BMD effect of replacement estradiol on these estrogen-deficient young women. A gain of 7.7% in the lumbar spine BMD over 3 years is remarkable. Our findings demonstrate not only an early and robust effect to bring bone resorption markers equivalent to levels seen in control women with normal ovarian function but also an impressive stimulus to bone formation markers significantly greater than present in controls. Clearly this study demonstrates that, in the absence of the confounding factors mentioned above, the bones of young women can heal from their prior estrogen-deficient state.
We conclude that in young women with spontaneous 46,XX POI addition of physiological transdermal T replacement does not significantly improve BMD as compared with transdermal estradiol and medroxyprogesterone acetate replacement alone. Importantly, however, 3 years of physiological transdermal estradiol replacement in combination with cyclic oral medroxyprogesterone acetate restores the mean femoral neck BMD to normal in this population. The findings provide important evidence for counseling patients with this condition with regard to hormone replacement therapy.
Acknowledgments
We thank all of the women who served as controls, the women with primary ovarian insufficiency who participated in the study, and the staff of the National Institutes of Health Clinical Center.
L.M.N. (principal investigator) had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
This study had a trial registration number of NCT00001951 and is registered at http://www.clinicaltrials.gov/ct2.
This work was supported by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, and Procter & Gamble under Cooperative Research and Development Agreement CRADA 00983.
Disclosure Summary: The authors have nothing to declare.
Footnotes
- BMD
- bone mineral density
- BMI
- body mass index
- EPT
- estradiol, progestin, and T
- EPP
- estradiol, progestin and placebo
- POI
- primary ovarian insufficiency.
References
- 1. Lewiecki EM. In the clinic. Osteoporosis. Ann Intern Med 2011;155(1):ITC1–15 [DOI] [PubMed] [Google Scholar]
- 2. Popat VB, Calis KA, Vanderhoof VH, et al. Bone mineral density in estrogen-deficient young women. J Clin Endocrinol Metab. 2009;94(7):2277–2283 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women's Health Initiative randomized controlled trial. JAMA. 2002;288(3):321–333 [DOI] [PubMed] [Google Scholar]
- 4. Riis BJ, Christiansen C, Johansen JS, Jacobson J. Is it possible to prevent bone loss in young women treated with luteinizing hormone-releasing hormone agonists? J Clin Endocrinol Metab. 1990;70(4):920–924 [DOI] [PubMed] [Google Scholar]
- 5. Metka M, Holzer G, Heytmanek G, Huber J. Hypergonadotropic hypogonadic amenorrhea (World Health Organization III) and osteoporosis. Fertil Steril. 1992;57(1):37–41 [DOI] [PubMed] [Google Scholar]
- 6. Leather AT, Studd JW, Watson NR, Holland EF. The prevention of bone loss in young women treated with GnRH analogues with “add-back” estrogen therapy. Obstet Gynecol. 1993;81(1):104–107 [PubMed] [Google Scholar]
- 7. Field CS, Ory SJ, Wahner HW, Herrmann RR, Judd HL, Riggs BL. Preventive effects of transdermal 17β-estradiol on osteoporotic changes after surgical menopause: a two-year placebo-controlled trial. Am J Obstet Gynecol. 1993;168(1 Pt 1):114–121 [DOI] [PubMed] [Google Scholar]
- 8. Castelo-Branco C, Rovira M, Pons F, et al. The effect of hormone replacement therapy on bone mass in patients with ovarian failure due to bone marrow transplantation. Maturitas. 1996;23(3):307–312 [DOI] [PubMed] [Google Scholar]
- 9. Grinspoon SK, Friedman AJ, Miller KK, Lippman J, Olson WH, Warren MP. Effects of a triphasic combination oral contraceptive containing norgestimate/ethinyl estradiol on biochemical markers of bone metabolism in young women with osteopenia secondary to hypothalamic amenorrhea. J Clin Endocrinol Metab. 2003;88(8):3651–3656 [DOI] [PubMed] [Google Scholar]
- 10. Gulekli B, Davies MC, Jacobs HS. Effect of treatment on established osteoporosis in young women with amenorrhoea. Clin Endocrinol (Oxf). 1994;41(3):275–281 [DOI] [PubMed] [Google Scholar]
- 11. Cohen A, Fleischer J, Freeby MJ, McMahon DJ, Irani D, Shane E. Clinical characteristics and medication use among premenopausal women with osteoporosis and low BMD: the experience of an osteoporosis referral center. J Womens Health (Larchmt ). 2009;18(1):79–84 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Albright F, Smith PH, Fraser R. A syndrome characterized by primary ovarian insufficiency and decreased stature. Am J Med Sci. 1942;204(5):625–648 [Google Scholar]
- 13. Coulam CB, Adamson SC, Annegers JF. Incidence of premature ovarian failure. Obstet Gynecol. 1986;67:604–606 [PubMed] [Google Scholar]
- 14. Kalantaridou SN, Calis KA, Vanderhoof VH, et al. Testosterone deficiency in young women with 46,XX spontaneous premature ovarian failure. Fertil Steril. 2006;86(5):1475–1482 [DOI] [PubMed] [Google Scholar]
- 15. Davis SR, McCloud P, Strauss BJ, Burger H. Testosterone enhances estradiol's effects on postmenopausal bone density and sexuality. Maturitas. 1995;21(3):227–236 [DOI] [PubMed] [Google Scholar]
- 16. Slemenda C, Longcope C, Peacock M, Hui S, Johnston CC. Sex steroids, bone mass, and bone loss. A prospective study of pre-, peri-, and postmenopausal women. J Clin Invest. 1996;97(1):14–21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Khosla S, Melton LJ, III, Atkinson EJ, O'Fallon WM, Klee GG, Riggs BL. Relationship of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: a key role for bioavailable estrogen. J Clin Endocrinol Metab. 1998;83(7):2266–2274 [DOI] [PubMed] [Google Scholar]
- 18. Anasti JN, Kalantaridou SN, Kimzey LM, Defensor RA, Nelson LM. Bone loss in young women with karyotypically normal spontaneous premature ovarian failure. Obstet Gynecol. 1998;91:12–15 [DOI] [PubMed] [Google Scholar]
- 19. Popat VB, Vanderhoof VH, Calis KA, Troendle JF, Nelson LM. Normalization of serum luteinizing hormone levels in women with 46,XX spontaneous primary ovarian insufficiency. Fertil Steril. 2008;89(2):429–433 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Southren AL, Gordon GG, Tochimoto S. Further study of factors affecting the metabolic clearance rate of testosterone in man. J Clin Endocrinol Metab. 1968;28(8):1105–1112 [DOI] [PubMed] [Google Scholar]
- 21. Abraham GE, Lobotsky J, Lloyd CW. Metabolism of testosterone and androstenedione in normal and ovariectomized women. J Clin Invest. 1969;48(4):696–703 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Nelson LM. Primary ovarian insufficiency. N Engl J Med. 2009;360(6):606–614 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Kalantaridou SN, Calis KA, Mazer NA, Godoy H, Nelson LM. A pilot study of an investigational testosterone transdermal patch system in young women with spontaneous premature ovarian failure. J Clin Endocrinol Metab 2005;90(12):6549–6552 [DOI] [PubMed] [Google Scholar]
- 24. Consensus Development Conference: Diagnosis, Prophylaxis, and Treatment of Osteoporosis. Am J Med 1993;94:646–650 [DOI] [PubMed] [Google Scholar]
- 25. Ferriman D, Gallwey JD. Clinical assessment of body hair growth in women. J Clin Endocrinol Metab 1961;21:1440–1447 [DOI] [PubMed] [Google Scholar]
- 26. Tagatz GE, Kopher RA, Nagel TC, Okagaki T. The clitoral index: a Bioassay of androgenic stimulation. Obstet Gynecol. 1979;54(5):562–564 [PubMed] [Google Scholar]
- 27. Cremoncini C, Vignati E, Libroia A. Treatment of hirsutism and acne in women with two combinations of cyproterone acetate and ethinylestradiol. Acta Eur Fertil. 1976;7(4):299–314 [PubMed] [Google Scholar]
- 28. Bonnick SL, Johnston CC, Jr, Kleerekoper M, et al. Importance of precision in bone density measurements. J Clin Densitom. 2001;4(2):105–110 [DOI] [PubMed] [Google Scholar]
- 29. Nelson JC, Tomei RT. Direct determination of free thyroxine in undiluted serum by equilibrium dialysis/radioimmunoassay. Clin Chem. 1988;34(9):1737–1744 [PubMed] [Google Scholar]
- 30. Scarabin PY, Henc-Gelas M, Plu-Bureau, Taisne P, Agher R, Aiach M. Effects of oral and transdermal estrogen/progesterone regimens on blood coagulation and fibrinolysis in postmenopausal women. A randomized controlled trial. Arterioscler Thromb Vasc Biol. 1997;17(11):3071–3078 [DOI] [PubMed] [Google Scholar]
- 31. Scarabin PY, Oger E, Plu-Bureau Differential association of oral and transdermal oestrogen-replacement therapy with venous thromboembolism risk. Lancet 2003; 362(9382):428–432 [DOI] [PubMed] [Google Scholar]
- 32. Canonico M, Oger E, Plu-Bureau, et al. Hormone therapy and venous thromboembolism among postmenopausal women: impact of the route of estrogen administration and progestogens: the ESTHER study. Circulation. 2007;115(7):840–845 [DOI] [PubMed] [Google Scholar]
- 33. Mishell DR, Jr, Nakamura RM, Crosignani PG, et al. Serum gonadotropin and steroid patterns during the normal menstrual cycle. Am J Obstet Gynecol. 1971;111(1):60–65 [DOI] [PubMed] [Google Scholar]
- 34. Crofton PM, Evans N, Bath LE, et al. Physiological versus standard sex steroid replacement in young women with premature ovarian failure: effects on bone mass acquisition and turnover. Clin Endocrinol (Oxf). 2010;73(6):707–714 [DOI] [PubMed] [Google Scholar]
- 35. Rochira V, Zirilli L, Madeo B, et al. Skeletal effects of long-term estrogen and testosterone replacement treatment in a man with congenital aromatase deficiency: evidences of a priming effect of estrogen for sex steroids action on bone. Bone. 2007;40(6):1662–1668 [DOI] [PubMed] [Google Scholar]
- 36. Misra M, Katzman D, Miller KK, et al. Physiologic estrogen replacement increases bone density in adolescent girls with anorexia nervosa. J Bone Miner Res. 2011;26(10):2430–2438 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Klibanski A, Biller BM, Schoenfeld DA, Herzog DB, Saxe VC. The effects of estrogen administration on trabecular bone loss in young women with anorexia nervosa. J Clin Endocrinol Metab. 1995;80(3):898–904 [DOI] [PubMed] [Google Scholar]