|
|
| |||||||||||||||||||||||||||||||
Background Analogues of gonadotropin-releasing hormone (GnRH) are often given to induce hypogonadism in women who have estrogen-dependent disorders such as endometriosis and uterine leiomyomas. Because estrogen deficiency causes bone loss, concern about premature osteoporosis has prevented long-term therapy with GnRH analogues. We conducted a study to determine whether parathyroid hormone could prevent bone loss in women receiving therapy with GnRH analogues.
Methods We administered human parathyroid hormone (40 µg subcutaneously daily) to 20 of 40 women with endometriosis who were being treated with nafarelin (200 µg intranasally twice daily) for six months; the remaining 20 women received only nafarelin. Cortical and trabecular bone density and biochemical markers of bone turnover were measured every three months during the six-month study period.
Results Serum estradiol concentrations fell to postmenopausal values in 36 of the 40 women. In the women who received nafarelin alone, the mean (±SE) bone density in the lumbar spine decreased by 2.8 ±0.5 percent (P<0.001) when measured in the anteroposterior projection and by 3.5 ±0.8 percent (P<0.001) when measured in the lateral projection. In contrast, bone density in the lumbar spine did not change when measured in the anteroposterior projection and increased by 3.4 ±1.2 percent when measured in the lateral projection (P = 0.01) in the women who also received parathyroid hormone. Bone density at the femoral neck decreased slightly and similarly in both groups. Radial bone density did not change in either group. Serum alkaline phosphatase and osteocalcin concentrations and urinary hydroxyproline and pyridinoline excretion increased (P<0.001) in the women who received nafarelin plus parathyroid hormone.
Conclusions Parathyroid hormone can prevent bone loss in the lumbar spine in young women with estrogen deficiency caused by treatment with GnRH analogues.
Although continuous administration of parathyroid hormone (PTH) decreases bone mass,10 intermittent administration of PTH increases bone formation more than it increases bone resorption,10,11,12,13 leading to increased bone mass13,14,15,16,17,18,19. Moreover, intermittent administration of PTH prevents short-term bone loss caused by castration in animals20,21. Thus, we hypothesized that the intermittent administration of low doses of PTH would prevent bone loss in young women with endometriosis who had hypogonadism induced by therapy with a GnRH analogue.
Methods
Study Subjects
We studied 50 women (48 white and 2 black) who were 20 to 44 years of age and who had symptomatic, laparoscopically proved endometriosis. Ovulation during the cycle before the study was confirmed by biphasic basal-body-temperature records and a luteal-phase serum progesterone concentration greater than 5 ng per milliliter (16 nmol per liter). All the women had normal serum calcium, phosphate, alkaline phosphatase, bilirubin, creatinine, thyroxine, thyrotropin, and prolactin concentrations. Women with a history of hyperthyroidism, hyperparathyroidism, Cushing's syndrome, hyperprolactinemia, anorexia nervosa, chronic renal or hepatic disease, or alcoholism and those who were taking a glucocorticoid or diuretic, doses of thyroxine that suppress serum thyrotropin concentrations below the normal range, or anticonvulsant drugs were excluded. Oral contraceptives and danazol were discontinued for at least two months and GnRH-analogue therapy was discontinued for at least nine months before the start of the study. We also studied 13 normal women, 23 to 45 years of age, without endometriosis who met all the other entrance criteria to determine whether bone density and markers of bone turnover were normal in women with endometriosis. The study was approved by the human studies committee of Massachusetts General Hospital, and all the women gave informed consent.
Protocol
The women were admitted to the Clinical Research Center between days 6 and 10 of their menstrual cycles for measurements of the bone mineral density of the radius, spine, and proximal femur. We also measured hydroxyproline, free pyridinoline plus deoxypyridinoline, and creatinine in the second urine specimens obtained in the morning after an overnight fast (to assess bone resorption) and alkaline phosphatase, osteocalcin, calcium, phosphate, PTH, calcidiol, calcitriol, total and fractionated cholesterol, and estradiol concentrations in fasting serum samples. The women's daily calcium intake, percentage of body fat, body-mass index (the weight in kilograms divided by the square of the height in meters), and percentage of the ideal body weight (according to Metropolitan Life Insurance Company tables) were determined by a research dietitian. Histories of the women's involvement in running, weight lifting, and aerobic exercise were obtained.
The women with endometriosis were randomly assigned to receive either the GnRH analogue nafarelin acetate (Synarel, Syntex Laboratories, Palo Alto, Calif.) at a dose of 200 µg intranasally twice a day for six months (group 1, 22 women) or nafarelin acetate plus the fragment of human PTH consisting of the first 34 amino acids (hPTH-[1-34]) at a dose of 40 µg (500 units) subcutaneously per day for six months (group 2, 28 women). The women were asked to maintain a daily calcium intake of approximately 1200 mg through diet or calcium carbonate supplements, but they were not asked to change their exercise habits.
Two women in group 1 and eight women in group 2 withdrew from the study before any follow-up measurements were made. Three of these women withdrew for reasons clearly related to the PTH injections: one because of inability to give herself injections, one because of discomfort from needle sticks, and one because of a local reaction to PTH. Other reasons for dropping out were headaches (two women), family problems (one woman), vasovagal reaction to phlebotomy (one woman), preexisting hypercalciuria (one woman), move to a residence out of state (one woman), and mood swings (one woman). Because the results are presented as change from base-line values, the data on these women are not included.
The remaining 40 women (20 in each group) were readmitted after three and six months of therapy for reevaluation, including assessment of side effects and reported relief of symptoms. Blood was drawn just before the morning doses of PTH and nafarelin or nafarelin alone. Compliance was assessed by means of medication diaries, counts of medication vials, and measurements of serum estradiol. If the serum estradiol concentrations remained 
40 pg per milliliter (147 pmol per liter) or if symptoms of endometriosis were not substantially relieved after three months, the nafarelin dose was increased by 200 or 400 µg per day. In group 2, 24-hour urinary calcium excretion and serum calcium were measured two to four weeks after the beginning of PTH therapy. If urinary calcium excretion exceeded 350 mg per day (8.7 mmol per day), dietary calcium intake was decreased by 50 percent. If the serum calcium concentration exceeded 10.5 mg per deciliter (2.62 mmol per liter), the dose of PTH and dietary calcium intake were both decreased by 50 percent. Measurements of calcium in serum or urine were repeated to document normality. More than 99 percent of the prescribed nafarelin was taken by all the women in group 1 and by 19 women in group 2. In group 2, 19 women received at least 93 percent of their PTH injections. Vial counts and serum estradiol measurements suggested noncompliance with the nafarelin and PTH regimen on the part of one woman in group 2.
Determination of Bone Mineral Density
Bone mineral density at the radius, lumbar spine, and femoral neck was determined by dual-energy x-ray absorptiometry (Hologic QDR-2000, Waltham, Mass.)22. Measurements of the nondominant radius were made at the junction of the proximal two thirds and the distal one third of the radial shaft; the coefficient of variation for this measurement was 1.5 percent23. The bone mineral density of the lumbar spine was assessed in both the anteroposterior and the lateral projection with the women supine. Because lateral measurements of bone density in the spine eliminate the contribution of the posterior vertebral elements, they estimate trabecular bone mass better than measurements made in the anteroposterior projection22. The standard deviations for the anteroposterior and lateral measurements were 0.010 and 0.013 g per cubic centimeter, respectively, in 50 subjects with paired measurements, and these values did not vary with bone density. The coefficient of variation for bone density at the femoral neck was 2.1 percent in 51 subjects with paired measurements24. All scans were reviewed by a physician who did not know the woman's treatment assignment. Follow-up scans were matched to base-line scans to ensure measurement of identical bone regions. Vertebrae with obvious deformities or areas of focal sclerosis were not analyzed, and those that had visible overlap with ribs or the pelvis were eliminated from the analysis of lateral spinal scans.
Biochemical Assays
Serum intact PTH and osteocalcin were measured with immunoradiometric assays (Nichols Institute, San Juan Capistrano, Calif.). The assay for intact PTH did not detect the 34-amino-acid PTH fragment administered in the study. Serum calcidiol (Incstar, Stillwater, Minn.), calcitriol (Nichols Institute), and estradiol concentrations were determined by radioimmunoassay. Urinary hydroxyproline excretion was measured by SmithKline Bioscience Laboratories (Van Nuys, Calif.) with an automated spectrophotometric analysis. Urinary free pyridinolines were measured with an enzyme-linked immunosorbent assay (Metra Biosystems, Palo Alto, Calif.). Except for the serum estradiol and urinary hydroxyproline measurements, all samples for each woman were analyzed in a single assay. The respective intraassay and interassay coefficients of variation were 3 percent and 6 percent for PTH, 6 percent and 16 percent for calcidiol, 5 percent and 9 percent for calcitriol, 5 percent and 7 percent for osteocalcin, and 5 percent and 8 percent for urinary free pyridinolines.
Statistical Analysis
A mixed-model analysis of variance was used to assess the effect of treatment on each variable. The random effects were the individual subject and the interaction of the time of measurement (0, 3, or 6 months) with the subject. The fixed effects were the treatment group, the time of the measurement, and the interaction of time and treatment group. The effect of treatment on each end point was measured by the effect of the interaction of time and treatment, with the effect of the interaction of time and subject as the error term. Serum calcium, phosphate, estradiol, PTH, calcidiol, calcitriol, and cholesterol concentrations were compared within each group at base line and at six months by means of paired t-tests. Base-line clinical characteristics (except for discontinuous variables) and laboratory values for all groups were compared by analysis of variance. Unless noted otherwise, the results are expressed as means ±SD.
Results
Clinical Characteristics
The clinical characteristics of the women are shown in Table 1. In group 1, 18 women had previously taken oral contraceptives for an average of 3 ±3 years; 4 had taken danazol; and 5 had taken a GnRH analogue. In group 2, 16 women had previously taken oral contraceptives for an average of 4 ±6 years; 4 had taken danazol; and 5 had taken a GnRH analogue. The base-line characteristics did not differ between the two groups or between the women with endometriosis and the 13 normal women without endometriosis.
|
During nafarelin therapy, serum estradiol concentrations decreased significantly and comparably in both groups (Table 2), reaching postmenopausal values in 19 of the 20 women in group 1 and 17 of the 20 women in group 2. In group 1, the daily nafarelin dose at the end of six months was 400 µg for 12 women, 600 µg for 3 women, and 800 µg for 5 women. In group 2, it was 400 µg for 14 women, 600 µg for 2 women, and 800 µg for 4 women. In group 1, 19 women reported a decrease of at least 75 percent in the symptoms of endometriosis; 9 of them reported complete relief of symptoms. In group 2, 18 women reported a decrease of at least 75 percent in the symptoms of endometriosis; 11 reported complete relief. The side effects of nafarelin treatment in the two groups are listed in Table 3. Hypercalciuria developed in two women (urinary calcium excretion, 410 and 430 mg per day [10.2 and 10.7 mmol per day]); these values normalized after their calcium intake was reduced. One woman had a serum calcium concentration of 10.7 mg per deciliter (2.67 mmol per liter) after three months that returned to normal when the dose of PTH was decreased to 20 µg daily. Another woman had a serum calcium concentration of 11.7 mg per deciliter (2.92 mmol per liter) at the final examination.
|
|
The mean (±SE) bone mineral density of the lumbar spine decreased by 2.8 ±0.5 percent when measured in the anteroposterior projection (P<0.001) and by 3.5 ±0.8 percent when measured in the lateral projection (P<0.001) in the women who received nafarelin alone (Figure 1). In contrast, the simultaneous administration of PTH and nafarelin increased spinal bone mineral density in the lateral projection by 3.4 ±1.2 percent (P = 0.01) and prevented the decrease in spinal bone mineral density in the anteroposterior projection. The rates of change in spinal bone mineral density in both anteroposterior and lateral projections differed significantly between the two groups (Figure 1). The bone mineral density of the femoral neck decreased slightly and similarly in both groups (P = 0.01 for group 1; P = 0.02 for group 2). Radial bone mineral density did not change in either group. Differences in exercise levels did not account for the beneficial effects of PTH on bone density.
|
In group 1, the serum alkaline phosphatase concentration and urinary excretion of free pyridinolines increased significantly (P = 0.02 for both comparisons), whereas the serum osteocalcin concentration and urinary hydroxyproline excretion did not change (Figure 2). Serum alkaline phosphatase and osteocalcin concentrations and urinary excretion of hydroxyproline and pyridinolines all increased significantly (P<0.001) in group 2 (Figure 2). The rate of change for each marker of bone turnover differed significantly between the two groups.
|
Discussion
We demonstrated that the intermittent administration of PTH prevents spinal bone loss in young women with endometriosis who have estrogen deficiency caused by therapy with nafarelin, a GnRH analogue. Spinal bone density, which reflects mainly the status of trabecular bone, decreased by 3 to 4 percent in women receiving nafarelin alone; it increased in the women who received PTH, despite severe estrogen deficiency. Increased bone density was accompanied by increases in biochemical markers of bone formation and resorption.
Bone loss caused by estrogen deficiency is an important problem as women reach menopause26 and in premenopausal women with hypogonadism1,2,3,27. If estrogen deficiency is sustained, bone loss becomes irreversible3,26,27. Because current therapies cannot reverse osteoporosis, it is crucial to prevent its development. Although agents that inhibit osteoclastic bone resorption, including estrogen, calcitonin, and calcium, can retard bone loss early after menopause,28 only estrogen reliably prevents both cortical and trabecular bone loss. Still, only 15 percent of postmenopausal women receive estrogen-replacement therapy, mainly because of concern about the risk of breast cancer and because cyclic regimens cause monthly menstrual bleeding28. Thus, alternative strategies to prevent bone loss caused by estrogen deficiency are needed. It is likely that PTH will prevent bone loss in women with other causes of estrogen deficiency, as it does in those with estrogen deficiency induced by GnRH analogues.
Agents such as estrogen, calcitonin, and calcium decrease bone formation and resorption, whereas the intermittent administration of PTH stimulates bone formation21. Data from studies in animals suggest that the intermittent administration of PTH does not increase bone resorption10,13. However, the biochemical measurements we performed suggest that the administration of PTH increases both processes. Furthermore, PTH preserves and may restore the internal structure of trabecular bone,29 whereas GnRH analogues disrupt trabecular microstructure30. Because bone density at all sites in the women who received nafarelin and PTH equaled or exceeded that in the women who received only nafarelin, it appears that increased bone formation offset the increase in bone resorption. Although the mechanism for the anabolic action on bone of PTH administered intermittently is poorly understood, growth hormone is required for this effect in animals31; this effect may be caused by increased local production of insulin-like growth factor I or other bone-growth factors32.
The effect of PTH on cortical bone is less clear. PTH consistently increases cortical bone mass in animals12,31,33,34. When administered with estrogen to women with postmenopausal osteoporosis, PTH increases both cortical and trabecular bone mass35. However, PTH administered alone or in combination with calcitriol accelerates cortical bone loss in older women18,19. Thus, the effects of PTH may differ in younger and older women. Cortical bone density did not change in either group in our study. Because cortical bone is lost more slowly than trabecular bone in estrogen-deficient women, additional follow-up is needed to determine whether the administration of PTH prevents cortical bone loss in women treated with GnRH analogues.
Although one study suggested that bone mass is decreased in women with endometriosis,36 in both our study and a recent large study base-line bone density was normal37. Bone loss is largely reversible when GnRH-analogue therapy is limited to six months,5,6,7,8 but the disorders for which these drugs are prescribed typically require long-term gonadal suppression. Bone loss can be minimized or prevented if estrogen or a progestin is administered with the GnRH analogue6. However, the addition of a progestin lowers serum HDL cholesterol concentrations4,6,8 and diminishes the effectiveness of GnRH analogues as a treatment for uterine leiomyomas4. Because therapy with PTH does not lower HDL cholesterol concentrations and stops bone loss without gonadal steroid treatment, it may allow long-term GnRH analogue therapy without the side effects that otherwise occur.
We conclude that the administration of PTH prevents spinal bone loss caused by acute estrogen deficiency for at least six months with minimal side effects. Although PTH therapy holds promise as a means to prevent bone loss caused by estrogen deficiency, its value is currently limited by the need for parenteral administration, its cost, and the need to measure serum and urinary calcium periodically.
Supported by grants (HD-21204, RR-1066, and 1 R29 DK/43341-01A2) from the National Institutes of Health (NIH) and by an NIH Clinical Associate Physician Award to Dr. Finkelstein.
We are indebted to Syntex Laboratories for supplying the nafarelin acetate (Synarel) used in this study; to Ellen Anderson, R.D., Ellen Williams, R.D., Ms. Tonya Walshe, Jane Hubbard, R.D., Kimberly Ziemer, R.D., and Ms. Kirti Raol for obtaining the dietary histories and body-fat measurements; to Ms. Robbin Cleary and Ms. Sarah Zhang for measuring the bone mineral densities; to Ms. Pamela Miller and Julia MacLaughlin for performing the biochemical assays; to Dr. David A. Schoenfeld for statistical advice; to Drs. Veronica Ravnikar, Najmosama Nikrui, Thomas Toth, Raja Sayegh, and Keith Isaacson for referring the women to our center; and to the nursing staff of the general clinical research center for their meticulous performance of the study and their dedicated care of the patients.
Source Information
From the Endocrine Unit (J.S.F., E.H.S., R.M.N.) and the Neuroendocrine Unit (A.K.), Department of Medicine, and the Department of Gynecology (I.S.), Massachusetts General Hospital, Boston; and the Department of Obstetrics and Gynecology, Brigham and Women's Hospital, Boston (M.D.H.).
Address reprint requests to Dr. Finkelstein at the Endocrine Unit, Bulfinch 327, Massachusetts General Hospital, Boston, MA 02114.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
