0021-972X/90/7104-0836$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1990 by The Endocrine Society
Vol. 71, No. 4 Printed in U.S.A.
17/?-Estradiol and Continuous Norethisterone: A Unique Treatment for Established Osteoporosis in Elderly Women CLAUS CHRISTIANSEN AND BENTE JUEL RIIS Department of Clinical Chemistry, Glostrup Hospital, University of Copenhagen, DK-2600 Glostrup, Denmark line phosphatase and plasma osteocalcin), decreased significantly (P < 0.001) in the group treated with hormones, but remained unchanged in the placebo group. The reduction in indices of bone resorption was more pronounced than that in bone formation after one year, indicating a positive bone balance. No further changes were seen in these bone turnover parameters during the second year of treatment. In the group treated with hormones, serum levels of triglycerides, total cholesterol, and low density lipoprotein cholesterol decreased by about 12% (P < 0.05-P < 0.01), whereas high density lipoprotein cholesterol decreased by about 8% (P < 0.001). The high density lipoprotein cholesterol/low density lipoprotein cholesterol ratio was unchanged. The hormone treatment did not produce any major side effects, and only minor bleedings were experienced by a few women. The present study demonstrates that treatment with female sex hormones in this particular combination is a realistic approach to the treatment of women with established postmenopausal osteoporosis. (J Clin Endocrinol Metab 7 1 : 836-841, 1990)
ABSTRACT. Forty women aged 64.7 ± 5.1 yr with established postmenopausal osteoporosis were blindly allocated to 1 yr's treatment with either continuous combined estrogen/progestogen therapy (2 mg estradiol + 1 mg norethisterone acetate + 500 mg calcium daily) or placebo + 500 mg calcium daily. In the group treated with hormones bone mineral density in the spine (dual photon absorptiometry) and bone mineral content in the ultradistal forearm (single photon absorptiometry) increased highly significantly by 8-10% during the 1 yr of treatment. Bone mineral content in the mid-shaft of the forearm (single photon absorptiometry) and the total body bone mineral (dual photon absorptiometry) increased by 3-5% when compared to that in the placebo group, which showed virtually unchanged values at all measurement sites. Seven of the women treated with hormones were examined after a further year of treatment. BMC increased by another 3-6%, reaching a 12% increase in bone mineral density in the spine after 2 yr of treatment. Biochemical estimates of bone resorption (fasting urinary calcium and hydroxyproline) and bone formation (serum alka-
P
OSTMENOPAUSAL osteoporosis is a disorder characterized by reduced amounts of bony tissue per unit volume of bone and by increased susceptibility to fractures (1). The clinical manifestations of osteoporosis include fractures and their complications. Osteoporosis is a common clinical disorder and a major public health problem, affecting as many as one out of two women in a 70-yr-old population (2). It is now considered established that postmenopausal estrogen therapy stops further bone loss (3, 4), and, if started in due time and continued long enough, decreases the number of osteoporotic fractures (5, 6). A number of studies have furthermore shown that estrogens in combination with different progestogens, have the same effect on the early postmenopausal bone loss as that of unopposed estrogen therapy (7, 8). We reported in an earlier study that treatment with
17/3-estradiol (17/3E2) in sequential combination with norethisterone acetate (NETA) induced a continuous increase in bone mineral content (BMC) of 1%/yr for at least 3 yr in early postmenopausal women (4), and for at least 1 yr in 70-yr-old women (9). These results have been confirmed by a recent study with the same estrogen/ progestogen regimen, which also produced an increased BMC in early postmenopausal women (10). We recently performed a short-term study with the aim of closely investigating biochemical estimates of bone resorption and bone formation during treatment of early postmenopausal women with sequential 17/3E2 + NETA. The results suggested that NETA in combination with 17jSE2 has a unique ability to stimulate bone formation (11). This theory was supported by a study in which premenopausal women with endometriosis were treated with an LHRH agonist, a treatment that blocks the ovarian production of estrogen. The premature bone loss which would otherwise have occurred, was almost completely prevented when the LHRH agonist was given in combination with NETA (12).
Received January 16, 1990. Address requests for reprints to: Claus Christiansen, M.D., Department of Clinical Chemistry, Glostrup Hospital, DK-2600 Glostrup, Denmark. 836
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 September 2015. at 01:32 For personal use only. No other uses without permission. . All rights reserved.
ESTROGEN/PROGESTOGEN THERAPY ON ELDERLY WOMEN WITH OSTEOPOROSIS To avoid the monthly withdrawal bleeding accompanying sequential combined therapy, the continuous combined 17/3E2/NETA therapy has recently been introduced. The rationale is that the continuous administration of progestogen keeps the endometrium atrophic, and thereby prevents withdrawal bleeding and the development of endometrial hyperplasia or neoplasia (13, 14). Treatment of symptomatic postmenopausal osteoporosis still remains an unsolved problem. Many treatment regimens have been proposed, including calcitonin (15), fluoride (16), diphosphonates (17), anabolic steroids (18), 1-a-vitamin D3 (19), and PTH (20), but the cost effectiveness of these treatments are still debated. Whereas the 85-yr old woman with severe osteoporosis is generally acknowledged to be out of therapeutic reach, the treatment of the 65-yr old woman showing the first signs of osteoporosis would certainly be worthwhile, if this was possible. The aim of the present study was therefore to assess the continuous combined estradiol/NETA regimen in the treatment of symptomatic postmenopausal osteoporosis. Patients and Methods Subjects This study was part of a large double-blind clinical controlled trial carried out at Glostrup Hospital from March 1985 to December 1986. All participants were informed verbally and in writing about the trial, and all gave their informed consent, in accordance with Helsinki Declaration II. The participants were selected according to the following procedures: 1) All spinal x-rays (4900) of women aged 55-75 yr taken during the preceding 5 yr at Glostrup Hospital were retrieved. Those carrying an original radiological diagnosis indicating osteoporosis (osteoporosis, osteopenia, vertebral deformity) were reassessed. Of these, 219 women were found to have had at least one vertebral fracture (wedge or compression). Eightyone were excluded because of malignant, gastrointestinal, metabolic, or rheumatic disease, or drug intake known to affect calcium metabolism. Former, transitory use of hormonal substitution was not an exclusion criterion. Of the remaining 138 patients, 44 agreed to undergo experimental therapy for osteoporosis. After the introductory investigation five patients dropped out. 2) All patients referred to Glostrup Hospital because of distal forearm fracture in the same period were approached and the same criteria were applied. Of 316 women 133 were excluded. Eighty of the remaining 183 agreed to participate. After an initial examination the women were randomly allocated to one of three main treatment groups: 1) estrogen/gestogen; 2) anabolic steroids (21); 3) nasal calcitonin (22). Randomization was performed as follows: the medication was coded outside the clinic with numbers from 1 to 40 for each of the three groups, but taking care that the number allocated to the active treatment and that to placebo were the same for the first 20 and the last 20 numbers (blocked randomization). The first 20 patients with Colles' fracture (nos. 1-20) and the first 20 patients with
837
spinal fractures (nos. 21-40) were assigned to the estrogen/ gestogen (or placebo) group. The next 20 patients with Colles' fracture (nos. 1-20) and the last 19 patients with spinal fractures (nos. 21-39) were assigned to the anabolic steroid (or placebo group). The last 40 patients with Colles' fractures were assigned to the calcitonin (or placebo) group. We report here the data of the 40 women who blindly received either 2 mg 17j8E2 + 1 mg NETA + 500 mg calcium (Sandoz) (n = 20; 10 with spinal fractures, 10 with Colles; fractures) or a placebo combined with the same calcium supplementation (n = 20; 10 with spinal fractures, 10 with Colles' fractures). The treatment was given for 1 yr, during which all patients were examined every 3 months (five examinations in all). Seven of the women who received hormones were examined after a further year of treatment and were compared with 10 women from the same background population who received no therapy, but who were followed up in parallel. Compliance to trial Of the 40 women who entered the trial, 31 (78%) completed the 1 yr of treatment. Four women in the hormone group left (two with Colles' fracture, two with spinal fracture): one woman was killed in a car accident, one because she started treatment with digitalis for decompensated heart disease, one woman because of unacceptable breast tenderness, and one because of a discovered breast tumour (benign). Five women left the placebo group (two with Colles' fracture, three with spinal fracture): one woman died from unknown causes, one because of aggravation of heart disease, one because of general indisposition for some unknown reason, one because of aggravation of eczema, and one for personal reasons. This left 16 women in the hormone group and 15 women in the placebo group. Methods Bone mass measurements. BMC of the forearms was measured by single photon absorptiometry, with I125. The method determines BMC in a proximal (BMCprox) and a distal (BMCdist) region of the forearm with a trabecular bone content of 13% and 55%. The software for the equipment is developed in our laboratory and the long term in vivo precision of proximal and distal BMC is 1% and 1.5% (23). In the case of forearm fracture, we measured the nonfractured forearm; where bilateral fractures had occurred at different times, we measured the arm with the oldest fracture; and where bilateral fractures had occurred at the same time, we measured both forearms. Bone mineral content of the lumbar spine (BMCspine) was measured by dual photon absorptiometry on a Lunar Radiation Corporation DP3 scanner. BMCspine was calculated as the BMC values obtained in vertebrae L2, L3, and L4. The long term in vivo precision is 4%, after correcting the values for the errors due to source change (23). Bone mineral content of the total skeleton (TBBM) was also measured by dual photon absorptiometry on a whole body scanner developed in our laboratory. The long-term in vivo precision of TBBM is about 2.5% (24), after correcting for source change. Biochemical estimates of calcium metabolism. Blood samples were taken and urine collected in the morning after an over-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 September 2015. at 01:32 For personal use only. No other uses without permission. . All rights reserved.
CHRISTIANSEN AND RIIS
838
night fast and tobacco abstinence. Serum alkaline phosphatase (AP) was measured enzymatically according to Scandinavian recommendations, and plasma bone Gla protein (BGP) by RIAs (25). Fasting urinary calcium and creatinine were measured on an SMA 6/60 auto-analyzer and fasting urinary hydroxyproline by a spectrophotometric method. Calcium and hydroxyproline excretions were corrected for creatinine excretion (FU Ca/Cr and FU Hpr/Cr). Radiopharmaceutical measurement of bone turnover. Whole body retention (WBR) of 99Technetium-diphosphonate (99mTc-DP) was measured with a wide-field 7-camera with diverging parallel (fishtail) collimator. Two microcuries of 99mTc-DP were injected iv and the whole body measurement was started 5 min later. Scanning was repeated at 24 h and WBR was calculated after subtracting background radiation and correcting for radioactive decay by using the mean count at 5 min as 100% and comparing this to the mean count at 24 h. WBR is expressed in percent of the retained radioactivity. The method has a precision of 4.8% (26). Serum lipids and lipoproteins. High density lipoprotein cholesterol (HDL-C) was separated from chylomicrons, and low density lipoprotein cholesterol (LDL-C) and very low density lipoprotein (VLDL-C) by the mg++/dextran sulfate precipitation technique (27). Total serum cholesterol and HDL-C were determined by an enzymatic technique with the use of an LKB Ultralab system. The intra- and interassay variations were 1% and 2% (total cholesterol) and 2% and 5% (HDL-C), respectively. Serum triglycerides were determined by an enzymatic technique with the use of an LKB 8600 reaction rate analyser. The intra- and interassay variations of this method are 5% and 8%. LDL-C was then estimated according to the procedure of Friedewald et al. (28). The intra- and interassay variations are 5% and 10%.
JCE & M • 1990 Vol71«No4
TABLE 1. Clinical data (mean ± 1 SD) Hormones (n = 16) Age (yr) Menopausal age (yr) Ht (cm) Wt (kg) No. of patients with spinal fractures (crush/wedge)
Significance
Placebo (n = 15)
-c 01
63.6 ± 5.0 16 ± 6
difference NS NS
64.8 ± 5.7 18 ± 7
159.7 ± 9.4 59.9 ± 8.3
NS NS
159.3 ± 6.1 65.2 ± 9.9
12
12
TABLE 2. Absolute values of bone mass before and after 1 yr of treatment with hormones or placebo (mean ± 1 SD) Hormones Before therapy BMCprox (units) 27.4 ± 5.1 BMCdist (units) 24.7 ± 4.3 BMC8pine (g) 35.4 ±8.4 TBBM (g) 2534 ± 558 0 6
Placebo Before therapy
After therapy 27.9 ± 5.0° 27.6 ± 6.26 37.0 ± 8.3° 2655 ± 590°
27.4 26.6 33.4 2586
± ± ± ±
6.2 5.6 6.4 496
After therapy 27.2 ± 6.1 26.7 ± 5.8 32.4 ± 6.8 2588 ± 469
P < 0.05. P < 0.01 (paired t test). Mainly trabecular bone
Mainly cortical bone
Uterine Bleeding. Bleeding was recorded on a day-to-day calendar supplied at each examination. -10-
Safety parameters
J
Serum aspartate aminotransferase (s-ASAT), serum creatinine (s-Crea), hemoglobin (hb), and blood pressure (BP) were measured by routine procedures.
-20
Statistical analysis
and the group with vertebral fractures the groups were pooled. Tables 1 and 2 give the clinical data and initial values of bone mineral measurements. By chance, the hormone group weighed about 5 kg less than the placebo group, but, the difference was not statistically significant. The two groups were otherwise well matched. Figure 1 shows the individual changes in the bone mineral content occurring over the 1 yr of treatment (percent values). At all four measurement sites, the mean bone mineral content was increased in the hormone group as compared with the placebo group (P < 0.05). BMCdist and BMCspine showed a difference between the hormone and the placebo groups of about 8%, whereas the differences in BMCprox and TBBM were
The required sample size was estimated to be about 40 participants, as calculated on the precision of BMCspine (4%) and an estimated difference of interest between the active treatment and placebo groups of 4%. Clinical data and initial values of the measured variables were compared by Student's t test for unpaired data (Table 1). All initial values were set at 100% and subsequent values were expressed in per cent of that value. Differences within groups were evaluated by Student's t test for paired data and between groups by Student's t test for unpaired data.
Results Because the response to treatment was not significantly different between the group with Colles' fractures
BMCprox
TBBM
FIG. 1. Individual changes in bone mass measurements after 1 yr of treatment with hormones (O) or placebo (•).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 September 2015. at 01:32 For personal use only. No other uses without permission. . All rights reserved.
ESTROGEN/PROGESTOGEN THERAPY ON ELDERLY WOMEN WITH OSTEOPOROSIS about 3% and 5%, respectively. Table 2 gives the absolute changes in bone mass, which give virtually the same statistical significance. Figure 2 shows the mean percent changes in the four bone compartments over the 2 yr of treatment. The women receiving the active treatment had on the average an increased bone mass of 3.6-13.1%; out of the 28 individual changes in the hormone group only one was below zero, whereas 34 out of the 40 values in the control group were negative. Figure 3 visualizes the changes in the estimates of bone formation and bone resorption. Compared with the placebo group, the hormone group had significant decreases in pBGP, sAP, FU Hpr/Cr, FU Ca/Cr and sphosphate. WBR was also decreased but only to borderline significance (P = 0.06). None of the biochemical variables changed significantly during the second year of treatment. Table 3 shows the initial values of serum lipids and lipoproteins and the cumulated changes during treatment. The initial values were similar in the two groups. The changes in the hormone group showed significant decreases in all serum lipids (P < 0.05-0.001). All values were virtually stable during the second year. Table 4 shows the initial values of and the percent changes in the safety parameters. The hormone group had a mean systolic blood pressure significantly lower than that of the placebo group (P < 0.05). Neither group had significant changes in any of the variables, except for serum calcium (P < 0.001) and the diastolic blood pressure (P < 0.05), both of which were significantly decreased in the hormone group. Six women in the hormone group experienced uterine bleeding; the bleedings were slight (not requiring pads), and lasted at most for 2 days. In four women this occurred during the first 3 months of treatment, in one during the fourth, and in one during the third, fourth, and sixth months of treatment. One woman in the placebo group BMCprox 18
«
100!
90r
TBBM
i
i
i
i
BMCspine
120110100-
12
24
0
6
Bone resorption FUHpr/Cr
months
FIG. 3. Mean changes in biochemical estimates of bone formation and resorption over 1 yr of treatment with hormones (O) or placebo (•).
had slight bleeding on three occasions during the last six treatment months and this was due to senile colpitis. The four women with bleeding in the first 3 months were given a gynecological examination, but curettage was not indicated. The two women with bleeding after this time had a curettage. All gynecological and histological examinations were normal. The woman in the placebo group who had bleeding was given a gynecological examination but not curettage.
Discussion
BMCdist
11018
839
12
24
months
FIG. 2. Mean changes in bone mass measurements over 2 yr of treatment with (O) and without hormones (•).
The women in the present study were selected on the original radiological diagnosis. The result was a relatively low number of women with vertebral deformities, probably because many radiologists are not aware of the importance of osteoporosis or vertebral deformities. It does not reflect the true prevalence. In an earlier study of 70-yr-old women (29), we found a much higher prevalence, similar to that reported by others (30). A large number of clinical studies have been performed in recent years in order to find a therapeutic agent able to increase bone mass in osteoporotic subjects. The study populations have often comprised relatively oldpostmenopausal women with spinal crush fractures and thus a severe degree of osteoporosis. The participants in the present study were selected on their moderate degree of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 September 2015. at 01:32 For personal use only. No other uses without permission. . All rights reserved.
840
CHRISTIANSEN AND RIIS
JCE&M-I99O Vol 71 • No 4
TABLE 3. Serum lipids and lipoproteins before and after 1 yr of treatment with hormones or placebo (mean ± 1 SD) Hormones
S-triglycerides mmol/L Total s-cholesterol mmol/L S LDL mmol/L S HDL mmol/L
Placebo
Before therapy (absolute values)
After therapy (% of initial values)
1.16 ± 0.38 6.82 ± 1.30 4.87 ± 1.25 1.45 ± 0.27
88.0 ± 14.7° 88.0 ± 9.0" 88.9 ± 11.3* 92.2 ± 14.3C
Before therapy (absolute values) 1.36 ± 7.86 ± 5.70 ± 1.55 ±
0.75 2.16 2.12 0.47
After therapy (% of initial values) 103.7 ± 27.1 99.3 ± 9.6 96.7 ± 11.3 110.7 ± 13.0c
a
P
Vol. 71, No. 4 Printed in U.S.A.
17/?-Estradiol and Continuous Norethisterone: A Unique Treatment for Established Osteoporosis in Elderly Women CLAUS CHRISTIANSEN AND BENTE JUEL RIIS Department of Clinical Chemistry, Glostrup Hospital, University of Copenhagen, DK-2600 Glostrup, Denmark line phosphatase and plasma osteocalcin), decreased significantly (P < 0.001) in the group treated with hormones, but remained unchanged in the placebo group. The reduction in indices of bone resorption was more pronounced than that in bone formation after one year, indicating a positive bone balance. No further changes were seen in these bone turnover parameters during the second year of treatment. In the group treated with hormones, serum levels of triglycerides, total cholesterol, and low density lipoprotein cholesterol decreased by about 12% (P < 0.05-P < 0.01), whereas high density lipoprotein cholesterol decreased by about 8% (P < 0.001). The high density lipoprotein cholesterol/low density lipoprotein cholesterol ratio was unchanged. The hormone treatment did not produce any major side effects, and only minor bleedings were experienced by a few women. The present study demonstrates that treatment with female sex hormones in this particular combination is a realistic approach to the treatment of women with established postmenopausal osteoporosis. (J Clin Endocrinol Metab 7 1 : 836-841, 1990)
ABSTRACT. Forty women aged 64.7 ± 5.1 yr with established postmenopausal osteoporosis were blindly allocated to 1 yr's treatment with either continuous combined estrogen/progestogen therapy (2 mg estradiol + 1 mg norethisterone acetate + 500 mg calcium daily) or placebo + 500 mg calcium daily. In the group treated with hormones bone mineral density in the spine (dual photon absorptiometry) and bone mineral content in the ultradistal forearm (single photon absorptiometry) increased highly significantly by 8-10% during the 1 yr of treatment. Bone mineral content in the mid-shaft of the forearm (single photon absorptiometry) and the total body bone mineral (dual photon absorptiometry) increased by 3-5% when compared to that in the placebo group, which showed virtually unchanged values at all measurement sites. Seven of the women treated with hormones were examined after a further year of treatment. BMC increased by another 3-6%, reaching a 12% increase in bone mineral density in the spine after 2 yr of treatment. Biochemical estimates of bone resorption (fasting urinary calcium and hydroxyproline) and bone formation (serum alka-
P
OSTMENOPAUSAL osteoporosis is a disorder characterized by reduced amounts of bony tissue per unit volume of bone and by increased susceptibility to fractures (1). The clinical manifestations of osteoporosis include fractures and their complications. Osteoporosis is a common clinical disorder and a major public health problem, affecting as many as one out of two women in a 70-yr-old population (2). It is now considered established that postmenopausal estrogen therapy stops further bone loss (3, 4), and, if started in due time and continued long enough, decreases the number of osteoporotic fractures (5, 6). A number of studies have furthermore shown that estrogens in combination with different progestogens, have the same effect on the early postmenopausal bone loss as that of unopposed estrogen therapy (7, 8). We reported in an earlier study that treatment with
17/3-estradiol (17/3E2) in sequential combination with norethisterone acetate (NETA) induced a continuous increase in bone mineral content (BMC) of 1%/yr for at least 3 yr in early postmenopausal women (4), and for at least 1 yr in 70-yr-old women (9). These results have been confirmed by a recent study with the same estrogen/ progestogen regimen, which also produced an increased BMC in early postmenopausal women (10). We recently performed a short-term study with the aim of closely investigating biochemical estimates of bone resorption and bone formation during treatment of early postmenopausal women with sequential 17/3E2 + NETA. The results suggested that NETA in combination with 17jSE2 has a unique ability to stimulate bone formation (11). This theory was supported by a study in which premenopausal women with endometriosis were treated with an LHRH agonist, a treatment that blocks the ovarian production of estrogen. The premature bone loss which would otherwise have occurred, was almost completely prevented when the LHRH agonist was given in combination with NETA (12).
Received January 16, 1990. Address requests for reprints to: Claus Christiansen, M.D., Department of Clinical Chemistry, Glostrup Hospital, DK-2600 Glostrup, Denmark. 836
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 September 2015. at 01:32 For personal use only. No other uses without permission. . All rights reserved.
ESTROGEN/PROGESTOGEN THERAPY ON ELDERLY WOMEN WITH OSTEOPOROSIS To avoid the monthly withdrawal bleeding accompanying sequential combined therapy, the continuous combined 17/3E2/NETA therapy has recently been introduced. The rationale is that the continuous administration of progestogen keeps the endometrium atrophic, and thereby prevents withdrawal bleeding and the development of endometrial hyperplasia or neoplasia (13, 14). Treatment of symptomatic postmenopausal osteoporosis still remains an unsolved problem. Many treatment regimens have been proposed, including calcitonin (15), fluoride (16), diphosphonates (17), anabolic steroids (18), 1-a-vitamin D3 (19), and PTH (20), but the cost effectiveness of these treatments are still debated. Whereas the 85-yr old woman with severe osteoporosis is generally acknowledged to be out of therapeutic reach, the treatment of the 65-yr old woman showing the first signs of osteoporosis would certainly be worthwhile, if this was possible. The aim of the present study was therefore to assess the continuous combined estradiol/NETA regimen in the treatment of symptomatic postmenopausal osteoporosis. Patients and Methods Subjects This study was part of a large double-blind clinical controlled trial carried out at Glostrup Hospital from March 1985 to December 1986. All participants were informed verbally and in writing about the trial, and all gave their informed consent, in accordance with Helsinki Declaration II. The participants were selected according to the following procedures: 1) All spinal x-rays (4900) of women aged 55-75 yr taken during the preceding 5 yr at Glostrup Hospital were retrieved. Those carrying an original radiological diagnosis indicating osteoporosis (osteoporosis, osteopenia, vertebral deformity) were reassessed. Of these, 219 women were found to have had at least one vertebral fracture (wedge or compression). Eightyone were excluded because of malignant, gastrointestinal, metabolic, or rheumatic disease, or drug intake known to affect calcium metabolism. Former, transitory use of hormonal substitution was not an exclusion criterion. Of the remaining 138 patients, 44 agreed to undergo experimental therapy for osteoporosis. After the introductory investigation five patients dropped out. 2) All patients referred to Glostrup Hospital because of distal forearm fracture in the same period were approached and the same criteria were applied. Of 316 women 133 were excluded. Eighty of the remaining 183 agreed to participate. After an initial examination the women were randomly allocated to one of three main treatment groups: 1) estrogen/gestogen; 2) anabolic steroids (21); 3) nasal calcitonin (22). Randomization was performed as follows: the medication was coded outside the clinic with numbers from 1 to 40 for each of the three groups, but taking care that the number allocated to the active treatment and that to placebo were the same for the first 20 and the last 20 numbers (blocked randomization). The first 20 patients with Colles' fracture (nos. 1-20) and the first 20 patients with
837
spinal fractures (nos. 21-40) were assigned to the estrogen/ gestogen (or placebo) group. The next 20 patients with Colles' fracture (nos. 1-20) and the last 19 patients with spinal fractures (nos. 21-39) were assigned to the anabolic steroid (or placebo group). The last 40 patients with Colles' fractures were assigned to the calcitonin (or placebo) group. We report here the data of the 40 women who blindly received either 2 mg 17j8E2 + 1 mg NETA + 500 mg calcium (Sandoz) (n = 20; 10 with spinal fractures, 10 with Colles; fractures) or a placebo combined with the same calcium supplementation (n = 20; 10 with spinal fractures, 10 with Colles' fractures). The treatment was given for 1 yr, during which all patients were examined every 3 months (five examinations in all). Seven of the women who received hormones were examined after a further year of treatment and were compared with 10 women from the same background population who received no therapy, but who were followed up in parallel. Compliance to trial Of the 40 women who entered the trial, 31 (78%) completed the 1 yr of treatment. Four women in the hormone group left (two with Colles' fracture, two with spinal fracture): one woman was killed in a car accident, one because she started treatment with digitalis for decompensated heart disease, one woman because of unacceptable breast tenderness, and one because of a discovered breast tumour (benign). Five women left the placebo group (two with Colles' fracture, three with spinal fracture): one woman died from unknown causes, one because of aggravation of heart disease, one because of general indisposition for some unknown reason, one because of aggravation of eczema, and one for personal reasons. This left 16 women in the hormone group and 15 women in the placebo group. Methods Bone mass measurements. BMC of the forearms was measured by single photon absorptiometry, with I125. The method determines BMC in a proximal (BMCprox) and a distal (BMCdist) region of the forearm with a trabecular bone content of 13% and 55%. The software for the equipment is developed in our laboratory and the long term in vivo precision of proximal and distal BMC is 1% and 1.5% (23). In the case of forearm fracture, we measured the nonfractured forearm; where bilateral fractures had occurred at different times, we measured the arm with the oldest fracture; and where bilateral fractures had occurred at the same time, we measured both forearms. Bone mineral content of the lumbar spine (BMCspine) was measured by dual photon absorptiometry on a Lunar Radiation Corporation DP3 scanner. BMCspine was calculated as the BMC values obtained in vertebrae L2, L3, and L4. The long term in vivo precision is 4%, after correcting the values for the errors due to source change (23). Bone mineral content of the total skeleton (TBBM) was also measured by dual photon absorptiometry on a whole body scanner developed in our laboratory. The long-term in vivo precision of TBBM is about 2.5% (24), after correcting for source change. Biochemical estimates of calcium metabolism. Blood samples were taken and urine collected in the morning after an over-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 September 2015. at 01:32 For personal use only. No other uses without permission. . All rights reserved.
CHRISTIANSEN AND RIIS
838
night fast and tobacco abstinence. Serum alkaline phosphatase (AP) was measured enzymatically according to Scandinavian recommendations, and plasma bone Gla protein (BGP) by RIAs (25). Fasting urinary calcium and creatinine were measured on an SMA 6/60 auto-analyzer and fasting urinary hydroxyproline by a spectrophotometric method. Calcium and hydroxyproline excretions were corrected for creatinine excretion (FU Ca/Cr and FU Hpr/Cr). Radiopharmaceutical measurement of bone turnover. Whole body retention (WBR) of 99Technetium-diphosphonate (99mTc-DP) was measured with a wide-field 7-camera with diverging parallel (fishtail) collimator. Two microcuries of 99mTc-DP were injected iv and the whole body measurement was started 5 min later. Scanning was repeated at 24 h and WBR was calculated after subtracting background radiation and correcting for radioactive decay by using the mean count at 5 min as 100% and comparing this to the mean count at 24 h. WBR is expressed in percent of the retained radioactivity. The method has a precision of 4.8% (26). Serum lipids and lipoproteins. High density lipoprotein cholesterol (HDL-C) was separated from chylomicrons, and low density lipoprotein cholesterol (LDL-C) and very low density lipoprotein (VLDL-C) by the mg++/dextran sulfate precipitation technique (27). Total serum cholesterol and HDL-C were determined by an enzymatic technique with the use of an LKB Ultralab system. The intra- and interassay variations were 1% and 2% (total cholesterol) and 2% and 5% (HDL-C), respectively. Serum triglycerides were determined by an enzymatic technique with the use of an LKB 8600 reaction rate analyser. The intra- and interassay variations of this method are 5% and 8%. LDL-C was then estimated according to the procedure of Friedewald et al. (28). The intra- and interassay variations are 5% and 10%.
JCE & M • 1990 Vol71«No4
TABLE 1. Clinical data (mean ± 1 SD) Hormones (n = 16) Age (yr) Menopausal age (yr) Ht (cm) Wt (kg) No. of patients with spinal fractures (crush/wedge)
Significance
Placebo (n = 15)
-c 01
63.6 ± 5.0 16 ± 6
difference NS NS
64.8 ± 5.7 18 ± 7
159.7 ± 9.4 59.9 ± 8.3
NS NS
159.3 ± 6.1 65.2 ± 9.9
12
12
TABLE 2. Absolute values of bone mass before and after 1 yr of treatment with hormones or placebo (mean ± 1 SD) Hormones Before therapy BMCprox (units) 27.4 ± 5.1 BMCdist (units) 24.7 ± 4.3 BMC8pine (g) 35.4 ±8.4 TBBM (g) 2534 ± 558 0 6
Placebo Before therapy
After therapy 27.9 ± 5.0° 27.6 ± 6.26 37.0 ± 8.3° 2655 ± 590°
27.4 26.6 33.4 2586
± ± ± ±
6.2 5.6 6.4 496
After therapy 27.2 ± 6.1 26.7 ± 5.8 32.4 ± 6.8 2588 ± 469
P < 0.05. P < 0.01 (paired t test). Mainly trabecular bone
Mainly cortical bone
Uterine Bleeding. Bleeding was recorded on a day-to-day calendar supplied at each examination. -10-
Safety parameters
J
Serum aspartate aminotransferase (s-ASAT), serum creatinine (s-Crea), hemoglobin (hb), and blood pressure (BP) were measured by routine procedures.
-20
Statistical analysis
and the group with vertebral fractures the groups were pooled. Tables 1 and 2 give the clinical data and initial values of bone mineral measurements. By chance, the hormone group weighed about 5 kg less than the placebo group, but, the difference was not statistically significant. The two groups were otherwise well matched. Figure 1 shows the individual changes in the bone mineral content occurring over the 1 yr of treatment (percent values). At all four measurement sites, the mean bone mineral content was increased in the hormone group as compared with the placebo group (P < 0.05). BMCdist and BMCspine showed a difference between the hormone and the placebo groups of about 8%, whereas the differences in BMCprox and TBBM were
The required sample size was estimated to be about 40 participants, as calculated on the precision of BMCspine (4%) and an estimated difference of interest between the active treatment and placebo groups of 4%. Clinical data and initial values of the measured variables were compared by Student's t test for unpaired data (Table 1). All initial values were set at 100% and subsequent values were expressed in per cent of that value. Differences within groups were evaluated by Student's t test for paired data and between groups by Student's t test for unpaired data.
Results Because the response to treatment was not significantly different between the group with Colles' fractures
BMCprox
TBBM
FIG. 1. Individual changes in bone mass measurements after 1 yr of treatment with hormones (O) or placebo (•).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 September 2015. at 01:32 For personal use only. No other uses without permission. . All rights reserved.
ESTROGEN/PROGESTOGEN THERAPY ON ELDERLY WOMEN WITH OSTEOPOROSIS about 3% and 5%, respectively. Table 2 gives the absolute changes in bone mass, which give virtually the same statistical significance. Figure 2 shows the mean percent changes in the four bone compartments over the 2 yr of treatment. The women receiving the active treatment had on the average an increased bone mass of 3.6-13.1%; out of the 28 individual changes in the hormone group only one was below zero, whereas 34 out of the 40 values in the control group were negative. Figure 3 visualizes the changes in the estimates of bone formation and bone resorption. Compared with the placebo group, the hormone group had significant decreases in pBGP, sAP, FU Hpr/Cr, FU Ca/Cr and sphosphate. WBR was also decreased but only to borderline significance (P = 0.06). None of the biochemical variables changed significantly during the second year of treatment. Table 3 shows the initial values of serum lipids and lipoproteins and the cumulated changes during treatment. The initial values were similar in the two groups. The changes in the hormone group showed significant decreases in all serum lipids (P < 0.05-0.001). All values were virtually stable during the second year. Table 4 shows the initial values of and the percent changes in the safety parameters. The hormone group had a mean systolic blood pressure significantly lower than that of the placebo group (P < 0.05). Neither group had significant changes in any of the variables, except for serum calcium (P < 0.001) and the diastolic blood pressure (P < 0.05), both of which were significantly decreased in the hormone group. Six women in the hormone group experienced uterine bleeding; the bleedings were slight (not requiring pads), and lasted at most for 2 days. In four women this occurred during the first 3 months of treatment, in one during the fourth, and in one during the third, fourth, and sixth months of treatment. One woman in the placebo group BMCprox 18
«
100!
90r
TBBM
i
i
i
i
BMCspine
120110100-
12
24
0
6
Bone resorption FUHpr/Cr
months
FIG. 3. Mean changes in biochemical estimates of bone formation and resorption over 1 yr of treatment with hormones (O) or placebo (•).
had slight bleeding on three occasions during the last six treatment months and this was due to senile colpitis. The four women with bleeding in the first 3 months were given a gynecological examination, but curettage was not indicated. The two women with bleeding after this time had a curettage. All gynecological and histological examinations were normal. The woman in the placebo group who had bleeding was given a gynecological examination but not curettage.
Discussion
BMCdist
11018
839
12
24
months
FIG. 2. Mean changes in bone mass measurements over 2 yr of treatment with (O) and without hormones (•).
The women in the present study were selected on the original radiological diagnosis. The result was a relatively low number of women with vertebral deformities, probably because many radiologists are not aware of the importance of osteoporosis or vertebral deformities. It does not reflect the true prevalence. In an earlier study of 70-yr-old women (29), we found a much higher prevalence, similar to that reported by others (30). A large number of clinical studies have been performed in recent years in order to find a therapeutic agent able to increase bone mass in osteoporotic subjects. The study populations have often comprised relatively oldpostmenopausal women with spinal crush fractures and thus a severe degree of osteoporosis. The participants in the present study were selected on their moderate degree of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 September 2015. at 01:32 For personal use only. No other uses without permission. . All rights reserved.
840
CHRISTIANSEN AND RIIS
JCE&M-I99O Vol 71 • No 4
TABLE 3. Serum lipids and lipoproteins before and after 1 yr of treatment with hormones or placebo (mean ± 1 SD) Hormones
S-triglycerides mmol/L Total s-cholesterol mmol/L S LDL mmol/L S HDL mmol/L
Placebo
Before therapy (absolute values)
After therapy (% of initial values)
1.16 ± 0.38 6.82 ± 1.30 4.87 ± 1.25 1.45 ± 0.27
88.0 ± 14.7° 88.0 ± 9.0" 88.9 ± 11.3* 92.2 ± 14.3C
Before therapy (absolute values) 1.36 ± 7.86 ± 5.70 ± 1.55 ±
0.75 2.16 2.12 0.47
After therapy (% of initial values) 103.7 ± 27.1 99.3 ± 9.6 96.7 ± 11.3 110.7 ± 13.0c
a
P
