| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Department of Medical Sciences, University Hospital, S-751 85 Uppsala, Sweden
Address all correspondence and requests for reprints to: Anna G. Johansson, M.D., Ph.D., Department of Medical Sciences, University Hospital, S-751 85 Uppsala, Sweden. E-mail: anna.johansson{at}medicin.uu.se
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
GH treatment in adults is accompanied by increased bone turnover, as reflected in the serum and urine by an elevation of biochemical markers for bone metabolism (11, 12, 13). After long term treatment, an accretion of bone mineral has been demonstrated, although the reported effects are not entirely congruent (14, 15, 16, 17, 18, 19, 20). A duration of treatment with GH of at least 18 months seems to be necessary before any increase in bone mineral content (BMC) or bone mineral density (BMD) can be detected. Recent data have shown that men with GHD are more responsive to recombinant human GH (rhGH) treatment than women with respect to changes in body composition and lipid metabolism (21), but few studies have addressed the possibility of gender as a prognostic factor for the effects on bone mass. One recent study has indicated that in terms of gain in bone density during GH replacement, women are more favored (19).
In the present study we have assessed the effects of long term treatment with rhGH on bone metabolism and bone mineral density in men and women with GHD. The male and female subgroups were similar with respect to the duration, severity, and age at onset of the disease.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A total of 36 patients (15 women and 21 men) participated in the
placebo-controlled part of the study. Baseline characteristics are
given in Table 1
. GH secretion was
evaluated as previously reported (23), and no patient had a GH response
greater than 3 µg/L during insulin-induced hypoglycemia (blood
glucose, 2.2 mmol/L). Men and women differed with regard to baseline
serum levels of IGF-I (Table 1), as previously reported (21). All
patients had at least one other pituitary deficiency, and in all
patients but two hypopituitarism was acquired in adulthood. Eight of
the women were receiving estrogen replacement therapy (oral
administration, n = 4; transdermal administration, n = 4),
and all men were receiving testosterone. The women who were receiving
estrogen replacement were younger than those who did not [median age:
estrogen replacement, 43 yr (range, 37–52); no estrogen replacement,
52 yr (range, 45–57); P = 0.01]. Other deficiencies
were being adequately replaced with levothyroxine, adrenal steroids,
and (nine cases) desmopressin. All replacement treatment had been
stable for at least 6 months before inclusion and was not changed
during the study. None of the patients had previously received GH.
|
An outline of the study protocol is shown in Fig. 1. The
patients were randomized to treatment with GH or placebo for 9 months,
and after 3 months of wash-out the other treatment was given for an
additional 9 months. Thereafter, 33 patients chose to continue in an
open study with GH treatment for up to a total of 45 months,
i.e. 36 months in addition to the initial 9 months of the
placebo-controlled study (Fig. 1). The mean dose of rhGH (Norditropin,
Novo Nordisk A/S, Copenhagen, Denmark) was 1.25
U/m2 at the end of the placebo-controlled part of the
study, after which the dose was adjusted according to the patient’s
report of side-effects and to the serum concentration of IGF-I to
maintain this within the normal reference range for the relevant age
group. Injections of rhGH or placebo were given sc at bedtime by the
patients themselves.
|
). The study design was approved by the Swedish Medical Products Agency and the ethical committee of the Faculty of Medicine, University of Uppsala (Uppsala, Sweden). The patients gave their informed consent to participate in the study and were told of their right to discontinue treatment at any time.
Bone mineral measurements
BMD and BMC were determined by dual energy x-ray absorptiometry (DXA; DPX-L equipment from Lunar Corp., Madison, WI) of the total body, the lumbar spine (vertebrae L2–L4; antero-posterior projection), and the femoral neck. The area of the femoral neck within a box of 12 x 50 mm and the area of the L2–L4 segment were also determined. The same box size was used for all neck analyses, and the same height of the L2–L4 vertebral segment was used for all spine evaluations via the compare scan mode. All analyses were performed by the same investigator (A.G.J.) at the end of the study. The coefficient of variation for DXA was less than 1%. The reference values provided by the manufacturer were used to calculate t-scores, which express individual BMD values as SD scores in relation to the normal mean and variation in BMD values in healthy young men or women (20–45 yr old in the femur reference group and 20–40 yr old in the antero-posterior spine reference group), and age-, sex-, and weight-adjusted z-scores.
At baseline, the t-score for total body BMD was for all
patients -0.47 ± 1.02 (range, -2.74 to 1.34); for the men it
was -0.62 ± 0.94, and for the women it was -0.24 ± 1.12
(P = 0.33 for difference between genders). The z-score
at baseline for total body BMD was -0.39 ± 0.93 (range, -2.79
to 1.42) for the whole patient group, -0.36 ± 0.89 for the men
and -0.44 ± 1.03 for the women (P = 0.83 for
difference between genders). After 15 months of GH treatment, when the
first DXA measurements of the hip and the spine were performed (Fig. 1), the t-score for the
femoral neck was -0.75 ± 1.00 (-2.58 to 1.03) for all patients,
-0.83 ± 1.17 for the male patients, and -0.62 ± 0.67 for
the female patients (P = 0.64 for difference between
genders). The z-score for the femoral neck was -0.35 ± 1.01
(range, -2.56 to 1.19) for all subjects, -0.44 ± 1.15 for the
men, and -0.18 ± 0.74 for the women (P = 0.57
for difference between genders). The t-score for the lumbar
spine (L2-L4) was -1.0 ± 1.20 (range, -3.16 to 1.99) for all
patients, -0.74 ± 1.15 for the men, and -1.46 ± 1.22 for
the women (P = 0.18 for difference between genders).
The z-score for the lumbar spine (L2–L4) was -0.60 ± 1.18
(range, -2.18 to 2.31) for all patients, -0.35 ± 1.13 for the
men, and -1.04 ± 1.20 for the women (P = 0.19
for difference between genders).
Biochemical analyses
Osteocalcin concentrations were determined in fasting morning serum by RIA with a commercial kit (CIS-Bio International, Oris Industries, Gif-Sur-Yvette, France). The intra- and interassay variations were less than 7%, and the reference range was 5–15 µg/L. Serum IGF-I was measured with a commercial RIA (Nichols Institute Diagnostics, San Juan Capistrano, CA) after extraction of binding proteins with acid-ethanol. The reference ranges in relevant age groups were 150–450 µg/L (20–40 yr) and 100–340 µg/L (41–60 yr). Plasma GH was assayed with a polyclonal RIA, with the lowest level of detection at 0.3 µg/L and intra- and interassay variations below 8%.
Statistics
Values are expressed as the mean ± SD in tables and as the mean ± SEM in figures. Student’s paired t test was used to test for changes within groups, and the unpaired t test was used for gender comparisons at 33 months. To limit the number of comparisons, paired comparisons with the baseline measurement were performed with the last time point of the treatment periods in the placebo-controlled study and with the last two time points of the follow-up study (study months 33 and 45).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the patients who received rhGH during the first 9 months, total
body BMD was significantly decreased at the end of this period and then
returned to baseline during the following placebo period (Fig. 2). In the patients who received placebo
first, total body BMD did not change until they received rhGH
treatment. During the following open treatment with rhGH up to a total
of 45 months (n = 16), total body BMD in the whole group of
patients showed no significant change compared to baseline.
|
). There was also a tendency for an increase in total body BMC
after 15 months in the group that received rhGH during the second
9-month period (P = 0.06). After a total of 45 months
of treatment with rhGH, the mean increase in total body BMC in the
whole group of patients was 5.1 ± 3.9% compared to baseline
(P = 0.0001).
Femoral neck BMD was increased in the whole group of patients by 6.7%
from 15–45 months of treatment with rhGH (Table 2). BMC and the area of the femoral neck
also increased during this period of treatment. In the spine, BMD was
significantly increased, and concomitant increases in BMC and the area
of the L2–L4 vertebrae were observed after 33 and 45 months of
treatment with rhGH (Table 2) compared to values at 15 months.
|
Gender differences in response to GH treatment
On completion of the placebo-controlled period, the dose of GH
adjusted for body surface area was similar in men and women (21) (Fig. 3). At this point the increase in serum
IGF-I was greater in men than in women (308 ± 133 vs.
206 ± 91 µg/L; P = 0.02). During the following
open study phase the dose was adjusted as described, and by this
procedure the doses were lowered, particularly in the male patients.
This resulted in a dose approximately 2 times higher in the women than
in the men, with total daily doses of 1.9 ± 1.1 and 1.0 ±
0.6 U, respectively, after a total of 33 months of GH treatment (Fig. 3). Despite this, the increase in serum IGF-I compared to the baseline
level was similar in the men and women (Fig. 3). The doses given to
women receiving and not receiving estrogen did not differ at any time
point during the study. Serum IGF-I at baseline was 62 ± 39
µg/L in women receiving estrogen and 60 ± 29 µg/L in women
who did not. The increase in serum IGF-I during treatment was similar
in the two groups of women.
|
). As previously reported, the
increments in the other serum markers of bone formation [bone alkaline
phosphatase and procollagen type I C-peptide (PICP)] and
bone resorption [telopeptide of collagen type I (ICTP)] also
differed between men and women given the same dose during the blinded
study phase (21), but there was no difference between the male and
female patients during the open study phase (data not shown).
The changes in BMD and BMC were analyzed with respect to gender after
33 months of GH treatment, i.e. 24 months after individual
dose adjustments based on IGF-I and/or side-effects. It was found that
the total body BMC, femoral neck BMD and BMC, and spine BMC had
increased significantly in the men but not in the women (Fig. 4).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
There is both clinical and experimental support for an interaction between sex hormones and GH at the peripheral level. GH replacement therapy in young men with childhood-onset GHD has been observed to stimulate the growth of androgen-dependent body hair without a concomitant change in circulating levels of free testosterone (22). This report corroborates an early finding that the dose of androgen required to induce secondary sex characteristics in sexually immature boys was significantly higher in GH-deficient than in GH-sufficient individuals (23). These observations suggest that androgens may potentiate the effects of GH. In contrast, administration of estrogens to postmenopausal women decreases the serum concentrations of IGF-I (24, 25), and high doses of estrogens were previously used to alleviate symptoms and metabolic effects in acromegalic patients (26). The similar serum concentrations of IGF-I in healthy men and women despite a 2- to 3-fold higher secretion of GH in women (27, 28, 29) also suggest an antagonism between GH and estrogen at the target site.
In castrated rabbits the expression of GH receptor messenger ribonucleic acid in both the liver and the growth plate has been reported to be increased by testosterone and decreased by estradiol (30). This could be a mechanism responsible for the gender difference in the serum IGF-I response to GH treatment and could be of relevance for the present finding of a greater gain in bone mass in the GHD men. On the other hand, in rats estradiol has been found to stimulate GH receptor expression (31), and testosterone did not influence GH-induced IGF-I expression (32). Osteoblasts contain receptors for both androgens and estrogens (33), and in bone cells, estradiol stimulated GH receptor expression (34). The discrepant results could be explained by tissue or species differences.
In several previous investigations of the effects of rhGH on bone density the results have varied markedly. In general, the studies in which more substantial increases in BMD have been found have included a relatively larger proportion of men than women (14, 15, 16, 18, 20, 35, 36, 37). The report of a significant difference in the change in total body BMD in women than in men with rhGH for 2 yr (19) remains to be explained, in particular as total body BMD was not significantly altered from baseline, and the difference seems to be due to on an initial reduction in total body BMD in the male patients. Furthermore, no gender difference was found when assessments were made in the hip and spine, regions that, in general, respond more to treatment with GH than total bone (19, 20).
When comparing the effects of GH on bone in men and women, other factors have to be accounted for, such as the severity of osteopenia, concomitant medication or other diseases that might modulate bone density, and also differences in hydrocortisone and thyroid replacement therapy. Although men and women were not stratified for possible confounding factors, there were no apparent differences between the two groups, and the number of women receiving estrogen as pills or patches were similar. The z and t-scores for BMD were not statistically different between men and women. However, contrary to another report (19), we did not find a relation between the severity of osteopenia per se, and bone density changes during treatment. Two of the men had onset of GHD during adolescence, but their bone densities were not more pronounced than those in the other patients; furthermore, one of them did not participate in the open study phase. In women, the route of estrogen administration (oral vs. transdermal) may also influence the GH/IGF-I axis (24).
Similar to Johannsson et al. (19), we did not detect a difference in bone acquisition or in markers of bone metabolism between women with and without estrogen replacement (data not shown). As the women had low androgen levels due to lack of stimulation of the adrenals, this might suggest that the combined effect of GH and testosterone is more important than an interaction between GH and estrogens. However, an evaluation of the impact of female sex steroids was hampered by the small number of women in each group and the fact that the women not receiving estrogen were older.
In a study of adult GHD men in whom bone histology was investigated after 12 months of treatment with rhGH, both the number of osteoclasts and the cortical thickness were increased, but trabecular bone volume did not change after the treatment (38). This is compatible with the finding in experimental animals that GH increases cortical thickness and has the potential to increase bone size (39, 40), possibly by enhancing periosteal bone formation. Indeed, an increased projected bone area was observed during treatment with rhGH in the present study in both the spine and hip. Such an increase in bone size may be expected to synergize with increased bone density with regard to bone strength.
In the present study, we found a reduction in BMD during the initial 9-month period of the treatment. Similar findings have previously been reported by others (12, 15, 17, 18, 36, 41, 42, 43, 44). The popular theory that this is due to an expansion of the remodeling space as a result of an increased activation frequency is corroborated by the histomorphometric findings of Bravenboer et al. (38). The complete restoration of total body BMD in the present study during wash-out and placebo treatment is compatible with that theory. If the rhGH treatment primarily increases periosteal bone formation, this could contribute to a reduction in areal bone density, as newly formed bone has a lower mineral density than older bone (45), but still increase the projected bone area.
In conclusion, the present study indicates a progressive gain in BMD and BMC during treatment with rhGH for as long as 45 months. A gender difference in the sensitivity to GH is demonstrated by similar increases in serum levels of IGF-I and bone biomarkers in the men and women despite 2-fold higher doses in the women and also by a more marked gain in bone mass in the men during long term treatment. To further clarify the impact of sex hormones on the effects of GH on bone, a controlled study including large numbers of GHD women with and without estrogen replacement therapy is awaited.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received July 10, 1998.
Revised December 2, 1998.
Revised February 24, 1999.
Accepted March 1, 1999.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
