Changes in bone mineral density and body composition during initial and long-term gonadotropin-releasing hormone agonist treatment for prostate carcinoma

Authors


Abstract

BACKGROUND

Initial treatment with a gonadotropin-releasing hormone (GnRH) agonist increases fat mass, decreases lean body mass, and decreases bone mineral density (BMD) in men with prostate carcinoma. To the authors' knowledge, little is known regarding either the long-term effects of treatment or predictors of treatment-related changes in BMD and body composition.

METHODS

The authors analyzed prospective 12-month data from 65 men during initial and long-term GnRH agonist treatment for prostate carcinoma. Relationships between baseline characteristics (age, treatment duration, body mass index, and baseline values for outcome of interest) and changes in lean mass, fat mass, and BMD were assessed using univariate and multivariate regression models.

RESULTS

Mean BMD of total hip decreased by 1.9 ± 2.7%, lean body mass decreased by 2.0 ± 3.3%, and fat mass increased by 6.6 ± 9.4% at 12 months (P < 0.001 for each comparison). Twenty-three men (35%) had received treatment with a GnRH agonist before study entry, and the mean (± standard deviation) duration of previous treatment was 35 ± 41 months. Longer duration of previous GnRH agonist treatment was found to independently predict less fat accumulation and less loss of lean body mass in multivariate models. In contrast, the duration of GnRH agonist treatment did not significantly predict changes in BMD. Other covariates did not appear to predict changes in body composition or BMD consistently.

CONCLUSIONS

The results of the current study showed that fat mass increased and lean body mass decreased mostly during initial GnRH agonist therapy whereas BMD decreased steadily during initial and long-term treatment. Cancer 2005. © 2005 American Cancer Society.

Gonadotropin-releasing hormone (GnRH) agonists decrease bone mineral density (BMD)1–4 and increase fracture risk5 in men with prostate carcinoma. GnRH agonists also increase fat mass and decrease lean body mass.6, 7 These treatment-related changes in body composition may contribute to fatigue, emotional distress, and decreased quality of life.8–10 Whereas the consequences of initial GnRH agonist on BMD and body composition are well characterized, less is known regarding the long-term, treatment-related, adverse effects. In addition, little is known concerning the predictors of treatment-related changes in BMD and body composition.

We analyzed prospective 12-month data from men during initial and long-term GnRH agonist treatment for nonmetastatic prostate carcinoma. The correlations between baseline characteristics and changes in lean body mass, fat mass, and BMD of the total hip were assessed using univariate and multivariate regression models.

MATERIALS AND METHODS

Subjects and Study Design

These analyses included subjects from the control groups from three randomized, controlled trials for treatment-related bone loss in men with nonmetastatic prostate carcinoma.2, 11, 12 All subjects in the analyses were either receiving treatment with a GnRH agonist for prostate carcinoma at the time of study entry or were initiating GnRH agonist treatment at study entry. Subjects had a radionuclide bone scan within 6 months of initiating GnRH agonist treatment or 6 months before study entry. Men with bone metastases or evidence of progressive disease (serum prostate-specific antigen level > 150% nadir value) were excluded. Men with metabolic bone disease, a history of treatment for osteoporosis, serum calcium level < 8.4 mg/dL or > 10.6 mg/dL, or serum creatinine concentration > 2.0 mg/dL (177 μmol/L) were also excluded.

All subjects continued treatment with a GnRH agonist throughout the study period. All subjects received calcium carbonate (500 mg daily) and a daily multivitamin containing 400 IU of vitamin D.

BMD and body composition were measured by dual-energy X-ray absorptiometry at baseline and 12 months. The institutional review board of Dana Farber Partners Cancer Care approved the study. Written informed consent was obtained from all subjects.

Study Endpoints

BMD of the proximal femur was determined by dual-energy X-ray absorptiometry using a Hologic QDR 4500A densitometer (Hologic, Inc., Waltham, MA). Lean and fat mass also were determined by dual-energy X-ray absorptiometry (software version 11.2) as described previously.7

Fasting subjects were weighed wearing a hospital gown and no shoes. Body weight was measured to the nearest 0.1 kg using a digital platform scale (Blue Bell BioMedical model 500; SR Instruments, Tonawanda, NY). Height was measured to the nearest 0.1 cm using a wall-mounted stadiometer. The mean of three height measurements was recorded.

Statistical Analyses

The primary outcomes were the percentage change in the BMD of the total hip, lean body mass, and fat mass from baseline to 12 months. Values are reported as means ± standard deviation (SD) unless indicated otherwise. Changes were tested for significance using one-sample Student t tests.

Univariate regression models were fit for changes in BMD, lean mass, and fat mass using each covariate as a single explanatory variable. Next, multivariate models were fit using all explanatory variables to identify those that were independently predictive. Stepwise backward elimination was performed to determine the simplest multivariate model that contained only explanatory variables that were independently significant and predictive of changes in BMD. The covariates considered in the regression analyses were age, duration of previous GnRH agonist treatment, body mass index (BMI), and baseline value for BMD, lean mass, or fat mass.

Statistical analyses were performed using SAS version 8.1 (SAS Institute Inc., Cary, NC). All P values were 2 sided and Ps < 0.05 were considered to be statistically significant.

RESULTS

Characteristics of the Subjects

Sixty-five men were included in the analyses (Table 1). Their mean (± SD) age was 66 ± 10 years. Eighty-nine percent were white. Twenty-three men (35%) had received treatment with a GnRH agonist before study entry. The mean (± SD) duration of previous treatment was 35 ± 41 months. The remaining 42 subjects initiated GnRH agonist treatment at study entry. All subjects were treated with a GnRH agonist throughout the 12-month study period.

Table 1. Subject Characteristics (n = 65)
CharacteristicsValue
  • GnRH: gonadotropin-releasing hormone.

  • a

    Values represent the subset of men with previous gonadotropin-releasing hormone agonist treatment.

Mean age in years66 ± 10
Percent race 
 Non-Hispanic white89%
 Non-Hispanic black6%
 Other5%
Body mass index in (kg/m2)27.5 ± 3.4
Percent receiving previous GnRH agonist treatment35%
Duration of previous GnRH agonist treatment in mosa 
 Mean35 ± 41
 Median18
 Range3–138

Changes in BMD and Body Composition

The mean (± SD) BMD of the total hip decreased by 1.9 ± 2.7% at 12 months (P < 0.001) (Table 2). The mean lean body mass decreased by 2.0 ± 3.3% (P < 0.001). The mean fat mass increased by 6.6 ± 9.4% (P < 0.001).

Table 2. Changes in BMD and Body Compositiona
OutcomeBaselineMo 12Percent changeP value
  • BMD: bone mineral density.

  • a

    Values are shown as the means ± standard deviation.

BMD total hip0.995 ± 0.1360.977 ± 0.138−1.9 ± 2.7%<0.001
Percent lean mass69.7 ± 4.768.2 ± 4.8−2.0 ± 3.3%<0.001
Percentage fat mass27.1 ± 4.928.8 ± 5.06.6 ± 9.4%<0.001

Regression Analyses

In univariate models, a longer duration of GnRH agonist treatment was associated with less fat accumulation (parameter estimate of 0.106 for each month increase in the duration of previous treatment; P = 0.007) and less loss of lean body mass (parameter estimate of 0.036 for each month increase in the duration of previous treatment; P = 0.01) (Table 3). A longer duration of GnRH agonist treatment predicted less fat accumulation (parameter estimate of 0.079 for each month increase in the duration of previous treatment; P = 0.034) and less loss of lean body mass (parameter estimate of 0.030 for each month increase in duration of previous treatment; P = 0.035) on multivariate models. In contrast, duration of GnRH agonist treatment did not significantly predict changes in BMD in either univariate or multivariate analyses (Table 3).

Table 3. Regression Analysis of Percent Changes in BMD and Body Composition
OutcomesCovariatesUnadjusted regression coefficientP valueAdjusted regression coefficientP value
  1. BMD: bone mineral density; NS: not statistically significant at the 5%; significance level; GnRH: gonadotropin-releasing hormone.

Total hip BMDAge in years−0.051NS−0.054NS
 GnRH agonist treatment in mos−0.002NS0.004NS
 Baseline BMD in g/cm21.220NS0.575NS
 Body mass index in kg/m2−0.002NS−0.052NS
Percent lean massAge in years0.075NS0.038NS
 GnRH agonist treatment in mos0.0360.010.0300.035
 Percent baseline lean mass %−0.116NS0.177NS
 Body mass index in kg/m2−0.014NS−0.157NS
Percent fat massAge in years−0.2570.03−0.147NS
 GnRH agonist treatment in mos−0.1060.007−0.0790.034
 Percent baseline fat mass %−0.7040.003−0.9470.002
 Body mass index in kg/m2−0.265NS0.609NS

A higher fat mass at baseline was associated with less fat accumulation in both univariate (parameter estimate of 0.704 for each percent increase in fat mass; P = 0.003) and multivariate (parameter estimate of 0.947 for each percent increase in fat mass; P = 0.002) models (Table 3). In contrast, baseline BMD and lean body mass were not associated with changes in BMD or lean body mass. Age and BMI were not associated with changes in fat mass, lean body mass, or BMD.

Outcomes by Treatment Duration

Figure 1 depicts changes in BMD and body composition in subgroups of men with no previous GnRH agonist treatment at study entry (n = 42), ≤ 18 months of previous treatment (n = 12, median duration of previous GnRH agonist treatment of 7 mos), and > 18 months previous treatment (n = 11, median duration of previous GnRH agonist treatment of 43 mos). BMD of the hip decreased significantly in all subgroups. In contrast, lean body mass decreased significantly and fat mass increased in men with no previous treatment but did not change significantly in either subgroup of men with previous GnRH agonist therapy.

Figure 1.

Outcomes according to treatment duration. Shown are changes in bone mineral density of the total hip, lean mass, and fat mass in subgroups of men with no previous gonadotropin-releasing hormone (GnRH) agonist treatment at the time of study entry (n = 42), ≤ 18 months of previous treatment (n = 12, median treatment duration of 7 mos), and > 18 months of previous treatment (n = 11, median treatment duration of 43 mos). Values are shown as the means ± standard error. P values are shown for change from baseline to Month 12. NS: not significant.

DISCUSSION

We found that a longer duration of previous GnRH agonist treatment independently predicted less fat accumulation and less reduction in lean body mass during GnRH agonist therapy. In contrast, the duration of previous treatment did not predict changes in BMD of the hip. These findings suggest that fat mass increases and lean body mass decreases mainly during initial GnRH agonist therapy whereas BMD decreases steadily during initial and long-term treatment.

In early menopause, women experience a short (2–3-yr) period of accelerated bone loss. In late menopause, BMD decreases more slowly.13, 14 In large prospective studies of postmenopausal women with osteoporosis, for example, BMD of the hip was reported to decrease by 0–1% per year.15, 16 In our GnRH agonist-treated subjects with prostate carcinoma, BMD of the hip decreased at similar rates during initial and long-term GnRH agonist treatment. Consistent with this observation, the duration of GnRH agonist treatment was found to independently predict fracture risk in men with prostate carcinoma.5

Estrogens play an important role in skeletal homeostasis in normal men. In older men, low serum estradiol levels are associated with decreased BMD and increased fracture risk.17–20 Estrogens contribute to the regulation of both bone formation and bone resorption in men.21, 22 Serum estradiol levels are higher in elderly men than in postmenopausal women. In men, GnRH agonists decrease serum estradiol levels by > 80%11 to levels lower than those in postmenopausal women. A greater severity of estrogen deficiency may account for higher rates of bone loss in GnRH agonist-treated men than in late postmenopausal women. Aromatase inhibitors decrease BMD and increase fracture risk in postmenopausal women with breast carcinoma.23 Severe estrogen deficiency in GnRH agonist-treated men and in aromatase inhibitor-treated postmenopausal women may explain the persistent high rates of bone loss in these situations. Notably, serum estradiol levels do not predict changes in BMD in men during long-term GnRH agonist treatment,12 most likely because estradiol levels are uniformly very low.

The key outcomes in these analyses were chosen to represent the bone, muscle, and fat compartments. BMD of the total hip was selected to represent the bone compartment for several reasons. First, the hip is the most clinically important site of fracture. Second, the hip represents best both cortical and trabecular bone. Third, measurement of BMD of the total hip is more precise than measurement of hip subregions. Fourth, measurement of hip BMD is less prone to the confounding factors associated with age and comorbid disease than measurement of the spinal BMD.24 Larger studies are needed to evaluate other outcomes including changes in BMD at other skeletal sites.

The study evaluated the relation between several covariates including duration of GnRH agonist therapy at study entry and 12-month changes in BMD and body composition. Prospective studies with long-term follow-up are necessary to confirm the observed relations between duration of GnRH agonist treatment and subsequent changes in BMD and body composition. Additional larger studies also are needed to assess other covariates.

Conclusions

Fat mass increases and lean body mass decreases primarily during initial GnRH agonist therapy. In contrast, BMD decreases steadily during initial and long-term treatment.

Acknowledgements

The authors thank the dedicated staffs of Mallinckrodt General Clinical Research Center and Massachusetts General Hospital Bone Density Center.

Ancillary