| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Clinical Research Centers |
Neuroendocrine Unit (S.G., C.C., T.S., L.K., A.K.), Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts 02114
Address all correspondence and requests for reprints to: Steven Grinspoon, M.D., Neuroendocrine Unit, Bulfinch 457B, Massachusetts General Hospital, Boston, Massachusetts 02114.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In this study we investigated the relationship between gonadal function and the GH-IGF-I axis cross-sectionally in eugonadal and hypogonadal HIV-infected men and, prospectively, in hypogonadal HIV-infected men before and after testosterone administration. Our data demonstrate increased basal GH secretion and pulse frequency in association with reduced IGF-I levels, consistent with GH resistance among both hypogonadal and eugonadal men with AIDS wasting. Testosterone administration decreases GH among hypogonadal men with AIDS wasting, in whom the decrease in GH is a direct inverse function of the gain in lean body mass. These data demonstrate that testosterone administration has a significant effect on the GH axis among men with the AWS. Decreased GH resulting from improvement in body composition, a marker of nutritional status, predominates over direct stimulatory effects of testosterone on GH secretion in this population.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Fifty-one HIV-positive subjects [age 42 ± 1 (mean ± SEM) yr] with wasting [weight <90% ideal body weight (IBW) or >10% involuntary weight loss from preillness weight] and decreased screening free testosterone level (<12.0 pg/mL; normal range of age group, 12.0–35.0 pg/mL) were enrolled in a randomized, placebo-controlled study of testosterone administration (7). Inclusion was not limited based on CD4 count. Subjects with significant diarrhea (more than six stools per day); hemoglobin below 8 g/dL; platelet count below 50,000 cells/mm3; creatinine above 2 mg/dL; new opportunistic infection within 6 weeks of screening; prior usage of testosterone, anabolic steroids, GH, ketoconazole, or systemic steroid therapy within 3 months of screening; or history of prostate malignancy were excluded. In addition, patients receiving antiretroviral agents, including protease inhibitors, were required to be on a stable regimen for at least 6 weeks before study entry. Ten of the subjects were receiving chronic megestrol acetate therapy.
Ten eugonadal HIV-infected men (age, 41 ± 3 yr; free testosterone, >12.0 pg/mL) with wasting (defined similarly as for the hypogonadal subjects) were sequentially recruited over the same time period for comparison of GH and IGF-I levels. None of the eugonadal subjects was receiving megestrol acetate therapy. In addition, data were compared with those obtained previously in 15 age-matched healthy male control subjects (age, 46 ± 2 yr) without known hormonal dysfunction and normal total testosterone levels (473 ± 46 ng/dL) on laboratory testing. All subjects gave written consent as approved by the human studies committee of the Massachusetts General Hospital.
Study design
The screening eligibility assessment included weight, testosterone levels, physical examination, and medication history. Eugonadal HIV-positive subjects with wasting were studied within 2 weeks of screening during a 3-day in-patient admission inpatient visit to the General Clinical Research Center at the Massachusetts General Hospital on an isocaloric, meat-free diet (see below) for assessment of body composition, nutritional status, and hormonal function, including overnight frequent sampling for GH. Hypogonadal subjects with wasting were studied during a similar 3-day in-patient baseline admission and were then randomized to receive either testosterone enanthate (300 mg; Bio-Technology General Corp., Iselin, NJ) or placebo im every 3 weeks by self-injection (7). Subjects were stratified for weight less than or greater than 90% of IBW and megestrol acetate use before randomization. Randomization was blinded to both patients and investigators. No subjects experienced the onset of a new opportunistic infection, other complication, or significant weight change between the screening and baseline visits. Patients unable to self-administer study drug received injections every 3 weeks by the nursing staff. Subjects in the longitudinal study returned at 6 months for assessment of weight, hormone levels, and body composition identical to that obtained at baseline. The effects of testosterone on body composition, nutritional status, and immunological parameters as well as partial GH data on a subset of the subjects were previously reported (6, 7). The follow-up study visit was timed to correspond to the midpoint between study drug injections. Study drug compliance was confirmed by history, medication diaries, out-patient injection records, empty vial counts, and serum testosterone levels. Healthy control subjects underwent GH frequent sampling and similar nutritional and body composition assessments.
Clinical end points
Hormonal assessment. The GH-IGF-I axis was assessed by frequent GH sampling performed every 20 min from 2000–0800 h to determine mean overnight GH levels and GH pulse frequency, amplitude, and interpulse interval in the HIV-infected patients. Frequent sampling was repeated at 6 months for the hypogonadal men enrolled in the longitudinal study. Frequent sampling was performed every 10 min over 24 h in the healthy controls, but pulse analysis was performed on a comparable subset of the sampling obtained every 20 min from 2000–0800 h using the Pulsar computer program. The cut-off parameters for accepting peaks one, two, three, four, and five points wide were 3.63, 2.12, 1.43, 1.00, and 0.70 times the intraassay SD (13). The mean overnight GH level, GH pulse frequency, amplitude (peak height minus the calculated baseline), interpulse interval, and basal GH level were determined. Subjects were not allowed to eat after 1800 h on the day of sampling. Fasting serum IGF-I, IGF-binding protein-3 (IGFBP-3), insulin, and GH levels were determined in all subjects at 0800 h immediately before standard arginine stimulation testing (iv administration of 0.5 g/kg arginine hydrochloride; maximum dose, 30 g) with GH sampling at 30, 60, and 90 min. Serum total and free testosterone levels were also assessed at each visit in the longitudinal study. Twenty-four-hour urinary free cortisol levels were assessed at baseline in the subjects with AIDS wasting.
Nutritional assessment and body composition analysis. Weight was determined on the first day of each visit after an overnight fast. Prealbumin levels were determined at baseline in the subjects with AIDS wasting. Caloric intake and serum albumin levels were determined at each study visit. Fat-free mass was determined by dual energy x-ray absorptiometry in all subjects. Determination was made at the baseline and 6 month visits for the hypogonadal men enrolled in the treatment study.
Methods
IGF-I and free IGF-I were assessed by a single fasting measurement in all subjects. Serum IGF-I was determined after acid-alcohol extraction using a RIA kit with an intraassay coefficient of variation of 2.4–3.0% (Corning Nichols Institute Diagnostics, San Juan Capistrano, CA). Free IGF-I was measured by radioimmunometric assay with intraassay coefficients of variation of 3.3–10.3% (Diagnostics Systems Laboratories, Inc., Webster, TX). GH was measured by two-site radioimmunometric assay with an intraassay coefficient of variation of 2.8–4.2% (Corning Nichols Institute Diagnostics). Insulin was assessed by RIA with a sensitivity of 2 µU/mL and an intraassay coefficient of variation of 1.6–4.4% (Linco Diagnostics, St. Louis, MO). IGFBP-3 was assessed by RIA with an intraassay coefficient of variation of 5.3–6.7% (Diagnostic System Laboratories). Serum prealbumin was determined by rate nephelometry (MetPath, Inc., Cambridge, MA). Urinary free cortisol was assessed by RIA (14). Serum total and free testosterone were measured using a RIA kit (Diagnostics Products, Los Angeles, CA) with intraassay coefficients of variation of 5–12% for total testosterone and 3.2–4.3% for free testosterone. CD4 cell counts and viral burden were determined using previously described methods (7).
The percent IBW was calculated based on standard height and weight tables (15). Fat and fat-free mass were determined by dual energy x-ray absorptiometry using a Hologic-2000 densitometer (Hologic, Inc., Waltham, MA). The technique has a precision error of 3% for fat and 1.5% for fat-free mass (16). Subjects were instructed on completion of a 4-day food record, which was analyzed for total calorie, fat, protein, and carbohydrate contents (version 8A/2.6, Minnesota Nutrition Data Systems, Minneapolis, MN) by the Clinical Research Center dietitian. Subjects received an isocaloric, meat-free, protein-substituted diet 3 days before and during the in-patient assessments at baseline and 6 months. A meal plan was designed for each patient to match total caloric, carbohydrate, protein, and fat intake reported from out-patient food records. Calorie intake was ad libitum in the control subjects.
Statistical analysis
Baseline comparisons between groups were performed using the Wilcoxon test. Correlation coefficients between the baseline variables were calculated by simple linear regression analysis. The treatment effect on the GH axis was determined by comparing the change between the treatment groups using Student’s t test. The primary end point was the change in the mean overnight GH level. Other parameters of GH secretion, such as pulse frequency, interpulse interval, and peak GH response to arginine, were analyzed to determine the mechanism of change in GH. The factors associated with baseline GH and the change in GH in response to testosterone were determined in separate stepwise regression analyses using forward selection with P < 0.05 to enter the model. Weight, testosterone, caloric intake, viral load, CD4, albumin, lean body mass, and fat mass were tested for inclusion in the model as the independent variables. Forty-one of the 51 hypogonadal patients enrolled completed the prospective study of testosterone administration. Seven patients died, and 2 patients dropped out of the study, equally distributed in the treatment groups (7). All available data from these patients were included in the analysis. Patients were also classified as to whether they were receiving protease inhibitors at baseline, started them during the study, or never received them. The proportions in these three groups was compared between treatment groups by Fisher’s exact test. Results are the mean ± SEM unless otherwise indicated.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The clinical characteristics of the study subjects are shown in
Table 1
. Age was not different among the
three study groups. Weight, indexes of nutritional status, CD4 count,
and viral burden were not different between the eugonadal and
hypogonadal HIV-positive subjects, but weight was significantly lower
in the HIV-positive subjects compared to that in the healthy controls.
Age and baseline weight, body composition, virological parameters,
hormonal parameters, GH, IGF-I, and GH pulse characteristics were not
different between the patients randomized to testosterone or placebo in
the longitudinal study (data not shown).
|
Mean overnight GH, IGF-I, and arginine-stimulated GH levels as
well as GH pulse frequency, amplitude, and interpulse interval were
similar between the eugonadal and hypogonadal HIV-positive groups
(Table 1
). Mean overnight GH levels were increased and IGF-I levels
were decreased in both groups compared to those in the control group of
healthy, age-matched patients. GH pulse frequency (Fig. 1) and basal GH levels were also
increased in the HIV-positive groups compared to the control
population. Peak GH stimulation in response to arginine was not
different between HIV-infected patients and normal controls.
|
Overnight GH levels were inversely correlated with IGF-I (r =
-0.42; P = 0.003; Fig. 2), percent IBW (r = -0.36;
P = 0.012), albumin (r = -0.37; P
= 0.012), and fat mass (r = -0.52; P = 0.0002) at
baseline before testosterone administration in men enrolled in the
prospective study. In contrast, IGF-I levels correlated positively with
free testosterone (r = 0.35; P = 0.011) and
caloric intake (r = 0.32; P = 0.023). Albumin
(P = 0.0028) and total testosterone (P
= 0.011) were the only significant variables in a stepwise regression
model and explained 49% of the variation in mean overnight GH levels,
with the final equation: mean GH (µg/L) = -1.82 x albumin
(g/dL) + 0.003 x total testosterone (µg/L) + 6.5.
|
Mean overnight GH levels decreased significantly in the
testosterone-treated patients compared to those in the placebo-treated
hypogonadal patients (Table 2 and Fig. 3). Pulse analysis demonstrated a trend
toward decreased GH pulse amplitude in the testosterone-treated
patients compared to the placebo-treated patients (P =
0.07) without a significant difference in the change in GH pulse
frequency or interpulse interval between the groups (Table 2). In
contrast, no differences in IGF-I, free IGF-I, or GH response to
arginine stimulation were observed in response to testosterone
administration. Fat-free mass increased significantly in the patients
randomized to testosterone vs. placebo (2.0 ± 0.3
vs. -0.6 ± 0.3 kg; P = 0.036) as
previously reported (7). The decrease in mean overnight GH was
inversely associated with the gain in fat-free mass in response to
testosterone administration (r = -0.49; P =
0.024; Fig. 4). In a stepwise regression
analysis, the only significant determinant of change in GH was change
in fat-free mass (P = 0.022), which explained 27% of
the variation in the change in GH, with the final equation: change GH
(µg/L) = -0.197 x fat-free mass (kg) - 0.53.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Multiple abnormalities in the GH-IGF-I axis have been demonstrated among subjects with AWS. IGF-I levels have been shown to decrease with stage of illness (17) and to correlate with both albumin and body cell mass as measured by total body potassium (18). In contrast, elevated GH levels (19) and a decreased IGF-I response to exogenous GH (12) suggest acquired resistance to GH, which correlates in part with caloric intake and other indexes of nutritional status. Furthermore, Frost et al. have shown a reduction in the acid-labile substance and IGF-I ternary complex that may result from enhanced IGFBP-3 proteolysis in HIV-infected children with failure to thrive (20).
In this study, mean overnight GH levels were increased and inversely correlated with decreased IGF-I, a pattern consistent with acquired GH resistance, among HIV-infected men with wasting. Pulse analysis of frequent GH sampling data demonstrates increased GH pulse frequency and basal GH secretion, rather than increased GH pulse amplitude, among these men. These data are in contrast to those reported by Heijligenberg et al., who demonstrated normal circadian characteristics of GH secretion and normal IGF-I in asymptomatic HIV-infected patients with stable weight (21). Therefore, the wasting syndrome, rather than HIV infection alone, affects the GH-IGF-I axis.
Our data suggest that undernutrition contributes significantly to the observed pattern of GH and IGF-I in men with AIDS wasting. GH levels were inversely correlated with weight, albumin, and fat mass and were most highly predicted by serum albumin in a stepwise regression model, but were not significantly correlated with lean body mass. Similarly, IGF-I was highly correlated with caloric intake. These data are consistent with the hypothesis that GH resistance is due to undernutrition. Nutritionally mediated GH resistance is a well established sequelae of caloric deprivation and/or selective micronutrient reduction as occurs with dietary protein restriction (22, 23, 24, 25). Down-regulation of the GH receptor as well as abnormal postreceptor signaling independent of GH binding are thought to contribute to decreased IGF-I and are reversible with nutritional repletion (26).
Androgen deficiency in addition to undernutrition may also contribute to changes in the GH-IGF-I axis among hypogonadal men with AWS. IGF-I and GH were both positively correlated with serum androgen levels. This association remained significant in a stepwise regression model controlling for nutritional factors, suggesting that the effect of androgen is independent of nutritional status. A potent effect of androgens to increase IGF-I levels has been shown in other populations, including men with isolated hypogonadotropic hypogonadism (8), prepubertal boys (27, 28), and normal men (29). IGF-I is increased in these populations as a result of increased GH secretion manifested in higher 24-h mean GH levels and GH pulse amplitude in response to testosterone (8, 9, 10, 27). Our cross-sectional data demonstrate a similar positive association among testosterone, GH, and IGF-I in hypogonadal men with AIDS wasting.
We hypothesized that testosterone administration might have stimulatory effects on GH and IGF-I secretion in our subjects, as observed in other populations of hypogonadal men. However, we observed a significant decrease in the mean overnight GH level in the testosterone-treated compared to the placebo-treated patients, potentially a result of decreased GH pulse amplitude rather than a change in pulse frequency. To further investigate the potential mechanism of decreased GH in response to testosterone, we performed a stepwise regression analysis that demonstrated that the change in lean body mass was a highly significant, inverse predictor of the change in GH. The model demonstrated that the greatest change in GH occurred in association with the largest increase in lean body mass, independent of weight or other nutritional and hormonal parameters. In contrast, neither IGF-I nor free IGF-I decreased in response to testosterone administration. Taken together, these data suggest that although mean GH is not significantly correlated with lean body mass before testosterone treatment, the change in lean body mass is nonetheless the strongest predictor of decreased GH in response to testosterone. It is therefore not known whether the change in lean body mass directly affects GH secretion or is a marker for the effects of improved nutritional status on the GH axis.
These data suggest that the primary effect of testosterone on the GH-IGF-I axis in hypogonadal men with AIDS wasting is to decrease GH as a result of increased lean body mass. Alternative explanations include a direct effect of testosterone to lower GH. This is unlikely given the preponderance of data in the literature and our own baseline data that suggest a significant positive correlation between GH and serum testosterone levels. Alternatively, testosterone might stimulate IGF-I directly, with decreased GH by feedback inhibition. The absence of any significant effect on either total or free IGF-I argues against this explanation. We do not believe our results to be a manifestation of an unrelated improvement in disease status in the testosterone vs. the placebo group. There were no differences in viral burden, CD4, or use of protease inhibitors between the groups in this randomized, double blind, placebo-controlled study (7). Our data suggest that increased lean body mass resulting from testosterone administration contributes to decreased GH secretion, independent of the observed changes in total or free IGF-I. Although our data suggest an effect of increased lean body mass on GH secretion, the mechanism of this effect and the relationship to changes in overall nutritional status remain unknown. Furthermore, it is not known whether a similar pattern of GH would result from androgen administration in eugonadal men with AWS.
These data are consistent with GH resistance among both hypogonadal and eugonadal men with AIDS wasting, characterized by increased basal GH secretion and pulse frequency in association with reduced IGF-I levels. Testosterone administration decreases GH among hypogonadal men with AIDS wasting, an opposite effect on GH secretion than that seen in other hypogonadal populations. This unique effect is probably related to the poor underlying nutritional status and partial restoration of lean body mass by testosterone administration. Further studies are necessary to better determine the mechanisms by which testosterone affects the GH axis in this population.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received July 6, 1998.
Revised August 19, 1998.
Accepted August 26, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Schurgin, S. Dolan, A. Perlstein, M. P. Sullivan, N. Aliabadi, and S. Grinspoon Effects of Testosterone Administration on Growth Hormone Pulse Dynamics in Human Immunodeficiency Virus-Infected Women J. Clin. Endocrinol. Metab., July 1, 2004; 89(7): 3290 - 3297. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Hadigan, C. Corcoran, T. Stanley, S. Piecuch, A. Klibanski, and S. Grinspoon Fasting Hyperinsulinemia in Human Immunodeficiency Virus-Infected Men: Relationship to Body Composition, Gonadal Function, and Protease Inhibitor Use J. Clin. Endocrinol. Metab., January 1, 2000; 85(1): 35 - 41. [Abstract] [Full Text]
|
|
|
|
 |
|
 C. Corcoran and S. Grinspoon Treatments for Wasting in Patients with the Acquired Immunodeficiency Syndrome N. Engl. J. Med., June 3, 1999; 340(22): 1740 - 1750. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
