| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Care |
Neuroendocrine Unit, Massachusetts General Hospital and Harvard Medical School (S.G., C.C., T.S., J.R., S.W.), 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. E-mail: sgrinspoon{at}partners.org
Abstract
Although prior studies suggest reduced androgen levels in women with acquired immune deficiency syndrome wasting, little is known regarding the regulation of adrenal and ovarian androgen secretion in such patients. We investigated ovarian and adrenal function in 13 human immunodeficiency virus-infected women with acquired immune deficiency syndrome wasting and 21 age- and body mass index-matched healthy control subjects studied in the early follicular phase. Subjects received hCG (5000 U, im) on d 1 and Cosyntropin (0.25 mg, iv) on d 3 after dexamethasone (1 mg, orally, at 2400 h) pretreatment on d 2. At baseline, human immunodeficiency virus-infected subjects demonstrated significantly reduced T [18 ± 2 vs. 25 ± 2 ng/dl (0.6 ± 0.1 vs. 0.9 ± 0.1 nmol/liter); P = 0.02], free T [1.5 ± 0.1 vs. 2.4 ± 0.2 pg/ml (5.3 ± 0.5 vs. 8.3 ± 0.6 pmol/liter); P = 0.001], androstenedione [119 ± 6 vs. 162 ± 14 ng/dl (4.16 ± 0.20 vs. 5.66 ± 0.48 nmol/liter); P = 0.02], and dehydroepiandrosterone sulfate [0.96 ± 0.17 vs. 1.55 ± 0.19 µg/ml (2.6 ± 0.5 vs. 4.2 ± 0.5 µmol/liter); P = 0.047] levels compared with the control subjects. T [8 ± 2 vs. 6 ± 2 ng/dl (0.3 ± 0.1 vs. 0.2 ± 0.1 nmol/liter); P = 0.48], free T [0.5 ± 0.2 vs. 0.4 ± 0.1 pg/ml (1.7 ± 0.7 vs. 1.5 ± 0.5 pmol/liter); P = 0.85], 17hydroxyprogesterone [0.5 ± 0.2 vs. 0.7 ± 0.2 µg/liter (1.6 ± 0.6 vs. 2.0 ± 0.6 nmol/liter); P = 0.63], and androstenedione [-1 ± 12 vs. 8 ± 11 ng/dl (-0.03 ± 0.42 vs. 0.28 ± 0.39 nmol/liter), P = 0.61] responses to hCG were not different between the groups. Cortisol responses were increased and dehydroepiandrosterone sulfate responses were decreased in the human immunodeficiency virus-infected vs. control subjects after ACTH stimulation. The ratio of DHEA to cortisol was significantly decreased at 60 (71 ± 11 vs. 107 ± 10; P = 0.02) and 90 (63 ± 8 vs. 102 ± 9; P = 0.004) min post-ACTH in the human immunodeficiency virus-infected patients compared with control subjects. Baseline urinary free cortisol levels were not different between the groups [36 ± 9 vs. 36 ± 5 µg/24 h (99 ± 26 vs. 100 ± 13 nmol/d)]. The DHEA to cortisol ratio correlated with the CD4 count (r = 0.67; P = 0.01). These data demonstrate significant shunting of adrenal steroid metabolism away from androgenic pathways and toward cortisol production in human immunodeficiency virus-infected women with the wasting syndrome. In contrast, our data suggest intact ovarian androgen responsivity to hCG stimulation. Further studies of the mechanism of adrenal steroid shunting and the efficacy of androgen replacement in human immunodeficiency virus-infected women are necessary.
REDUCED SERUM FREE T levels have been reported in women with acquired immune deficiency syndrome (AIDS) wasting and may contribute to reduced muscle mass and quality of life in this population (1, 2, 3, 4). A number of potential explanations exist to explain low serum free T levels in women with AIDS wasting. Illness per se (5, 6) and protein calorie malnutrition are known to affect both hypothalamic GnRH and pituitary gonadotropin secretion as well as responsiveness to GnRH (6, 7, 8, 9). Adrenal androgen secretion may also be affected in severe acute illness (10). Shunting of adrenal steroid biosynthesis toward cortisol has been reported in men with AIDS wasting (11, 12), but prior studies have investigated neither adrenal nor gonadal androgen metabolism in human immunodeficiency virus (HIV)-infected women. In this study we investigated both ovarian and adrenal androgen function in HIV-infected women with wasting.
Materials and Methods
Thirteen women with AIDS wasting (weight <90% of ideal
body weight and weight loss >10% from preillness baseline weight)
were investigated and compared with 21 healthy age- and body mass index
(BMI)-matched eumenorrheic HIV-negative control subjects. Subjects were
between 18–45 yr of age and were excluded if they were pregnant or
receiving E, progesterone, megestrol acetate, glucocorticoid,
mineralocorticoid, androgens, or any drugs known to affect the
hypothalamic-pituitary-gonadal or adrenal axes within 3 months of the
study. Subjects were also excluded if they were diagnosed with a new
opportunistic infection within 4 wk of the study, had begun a new
antiretroviral treatment within 6 wk of the study, had been previously
diagnosed with significant liver disease, or demonstrated a serum
creatinine greater than 2.0 mg/dl (180 µmol/liter) or a hemoglobin
level less than 8 g/dl. Control subjects had normal menstrual function
and no history of acute or chronic medical illnesses. Subjects were
admitted to the General Clinical Research Center of the Massachusetts
General Hospital for 3 consecutive days for the purposes of determining
adrenal and gonadal androgen secretory patterns. Testing was performed
fasting at 0800 h on consecutive days in the early follicular
phase within 7 d of the onset of menses. The testing sequence and
strategy were as follows (Fig. 1
).
|
At 0800 h, measurements were made of baseline T, free T, E2, SHBG, LH, FSH, androstenedione, 17-hydroxyprogesterone (17-OHP), cortisol, dehydroepiandrosterone sulfate (DHEAS), dehydroepiandrosterone (DHEA), and urinary hCG. hCG (5000 U, im) was administered. A collection of 24-h urine was made for urinary free cortisol and creatinine determinations.
Day 2
At 0800 h, repeat measurements of T, free T, androstenedione, and 17-OHP were made. Dexamethasone (1 mg) was given at 2400 h.
Day 3
At 0800 h, measurements of cortisol, DHEA, DHEAS, and aldosterone were made. Cosyntropin [ACTH-(1–25); 250 µg, iv) was given, and repeat measurements of hormone levels at 30, 60, and 90 min were made.
Caloric, protein and micronutrient intake were determined by the nutrition staff of the Massachusetts General Hospital General Clinical Research Center based on food intake questionnaires. Dual energy x-ray absortiometry was performed using a model 4500 scanner (Hologic, Inc., Waltham, MA). The precision of this technique in the determination of whole body fat and lean mass is 3% for fat mass and 1.5% for fat-free mass (13). Pretreatment with dexamethasone before ACTH testing was used to permit determination of the change in response to ACTH as evidence for steroid shunting. Written consent was obtained from all subjects, and the study was approved by the Subcommittee on Human Studies of the Massachusetts General Hospital. Subjects with a free T level less than 3.0 pg/ml (<10.4 pmol/liter) were subsequently enrolled into a treatment study of T after all testing had been completed.
Statistics
Baseline comparisons were made between the groups by t test. Stimulatory responses between the two groups (HIV-wasted and HIVnegative healthy females) were compared by t test. Univariate regression analyses were performed between biochemical parameters, CD4, and viral load. Univariate regression analysis was also used to compare baseline DHEAS and DHEAS AUC to ACTH. Analyses were repeated in a comparison of the eumenorrheic HIV-infected to healthy control subjects and also by protease inhibitor (PI), nucleoside reverse transcriptase inhibitor (NRTI), and non-nucleoside reverse transcriptase inhibitor (NNRTI) status. Results are the mean ± SEM unless otherwise indicated.
Assay methods
All samples from the same patient were run in duplicate in the same assay. The free T concentration was determined as the product of the percent free T, measured by equilibrium dialysis, and the total T concentration (Endocrine Sciences, Inc., Calabasas Hills, CA). The intraassay coefficient of variation of free T is 6.9%, and the intraassay coefficient of variation for total T is less than 8.1%. The intraassay coefficients of variation (CVs) were developed using pooled sera covering the range of the assay. The normal range for total T is 10–55 ng/dl (0.4–1.9 nmol/liter), and that for free T is 1.1–6.3 pg/ml (3.8–21.8 pmol/liter) in adult females. The interassay CV for T is 8–15%, and that for free T, 8.9–11.9%. The sensitivity of the total T assay is 3 ng/dl (0.1 nmol/liter). The sensitivity of the determination of percent free T by this method is 0.1%. RIAs were performed for cortisol (intraassay CV, 6.6–7.7%; interassay CV, 8.8–9.8%; DiaSorin, Inc., Stillwater, MN), DHEAS (intraassay CV, 3.8–5.3%; interassay CV, 6.3–11.0%; Diagnostic Products, Los Angeles, CA), DHEA (intraassay CV, 5.6–10.6%; interassay CV, 7.0–10.2%; Diagnostics Systems Laboratories, Inc.), aldosterone (intraassay CV, 3.6–8.3%; interassay CV, 7.3–10.4%; Diagnostics Systems Laboratories, Inc.), androstenedione (intraassay CV, 2.8–5.6%; interassay CV, 6.0–9.8%; Diagnostics Systems Laboratories, Inc.), 17-OHP (intraassay CV, 3.5–7.1%; interassay CV, 5.0–11.0%; Diagnostic Products), LH (intraassay CV, 2.6%; interassay CV, 4.5–5.4%; Nichols Institute Diagnostics, San Juan Capistrano, CA), FSH (intraassay CV, 1.6–2.3%; interassay CV, 3.2–3.8%; Nichols Institute Diagnostics), and E2 (intraassay CV, 7.5–12.2%; interassay CV, 6.5–8.9%; Diagnostics Systems Laboratories, Inc.). SHBG was performed by immunoradiometric assay with an intraassay CV of less than 4% and an interassay CV of 7.8–10.6% (Endocrine Sciences, Inc.). Urinary free cortisol was measured by standardized techniques (14).
CD4 counts were measured by flow cytometry using FACS lysing solution and a FACScan analyzer (Becton Dickinson and Co., San Jose, CA; normal range, 537-1571 CD4 cells/mm3). HIV-1 RNA was quantified using a sandwich nucleic acid hybridization procedure, the Quantiplex HIV-RNA assay (Chiron Corp., Emeryville, CA).
Results
The baseline clinical characteristics are shown in Table 1. Weight loss was, on the average,
17 ± 2% from the preillness baseline. Of the 13 HIV-infected
patients, 10 were eumenorrheic, 1 was oligomenorrheic, and 2 were
amenorrheic. FSH levels were not increased in any subject. None of the
HIV-infected subjects demonstrated any clinical characteristic of the
lipodystrophy syndrome, including evidence of excess truncal fat or
loss of peripheral or facial sc fat. Three HIV-infected subjects were
receiving therapy with a protease inhibitor, 7 were being treated with
a nucleoside reverse transcriptase inhibitor, and 5 were receiving
treatment with a nonnucleoside reverse transcriptase inhibitor. Five
subjects were not receiving antiretroviral treatment for HIV disease.
The women with AIDS wasting were matched for age and BMI with the
healthy control subjects and were not significantly different in
minority composition from the control subjects. Neither percent fat,
total body fat, or caloric intake differed between the groups. The mean
Karnofsky score was 91.5 ± 2.2 in the HIV-infected women.
|
). The difference in total T
between HIV-infected and control subjects remained significant when
controlling for the differences in SHBG between the groups
(P < 0.0001). E2 levels did not differ when
controlling for the differences in SHBG (P = 0.73) or
when comparing the free E2 index (15) (E2/SHBG, 0.26
± 0.06 vs. 0.30 ± 0.06; P = 0.60).
Reduced androgen levels were also seen in a subanalysis comparing
eumenorrheic HIV-infected subjects to control subjects [free T
1.6 ± 0.2 vs. 2.4 ± 0.2 pg/ml (5.4 ± 0.6
vs. 8.2 ± 0.6 pmol/liter); P =
0.007]. There were no significant differences in total and free T, E2,
17-OHP, or gonadotropin concentrations within the HIV-infected group
when analyzed by antiretroviral drug class (PI vs. no PI,
NRTI vs. no NRTI, NNRTI vs. no NNRTI, or any
antiretroviral agent vs. no antiretroviral agent; data not
shown).
In response to hCG, there were no significant differences in the
responses of T, free T, 17-OHP, or androstenedione (Table 2). Androgen responses to hCG did not
differ by PI, NRTI, or NNRTI status (data not shown).
|
]. Urinary free cortisol levels
were also not significantly different between the groups [36 ± 9
vs. 36 ± 5 µg/24 h (99 ± 26 vs.
100 ± 13 nmol/d); P = 0.98; Table 1].
|
and Fig. 2). The results of hCG and ACTH testing
were similar in a subanalysis including only HIV-infected patients with
normal menstrual function. Similar results were obtained comparing the
change from baseline (data not shown). Baseline DHEAS was highly
correlated with overall DHEAS response to ACTH in both HIV-infected
patients (r = 0.98; P < 0.0001; Fig. 3) and in control subjects (r =
0.98; P < 0.0001). The DHEA to cortisol
ratio was associated with the CD4 count at baseline (r = 0.67;
P = 0.01; Fig. 4), but
not with BMI (r = -0.23; P = 0.44), weight loss
(r = -0.09; P = 0.76), or Karnofsky score (r
= 0.01; P = 0.97). The AUCs for the ratio of
DHEA to cortisol (7,763 ± 1098 vs.
11,041 ± 869; P = 0.03) and aldosterone to
cortisol (1,602 ± 282 vs. 2,980 ± 410;
P = 0.02) were significantly reduced in the
HIV-infected patients compared with the control subjects. The reduced
DHEA to cortisol and aldosterone to cortisol ratios were
not significantly different in subgroup analyses by PI, NRTI, or NNRTI
status.
|
|
|
In this study we investigated androgen physiology in ambulatory women with the AIDS wasting syndrome compared with that in age- and BMI-matched healthy control subjects. Significant androgen deficiency is common among women with HIV disease, occurring in 66% of HIV-infected women with wasting (1). We have previously shown that both free T and DHEAS concentrations correlate with muscle mass, and that physiological androgen replacement in such patients increases weight and quality of life indexes (3). In the current study we demonstrate a significant reduction in T, DHEAS, and androstenedione levels in women with AIDS wasting compared with age- and weight-matched control subjects and show altered adrenal steroid metabolism with shunting toward cortisol and away from adrenal androgen production in this population. In contrast, stimulatory testing with hCG demonstrated no differences in ovarian androgen responses.
Baseline androgen concentrations, including DHEAS, androstenedione, and free T, were significantly reduced in the HIV-infected patients compared with the control subjects. In contrast, E2 concentrations were not different between the groups, and this result was not affected in a secondary analysis adjusting for SHBG. We analyzed free T by an SHBG-independent method. Furthermore, DHEAS is not bound to any significant degree by SHBG. The specific etiology of increased SHBG in HIV-infected patients remains unknown (2, 16).
The 17-OHP and androstenedione levels achieved in response to hCG were used as an index of ovarian androgen secretory reserve in this study. hCG is not known to stimulate adrenal androgen secretion and is therefore useful as a selective ovarian androgen secretagogue (17). Furthermore, a single dose of hCG results in consistent increments in ovarian androgen metabolites after 24 h in normal premenopausal women (18). Our data demonstrate no significant differences in the androgen responses to hCG in the HIV-infected vs. normal control subjects, suggesting intact ovarian androgen secretory reserve. Similar results were shown in a subanalysis limited to patients with normal menstrual function. Abnormal conversion of androstenedione to T may also contribute to decreased measured free T levels in women with the AIDS wasting syndrome. Conversion of androstenedione to T is a function of many factors, including 17ß-reductase function in the ovary and adrenal as well as peripheral conversion in the liver, skin, and adipose tissue. However, neither the baseline nor the change in the ratio of androstenedione to T was different between the groups, suggesting normal 17-ketosteroid reductase activity in this population. Although androgen responses to stimulation with hCG were not different from those in the control population, subtle abnormalities in ovarian androgen function may exist that were not detected in our testing schema.
GnRH was not used to test ovarian androgen secretion in this study, because undernutrition is known to affect hypothalamic GnRH secretion and pituitary gonadotropin responsiveness to GnRH (6, 7, 8, 9). Prolonged testing with pulsatile GnRH is necessary in the case of potential hypothalamic dysfunction and is not an appropriate or direct test of ovarian androgen secretion in this population. In addition, pituitary gonadotropin function may be abnormal due to illness, medications, or other factors in this population, further decreasing the utility of GnRH as a test of ovarian androgen secretion. In this study, although the patients were BMI-matched, subjects with wasting had lost a significant amount of weight, which might have affected hypothalamic-pituitary-gonadal and adrenal function. Indeed, FSH, but not LH, concentrations were significantly reduced in the HIVinfected patients compared with the control subjects, and abnormal hypothalamic-pituitary function may contribute to reduced androgen concentrations in HIV-infected women. A selective reduction in FSH has previously been shown in studies of non-HIV-infected patients during short-term fasting and with weight loss resulting in hypothalamic amenorrhea (9, 19). Therefore, although the subjects were matched in terms of weight, the large weight loss may have contributed to a selective reduction in FSH in the HIV-infected patients. None of the patients demonstrated any signs or symptoms of pituitary mass effect to suggest a secondary cause of hypogonadism, and it is unlikely that selective suppression of FSH would be due to mass effect. Similarly, none of the patients was taking any medications known to affect hypothalamic-pituitary-gonadal function. In this regard, the use of megestrol acetate was a specific exclusion criterion for the study.
DHEAS, a specific marker of adrenal androgen secretion, was reduced at baseline in HIV-infected patients compared with control subjects. We further tested adrenal steroid metabolism with ACTH after dexamethasone pretreatment as an index of adrenal androgen secretory reserve (20). Pretreatment with dexamethasone at 1 mg was used to reduce intraindividual variation in pre-ACTH hormone levels and also to permit a determination of the change in response to ACTH as evidence of steroid shunting. The dose of dexamethasone used has not been shown to decrease the absolute response to ACTH (20). Baseline 24-h urine free cortisol and postdexamethasone cortisol concentrations were not significantly different between the groups.
We compared adrenal steroid hormones, including the DHEA to cortisol ratio and the aldosterone to cortisol ratio after ACTH stimulation between the HIV-infected and control patients. Our data demonstrate significant differences in DHEA and cortisol responses to ACTH between the groups. The HIV-infected subjects demonstrated reduced DHEA and increased cortisol responses to ACTH. The ratio of the DHEA to cortisol response, an index of shunting away from adrenal androgen secretion and toward increased cortisol production, was significantly decreased in the women with AIDS wasting syndrome compared with control subjects. Again, similar results were demonstrated in a comparison limited to HIV-infected patients with normal menstrual function. Evidence for shunting away from androgen synthesis has been shown by Findling et al. and others (11, 12, 21, 22) in men with HIV disease, but has not previously been investigated in women with AIDS wasting. A similar pattern of shunting toward cortisol and away from adrenal androgens has also been seen in acute and chronic illness in non-HIV-infected patients. Decreased adrenal 17,20-lyase function is thought to underlie the shift away from adrenal androgen synthesis in acute and chronic illness and undernutrition (10, 23, 24). The abnormalities of cortisol and adrenal steroid metabolism shown in this study are unlikely to be related to antiviral therapy per se, as therapy varied widely with respect to class of drug and specific agents. Furthermore, no significant differences were observed in subanalyses by class of antiretroviral therapy. Taken together, our data suggest significant shunting away from androgen metabolism and toward cortisol metabolism in women with the AIDS wasting syndrome, resulting in a low baseline DHEAS level, decreased adrenal androgen responses to ACTH, and increased cortisol responses to ACTH. Basal DHEAS and area under the curve (AUC) DHEAS responses to ACTH were nearly linearly related.
The pre-ACTH aldosterone to cortisol ratio and the post-ACTH AUC for the aldosterone to cortisol ratio were reduced in the HIV-infected compared with control subjects. In a prior study, Findling et al. (12) demonstrated no difference in aldosterone responses to ACTH in HIV-infected patients compared to control subjects, but showed that aldosterone responses tended to be lower with progressive HIV disease stage. Seventeen deoxysteroid levels were lower among HIV-infected patients after ACTH compared with control subjects in a study by Membreno et al. (11). In contrast to our study, the majority of patients investigated by Findling et al. (12) and Membreno et al. (11) were male (91% and 99% male, respectively), and responses to ACTH stimulation postdexamethasone were not studied. Furthermore, in contrast to prior studies, we also determined 24-h urinary free cortisol excretion as an index of overall cortisol metabolism. Taken together, the data from this study suggest simultaneous shunting away from mineralocorticoid and androgen production toward cortisol production in HIV-infected women with AIDS wasting. Shunting toward cortisol production may be an adaptive response to stress.
Similar to the observation of Findling et al. (12) and Membreno et al. (11), we demonstrated a significant correlation between the CD4 count and the ratio of DHEA to cortisol, suggesting a relationship between immune function, as measured by CD4 count, and altered steroid enzyme metabolism, characterized by greater cortisol and reduced androgen production. In contrast, we observed no relationship between weight, weight loss, or Karnofsky score and adrenal steroid concentrations or between weight and CD4 count. Furthermore, subjects were matched for weight and age with the comparison group. The clinical consequences of shunting away from adrenal androgen synthesis and toward cortisol in HIV-infected women are unknown, but such shunting may contribute to the severe degree of androgen deficiency in this population.
HIV disease is increasing in women, yet no previous studies have been performed exclusively in women to investigate abnormal androgen physiology in this emerging population of patients. Our data highlight the severe degree of androgen deficiency in women with AIDS wasting compared with healthy control subjects and strongly suggest shunting of adrenal steroid metabolism toward cortisol and away from androgen production. We simultaneously investigated ovarian androgen secretion and demonstrate intact ovarian androgen secretion in response to hCG. Taken together, our data provide new information on the physiological regulation of androgen metabolism in HIV-infected women with AIDS wasting. Further studies of the mechanism of adrenal androgen shunting in this population are needed. Furthermore, clinical studies of androgen replacement therapy to improve lean body mass, quality of life, and other androgen-dependent processes are needed in HIV-infected women.
Acknowledgments
We thank Jeff Breu of the Massachusetts Institute of Technology General Clinical Research Center Core laboratory for his help with the performance of the RIAs, and the nursing staff of the Massachusetts General Hospital General Clinical Research Center for their dedicated patient care.
Footnotes
This work was supported in part by NIH Grants R01-DK-54167, M01-RR-01066, and T32-DK-07703.
Abbreviations: AIDS, Acquired immune deficiency syndrome; AUC, area under the curve; BMI, body mass index; CV, coefficient of variation; DHEA, dehydroepiandrosterone; DHEAS, dehydroepiandrosterone sulfate; HIV, human immunodeficiency virus; NNRTI, non-nucleoside reverse transcriptase inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; 17-OHP, 17-hydroxyprogesterone; PI, protease inhibitor.
Received February 13, 2001.
Accepted May 15, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  P. Koutkia, J. Berry, K. Eaton, J. Breu, and S. Grinspoon Increased adrenal androgen secretion with inhibition of 11beta-hydroxylase in HIV-infected women Am J Physiol Endocrinol Metab, May 1, 2006; 290(5): E808 - E813. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Johnsen, S. E. Dolan, K. V. Fitch, K. M. Killilea, J. L. Shifren, and S. K. Grinspoon Absence of Polycystic Ovary Syndrome Features in Human Immunodeficiency Virus-Infected Women Despite Significant Hyperinsulinemia and Truncal Adiposity J. Clin. Endocrinol. Metab., October 1, 2005; 90(10): 5596 - 5604. [Abstract] [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 |
, P < 0.05; 
