Testosterone deficiency in men is manifested typically by symptoms of hypogonadism, including decreases in erectile function and libido. Testosterone also has an important role in the regulation of normal growth, bone metabolism and body composition. Specifically, testosterone deficiency is an important risk factor for osteoporosis and fractures in men. In men older than 65 years of age, the incidence of hip fracture is 4-5/1000 and approximately 30% of all hip fractures occur in men. Men with testosterone deficiency have significant decreases in bone density, particularly in the trabecular bone compart-ment. Testosterone defi-ciency has been reported in over half of elderly men with a history of hip fracture. Men with testosterone deficiency also have alterations in body composition that include an increase in body fat. Using quantitative CT scans to assess fat distribution, we have shown that testosterone deficiency is associated with an alteration in site-specific adipose deposition with increased deposits in all areas, particularly in the subcutane-ous and muscle areas. Be-cause truncal fat correlates with glucose intolerance and cardiovascular risk, hypogo-nadism may have important implications with regard to overall health and mortality. In one study, the alteration in skeletal muscle composition was associated with a de-crease in muscle strength. Therefore, testosterone deficiency is associated with an enhanced risk for osteopo-rosis, altered body composi-tion including increases in truncal fat, and, possibly, decreases in muscle perform-ance.

Administration of testosterone replacement therapy leads to improve-ments in libido and erectile function. Following testos-terone replacement, men note an increase in energy and mood, which may reflect either direct behavioral effects of androgens, or, an elevation of hematocrit due to rising testosterone levels. Testos-terone therapy also leads to important beneficial effects on the skeleton and lean tissue mass. Testosterone replace-ment increases bone density in hypogonadal men with the most dramatic effects seen in the trabecular bone compart-ment. These effects may be seen as early as 6 months following initiation of testos-terone therapy. In one recent study of the long-term bene-fits of testosterone therapy, the greatest benefits in trab-ecular bone were seen in the first several years of therapy. With regard to body compo-sition, testosterone replace-ment therapy results in a dramatic reduction in adipose content, with the greatest effects seen in the subcutane-ous and skeletal muscle areas. Androgen therapy leads to a significant increase in lean skeletal muscle mass and strength. Therefore, there are beneficial effects of testoster-one replacement on body composition and bone mineral density in adult hypogonadal men that may serve as indica-tions for therapy in addition to that of libido and sexual function.

There are several modes of administration available for testosterone replacement. The traditional form of testosterone therapy consists of intramuscular injections of testosterone esters given at 2 to 4 week intervals. This mode of therapy leads to an increase in testosterone levels, but there are marked oscillations in serum testosterone levels with early peak, supraphysiologic levels followed by levels that fall in the subtherapeutic range. Therefore, men may note an improvement in sexual function only in the immediate period following the injection. Also, men may describe mood swings and behavioral alterations that may reflect these changing testosterone levels. More recent advances had led to the development of novel delivery systems for androgens such as transdermal preparations of testosterone. The major benefit of these patches includes the attainment of more physiologic testosterone replacement with serum testosterone levels in the normal range throughout the day. There are transdermal systems applied to the scro-tum (Testoderm) and skin (Androderm) that are avail-able. Although these patches lead to normal testosterone levels, they are somewhat limited by difficulty with adherence and need for shav-ing (Testoderm) and local irritation (Androderm). A new system, Testoderm TTS (ALZA Pharmaceuticals) has been available on the market as of March, 1998. This new form of testosterone replace-ment consists of a single patch that is applied to non-scrotal skin and appears be have a low incidence of local irritation. Therefore, Testo-derm TTS may have an advantage over currently available systems.

There are several adverse effects of testosterone administration that need to be closely monitored, including clinically significant benign prostatic hypertrophy (BPH) and prostate cancer. Despite the theoretical considerations that androgens will augment prostate size, there is no evidence that androgen re-placement in elderly men will lead to the development of hyperplasia or aggravate its clinical status. There is also a concern that prostate cancer may develop during androgen therapy. There are no data available as to whether andro-gen therapy will enhance the progression of preclinical to clinical cancer. However, androgens may stimulate the growth of clinically diagnosed prostate cancer. Therefore, prior to testosterone initiation, patients should be screened for BPH and prostate cancer with a clinical history, digital exam, and PSA (prostate specific antigen) level. Because androgens may stimulate erythropoiesis and precipitate sleep-related breathing disor-ders, a cbc should be followed and subjects queried for the presence of sleep apnea.

References

1. Katznelson L, Finkel-stein JS, Schoenfeld DA, Rosenthal DI, Anderson EJ, Klibanski A. 1996 In-crease in bone density and lean body mass during testosterone administration in men with acquired hy-pogonadism. J Clin Endo-crinol Metab. 81:4358-65.
2. Simon D, Charles M, Nahoul K, et al. 1997 As-sociation between plasma total testosterone and car-diovascular risk factors in healthy adult men: the Telecom study. J Clin En-docrinol Metab. 82:682-5.
3. Swerdloff RS, Wang C. 1993 Androgen deficiency and aging in men. West J Med. 159:579-585.
4. Snyder PJ. 1984 Clini-cal use of androgens. Ann Rev Med. 35:207-17.
5. Bhasin S. 1992 Clinical review: Androgen treat-ment of hypogonadal men. J Clin Endocrinol Metab. 74:1221-5.

The novel occurrence of dorsocervical fat accumulation, centripetal obesity and muscle wasting has been described recently among a subset of HIV-infected patients. The ex-act prevalence of the syn-drome is unknown and es-timates vary from over 50% to only a small minor-ity of patients treated with aggressive antiretroviral therapy. The most promi-nent reported feature is striking fat deposition in a classical "buffalo hump" pattern. In addition, insulin resistance, abnormal glu-cose homeostasis and tri-glyceride elevations are also reported in some but not all subjects. The clini-cal course of four such pa-tients is outlined in detail in Table I and the prominent dorsocervical fat accumu-lation in one patient is shown in Figure 1. HIV-infected patients with this syndrome are increasingly referred to the endocri-nologist to rule out Cush-ing's syndrome. However, such patients are most ap-propriately categorized as having pseudo-Cushing's syndrome because of the absence of a known tumor resulting in hypercorti-solism and appropriate re-sponse to standard dexa-methasone suppression testing.

Pseudo-Cushing's syndrome is manifested by partial or complete clinical features of cortisol excess and characterized by nor-mal or abnormal biochemi-cal abnormalities in gluco-corticoid testing. For ex-ample, alcoholic patients may demonstrate clinical and biochemical features indistinguishable from true Cushing's syndrome. In contrast, patients with de-pression may demonstrate non-suppressibility of corti-sol levels in the absence of a Cushingoid habitus. Among a subset of HIV-infected patients respond-ing to aggressive antiretro-viral therapy, striking dor-socervical fat accumula-tion, centripetal obesity and, to a lesser degree, pe-ripheral muscle wasting in association with insulin re-sistance, suggest a state of partial or mild cortisol ex-cess. However, other clas-sical features of cortisol excess, including striae, muscle weakness and bruising are most often ab-sent, evidence that the syn-drome is not a true state of cortisol excess. Urine free cortisol levels may be ele-vated in some patients, but suppress adequately during formal dexamethasone testing. In addition, mini-mal cortisol and ACTH stimulation in response to CRH in the combined dexamethasone-CRH test argue against true Cush-ing's syndrome among pa-tients with elevated base-line cortisol levels.

The mechanisms of pseudo-Cushing's syn-drome in HIV-infection are unclear. The observed pattern of fat accumulation, centripetal obesity and in-sulin resistance suggest that systemic or local glu-cocorticoid excess may play a role in the syndrome. Activation of the hypotha-lamic-pituitary adrenal axis is known to occur in HIV-infected patients, suggest-ing a potential mechanism of the observed phenotypic abnormalities. HlV-infection is associated with higher than normal cortisol levels, even in the early stages of the disease. The source of the hypercorti-solism associated with HIV infection is unknown and may result from the stress of the illness per se or a cytokine effect. Variable levels of ACTH have been reported in association with increased cortisol levels in HIV disease suggesting that both central and pe-ripheral activation of the hypothalamic-pituitary-adrenal axis may occur. However, cortisol levels are often only modestly increased among HlV-infected patients, and the diurnal rhythm of cortisol is maintained, unlike the case in true Cushing's syndrome. Furthermore, abnormal fat deposition and other symptoms of glucocorti-coid excess have not previ-ously been reported among HlV-infected patients.

Notably, the recent reports of abnormal fat deposition in HIV infection coincide with the develop-ment of successful new strategies to rapidly and dramatically improve the immunologic status of the HIV-infected patient. In this regard, the syndrome is most often associated with recovery from the wasting syndrome, weight gain and marked improvement in immune function, suggest-ing that excess fat deposi-tion may be related to dis-ease recovery independent of medication usage. Alter-natively, the temporal as-sociation of the syndrome with the use of a specific class of agents known as protease inhibitors, sug-gests that some element of the syndrome may be a di-rect result of this particular therapy.

Table 1: Clinical, Hormonal and Virologic Parameters in Four HIV-Infected Patients with Pseudo-Cushing's Syndrome
(Adapted from Miller et al.)

	patient 1	patient 2	patient 3	patient 4	normal range
Sex	male	male	male	female
Age	44	42	46	39
Clinical signs	dorsocervical fat pad submandibular fat central obesity muscle wasting	dorsocervical fat pad submandibular fat central obesity muscle wasting facial plethora	dorsocervical fat pad subniandibular fat central obesity facial plethora	dorsocervical fat pad central obesity
Weight gain (lbs)	20	20	33	11	not applicable
Predisposing factors (alcoholism, glucocorticoid or megesterol acetate)	none	none	none	none
CD4 count (ceffs/mm3) Pre pseudo-Cushing's With pseudo-Cushing's	97 267	22 200	130 280	127 244	>800
Viral load (copies/mL) Pre pseudo-Cushing's With pseudo-Cushing's	8900 <400*	66,435 <400*	130,800 2539	122,851 <500*	undetectable
ACTH (pmol/L)	9.5	7.9			1.3-16.7
Urine free cortisol** (nmol/day)	30*** 55*** 47***	223*** 242*** 251***	361****	<14*****	see footnotes below
One mg overnight dexamethasone suppression test		<28	<33		serum cortisol <140 mnol/L at 0800h the following morning
2-day low dose dexamethasone suppression test		17			urine free cortisol <55 mmol/d
CRH-low dose dexamethasone test		27.6			serum cortisol < 38.6 mmol/L excludes Cushing's syndrome

Footnotes:

*Undetectable range for specific assay used;
**An adequate collection was demonstrated by urine creatinine levels;
***Normal range: 55-193 nmol/day;
****Normal range: 55-248 mnol/day;
*****Normal range: 0-138 mnol/day;

The primary feature in virtually all reported cases of the pseudo-Cushing's syndrome asso-ciated with HIV-infection is a striking excess of fat deposition, primarily in a dorsocervical pattern. In the setting of weight gain, abnormal partitioning into fat may result from an al-tered steroid or cytokine milieu, or a paracrine ac-tion of the virus itself on the glucocorticoid recep-tor. For example, a direct stimulatory effect of viral proteins on the glucocorti-coid receptor has been demonstrated and may re-sult in enhanced glucocor-ticoid action and fat depo-sition during weight recov-ery. In addition, fatty acids and other liposoluble sub-stances in serum from HIV-positive patients may affect the binding of corti-sol to the glucocorticoid receptor. Tissue-specific paracrine abnormalities in cortisol metabolism may explain excess fat deposi-tion during weight recov-ery in the absence of changes in systemic steroid levels. Alternatively, pref-erential deposition of fat may occur upon recovery due to increased triglycer-ide production associated with HlV-infection.

Figure 1. Forty-four year old gentleman presenting with the onset of excess dorsocervical and subman-dibular fat, central obesity and peripheral muscle wasting over 4 months in the setting of a 20 pound weight gain and dramatic improvement in immuno-logic function upon the ini-tiation of protease inhibitor therapy.

In addition, fat deposition may be the re-sult of altered P450 me-tabolism of endogenous glucocorticoid or other as yet unknown effects of protease inhibitor therapy. A significant number of patients described in recent reports were taking combi-nations of new antiretrovi-ral therapies, most often including indinavir, and had experienced dramatic im-provements in disease status and significant dec-rements in viral load in as-sociation with the devel-opment of pseudo-Cushing's syndrome. A dramatic effect of protease inhibitors on P450 metabo-lism has recently been demonstrated. Such medi-cines may also have tissue-specific paracrine effects to increase glucocorticoid concentration in fat cells, independent of any meas-urable change in circulating steroid hormone levels. Protease inhibitors and/or other new antiretroviral agents may therefore alter metabolism in such a way that patients preferentially gain fat mass during weight recovery.

There is no known treatment for the pseudo-Cushing's syndrome asso-ciated with HIV-infection. Anecdotal evidence sug-gest that discontinuation of protease inhibitor therapy in some cases will result in decreased symptomatolgy. Liposuction of the poste-rior cervical fat pad is a cosmetic alternative that may allow significant im-provement in appearance as well as functional en-hancement of neck exten-sion in some patients. In addition, ketoconazole, a medication which inhibits adrenal steroidogenesis and is used to treat Cushing's syndrome may poentially be of some utility in a sub-set of patients with HIV-associated pseudo-Cushing's syndrome. For example, moderate im-provement in dorsocervical fat was recently described in one patient receiving ketoconazole. In this case, the patient had an elevated baseline urine free cortisol level, suggesting that keto-conazole might be of po-tential benefit. However, the use of ketoconazole for this indication should be considered experimental and most appropriately re-served for patients with increased urine free cortisol levels. It must be empha-sized that ketoconazole administration is associated with adrenal insufficiency, hypogonadism and poten-tial liver function abnor-malities and should be used with caution in such pa-tients. Controlled pro-spective studies are needed to establish the efficacy of ketoconazole administra-tion and/or other therapies in this population.

The development of pseudo-Cushing's syn-drome among a relatively large subset of HIV-infected patients treated with aggressive antiretrovi-ral therapy is an important and unexpected new devel-opment which may be a direct result of improve-ment in the immunologic and clinical status of such patients. The striking similarities in phenotypic expression of the syndrome among affected patients suggests a common, but still unknown, etiologic factor which may relate to specific medication effects or more generally to dis-ease recovery. Clinicians treating patients with HIV disease are increasingly likely to encounter patients with these symptoms as ever more powerful antivi-ral therapies are used to improve immune function. Important, as yet unan-swered, questions exist as to the mechanism of syn-drome and its clinical im-plications in terms of re-covery from the wasting syndrome and the potential prohibitive effect of potent new antiretroviral therapies in this regard. Is the syn-drome simply a cosmetic issue, or do the associated metabolic and phenotypic abnormalities have real im-plications for the long-term health of such patients? Finally, the potential treat-ment of such patients is now limited but may in-clude potential endocrine manipulation with steroid inhibitors in the near fu-ture. Further research into the mechanisms of HIV-associated pseudo-Cushing's syndrome is needed to allow treatment and/or prevention of this syndrome in the future.

References

1. Carr A, Samaras K, Law M, Freund J, Chisholm DJ, Cooper DA. A Syndrome of peripheral lipodystrophy, hyperlipi-demia, and insulin resis-tance in patients receiving HIV protease inhibitors. AIDS. 1998 (In Press).
2. Christeff N, Gharakha-nian S, Thobie N, Rozen-baum W, Nunez E. Evi-dence for changes in adre-nal and testicular steroids during HIV infection. J Acq Immune Def Syndr.1992;5:841-6.
3. Debord M, Levi F. Cir-cadian variations in plasma levels of hypophyseal, adrenocortical and testicu-lar hormones in men in-fected with human immu-nodeficiency virus. J Clin Endocrinol Metab. 1990;70:572-7.
4. Eagling VA, Back DJ, Barry MG. Differential in-hibition of cytochrome P450 isoforms by the pro-tease inhibitors, ritonovir, saquinivir and indinavir. Br J Clin Pharmacol. 1997;44:190-4.
5. Grinspoon SK, Bil-ezikian JB. HIV disease and the endocrine system. N Engl J Med. 1992;327: 1360-5.
6. Kino T, Kopp JB, Chrousos GP. The HIV- 1 VPR gene product en-hances the transactivating effects of glucocorticoids in human lymphoid and muscle-derived cell lines. The Annual Meeting of the Endocrine Society. Min-neapolis; 1997.
7. Membreno L, Irony I, Dere W, Klein R, Biglieri EG, Cobb E. Adrenocorti-cal function in Acquired Immuno-Deficiency Syn-drome. J Clin Endocrinol Metab. 1987;65:4827.
8. Miller K, Daily P, Sen-tochnik D, Doweiko J, Samore M, Basgoz N, Grinspoon S. Pseu-docushing's syndrome in HIV-infected patients. Clin Infectious Dis. 1998 (In Press).
9. Verges B, Chavanet P, Desgres J, et al. Adrenal function in HIV-infected patients. Acta Endocrinol. 1989; 121 :633-7.