The Effect of a Pure Antiandrogen Receptor Blocker, Flutamide, on the ...

0021-972X/98/$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1998 by The Endocrine Society

Vol. 83, No. 8 Printed in U.S.A.

The Effect of a Pure Antiandrogen Receptor Blocker, Flutamide, on the Lipid Profile in the Polycystic Ovary Syndrome EVANTHIA DIAMANTI-KANDARAKIS, ASIMINA MITRAKOU, SOTOS RAPTIS, GEORGE TOLIS, ANTONI J. DULEBA

AND

Laiko Hospital, First Department of Internal Medicine, University of Athens (E.D.-K., G.T.), and Evangelismos Hospital, Second Department of Internal Medicine, University of Athens (A.M., S.R.), Athens, Greece; and the Department of Obstetrics and Gynecology, Yale University School of Medicine (A.J.D.), New Haven, Connecticut 06520 ABSTRACT Polycystic ovary syndrome (PCOS) is one of the most common endocrinopathies affecting women of reproductive age; it is associated with hyperandrogenism, hyperinsulinemia, and dyslipidemia. This study was designed to assess the long term effects of a pure androgen receptor blocker, flutamide, on the lipid profile in women with PCOS and to examine the possible mechanisms by which androgens may exert their influence. Seventeen women with PCOS (10 obese and 7 lean) were studied. All subjects received a 12-week course of oral flutamide (500 mg/day). The baseline and posttreatment evaluations included lipid profile, androgen levels, insulin sensitivity, and serum catecholamine determinations. The primary outcome was the change in the ratio of low density lipoproteins (LDL) to high density lipoproteins (HDL). Treatment with flutamide was associated with a signif-

P

OLYCYSTIC ovary syndrome (PCOS) is one of the most common endocrinopathies, affecting 5–10% of premenopausal women (1). Hyperandrogenism in women with PCOS is associated with metabolic aberrations of much greater importance than cosmetic problems such as hirsutism or acne. These aberrations include insulin resistance with accompanying hyperinsulinemia and dyslipidemia (2– 8). To our knowledge, there is no definitive prospective study demonstrating that PCOS represents an independent of obesity risk factor for development of cardiovascular disease. However, several retrospective and case-control studies indicated that, in the long term, women with a history of PCOS may be at increased risk for developing hypertension, noninsulindependent diabetes mellitus, atherosclerosis, coronary artery disease, and myocardial infarction (6, 9 –11). Most reports suggest that the lipid profile of women with PCOS is characterized by elevated serum levels of cholesterol, low density lipoproteins (LDL), very low density lipoproteins (VLDL), and triglycerides, with concomitantly reduced concentration of high density lipoproteins (HDL) (5, 6, 8, 12–15). These abnormalities in lipid levels have serious atherogenic consequences; in particular, as observed in the Received December 18, 1997. Revision received April 3, 1998. Rerevision received May 7, 1998. Accepted May 7, 1998. Address all correspondence and requests for reprints to: Dr. Evanthia Diamanti-Kandarakis, First Department of Internal Medicine, Endocrinology Section, University of Athens Medical School, 17 Ag. Thomas Street, Athens 115–27, Greece.

icant decrease in the LDL/HDL ratio by 23% (P 5 0.005), in total cholesterol by 18% (P , 0.0001), in LDL by 13% (P 5 0.002), and in triglycerides by 23% (P 5 0.002). Flutamide treatment was also associated with a trend toward an increase in HDL (by 14%; P 5 0.14). The effects on lipid profile were found regardless of obesity and were not associated with a change in weight. Furthermore, actions of flutamide on lipid metabolism were not associated with significant changes in circulating adrenaline or noradrenaline, glucose metabolism, or insulin sensitivity. This report has demonstrated for the first time that treatment with the pure antiandrogen, flutamide, may improve the lipid profile and that this effect may be due to direct inhibition of androgenic actions. (J Clin Endocrinol Metab 83: 2699 – 2705, 1998)

Framingham study, high LDL and low HDL predict the development of coronary artery disease (16, 17). Dyslipidemia typically occurs within a cluster of several interrelated cardiovascular risk factors, including obesity, high waist to hip ratio (WHR), hyperinsulinemia, and hyperandrogenism. The elucidation of cause and effect relationships among these factors and hence the identification of independent risk factors are exceedingly difficult. In light of recent studies, it appears that although obesity is a major contributor to dyslipidemia in women with PCOS, some abnormalities of lipid profile are independent of obesity. Graf et al. observed that obesity was associated with a decrease in HDL levels in both PCOS and control subjects (18). Subsequently, Robinson et al. observed that low HDL in women with PCOS cannot be explained solely by obesity, indicating, therefore, that other inherent features of PCOS predispose to dyslipidemia (19). Potential mechanisms of dyslipidemia in women with PCOS include hyperinsulinemia and hyperandrogenism. Alternatively, genetic variation in each of these mechanisms modified by environment may be occurring. The roles of insulin and sex hormones in the regulation of lipid metabolism have been well recognized. In particular, levels of HDL2 are regulated by lipoprotein lipase and hepatic lipase; these activities are responsive to insulin and sex steroids, respectively (20, 21). Ek et al. observed that women with PCOS had an impairment of catecholamine-induced adipocyte lipolysis due to defects such as a decreased number of

2699

2700

DIAMANTI-KANDARAKIS ET AL.

b2-adrenoreceptors as well as postreceptor dysfunction of the protein kinase A or hormone-sensitive lipase (22). These defects in the adrenergic regulation of lipolysis may be attributable to insulin resistance and hyperinsulinemia. Elevation of androgens offers another plausible explanation of metabolic disturbances in insulin resistance syndromes such as PCOS (23, 24). Hyperandrogenism is associated with upper body obesity (expressed as WHR) independently of weight (25–27). Furthermore, administration of exogenous androgen to women leads to increased visceral fat accumulation and decreased serum HDL (28). However, the mechanisms of androgen actions on lipid metabolism are still poorly understood. There is growing evidence that androgens may influence the predominant site of body fat deposition and muscle morphology, possibly in relation to alterations in splanchnic insulin metabolism and insulin sensitivity (29 –32). It appears that androgens may also adversely affect lipid metabolism by direct modulation of lipoprotein lipase and lipolysis (33). The effects of androgens on lipid profile may be sex dependent; studies in hypogonadal men are inconsistent and indicate that androgen replacement may have either an adverse or a beneficial effect (34, 35). The present study was undertaken to assess the long term effects of a pure androgen receptor blocker, flutamide, on the lipid profile of women with PCOS and to examine the possible mechanisms by which androgens may exert their influence. To our knowledge, this is the first report demonstrating that flutamide exerts a beneficial effect on serum lipids. Subjects and Methods Subjects Seventeen women with PCOS (10 obese and 7 lean) were studied (Table 1). The diagnosis of PCOS was made in the presence of chronic anovulation, and hyperandrogenism was confirmed by an elevation of at least 2 of the following plasma androgens: total testosterone, free testosterone, androstenedione, and 3a-androstanediol glucuronide. Hirsutism and acne were present in 9 subjects, 6 subjects had only hirsutism, and 2 subjects had only acne. The diagnosis of PCOS was further substantiated by the finding of thickened stroma and multiple subcapsular cysts on ovarian ultrasonographic examination. Subjects with diabetes mellitus, adrenal disorders, hyperprolactinemia, or other endocrine disorders were excluded. Congenital adrenal hyperplasia was ruled out by evaluation of 17-hydroxyprogesterone and, when appropriate, ACTH stimulation test. None of the patients had hypertension or overt heart disease. The subjects were not taking any drugs known to affect gonadal function or carbohydrate metabolism. TABLE 1. Profiles of obese and lean participants in the study Obese (n 5 10)a

Age Wt (kg)b Pretreatment Posttreatment Ht (m) BMI (kg/m2) WHR

Lean (n 5 7)

22.7 6 1.0

22.0 6 1.7

81.6 6 4.6 81.5 6 4.6 1.63 6 0.02 30.6 6 1.3 0.89 6 0.03

55.3 6 3.3 56.0 6 3.1 1.62 6 0.03 20.9 6 0.8 0.81 6 0.04

Results are presented as the mean 6 SEM. a Subjects were considered obese when their BMI was at least 25 kg/m2. b Weight posttreatment was not significantly different from baseline weight.

JCE & M • 1998 Vol 83 • No 8

Body mass index (BMI) was calculated as weight (kilograms)/height (meters)2. Patients were considered lean when their BMI was below 25 kg/m2 and obese when their BMI was at least 25 kg/m2. Body fat distribution was assessed by measuring the WHR as described by others (36). After the completion of the baseline studies (see below), all subjects received a 12-week course of oral flutamide (Flucinom, Schering-Plough, Kenilworth, NJ) at a dose of 500 mg/day. Flutamide treatment was initiated on the first day of the menstrual cycle. The patients were advised not to change their eating habits or their activities throughout the study period. All participants were nonsmokers and nondrinkers; they received regular meals with a weight-maintaining diet. The study was approved by the local ethics board. Informed consent was obtained from all patients. Throughout the study, the subjects were using barrier contraception.

Study protocol All evaluations were conducted within 10 days from the onset of menstrual flow. In the absence of spontaneous menstruation, periods were induced by medroxyprogesterone acetate withdrawal. The baseline evaluations were performed as follows: blood samples were collected at 0800 h after an overnight fast to determine serum levels of steroids (testosterone, free testosterone, androstenedione, and 3aandrostanediol glucuronide), sex hormone-binding globulin (SHBG), and lipids (total cholesterol, triglycerides, HDL, and LDL). Glucose metabolism and insulin sensitivity were evaluated using an oral glucose tolerance test (OGTT) and hyperinsulinemic-euglycemic clamp procedure. OGTT was performed by collection of baseline blood samples followed by ingestion of a 75-g glucose load and subsequent blood sampling at 30-min intervals for 2 h to determine serum levels of insulin and glucose; the test was used to determine the area under the curve (AUC) for glucose and insulin. Insulin sensitivity was determined (5–7 days after OGTT) by measuring glucose uptake during the hyperinsulinemic euglycemic clamp procedure, as described previously (23). Briefly, after an overnight fast, catheters were inserted into a dorsal hand vein and an antecubital vein. Crystalline human insulin (Actrapid, Novo-Nordisk, Athens, Greece) was infused via a Harvard pump (Harvard Apparatus, Millis, MA) at a rate of 287 pmol (40 mU)/m2zmin for 180 min to increase the plasma insulin level to approximately 500 pmol/L (;75 mU/mL). Serum glucose was kept constant at the fasting level with the aid of bedside serum glucose determinations every 5 min and appropriate adjustment of a variable infusion of 20% glucose. Serum adrenaline and noradrenaline were assessed at a baseline (from three pooled samples collected every 5 min) and at the end of the hyperinsulinemiceuglycemic clamp study (from pooled samples collected at 170, 175, and 180 min). All of the above studies were repeated at the end of the 12-week course of flutamide.

Analytical procedures Plasma glucose was determined with a Beckman glucose analyzer (Palo Alto, CA), using a glucose oxidase method. A solid phase 125I RIA was used for quantitative measurement of serum insulin levels as described by others (29). Serum samples were analyzed using commercial RIA kits to determine levels of total testosterone (Byk-Sangtec Diagnostica, Dietzenbach, Germany) and free testosterone (Diagnostic Products Corp., Los Angeles, CA). Androstenedione, dehydroepiandrosterone sulfate (DHEAS), and 3a-androstanediol glucuronide were measured using kits from Diagnostic Systems Laboratories (Webster, TX). For all determinations, the intra- and interassay coefficients of variation were 5–7% and 8 –11%, respectively. Total cholesterol, triglycerides, and HDL were determined using techniques of Spectrum FPX (Abbott Laboratories, North Chicago, IL), whereas LDL was calculated by the Friedewald equation. Adrenaline and noradrenaline were determined using kits from DDN Diagnostika (Mirpurg, Germany).

Statistical analysis The values were expressed as the mean 6 sem. Comparison of means was performed using Student’s t test or ANOVA, as appropriate. Baseline and posttreatment levels of lipids and hormones were compared using repeat measures ANOVA with one grouping factor (obese vs.

EFFECT OF FLUTAMIDE ON LIPID PROFILE lean). When appropriate, log transformation of variables was performed.

Results Baseline and posttreatment levels of lipids

Treatment with flutamide resulted in marked changes in the lipid profile. Comparisons of lipid levels before treatment (baseline) and after the 12-week course of flutamide (posttreatment) are presented in Table 2. The results were analyzed using repeated measures ANOVA; this approach allows simultaneous evaluation of the effects of treatment, obesity, and the role of interaction between treatment and obesity. In the context of this study, the interaction component may be interpreted as a determination of whether the effect of treatment was altered by obesity. Flutamide treatment was associated with an average decline in cholesterol by 18%, in triglycerides by 23%, and in LDL by 13%. Treatment with flutamide was also associated with a trend (albeit not statistically significant) toward an increase in HDL by 14%. Ultimately, the LDL/HDL ratio declined by 23%. Effects of obesity on lipid profile were not statistically significant; nevertheless, there was a trend among obese patients to have higher levels of triglycerides and LDL. Comparable effects of flutamide were observed in both obese and lean patients. The effects of a 12-week course of flutamide on lipid profile in individual patients are presented in Fig. 1. Flutamide treatment was associated with a decline in the cholesterol level in all subjects. Triglycerides decreased in 9 of 10 obese patients and in all 7 lean patients. LDL declined in 7 of 10 obese patients and in 6 of 7 lean patients. HDL increased in 6 of 10 obese patients and in 6 of 7 lean patients.

2701

amide treatment had no significant effect on fasting glucose levels, glucose AUC in response to oral glucose load, fasting insulin, insulin AUC in response to the glucose load, and glucose uptake during hyperinsulinemic, euglycemic clamp studies. Power analysis revealed that the present study had sufficient power (type II error ,0.2 for type I error ,0.05) to detect a 30% change in catecholamine levels or insulin sensitivity. Obese women had, on the average, 61% higher free testosterone levels than lean subjects. There was no significant difference in the levels of 3a-androstanediol glucuronide, androstenedione, DHEAS, total testosterone, estradiol, or SHBG between obese and lean women. Furthermore, obese and lean subjects had comparable levels of adrenaline and noradrenaline. Obesity had no significant effect on fasting glucose and glucose AUC. However, obese women, compared to lean subjects, had higher fasting insulin (by 155%), higher insulin AUC (by 46%), and lower glucose uptake (by 31%) as determined by hyperinsulinemic, euglycemic clamp studies. It should be noted that due to the relatively small number of subjects, this study had limited power to evaluate the effects of obesity. Patients were categorized as lean or obese to better assess the effects of flutamide while accounting for a potentially confounding role of obesity. There was no significant interaction between the effects of treatment and obesity on any of the studied endocrine parameters. This observation indicates that flutamide treatment resulted in comparable effects in both obese and lean subjects. Flutamide treatment had no significant effect on the ovulatory status of the subjects; serum progesterone levels remained in the nonovulatory range. Hepatic function

Baseline and posttreatment endocrine profiles

Table 3 presents comparisons of baseline and posttreatment levels of individual endocrine parameters. Treatment with flutamide was associated with an average decline in 3a-androstanediol glucuronide by 21%, in androstenedione by 48%, and in DHEAS by 37%. Flutamide treatment had no significant effect on the level of total or free testosterone or estradiol. SHBG declined during flutamide treatment, on the average, by 18%. Flutamide had no demonstrable effect on levels of adrenaline and noradrenaline. Furthermore, flut-

Serum levels of hepatic enzymes were evaluated before and at the end of the treatment period (Table 4). Flutamide treatment was associated with modest increases in the results of liver function tests; however, these changes were not clinically significant. Discussion

The results of this study indicate that 1) flutamide treatment is associated with a significant reduction in total

TABLE 2. Baseline and posttreatment lipid levels Obese

P valuea

Lean

Variable

Baseline

Posttreatment

Baseline

Posttreatment

Effect of treatment

Effect of obesity

Interaction: treatment vs. obesity

Cholesterol (mmol/L) Triglycerides (mmol/L) HDL (mmol/L) LDL (mmol/L) LDL:HDL ratio

4.50 6 0.38 0.88 6 0.19 0.92 6 0.08 2.87 6 0.35 3.46 6 0.29

3.85 6 0.33 0.73 6 0.14 1.03 6 0.10 2.50 6 0.30 2.66 6 0.39

4.49 6 0.35 0.56 6 0.08 1.00 6 0.06 2.31 6 0.17 2.35 6 0.19

3.43 6 0.08 0.33 6 0.03 1.14 6 0.09 1.99 6 0.19 1.79 6 0.19

,0.0001 0.002 0.14 0.002 0.005

0.64 0.05 0.82 0.27 0.18

0.20 0.10 0.81 0.88 0.78

Results are presented as the mean 6 SEM. Lipid levels are presented in Systeme International units. Normal ranges are: cholesterol, less than 5.2 mmol/L (multiply by 38.7 to convert to milligrams per dL); triglycerides, less than 1.8 mmol/L (multiply by 88.6 to convert to milligrams per dL); LDL, less than 3.36 mmol/L (multiply by 38.7 to convert to milligrams per dL); HDL, more than 1.29 mmol/L (multiply by 38.7 to convert to milligrams/dL). a Comparison of baseline and posttreatment levels determined using repeated measures ANOVA with one grouping factor (obese vs. lean). Interaction component of statistical analysis may be interpreted as a determination of whether the effect of treatment was different among obese vs. lean patients.

2702

DIAMANTI-KANDARAKIS ET AL.

JCE & M • 1998 Vol 83 • No 8

FIG. 1. Effect of a 12-week treatment with flutamide on lipid profile in lean and obese women with PCOS. Each line joins baseline and posttreatment values for each individual patient. Lipid levels are presented in Systeme International units. Normal ranges are: cholesterol, less than 5.2 mmol/L (multiply by 38.7 to convert to milligrams per dL); triglycerides, less than 1.8 mmol/L (multiply by 88.6 to convert to milligrams per dL); LDL, less than 3.36 mmol/L (multiply by 38.7 to convert to milligrams per dL); and HDL, more than 1.29 mmol/L (multiply by 38.7 to convert to milligrams per dL).

cholesterol, LDL, and triglycerides in young women with PCOS; 2) these effects occur in both obese and lean subjects; 3) the greatest effects of flutamide may be found in those with highest baseline levels of SHBG and androstenedione; and 4) the actions of flutamide on lipid metabolism appear not to be related to changes in circulating adrenaline and noradrenaline levels, glucose metabolism, or insulin sensitivity. The foremost importance of the present findings is the potential for the development of new therapeutic strategies for the treatment of dyslipidemia. The effects of flutamide on lipid levels are consistent with a significant decline in the risk for development of atherosclerosis and consequent cardiovascular disease. Most of the subjects in the present study did not have overt dyslipidemia. The potential therapeutic value of flutamide in the treatment of dyslipidemia has yet to be assessed in a broad population of subjects, over a longer treatment period, and with careful evaluation of possible adverse side-effects.

Although significant complications due to flutamide use are uncommon, patients may develop elevations of liver transaminases; furthermore, isolated cases of cholestatic hepatitis and even liver failure have been documented in elderly patients treated for prostatic cancer (37). Furthermore, a case of serious hepatotoxicity was reported in a women treated with flutamide for hirsutism (38). In this study, flutamide produced no significant side-effects. Flutamide is considered to be a “pure” androgen antagonist, acting by competitive inhibition of androgen receptors (39). Therefore, its actions on lipid profile may be most likely attributed to direct blockage of androgenic effects. As androgens are known to promote dyslipidemia, it is reasonable to expect that flutamide may promote a favorable lipid status by inhibiting these adverse actions of androgens. This line of reasoning is also supported by the findings of regression analysis, which indicate that the greatest improvement of total cholesterol may be found in patients with the highest pretreatment levels of androstenedione. In other words, the

EFFECT OF FLUTAMIDE ON LIPID PROFILE

2703

TABLE 3. Baseline and posttreatment endocrine profiles Obese

P valuea

Lean

Variable

Baseline

Posttreatment

Baseline

Posttreatment

Effect of treatment

Effect of obesity

Interaction: treatment vs. obesity

3a-Androstanediol glucuronide (nmol/L) Androstenedione (nmol/L) DHEAS (mmol/L) Free testosterone (pmol/L) Total testosterone (nmol/L) Estradiol (pmol/L) SHBG (nmol/L) Adrenaline, 0 min (ng/L) Adrenaline, 180 min (ng/L) Noradrenaline, 0 min (ng/L) Noradrenaline, 180 min (ng/L) Fasting glucose (mmol/L) Glucose: AUCb Fasting insulin (mU/L) Insulin: AUCb Glucose uptake (mmol/kgzmin)c

18.5 6 3.1

15.1 6 3.1

15.5 6 3.5

11.6 6 1.9

0.047

0.43

0.88

18.3 6 2.9 11.1 6 2.0 16.7 6 1.8 3.8 6 0.8 279 6 44 57.9 6 12 30.8 6 1.2 31.6 6 1.8 106.6 6 10.7 102.3 6 9.0 5.0 6 0.2 14.8 6 0.8 20.2 6 3.7 150 6 13 254 6 21

10.4 6 1.0 7.1 6 1.0 14.8 6 2.2 3.6 6 0.7 262 6 56 45.2 6 9 31.0 6 1.3 30.6 6 1.3 94.0 6 8.3 97.9 6 9.4 5.0 6 0.2 15.0 6 0.8 19.3 6 3.1 158 6 10 271 6 26

18.3 6 4.9 10.6 6 2.1 10.8 6 1.0 2.1 6 0.3 306 6 52 88.0 6 22.4 32.4 6 3.3 32.1 6 2.2 118.5 6 19.2 96.0 6 10.6 5.1 6 0.4 15.6 6 0.9 7.9 6 0.3 103 6 19 369 6 42

8.4 6 0.9 6.4 6 1.6 9.3 6 1.7 2.4 6 0.6 310 6 97 75.3 6 15.5 37.9 6 9 38.5 6 5.1 110.4 6 12.5 99.1 6 9.7 4.7 6 0.2 13.1 6 1.0 9.4 6 0.5 112 6 14 383 6 39

0.006 0.02 0.36 0.90 0.92 0.04 0.09 0.15 0.27 0.88 0.41 0.13 0.76 0.21 0.16

0.74 0.76 0.02 0.12 0.54 0.15 0.19 0.21 0.37 0.85 0.59 0.59 0.02 0.03 0.02

0.72 0.94 0.81 0.62 0.88 0.99 0.10 0.06 0.81 0.40 0.41 0.09 0.18 0.95 0.87

Results are presented as the mean 6 SEM. a Comparison of baseline and posttreatment levels determined using repeated measures ANOVA with one grouping factor (obese vs. lean). Interaction component of statistical analysis may be interpreted as a determination whether the effect of treatment was different among obese vs. lean patients. b The area under the curve (AUC) for glucose and insulin was determined during a 2-h oral glucose tolerance test (OGTT). c Glucose uptake (moles per L) determined during euglycemic, hyperinsulinemic clamp studies represents a measure of insulin sensitivity. TABLE 4. Liver function tests before and at the end of flutamide treatment Test

Aspartate aminotransferase (normal range, 5– 40 IU/L) Baseline End of treatment Alanine aminotransferase (normal range, 5– 40 IU/L) Baseline End of treatment g-Glutamyl transpeptidase (normal range, 10 – 40 IU/L) Baseline End of treatment Alkaline phosphatase (normal range, 30 –125 IU/L) Baseline End of treatment

Range

Median

Mean 6

SEM

8–19 7– 69

13 15

13 6 1 22 6 4

7– 46 6– 87

13 13

15 6 2 24 6 6

4– 48 3– 68

8 9

14 6 3 17 6 4

30– 67 10– 82

45 48

45 6 3 49 6 5

greatest improvement may be expected in those with the greatest initial androgenic effect. Lipid metabolism may be affected by various interlinked and interdependent mechanisms (18, 40). Thus, androgens may affect lipids not only directly, but also by affecting obesity, catecholamines, and insulin. In this study, treatment with flutamide had little or no effect on weight; furthermore, the improvement of lipid profile was observed in both obese and lean patients and regardless of the WHR. Consequently, although obesity and high WHR are important risk factors for dyslipidemia, the actions of flutamide cannot be attributed to the effects on total body fat and its distribution. Treatment with flutamide also had no significant effect on

adrenaline or noradrenaline; thus, it is unlikely that flutamide may have acted via modulation of metabolism of catecholamines. Furthermore, flutamide had no effect on fasting or post-OGTT insulin levels and glucose uptake during euglycemic clamp studies. Consequently, the effects of flutamide cannot be explained by the alterations in insulin sensitivity and its levels. Interestingly, Lovejoy et al. observed that administration of exogenous androgen to women led to increased visceral fat accumulation and decreased serum HDL without a change in fasting glucose or insulin sensitivity (28). Thus, androgens may affect lipid metabolism and fat deposition by mechanisms not involving insulin. Actions of flutamide on lipid metabolism may not be universal to all pure antiandrogens. Casodex, another nonsteroidal agent that blocks androgen receptors, had no significant effect on total cholesterol, HDL, or LDL in men (41). In other studies on men, the use of cyproterone acetate, a synthetic steroid antiandrogen, resulted in adverse changes in the lipid profile, most notably a decrease in HDL (42, 43). These effects of cyproterone acetate may be due to its progestogenic properties. Although flutamide is considered a pure antiandrogen, there is evidence that it has other biological activities and may, for example, inhibit adrenal 17–20-lyase (44). This activity may explain our present observation that flutamide treatment resulted in decreases in 3a-androstanediol glucuronide, androstenedione, and DHEAS. The present study revealed a complex and unexpected interrelationship between flutamide and SHBG. Treatment with flutamide led to a decrease in SHBG levels in both lean and obese patients; furthermore, high baseline SHBG level has been identified as one of the predictors of a greater decline in the total cholesterol level. Previous reports have demonstrated that hepatic production of SHBG may be in-

2704

DIAMANTI-KANDARAKIS ET AL.

hibited by insulin and androgens (45, 46). Yet in this study, flutamide had no detectable effect on insulin. Furthermore, antiandrogenic properties of flutamide would be expected to increase, rather than decrease, the SHBG level. Thus, an observed decline in SHBG may be due to flutamide acting via mechanisms independent of its antiandrogenic properties. Another explanation of the suppression of SHBG involves the possibility that some actions of androgens may be independent of androgen receptors and thus not be inhibited by flutamide. Indeed, Brown et al. observed that testosterone may act on the liver by a mechanism independent of androgen receptors (47). In this context, it is important to note that although flutamide blocks androgen receptor-mediated effects, it has no significant effect on the levels of free and total testosterone. Thus, it is possible that in the presence of flutamide, testosterone may maintain suppressive activity on hepatic SHBG production. This explanation may also apply to the study by Winneker et al., who found that flutamide failed to increase SHBG after its suppression by androgen treatment (48). The present study was limited to evaluation of a single dose of flutamide and did not include a placebo group. The dose of flutamide was selected on the basis of previous studies demonstrating its effectiveness in the treatment of hyperandrogenic conditions such as hirsutism, acne, and hair loss (23, 49, 50). Ideally, the effects of flutamide on lipid profile should be reevaluated in a placebo-controlled trial assessing several doses of the drug. In conclusion, this report has demonstrated for the first time that treatment with the pure antiandrogen, flutamide, may improve the lipid profile. The beneficial actions of flutamide appear to be independent of obesity, catecholamine metabolism, and insulin resistance. References 1. Dunaif A, Givens JR, Haseltine F, Merriam GR. 1992 The Polycystic ovary syndrome. Cambridge: Blackwell. 2. Burghen GA, Givens JR, Kitabchi AE. 1980 Correlation of hyperandrogenism with hyperinsulinism in polycystic ovarian disease. J Clin Endocrinol Metab. 50:113–116. 3. Chang RJ, Nakamura RM, Judd HL, Kaplan SA. 1983 Insulin resistance in nonobese patients with polycystic ovarian disease. J Clin Endocrinol Metab. 57:356 –359. 4. Dunaif A, Graf M, Mandeli J, Laumas V, Dobrjansky A. 1987 Characterization of groups of hyperandrogenic women with acanthosis nigricans, impaired glucose tolerance, and/or hyperinsulinemia. J Clin Endocrinol Metab. 65:499 –507. 5. Wild RA, Painter PC, Coulson PB, Carruth KB, Ranney GB. 1985 Lipoprotein lipid concentrations and cardiovascular risk in women with polycystic ovary syndrome. J Clin Endocrinol Metab. 61:946 –951. 6. Conway GS, Agrawal R, Betteridge DJ, Jacobs HS. 1992 Risk factors for coronary artery disease in lean and obese women with the polycystic ovary syndrome. Clin Endocrinol (Oxf). 37:119 –125. 7. Diamanti-Kandarakis E, Dunaif A. 1996 New perspectives in polycystic ovary syndrome. Trends Endocrinol Metab. 7:267–271. 8. Mahabeer S, Naidoo C, Norman RJ, Jialal I, Reddi K, Joubert SM. 1990 Metabolic profiles and lipoprotein lipid concentrations in non-obese and obese patients with polycystic ovarian disease. Horm Metab Res. 22:537–540. 9. Dahlgren E, Johansson S, Lindsted G, et al. 1992 Women with polycystic ovary syndrome wedge resected in 1956 to 1965: a long-term follow-up focusing on natural history and circulating hormones. Fertil Steril. 57:505–513. 10. Dahlgren E, Janson PO, Johansson S, Lapidus L, Oden A. 1992 Polycystic ovary syndrome and risk for myocardial infarction. Acta Obstet Gynecol Scand. 71:599 – 604. 11. Guzick DS, Talbott EO, Sutton-Tyrrell K, Herzog HC, Kuller LH, Wolfson Jr SK. 1996 Carotid atherosclerosis in women with polycystic ovary syndrome: initial results from a case-control study. Am J Obstet Gynecol. 174:1224 –1229. 12. Wild RA, Applebaum-Bowden D, Demers LM, et al. 1990 Lipoprotein lipids

13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

25. 26. 27. 28.

29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

JCE & M • 1998 Vol 83 • No 8

in women with androgen excess: independent associations with increased insulin and androgens. Clin Chem. 36:283–289. Senoz S, Ozaksit G, Turhan NO, Gulekli B, Gokmen O. 1994 Lipid profiles in women with hirsutism and polycystic ovaries. Gynecol Endocrinol. 8:33–37. Holte J, Bergh T, Berne C, Lithell H. 1994 Serum lipoprotein lipid profile in women with the polycystic ovary syndrome: relation to anthropometric, endocrine and metabolic variables. Clin Endocrinol (Oxf). 41:463– 471. Talbott E, Guzick D, Clerici A, et al. 1995 Coronary heart disease risk factors in women with polycystic ovary syndrome. Arterioscler Thromb Vasc Biol. 15:821– 826. Gordon T, Castelli WP, Hjortland MC, Kannel WB, Dawber TR. 1977 High density lipoprotein as a protective factor against coronary heart disease. The Framingham Study. Am J Med. 62:707–714. Gordon T, Kannel WB, Castelli WP, Dawber TR. 1981 Lipoproteins, cardiovascular disease, and death. The Framingham study. Arch Intern Med. 141:1128 –1131. Graf MJ, Richards CJ, Brown V, Meissner L, Dunaif A. 1990 The independent effects of hyperandrogenaemia, hyperinsulinaemia, and obesity on lipid and lipoprotein profiles in women. Clin Endocrinol (Oxf). 33:119 –131. Robinson S, Henderson AD, Gelding SV, et al. 1996 Dyslipidaemia is associated with insulin resistance in women with polycystic ovaries. Clin Endocrinol (Oxf). 44:277–284. Sorva R, Kuusi T, Dunkel L, Taskinen MR. 1988 Effects of endogenous sex steroids on serum lipoproteins and postheparin plasma lipolytic enzymes. J Clin Endocrinol Metab. 66:408 – 413. Yki-Jarvinen H, Taskinen MR, Koivisto VA, Nikkila EA. 1984 Response of adipose tissue lipoprotein lipase activity and serum lipoproteins to acute hyperinsulinaemia in man. Diabetologia. 27:364 –369. Ek I, Arner P, Bergqvist A, Carlstrom K, Wahrenberg H. 1997 Impaired adipocyte lipolysis in nonobese women with the polycystic ovary syndrome: a possible link to insulin resistance? J Clin Endocrinol Metab. 82:1147–1153. Diamanti-Kandarakis E, Mitrakou A, Hennes MM, et al. 1995 Insulin sensitivity and antiandrogenic therapy in women with polycystic ovary syndrome. Metabolism. 44:525–531. Moghetti P, Tosi F, Castello R, et al. 1996 The insulin resistance in women with hyperandrogenism is partially reversed by antiandrogen treatment: evidence that androgens impair insulin action in women. J Clin Endocrinol Metab. 81:952–960. Pasquali R, Casimirri F, Balestra V, et al. 1991 The relative contribution of androgens and insulin in determining abdominal body fat distribution in premenopausal women. J Endocr Invest. 14:839 – 846. Pasquali R, Casimirri F, Cantobelli S, et al. 1993 Insulin and androgen relationships with abdominal body fat distribution in women with and without hyperandrogenism. Horm Res. 39:179 –187. Seidell JC, Cigolini M, Deurenberg P, Oosterlee A, Doornbos G. 1989 Fat distribution, androgens, and metabolism in nonobese women. Am J Clin Nutr. 50:269 –273. Lovejoy JC, Bray GA, Bourgeois MO, et al. 1996 Exogenous androgens influence body composition and regional body fat distribution in obese postmenopausal women–a clinical research center study. J Clin Endocrinol Metab. 81:2198 –2203. Peiris AN, Mueller RA, Smith GA, Struve MF, Kissebah AH. 1986 Splanchnic insulin metabolism in obesity. Influence of body fat distribution. J Clin Invest. 78:1648 –1657. Evans DJ, Hoffmann RG, Kalkhoff RK, Kissebah AH. 1983 Relationship of androgenic activity to body fat topography, fat cell morphology, and metabolic aberrations in premenopausal women. J Clin Endocrinol Metab. 57:304 –310. Holmang A, Svedberg J, Jennische E, Bjorntorp P. 1990 Effects of testosterone on muscle insulin sensitivity and morphology in female rats. Am J Physiol 259:E555–E560. Holmang A, Larsson B-M, Brzezinska Z, Bjorntorp P. 1992 Effects of shortterm testosterone exposure on insulin sensitivity of muscles in female rats. Am J Physiol 262:E851–E855. Rebuffe-Scrive M, Marin P, Bjorntorp P. 1991 Effect of testosterone on abdominal adipose tissue in men. Int J Obesity. 15:791–795. Bhasin S, Swerdloff RS, Steiner B, et al. 1992 A biodegradable testosterone microcapsule formulation provides uniform eugonadal levels of testosterone for 10 –11 weeks in hypogonadal men. J Clin Endocrinol Metab. 74:75– 83. Zgliczynski S, Ossowski M, Slowinska-Srzednicka J, et al. 1996 Effect of testosterone replacement therapy on lipids and lipoproteins in hypogonadal and elderly men. Atherosclerosis. 121:35– 43. Kissebah AH, Vydelingum N, Murray R, et al. 1982 Relation of body fat distribution to metabolic complications of obesity. J Clin Endocrionol Metab. 54:254 –260. Wysowski DK, Freiman JP, Tourtelot JB, Horton ML. 1993 Fatal and nonfatal hepatotoxicity associated with flutamide. Ann Intern Med. 118:860 – 864. Wallace C, Lalor EA, Chik CL. 1993 Hepatotoxicity complicating flutamide treatment of hirsutism. Ann Intern Med. 119:1150. Diamanti-Kandarakis E, Tolis G, Duleba AJ. 1995 Androgens and therapeutic aspects of antiandrogens in women. J Soc Gynecol Invest. 2:557–592. Fain JN, Garcia-Sainz A. 1983 Adrenergic regulation of adipocyte metabolism. J Lipid Res. 24:945–966.

EFFECT OF FLUTAMIDE ON LIPID PROFILE 41. Eri LM, Urdal P. 1995 Effects of the nonsteroidal antiandrogen casodex on lipoproteins, fibrinogen and plasminogen activator inhibitor in patients with benign prostatic hyperplasia. Eur Urol. 27:274 –279. 42. Wallentin L, Varenhorst E. 1980 Plasma lipoproteins during treatment with cyproterone acetate in men with prostatic carcinoma. J Clin Endocrinol Metab. 51:1118 –1122. 43. Gillatt DA, Bolton CH, Chadwick D, Downs LG, Hopton MI, Gingell JC. 1993 Lipoprotein levels following treatment with cyproterone acetate or LHRH analogues. Br J Urol. 71:728 –730. 44. Balzano S, Cappa M, Migliari R, et al. 1988 The effect of flutamide on basal and ACTH-stimulated plasma levels of adrenal androgens in patients with advanced prostate cancer. J Endocrinol Invest. 11:693– 696. 45. Nestler JE, Powers LP, Matt DW, et al. 1991 A direct effect of hyperinsulinemia on serum sex hormone-binding globulin levels in obese women with the polycystic ovary syndrome. J Clin Endocriol Metab. 72:83– 89. 46. Friedl KE, Jones RE, Hannan CJ, Plymate SR. 1989 The administration of

47.

48.

49.

50.

2705

pharmacological doses of testosterone or 19-nortestosterone to normal men is not associated with increased insulin secretion or impaired glucose tolerance. J Clin Endocrinol Metab. 68:971–977. Brown TR, Greene FE, Bardin CW. 1976 Androgen receptor dependent and independent activities of testosterone on hepatic microsomal drug metabolism. Endocrinology. 99:1353–1362. Winneker RC, Wagner MM, Shaw CJ, Snyder BW. 1989 Antiandrogens do not reverse androgen-induced inhibition of sex hormone-binding globulin levels in adult female rhesus monkeys. Endocrinology. 125:715–719. Cusan L, Dupont A, Gomez J-L, Tremblay RR, Labrie F. 1994 Comparison of flutamide and spironolactone in the treatment of hirsutism: a randomized controlled trial. Fertil Steril. 61:281–287. Pucci E, Genazzani AD, Monzani F, et al. 1995 Prolonged treatment of hirsutism with flutamide alone in patients affected by polycystic ovary syndrome. Gynecol Endocrinol. 9:221–228.