Effect of Ketoconazole on Plasma Sex Hormones, Lipids, Lipoproteins, Apolipoproteins in Hyperandrogenic Women Jorge Cedeno,
Soaira G. Mendoza,
Elsy Velazquez,
Humberto
and
Nucete, James Speirs, and Charles J. Glueck
The aim of the current study of 18 hyperandrogenic women was to determine the affects of ketoconazole (KTZ), an oral synthetic antifungal imidazole derivative that inhibits gonadal and adrenal steroidogenesis, on lipids, lipoprotein cholesterols, apolipoproteins, endogenous sex steroid hormones. and their interactions. Eighteen hyperandrogenic women, ages 18 to 35, with a history of severe acne and/or hirsutism, were randomly divided into two groups of nine, both receiving KTZ (group 1,4.00 mg/d; group 2,800 mg/d) for 10 days. In groups 1 and 2, KTZ therapy reduced cholesterol (10%. P c .Ol 19%. P 5 .05) and low-density lipoprotein (LDL)-cholesterol (13%. P 5 .05; 33%. P s .025). and increased apoplipoprotein (ape) Al (7%. P 5 905; 13%. P 5 .Ol). KTZ, 800 mg/d, decreased apo 6 (21%. P 5 905). and lowered the ratio of LDL-cholesterol to high-density lipoprotein (HDL)-cholesterol (40%. P 5 .Ol 1. KTZ therapy more than doubled the levels of estradiol (E2) in both groups (138%. P 5 .Ol ; 171%. P 5 .Ol 1 and, in the high-dose group, decreased the levels of free testosterone (FT) (48%. P 5 .05) and dehydroepiandrosterone-sulfate (DHEA-S) (38%. P ZG.005). The reductions of total and LDL-cholesterol appear to be attributable to the increases in E2 and possibly to the decrease in FT. KTZ therapy may have beneficial effects on atherogenic lipid and lipoprotein patterns in women with hyperandrogenicity. @ 1990 by W.B. Saunders Company.
;
K
ETOCONAZOLE (KTZ) is an oral synthetic antifungal imidazole derivative.’ The development of gynecomastia in some patients taking the drug prompted study of the effect of KTZ on steroid secretion.* Recently, KTZ was found to be a potent inhibitor of human gonadal’ and adrena14,’ steroidogenesis. It inhibits a wide variety of cytochrome P-450dependent enzymes, including 17,20desmolase that catalyzes conversion of C-21 steroidal precursors to C- 19 sex steroids. Steroidogenesis in the testis is much more sensitive to inhibition by KTZ compared with that in the adrenal, since production of testicular C-19 steroids is profoundly inhibited by KTZ at doses that do not inhibit basal production of cortisol and other C-21 adrenal steroids.3.4 Effects of KTZ on female steroid production are less well defined than for men. There is limited data on the effects of KTZ on plasma lipoprotein metabolism. The drug blocks the conversion on lanosterol to ergosterol in yeast by inhibiting the cytochrome P-45O-dependent step of C-14 demethylation6 Recently, low plasma cholesterol levels have been reported in a patient with chronic myelogenous leukemia receiving KTZ.’ Kraemer and Pant* found a 27% reduction in total serum cholesterol in seven patients with prostatic cancer treated with high doses of KTZ., primarily attributable to a 38% reduction in low-density lipoprotein (LDL)-cholesterol levels without changes in high-density lipoprotein (HDL)-cholesterol levels. Rollman et al9 reported transient reductions in plasma cholesterol levels followed by increased levels in patients receiving an I-month course of oral KTZ. Triglyceride levels also increased in those patients. Both endogenous and exogenous sex hormones play major roles in plasma lipoprotein and apolipoprotein metabolism.“-‘* Patients with polycystic ovary syndrome (PCO) have hyperandrogenism associated with hypertriglyceridemia and low levels of plasma HDL-cholesterol.‘3*‘4 If endogenous sex steroids modulate lipoprotein metabolism in hyperandrogenism,‘3’4 the administration of KTZ might modify plasma cholesterol, triglycerides, LDL-C and HDLC, and apolipoprotein (apo) Al and apo B in hyperandrogenie patients by reducing androgenic steroid levels or by increasing estradiol (E2) levels. The aim of the current study Metabolism,
Vol39,
No 5 (May), 1990: pp 5 1 l-5 17
of 18 hyperandrogenic women was to determine the effects of KTZ on lipids, lipoprotein cholesterols, apolipoproteins, endogenous sex steroid hormones, and their interactions. MATERIALS AND METHODS Patients
Eighteen women, age 18 to 34 years, were selected from the outpatient clinic, Endocrinology Unit, Hospital Universitario of Los Andes, Merida, Venezuela. All of them had a history of severe acne and/or hirsutism, with or without disturbances in menses or obesity, and were thought to have PC0 with hyperandrogenism. The diagnosis of PC0 was based on the presence of three or more of the following criteria: (1) persistent menstrual irregularities; (2) hirsutism or acne; (3) multicystic ovaries on ultrasonographic exploration; (4) luteinizing hormone/follicle-stimulating hormone (LH/FSH) ratio 2 25. Patients with other endocrine disorders were excluded. None of the subjects were receiving any medication. No subjects had clinical evidence of acanthosis nigricans. Height and weight were measured, and the Quetelet index (QI[wt(kg)/ht’ (cm) x lOO]) was calculated (Table 1); measures of peripheral versus truncal obesity were not taken. Signed informed consent was obtained for this 1O-day study. To provide a normal frame of reference for the 18 women with polycystic ovary syndrome, 28 healthy, non-hirsute, non-obese women with regular menses, age 18 to 32 years, were studied between days 7 and 9 of the menstrual cycle, fasting (12 hours) at 8 AM, taking no exogenous sex steroids. Study
Protocol
The PC0 women were randomly divided into two groups of nine, both receiving ketoconazole (Nizoral; Janssen, Piscataway, NJ) for 10 days. Group 1 received 400 mg/d and group 2,800 mg. During the lo-day outpatient treatment program, the patients followed their From the Endocrinology Unit, Department of Internal Medicine, University Hospital of Los Andes, Merida. Venezuela; and the Cholesterol Center, Jewish Hospital, Cincinnati, OH. Supported by CONICITSI-1555, Caracas, Venezuela. and by the Jewish Hospital Medical Research Council. Address reprint requests to Charles J. Glueck. MD, Cholesterol Center, Jewish Hospital, 3200 Burner Ave. Cincinnati. OH 45229. (D1990 by W.B. Saunders Company. 0026-0495/90/3905-001 OU.OOjO
511
512
CEDENO ET AL
Table 1. Patient Characteristics
in the l&Women
With
PC0 Syndrome QI
Age Patient
Weoght
Height
(kg1
(cm)
IV)
([wt/ht’ bll)]
x 100)
Group I 20
59.0
161
2.28
27
50.0
158
2.00
23
62.8
152
2.72
19
53.0
158
2.13
21
78.7
183
2.96
antigen for ape B assay was prepared in our laboratory according to Albers et a1.l9 The antigen used was LDL isolated from normal human plasma using sequential density preparative ultracentrifugation as described previously.” The protein concentration of “narrowcut” LDL was used as the primary standard concentration of apo B. Secondary standards were calibrated against the primary standard. Rocket heights between 2 and 5 cm were obtained using 2% rabbit antiserum in the gel buffered as described by Laurel!.” Samples were diluted to give rocket heights in the rectilinear part of the
standard curve. An aliquot of pooled plasma was included in each assay for quality control. Plasma apt B was measured in the patients in whole plasma. The electroimmunassays for apo A! measurements were developed along similar lines as for apo B, with specific modifications as required for these apoproteins.
18
60.3
156
2.48
20
56.2
153
2.40
8
27
61.3
164
2.28
9
20
48.6
164
2.05
Statistical Methods
18
40.5
151
1.78
26
63.0
158
2.52
After determination that the data were normally distributed (Shapiro-Wilk test), thus allowing use of parametric statistics, effects of the KTZ intervention (baseline v day 10) were assessed using Student’s paired t test. Pearson’s correlations were used to assess relationships between (KTZ induced) changes in endogenous sex steroids, and in lipids, lipoprotein cholesterols, and apolipoproteins. Clinical measures in the I8 PC0 women and 28 normal women were compared by Student’s t test.
Group I/
31
79.5
164
2.96
19
57.0
151
2.50
21
63.3
164
2.35
19
77.0
170
2.66
20
57.0
180
2.23
22
80.6
174
2.66
34
63.4
181
2.45 RESULTS
Mean k SEM
22.5
f 1.1
61.7
+ 2.6
160 k
1
2.411
+ 0.75
usual diets, exercise,
smoking, and alcohol intake patterns. Before and during the IO-day KTZ trial, the patients took no exogenous sex steroids, or other drugs which might affect lipoprotein metabolism. All blood samples were withdrawn before and after the lo-day trial of KTZ (at 8 AM after a 12-hour fast) for determination of lipids, lipoproteins, apoproteins, and the following sex hormones: free testosterone (FT), dehydroepiandrosterone-sulfate (DHEA-S), A4-androstenedione (A4-And), and E2. The basal evaluations were performed between days 7 and 9 of the menstrual cycle, or at any time if amenorrhea was present. An additional blood sample was obtained 4 hours after the initial KTZ dose in five patients of each group (groups 1-A and 2-A, respectively), in order to evaluate the acute effects of the drug. Fasting insulin and insulin response to glucose challenge were not measured, nor, given the short IO-day trial of KTZ, were plasma cortisol levels or tracsaminase levels. FT and A4-And were not measured in the normal female control subjects.
Measurements of Hormones, Lipids, Lipoproteins, and Apoproteins Blood samples were collected in tubes containing EDTA (1 mg/mL) and in plain tubes. Plasma was used for quantitation of lipids and lipoproteins. Serum was stored at -20% until hormones and apoproteins were assayed. FT, A4-And, DHEA-S, and E2 were measured by radioimmunoassay using commercial kits (Diagnostic Products, Los Angeles, CA), with an interassay variation of less than 8% and an intraassay variation of less than 5%. Total cholesterol and triglycerides were measured by enzymatic procedures using a bichromatic analyzer, Abbott ABA-50, and reagents from Boehringer-Mannheim GmbH, FRG. HDL-C was analyzed by heparin-manganese precipitation.‘5 LDL-C was estimated using the formula of Friedewald et al.lb Ape B and apo Al were measured by electroimmunoassay following the method of Laurell.” The antigens for ape Al assays, obtained by isoelectrofocusing and column chromatography,” were donated by M.L. Kashyap, Lipid Research Center, University of Cincinnati, OH. The
Clinical Measures: Women With PC0 and Normal Women Table I displays age, weight, height, and QI for the 18 PC0 women, who had higher relative ponderosity (QI) than the normal subjects, substantially higher apo B and DHEA-S levels, and much lower E2 (Table 2). Lipids, Lipoprotein Cholesterols, Apolipoproteins; Baseline and After Treatment With KTZ Group 1 (on 400 mg KTZ/d) had a significant (17 mg/dL; 10%) decrease in total cholesterol (164 v 147, P 5 .O!) and a iS mg/dL (13%) decrease in LDL-C (96 v 81, P 5 0.05) (Table 3). Apo Al was increased on therapy (13 1 v 140 mg/dL; 7%; P 5 .005) (Table 3). No significant changes were found in triglycerides, HDL-C, the LDL-C/HDL-C ratio, and apo B levels (Table 3). The higher dose of KTZ (800 mg/d) produced similar, but larger changes in lipids, lipoprotein cholesterols, and apoliTable 2. Clinical Measures,
Women With PC0 Syndrome and
Normal Women
(mean + SEMI
PC0 Women (n =
Age (yr) 01
18)
23 + 1 2.41
k 0.75’
Cholesterol (mg/dL)
173 k 7
Triglyceride (mg/dL)
112 + 9
Normal Women (n =
28)
25 + 1 2.23
+ 0.47
172 ? 7 92 + 5.7
HDL-C (mg/dL)
46 + 3
49 * 2
LDL-C tmg/dL)
105 * 7
103 k 7
Apo A 1 (mg/dL)
134 + 4*
119 + 4
109 + 6t
87 + 4
APO
8 (mg/dL)
E2 (pg/mL) DHEA-S (Ng/mL) lP 5 .05. tps.01. $P 5 .005.
44 + 47 444 -c 61$
78 k 12 234 + 29
KETOCONAZOLE.
SEX HORMONES,
LIPOPROTEINS
poproteins (Table 4). Total cholesterol decreased 39 mg/dL (19%) (182 v 143 mg/dL, P 5 .OS), and LDL-C fell 42 mg/dL (33%) (114 v 72 mg/dL, P 5 .025) (Table 4). On therapy, the LDL-C/HDL-C ratio decreased 40% (2.85 v 1.62, P:s .Ol), and apo B fell 21% (120 v 94 mg/dL, P I ,005). Triglyceride levels fell slightly, but not significantly. Apo Al increased on therapy (137 v 154 mg/dL, P 5 .Ol). A nonsignificant increase was observed in HDL-C levels. Endogenous Sex Steroid Hormones, Baseline and After Treatment With KTZ
Tables 5 through 7 summarize changes in endogenous sex steroid hormones on the two therapies with KTZ. Endogenous FT decreased during therapy on both KTZ doses; the 48% reduction (15 v 4 pg/mL) was significant (P I .05) in group 2 on 800 mg KTZ (Tables 5 and 6). On both the 400and 800-mg doses, E2 levels more than doubled, with the increases being 136% (P 5 .Ol) and 171% (P 5 .Ol), respectively (Tables 5 and 6). On KTZ therapy, DHEA-S levels decreased in both groups, with the decrease (36%) being significant (P 5 .005) on KTZ 800 mg/d (Tables 5 and 6). Androstenedione levels fell, but not significantly, on KTZ therapy, although the decrease was somewhat greater on the 800-mg dose (Tables 5 and 6). Four hours after KTZ, FT decreased in both groups, significantly in group 2-A (31%, P 5 .05) (Table 7). E2 increased in both groups, significantly in group 1-A (48%, P 5 .02S) (Table 7). DHEA-S decreased in group 2-A by 12% (P 5 .05) (Table 7). Relationships Between Changes in Endogenous Sex Hormone Steroids and Lipids, Lipoproteins, and Apofipoproteins
There were several significant associations between changes in endogenous sex steroid hormones and changes in lipids, lipoproteins, and apolipoproteins. On KTZ, 800 mg/d, as E2 levels increased, total cholesterol (r = -.72, P < .05) and LDL-C (r = - .70, P -c .05) levels decreased. On KTZ, 800 mg/d, the greater the decrease in A4-And, the greater was the decrease in total cholesterol (r = .75, P c .025) and LDL-C (r = .80, P -c .025). On KT’Z, 400 mg/d, the greater the decrease in DHEA-S, the smaller the decrease in plasma cholesterol (r = -.69, P < ,051. On KTZ 400 mg/d, the larger the decrease in FT, the larger the decrease in the ratio of LDL-C/HDL-C (r = .75, P < .05). DISCUSSION
In the 18 hyperandrogenic women, KTZ therapy uniformly and significantly elevated E2 levels by 136% (on 400 mg/d), and 17 1% (on 800 mg/d), and reduced mean levels of FT and DHEA-S, with these latter changes being significant in the upper-dose group. Although androstenedione levels fell in both groups, these changes were not significant. These changes in endogenous sex steroid hormones were temporally associated with significant reductions of total cholesterol and LDL-C and significant increases in apo Al in both KTZ
513
groups. There were also significant decreases in the ratio of LDL-C to HDL-C and reductions in apo B on KTZ, 800 mg/d. Thus, these results confirm earlier studies,* indicating that the predominant reduction in plasma total cholesterol was due to a reduction in LDL-C. KTZ reduced gonadal and adrenal steroidal genesis in a fashion comparable to that reported previously.3-5 The apparent primary mechanism of action of KTZ in the current study appears to be through its increase in endogenous E2, which was associated with a decrease in both total cholesterol and LDL-C. On 800 mg KTZ/d, as E2 levels increased, total and LDL-C levels decreased. Also, on 800 mg KTZ/d, as ACAnd decreased, total and LDL-C both decreased. On 400 mg KTZ/d, the larger the decrease in FT, the larger the decrease of the ratio of LDL-C to HDL-C. The increase in endogenous E2 may be due to an increase in aromatase activity, which could be directly affected by KTZ. Qualitatively similar to the cholesterol and LDL-C lowering effects of increased endogenous estrogens, as in the current study, administration of exogenous estrogens to normal postmenopausal women will usually reduce total cholesterol and LDL-C.*’ Since endogenous FT has previously been shown to be positively associated with LDLC,‘“,‘3,‘4 it is also not surprising that the current study revealed that as FT decreased, the ratio of LDL-C to HDL-C also decreased. Although the lowest baseline total cholesterol, lowest triglyceride, and lowest LDL-C in group 2 in the current study was in subject 5 who also had the highest DHEA-S level, on KTZ therapy both DHEA-S and cholesterol levels decreased; the greater the reduction in DHEA-S, the smaller was the decrease in plasma cholesterol. In contrast, when exogenous DHEA-S was given to normal men, producing a 2.5- to 3.5-fold increase in serum DHEA-S, total cholesterol and LDL-C decreased.*’ In the PC0 syndrome, obesity, hyperinsulinemia, insulin resistance, and a positive correlation between hyperinsulinism and hyperandrogenism have been reported.22.23 It has been suggested that there may be a causal relationship between hyperinsulinemia, hypertriglyceridemia, and low HDL-C levels.24.25Separate from the current report, we have recently studied 21 obese and 27 non-obese women with the PC0 syndrome26 and 28 normal women. Maximal insulin response to an oral glucose load was significantly increased in both PC0 groups. Both maximal insulin response and basal insulin levels positively correlated with triglycerides. Basal insulin levels were inversely correlated with HDL-C (r = -.41, P < .005).26 However, no significant correlations between endogenous serum androgens and insulin response were found. It is possible that insulin resistance is a primary event in PC0 patients and may be unrelated to ovarian hyperandrogenism.26 Hyperinsulinism persists despite inhibition of ovarian steroid secretion with a long-acting gonadotrophin-releasing hormone agonist,*’ and insulin resistance remains after removal of the ovaries in women with hyperthecosis even though androgen levels return to normal.28 On the other hand, insulin binds to specific receptors in ovarian tissue,29 stimulates ovarian androgen production,30 and results in menstrual disturbances and hirsutism.
161
129
140
158
124
160
190
136
147t + 7.4
160
156
160
165
140
162
203
185
164 r 8.5
zk2.9
123
144
-10
10 Days
Base
84 + 8.9
5
.Ol.
.005
,025.
tP
.Ol.
,005.
SPS
§Pc
5
.05.
49 i:4.3
69
61
55
35
39
45
39
61
35
10 Days
-2.2 + 6.1
50 k 5.9
76
47
71
27
36
41
44
69
36
BaS.2
HDL-C
81' + 6.5
57
118
83
74
106
69
69
84
70
10 Days
-13.4 f 6.9
96 ir7.2
96
142
71
101
110
91
91
69
91
BBS.2
LDL-C
140$ + 7.4
131 t 6.2 6.8 * 1.9
157
147
184
128
132
144
128
136
104
10 Days
147
140
168
128
124
132
112
120
108
BS.3
ApoAl
2.21 -t0.33
93 f 5.5
133
99
187
138
123
152
141
115
181
195
204
202
81
106
-19
+ 9.8
133 + 13.9 112 + 25
63
111
119
125
193
57
162
85
303
so
106
158
167 197
108
134
117
73
10 Days
248
TG
95
Base
182
10 Days
183
TC
55 105
175 110 75 48
40 35 22 59
27 50 52
128
58 48
34
17.6 + 10.7
42 k 3.1 48 t 4.5
88
52
-33.4
r 8.3
13 i 4.5
154$ + 10.4 137 z+4.3
137
157
137
168
148
152
156
176
10 Days
157
ApoAl
152
127
113
137
99 58
148
140
152
132
136
Bat%3
29
114 + 12.6 72t f 9.9
146
58
53
97 112
63
44
36
46
50
114
131
33
53
LDL-C 10 Days
IO Days
BEiS
HOL-C Base
Apo B
144
110
77
124
75
96
69
80
69
10 Days
2.42
4.29
2.55
1.87
0.96
2.78
3.06
3.80
3.97
-21
+ 3.6
1.83
0.90
1.00
1.57
0.49
2.27
3.00
1.38
2.15
10 Days
-39.8 +_8.7
1.62$ f 0.26
LDL-C/HDL-C BaSO
120 + 10.4 943 t 8.9 2.85 + 0.36
172
158
131
137
96
107
89
so
96
BaS8
1.75 + 0.18
0.83
1.93
1.51
2.11
2.72
1.53
1.77
1.38
2.00
10 Days
-11.4 zk 11.1
1.26
3.02
1.00
3.74
3.06
2.22
2.07
1.00
2.53
Base
LDL-W-IDL-C
124
115
90
96
83
96
82
77
78
10 Days
-3.9 f 3.7
98 + 6.5
115
119
95
103
101
125
77
78
72
BaM,
Apo B
Table 4. Plasma Lipids, Lipoproteins, and Apoproteins (mg/dL) and Percent Change Pre- and Posttreatment With Ketoconazol, 800 ma/d, Grouo 2
182 + 12.8 143* 2 9.6 -19 t 64
'Ps
Mean 2 SE A%
iP
Soaira G. Mendoza,
Elsy Velazquez,
Humberto
and
Nucete, James Speirs, and Charles J. Glueck
The aim of the current study of 18 hyperandrogenic women was to determine the affects of ketoconazole (KTZ), an oral synthetic antifungal imidazole derivative that inhibits gonadal and adrenal steroidogenesis, on lipids, lipoprotein cholesterols, apolipoproteins, endogenous sex steroid hormones. and their interactions. Eighteen hyperandrogenic women, ages 18 to 35, with a history of severe acne and/or hirsutism, were randomly divided into two groups of nine, both receiving KTZ (group 1,4.00 mg/d; group 2,800 mg/d) for 10 days. In groups 1 and 2, KTZ therapy reduced cholesterol (10%. P c .Ol 19%. P 5 .05) and low-density lipoprotein (LDL)-cholesterol (13%. P 5 .05; 33%. P s .025). and increased apoplipoprotein (ape) Al (7%. P 5 905; 13%. P 5 .Ol). KTZ, 800 mg/d, decreased apo 6 (21%. P 5 905). and lowered the ratio of LDL-cholesterol to high-density lipoprotein (HDL)-cholesterol (40%. P 5 .Ol 1. KTZ therapy more than doubled the levels of estradiol (E2) in both groups (138%. P 5 .Ol ; 171%. P 5 .Ol 1 and, in the high-dose group, decreased the levels of free testosterone (FT) (48%. P 5 .05) and dehydroepiandrosterone-sulfate (DHEA-S) (38%. P ZG.005). The reductions of total and LDL-cholesterol appear to be attributable to the increases in E2 and possibly to the decrease in FT. KTZ therapy may have beneficial effects on atherogenic lipid and lipoprotein patterns in women with hyperandrogenicity. @ 1990 by W.B. Saunders Company.
;
K
ETOCONAZOLE (KTZ) is an oral synthetic antifungal imidazole derivative.’ The development of gynecomastia in some patients taking the drug prompted study of the effect of KTZ on steroid secretion.* Recently, KTZ was found to be a potent inhibitor of human gonadal’ and adrena14,’ steroidogenesis. It inhibits a wide variety of cytochrome P-450dependent enzymes, including 17,20desmolase that catalyzes conversion of C-21 steroidal precursors to C- 19 sex steroids. Steroidogenesis in the testis is much more sensitive to inhibition by KTZ compared with that in the adrenal, since production of testicular C-19 steroids is profoundly inhibited by KTZ at doses that do not inhibit basal production of cortisol and other C-21 adrenal steroids.3.4 Effects of KTZ on female steroid production are less well defined than for men. There is limited data on the effects of KTZ on plasma lipoprotein metabolism. The drug blocks the conversion on lanosterol to ergosterol in yeast by inhibiting the cytochrome P-45O-dependent step of C-14 demethylation6 Recently, low plasma cholesterol levels have been reported in a patient with chronic myelogenous leukemia receiving KTZ.’ Kraemer and Pant* found a 27% reduction in total serum cholesterol in seven patients with prostatic cancer treated with high doses of KTZ., primarily attributable to a 38% reduction in low-density lipoprotein (LDL)-cholesterol levels without changes in high-density lipoprotein (HDL)-cholesterol levels. Rollman et al9 reported transient reductions in plasma cholesterol levels followed by increased levels in patients receiving an I-month course of oral KTZ. Triglyceride levels also increased in those patients. Both endogenous and exogenous sex hormones play major roles in plasma lipoprotein and apolipoprotein metabolism.“-‘* Patients with polycystic ovary syndrome (PCO) have hyperandrogenism associated with hypertriglyceridemia and low levels of plasma HDL-cholesterol.‘3*‘4 If endogenous sex steroids modulate lipoprotein metabolism in hyperandrogenism,‘3’4 the administration of KTZ might modify plasma cholesterol, triglycerides, LDL-C and HDLC, and apolipoprotein (apo) Al and apo B in hyperandrogenie patients by reducing androgenic steroid levels or by increasing estradiol (E2) levels. The aim of the current study Metabolism,
Vol39,
No 5 (May), 1990: pp 5 1 l-5 17
of 18 hyperandrogenic women was to determine the effects of KTZ on lipids, lipoprotein cholesterols, apolipoproteins, endogenous sex steroid hormones, and their interactions. MATERIALS AND METHODS Patients
Eighteen women, age 18 to 34 years, were selected from the outpatient clinic, Endocrinology Unit, Hospital Universitario of Los Andes, Merida, Venezuela. All of them had a history of severe acne and/or hirsutism, with or without disturbances in menses or obesity, and were thought to have PC0 with hyperandrogenism. The diagnosis of PC0 was based on the presence of three or more of the following criteria: (1) persistent menstrual irregularities; (2) hirsutism or acne; (3) multicystic ovaries on ultrasonographic exploration; (4) luteinizing hormone/follicle-stimulating hormone (LH/FSH) ratio 2 25. Patients with other endocrine disorders were excluded. None of the subjects were receiving any medication. No subjects had clinical evidence of acanthosis nigricans. Height and weight were measured, and the Quetelet index (QI[wt(kg)/ht’ (cm) x lOO]) was calculated (Table 1); measures of peripheral versus truncal obesity were not taken. Signed informed consent was obtained for this 1O-day study. To provide a normal frame of reference for the 18 women with polycystic ovary syndrome, 28 healthy, non-hirsute, non-obese women with regular menses, age 18 to 32 years, were studied between days 7 and 9 of the menstrual cycle, fasting (12 hours) at 8 AM, taking no exogenous sex steroids. Study
Protocol
The PC0 women were randomly divided into two groups of nine, both receiving ketoconazole (Nizoral; Janssen, Piscataway, NJ) for 10 days. Group 1 received 400 mg/d and group 2,800 mg. During the lo-day outpatient treatment program, the patients followed their From the Endocrinology Unit, Department of Internal Medicine, University Hospital of Los Andes, Merida. Venezuela; and the Cholesterol Center, Jewish Hospital, Cincinnati, OH. Supported by CONICITSI-1555, Caracas, Venezuela. and by the Jewish Hospital Medical Research Council. Address reprint requests to Charles J. Glueck. MD, Cholesterol Center, Jewish Hospital, 3200 Burner Ave. Cincinnati. OH 45229. (D1990 by W.B. Saunders Company. 0026-0495/90/3905-001 OU.OOjO
511
512
CEDENO ET AL
Table 1. Patient Characteristics
in the l&Women
With
PC0 Syndrome QI
Age Patient
Weoght
Height
(kg1
(cm)
IV)
([wt/ht’ bll)]
x 100)
Group I 20
59.0
161
2.28
27
50.0
158
2.00
23
62.8
152
2.72
19
53.0
158
2.13
21
78.7
183
2.96
antigen for ape B assay was prepared in our laboratory according to Albers et a1.l9 The antigen used was LDL isolated from normal human plasma using sequential density preparative ultracentrifugation as described previously.” The protein concentration of “narrowcut” LDL was used as the primary standard concentration of apo B. Secondary standards were calibrated against the primary standard. Rocket heights between 2 and 5 cm were obtained using 2% rabbit antiserum in the gel buffered as described by Laurel!.” Samples were diluted to give rocket heights in the rectilinear part of the
standard curve. An aliquot of pooled plasma was included in each assay for quality control. Plasma apt B was measured in the patients in whole plasma. The electroimmunassays for apo A! measurements were developed along similar lines as for apo B, with specific modifications as required for these apoproteins.
18
60.3
156
2.48
20
56.2
153
2.40
8
27
61.3
164
2.28
9
20
48.6
164
2.05
Statistical Methods
18
40.5
151
1.78
26
63.0
158
2.52
After determination that the data were normally distributed (Shapiro-Wilk test), thus allowing use of parametric statistics, effects of the KTZ intervention (baseline v day 10) were assessed using Student’s paired t test. Pearson’s correlations were used to assess relationships between (KTZ induced) changes in endogenous sex steroids, and in lipids, lipoprotein cholesterols, and apolipoproteins. Clinical measures in the I8 PC0 women and 28 normal women were compared by Student’s t test.
Group I/
31
79.5
164
2.96
19
57.0
151
2.50
21
63.3
164
2.35
19
77.0
170
2.66
20
57.0
180
2.23
22
80.6
174
2.66
34
63.4
181
2.45 RESULTS
Mean k SEM
22.5
f 1.1
61.7
+ 2.6
160 k
1
2.411
+ 0.75
usual diets, exercise,
smoking, and alcohol intake patterns. Before and during the IO-day KTZ trial, the patients took no exogenous sex steroids, or other drugs which might affect lipoprotein metabolism. All blood samples were withdrawn before and after the lo-day trial of KTZ (at 8 AM after a 12-hour fast) for determination of lipids, lipoproteins, apoproteins, and the following sex hormones: free testosterone (FT), dehydroepiandrosterone-sulfate (DHEA-S), A4-androstenedione (A4-And), and E2. The basal evaluations were performed between days 7 and 9 of the menstrual cycle, or at any time if amenorrhea was present. An additional blood sample was obtained 4 hours after the initial KTZ dose in five patients of each group (groups 1-A and 2-A, respectively), in order to evaluate the acute effects of the drug. Fasting insulin and insulin response to glucose challenge were not measured, nor, given the short IO-day trial of KTZ, were plasma cortisol levels or tracsaminase levels. FT and A4-And were not measured in the normal female control subjects.
Measurements of Hormones, Lipids, Lipoproteins, and Apoproteins Blood samples were collected in tubes containing EDTA (1 mg/mL) and in plain tubes. Plasma was used for quantitation of lipids and lipoproteins. Serum was stored at -20% until hormones and apoproteins were assayed. FT, A4-And, DHEA-S, and E2 were measured by radioimmunoassay using commercial kits (Diagnostic Products, Los Angeles, CA), with an interassay variation of less than 8% and an intraassay variation of less than 5%. Total cholesterol and triglycerides were measured by enzymatic procedures using a bichromatic analyzer, Abbott ABA-50, and reagents from Boehringer-Mannheim GmbH, FRG. HDL-C was analyzed by heparin-manganese precipitation.‘5 LDL-C was estimated using the formula of Friedewald et al.lb Ape B and apo Al were measured by electroimmunoassay following the method of Laurell.” The antigens for ape Al assays, obtained by isoelectrofocusing and column chromatography,” were donated by M.L. Kashyap, Lipid Research Center, University of Cincinnati, OH. The
Clinical Measures: Women With PC0 and Normal Women Table I displays age, weight, height, and QI for the 18 PC0 women, who had higher relative ponderosity (QI) than the normal subjects, substantially higher apo B and DHEA-S levels, and much lower E2 (Table 2). Lipids, Lipoprotein Cholesterols, Apolipoproteins; Baseline and After Treatment With KTZ Group 1 (on 400 mg KTZ/d) had a significant (17 mg/dL; 10%) decrease in total cholesterol (164 v 147, P 5 .O!) and a iS mg/dL (13%) decrease in LDL-C (96 v 81, P 5 0.05) (Table 3). Apo Al was increased on therapy (13 1 v 140 mg/dL; 7%; P 5 .005) (Table 3). No significant changes were found in triglycerides, HDL-C, the LDL-C/HDL-C ratio, and apo B levels (Table 3). The higher dose of KTZ (800 mg/d) produced similar, but larger changes in lipids, lipoprotein cholesterols, and apoliTable 2. Clinical Measures,
Women With PC0 Syndrome and
Normal Women
(mean + SEMI
PC0 Women (n =
Age (yr) 01
18)
23 + 1 2.41
k 0.75’
Cholesterol (mg/dL)
173 k 7
Triglyceride (mg/dL)
112 + 9
Normal Women (n =
28)
25 + 1 2.23
+ 0.47
172 ? 7 92 + 5.7
HDL-C (mg/dL)
46 + 3
49 * 2
LDL-C tmg/dL)
105 * 7
103 k 7
Apo A 1 (mg/dL)
134 + 4*
119 + 4
109 + 6t
87 + 4
APO
8 (mg/dL)
E2 (pg/mL) DHEA-S (Ng/mL) lP 5 .05. tps.01. $P 5 .005.
44 + 47 444 -c 61$
78 k 12 234 + 29
KETOCONAZOLE.
SEX HORMONES,
LIPOPROTEINS
poproteins (Table 4). Total cholesterol decreased 39 mg/dL (19%) (182 v 143 mg/dL, P 5 .OS), and LDL-C fell 42 mg/dL (33%) (114 v 72 mg/dL, P 5 .025) (Table 4). On therapy, the LDL-C/HDL-C ratio decreased 40% (2.85 v 1.62, P:s .Ol), and apo B fell 21% (120 v 94 mg/dL, P I ,005). Triglyceride levels fell slightly, but not significantly. Apo Al increased on therapy (137 v 154 mg/dL, P 5 .Ol). A nonsignificant increase was observed in HDL-C levels. Endogenous Sex Steroid Hormones, Baseline and After Treatment With KTZ
Tables 5 through 7 summarize changes in endogenous sex steroid hormones on the two therapies with KTZ. Endogenous FT decreased during therapy on both KTZ doses; the 48% reduction (15 v 4 pg/mL) was significant (P I .05) in group 2 on 800 mg KTZ (Tables 5 and 6). On both the 400and 800-mg doses, E2 levels more than doubled, with the increases being 136% (P 5 .Ol) and 171% (P 5 .Ol), respectively (Tables 5 and 6). On KTZ therapy, DHEA-S levels decreased in both groups, with the decrease (36%) being significant (P 5 .005) on KTZ 800 mg/d (Tables 5 and 6). Androstenedione levels fell, but not significantly, on KTZ therapy, although the decrease was somewhat greater on the 800-mg dose (Tables 5 and 6). Four hours after KTZ, FT decreased in both groups, significantly in group 2-A (31%, P 5 .05) (Table 7). E2 increased in both groups, significantly in group 1-A (48%, P 5 .02S) (Table 7). DHEA-S decreased in group 2-A by 12% (P 5 .05) (Table 7). Relationships Between Changes in Endogenous Sex Hormone Steroids and Lipids, Lipoproteins, and Apofipoproteins
There were several significant associations between changes in endogenous sex steroid hormones and changes in lipids, lipoproteins, and apolipoproteins. On KTZ, 800 mg/d, as E2 levels increased, total cholesterol (r = -.72, P < .05) and LDL-C (r = - .70, P -c .05) levels decreased. On KTZ, 800 mg/d, the greater the decrease in A4-And, the greater was the decrease in total cholesterol (r = .75, P c .025) and LDL-C (r = .80, P -c .025). On KT’Z, 400 mg/d, the greater the decrease in DHEA-S, the smaller the decrease in plasma cholesterol (r = -.69, P < ,051. On KTZ 400 mg/d, the larger the decrease in FT, the larger the decrease in the ratio of LDL-C/HDL-C (r = .75, P < .05). DISCUSSION
In the 18 hyperandrogenic women, KTZ therapy uniformly and significantly elevated E2 levels by 136% (on 400 mg/d), and 17 1% (on 800 mg/d), and reduced mean levels of FT and DHEA-S, with these latter changes being significant in the upper-dose group. Although androstenedione levels fell in both groups, these changes were not significant. These changes in endogenous sex steroid hormones were temporally associated with significant reductions of total cholesterol and LDL-C and significant increases in apo Al in both KTZ
513
groups. There were also significant decreases in the ratio of LDL-C to HDL-C and reductions in apo B on KTZ, 800 mg/d. Thus, these results confirm earlier studies,* indicating that the predominant reduction in plasma total cholesterol was due to a reduction in LDL-C. KTZ reduced gonadal and adrenal steroidal genesis in a fashion comparable to that reported previously.3-5 The apparent primary mechanism of action of KTZ in the current study appears to be through its increase in endogenous E2, which was associated with a decrease in both total cholesterol and LDL-C. On 800 mg KTZ/d, as E2 levels increased, total and LDL-C levels decreased. Also, on 800 mg KTZ/d, as ACAnd decreased, total and LDL-C both decreased. On 400 mg KTZ/d, the larger the decrease in FT, the larger the decrease of the ratio of LDL-C to HDL-C. The increase in endogenous E2 may be due to an increase in aromatase activity, which could be directly affected by KTZ. Qualitatively similar to the cholesterol and LDL-C lowering effects of increased endogenous estrogens, as in the current study, administration of exogenous estrogens to normal postmenopausal women will usually reduce total cholesterol and LDL-C.*’ Since endogenous FT has previously been shown to be positively associated with LDLC,‘“,‘3,‘4 it is also not surprising that the current study revealed that as FT decreased, the ratio of LDL-C to HDL-C also decreased. Although the lowest baseline total cholesterol, lowest triglyceride, and lowest LDL-C in group 2 in the current study was in subject 5 who also had the highest DHEA-S level, on KTZ therapy both DHEA-S and cholesterol levels decreased; the greater the reduction in DHEA-S, the smaller was the decrease in plasma cholesterol. In contrast, when exogenous DHEA-S was given to normal men, producing a 2.5- to 3.5-fold increase in serum DHEA-S, total cholesterol and LDL-C decreased.*’ In the PC0 syndrome, obesity, hyperinsulinemia, insulin resistance, and a positive correlation between hyperinsulinism and hyperandrogenism have been reported.22.23 It has been suggested that there may be a causal relationship between hyperinsulinemia, hypertriglyceridemia, and low HDL-C levels.24.25Separate from the current report, we have recently studied 21 obese and 27 non-obese women with the PC0 syndrome26 and 28 normal women. Maximal insulin response to an oral glucose load was significantly increased in both PC0 groups. Both maximal insulin response and basal insulin levels positively correlated with triglycerides. Basal insulin levels were inversely correlated with HDL-C (r = -.41, P < .005).26 However, no significant correlations between endogenous serum androgens and insulin response were found. It is possible that insulin resistance is a primary event in PC0 patients and may be unrelated to ovarian hyperandrogenism.26 Hyperinsulinism persists despite inhibition of ovarian steroid secretion with a long-acting gonadotrophin-releasing hormone agonist,*’ and insulin resistance remains after removal of the ovaries in women with hyperthecosis even though androgen levels return to normal.28 On the other hand, insulin binds to specific receptors in ovarian tissue,29 stimulates ovarian androgen production,30 and results in menstrual disturbances and hirsutism.
161
129
140
158
124
160
190
136
147t + 7.4
160
156
160
165
140
162
203
185
164 r 8.5
zk2.9
123
144
-10
10 Days
Base
84 + 8.9
5
.Ol.
.005
,025.
tP
.Ol.
,005.
SPS
§Pc
5
.05.
49 i:4.3
69
61
55
35
39
45
39
61
35
10 Days
-2.2 + 6.1
50 k 5.9
76
47
71
27
36
41
44
69
36
BaS.2
HDL-C
81' + 6.5
57
118
83
74
106
69
69
84
70
10 Days
-13.4 f 6.9
96 ir7.2
96
142
71
101
110
91
91
69
91
BBS.2
LDL-C
140$ + 7.4
131 t 6.2 6.8 * 1.9
157
147
184
128
132
144
128
136
104
10 Days
147
140
168
128
124
132
112
120
108
BS.3
ApoAl
2.21 -t0.33
93 f 5.5
133
99
187
138
123
152
141
115
181
195
204
202
81
106
-19
+ 9.8
133 + 13.9 112 + 25
63
111
119
125
193
57
162
85
303
so
106
158
167 197
108
134
117
73
10 Days
248
TG
95
Base
182
10 Days
183
TC
55 105
175 110 75 48
40 35 22 59
27 50 52
128
58 48
34
17.6 + 10.7
42 k 3.1 48 t 4.5
88
52
-33.4
r 8.3
13 i 4.5
154$ + 10.4 137 z+4.3
137
157
137
168
148
152
156
176
10 Days
157
ApoAl
152
127
113
137
99 58
148
140
152
132
136
Bat%3
29
114 + 12.6 72t f 9.9
146
58
53
97 112
63
44
36
46
50
114
131
33
53
LDL-C 10 Days
IO Days
BEiS
HOL-C Base
Apo B
144
110
77
124
75
96
69
80
69
10 Days
2.42
4.29
2.55
1.87
0.96
2.78
3.06
3.80
3.97
-21
+ 3.6
1.83
0.90
1.00
1.57
0.49
2.27
3.00
1.38
2.15
10 Days
-39.8 +_8.7
1.62$ f 0.26
LDL-C/HDL-C BaSO
120 + 10.4 943 t 8.9 2.85 + 0.36
172
158
131
137
96
107
89
so
96
BaS8
1.75 + 0.18
0.83
1.93
1.51
2.11
2.72
1.53
1.77
1.38
2.00
10 Days
-11.4 zk 11.1
1.26
3.02
1.00
3.74
3.06
2.22
2.07
1.00
2.53
Base
LDL-W-IDL-C
124
115
90
96
83
96
82
77
78
10 Days
-3.9 f 3.7
98 + 6.5
115
119
95
103
101
125
77
78
72
BaM,
Apo B
Table 4. Plasma Lipids, Lipoproteins, and Apoproteins (mg/dL) and Percent Change Pre- and Posttreatment With Ketoconazol, 800 ma/d, Grouo 2
182 + 12.8 143* 2 9.6 -19 t 64
'Ps
Mean 2 SE A%
iP
