UDK 618.213:616-008.842.6 Izvorni z~zanstvewi E h k Original Scien.t$c Paper
SERUM ANDROGENS IN WOMEN WITH HIRSUTISM
Hormone laboratol-y - Iustitutc for Clinical Cl~cmistry a d Clinical Biochemistry, Njcg-oSeva 4, Mcdical Ccnter Ljubljaiia, Slovcnia
' Clinic Pol. Endocri~iolog): ZaloSka 7, Mcdical Center Ljubljana, Slovenia
S U M M A R Y - The scope of the rum%. was lo estnblish the correlation between thx degree of hirsziitinn nnd serum levels o j testoste?-me (T), free testosterone (JT), mzdmster~etiione (A), dehyd,aepiandroste?.one sulphate (DFIEA-S), androstanediol glucwronide (ADIOL-G), sex horrn0n.e binding globu,lin (SHBG) and salivary test- osloane (saT). The body moss index (BIMI) rum a.lso taken into a.ccow~t. Serum S H B G mdJT levels showed the highest correht im with the degree of himitism, while the cot-rehliw~ for serum ?; saT and ADIOL-G was a little loruer, but greater th,ar~ for A and DFIEA-S. For hi,sute women, a significant correhtion between se- ,rm coadmgen levels and BMI rum found, as opposed to normal p i p .
Key words: androgens, liirsutis~n
INTRODUCTION
Clinical manifestations of androgenization in women include hirsutism, acne and baldness of the masculine type. In severe cases, signs of virilization, like deep- ening of the voice, increased muscle mass and cl*toromegaly are observed. The most common manifestation is hirsutism.
Hirsutism is more than a cosmetic problem that can attack a woman. It is the hormonal problenl resulting from an excess of androgens and abnormal metabo- lism of androgens in hair folicle, caused by disorders in the ovaries or adrenal glands.
Androgens are C-19 steroids (containing 19 carbons) derived from cholesterol (Fig. 1) and secreted by the adrenal cortex and ovaries. Androgens may also be derived (not secreted) from the conversion of other steroids by the liver and some peripheral tissues (1). Principal androgens include testosterone (T) and its me- tabolite dihydrotestosterone (DHT), androstendione (A), dehydroepiandrosterone (DHEA) and its metabolite del~ydroepiandrosterone sulphate (DHEA-S) (Fig. 2). An important metabolite of T and DHT is 3a,l7P-androstanediol glucuronide
BIOCHEMIA MEDICA god. 5, 61: 2-3, 1995. 69
Androgeils in hirsutisin J. Osredkar, A. Kociiantit
\/-. /',/ I
'&-A/ CHOLESTEROL
Qi3 N * - C-0
I' &dL/ OH
PREGNENOLONE 17 alpha-OH pregnenolory beta pregnantriol
CORTICOSTERONE DHEA
ALDOSTERONE
ESTRONE
CORTISOL ., ANDROSTENEDIONE b ' 4,
. I / ! a3/-+>?,\/'
TESTOSTERONE ESTRADIOL
F i g u r e 1 . - Synthesis of androge1zsfim ch,obsterol S l i k a 1 . - Sinnteza androgenn iz kolesterola.
BlOCHEMrA MEDICA god. 5, h 2-3, 1995.
J. Osrcclkal., A. KocijanCiE Androgens in hirsutism
TESTOSTERONE-17-G DHEA
j
DIHYDROTESTOSTERONE
i i k I r ANDROSTANEDIONE
DHT- 17 -G !I 0 al I I i
3 alpha ADIOL-17-G OH
t,I"UIl.".,
3 alpha ADIOL-3-G
F i g u r e 2. - Metabolic pathzoays of testoste~one dxgradation S 1 i k a 2 . - ~Vletaboli~ki ;b.lrtoui ~azgradnje te~tostero'n~a
BIOCHEMIA MEDICA god. 5, bx 2-3, 1995. 71
Androgens in 11il~sutisi1~ J. Osredkar, A. Kocijantit
(ADIOL-G). Although not androgenic itself, ADIOL-G serves as an effective marker of androgen action (2).
Several reports indicate that women with hirsutisin have elevated levels of ADIOL-G in both seruin and urine (3 - 5).
Since the plasma circulating androgens are the most readily demonstrable and the only obtainable biochemical parameters in routine evaluation of hirsute patients, we decided to study the correlation between the degree of hirsutism and seruin levels of T, free testrosterone (fT), A, DHEA-S, ADIOL-G, sex lior- mone binding globulin (SHBG) and salivary testosterone(saT).
The body mass index (BMI) was also considered, since androgen levels in obese women with mild hirsutism had been reported (6) to be higher than those in severely hirsute non-obese women.
PATIENTS AND METHODS
Our study included 46 women aged 17-39 years, mean age 25.2+3.2, referred to our outpatient clinic for hirsutism. The severity of hirsutism was assessed semiquantitatively as described by Ferryman and Gallwey (7). Blood samples for androgen determinations were drawn in the morning between second and sixth day of the menstrual cycle: tliree equal amounts of blood samples, drawn every 20 minutes, were pooled. The blood was centrifuged and the seruin frozen until analysis at -20 OC. Salivary samples were taken between the collection of blood samples without any additives for promotion of salivation.
DHT
vo1 (rnl) 4 4 4 5 7
1000
5 0 0
Isooctahe
F i g u r e 3. - Elr~aliol~ oJd@ewnt an,drogensfiom Celite col~~rnn S 1 i k a 3 . - E11~nczj'a~ razlititih androgenn s celitne kolone
-- ANDROSTANEDIONE!
ANDROSTANEDIOL
"
', i '\
/ 100% 9 0 % 85% 409 2 5 %
BlOCHEMlA MEDICA god. 5, br: 2-3, 1995.
Benzene 10 % 15% 6 0 % 7 5 %
1 . Osrcdkar, A. KocijanCi? Androgens in hirsutism
For all hormone measurements except ADIOL-G, conventional RIA meth- ods using coinmercial kits were employed, as follows: KIA by Immunotech (Marseille, France) for T, A, DHEA-S, KIA by DPC (USA) for fT, RIA by Milab (Malmo, Sweden) for SHBG, and RIA by Sorin (Sallugia, Italy) for saT (8). Se- rum ADIOL-G was measured by the method of Samojlik (9) and is presented in short as follows: 1 inL of serum was extracted with ether to remove free steroids: the serum was then incubated a1 45 "C for 48 11 with P-glucuronidase (Helix pomatia). After incubation, approximately 1000 cpm ("H) androstanediol were added to correct for procedural losses, and serum was extracted twice with water and evaporated to dryness. The resuspendecl extract was transferred to a Celite column impregnated with propylene glycol and eluted with isooctane progres- sively enriched with benzene. This system clearly separated A, T, and DHT from ADIOL (Fig. 3). The benzene-isooctane fraction (75:25) containing the ADIOL fraction was collected, and aliquots were taken to determine the procedural loss. The remainder was analyzed by RIA using a specific antibody raised against 3a- diol hemisuccinate conjugated to bovine serum albumin (BSA).
Normal hormone levels for women, established in our laboratory are: T 0.3 - 3.0 ninoVL, A 0.6 -10.2 nmoVL, DHEA-S 3.6 - 13.4 pmol/L, fT 3.8 - 10.4 pmol/ L, ADIOL-G 1.7 - 7.6 nmol/L, SHBG 27 - 58 nmol/L and 58 - 71 pmol/L for ST. BMI was calculated by dividing body weight (in kg) by the square of body height (in in). For statistical evaluation, multiple regression analysis and correlation analysis were used.
RESULTS
Serum androgen levels in 46 hirsute women are shown in Table 1.
Although the mean concentration of each androgen was elevated, only 17 (36.9%) had serum T elevated in excess of 3 nmol/L. Serum fT concentrations were elevated in 28 (60.9%) cases. Serum A concentrations in excess of 10.8 nmoV L were found in 33 (71.7%) hirsute women, and elevated serum DHEA-S levels in
T a b l e 1 . - Data summary
I S-SHBG ( ~ u n o V L ) I 26 ( 10 ( 8-47 1 27-51 1 s I
S-T (n~nol/L)
S-fT (plnoVL)
S-A ( I I I ~ o V L )
S-DHEA-S (pnoVL)
BIOCHEMIA MEDICA god. 5, 67: 2-3, 1995. 73
3.4
12.2
16.9
7.9
S-ADIOL-G (mnol/L)
Sa-T ( p l o V L )
1.5
5.8
7.2
2.2
8.7
91
1.5-8.8
3.5-28.1
5.6-33.7
4.1-15.1
4.7
51
0.3-3.0
3.4-10.4
0.7-10.8
3.9-9.6
1.8-20.8
32-310
s . S
S
NS
1.2-7.6
33-71
S
S
Androgens in hirsutisin J . Osrcdkal-, A. KocijantiE
only 2 (4.3%) hirsute women. By contrast, serum SHBG values were lower than the lower nornlal limit (27 nmoVL) in 27 out of 46 (58.7%) hirsute women. The mean serum ADIOL-G level in 46 hirsute women was 8.7 nmol/L, and serum ADIOL-G values were higher than the upper normal limit (7.6 nmoVL) in 22 out of 46 (47.8%) hirsute women. SaT values were elevated in 24 (52.2%) cases.
Because the mean BMI in the women with hirsutism was greater than in normal women, we examined the relationship between serum androgen levels and BMI. No correlation was found for normal women, but for women with id- iopathic hirsutism a weak but statistically significant correlation was observed.
If serum androgen levels do serve as markers of peripheral androgen ine- tabolism, then correlations between serum levels and FG score, an index of the extent of hirsutism, might be expected.
DISCUSSION
As already mentioned, the concentrations of circulating androgens represent only one of the parameters determining the clinical picture of hirsutism. In our group, 6 patients (13%) had normal androgen levels, although all of them were hirsute. The determination of serum SHBG and FT showed the highest correla- tion with the degree of hirsutism (~0.511; -0.493) in our patients, in contrast to a previous study of ours(l0). The correlation of serum T, saT and ADIOL-G was a little lower ( ~ 0 . 2 8 4 ; ~ 0 . 2 3 8 ; F-0.207), but greater than the determina- tions of serum A and DHEA-S (r=0.121; I=-0.127). On the other hand, we also showed a significant correlation between hirsutism and BMI. Our finding of a weak, but statistically significant, correlation between serum ADIOL-G levels and BMI suggested that body weight fie?. se, rather than llirsutism, might be an im- portant factor in determining serum ADIOL-G levels. The correlations of an- drogens determined and FG score showed a little different picture. The best cor- relations were found between saT ( ~ 0 . 4 0 7 ) and fT (~0 .414) us. FG score and the weakest between DHEA-S (I--0.087) and ADIOL-G ( ~ 0 . 2 4 8 ) .
Hirsutism cannot be regulated only as a function of circulating androgens, but may result from the processes occurring in peripheral tissues. The hair fol- licle sensitivity has been shown to be lligher in some women with hirsutism than in non-hirsute women (11). Much of the to date research supported the concept that serum ADIOL-G levels are related to Sa-reductase activity (2). In our group serum ADIOL-G levels were higher than those obtained by Morton (4) and Scanlon (12), and lower than that reported by Gompel (13). Determination of se- rum T, fT, A, DHEA-S, SHBG, ADIOL-G and saT represents the profile for clinical evaluation of hirsutism.
BIOCHEMIA MEDICA god. 5, b~ 2-3, 1995.
J. Osrcdkar, A. KocijanCiC Androgens in I~irsutism
KONCENTRACIJE ANDROGENIH HORMONA U ~ E N A S HIRSUTIZMOM
S A ~ E T A K - Cilj rada bio je ustaaor~ili korelncij~i izttledzr s l ~ i n j a , hi,zzi/izma. i serutnskih koncenhci jn ~ h ~ p t z o g trslosleronn (T), slobod~nog lestosleronn (JT), a~rdros~endianu (A), dehidroepia~~~droslerm si~lfala (DHEA-S) , anniroslandiol glukt~ro.nida (ADZOL-GJ, globu.li.nn hoji 71c2 spolne hormone (SHBG) le koncenttacije lesloslerona. I L slini (sa.T). Pralorn je uzel u obzir i i.ndelts Cjelesna m a s . ( B M I ) Naj71iSi korelncijski koqicijcnl 11 odnoszi na slupa,nj hirz~i/iz.ma naden je zn ko,zcrn~taci;iu SHBG i (fll u set-zm~c, ,neS/o slabija, korda.c+ rm&.rul je za. 7: saT i ADIOL- G , a .r~ajma;nja m A i DHEII-S. Izrnedu raz.t'tta serwmskih a,nrlrogena i BMI na&tm je statislilki zmCajtm korclacija u grupi hitz.u/~rih isna.
KljuEne rijeEi: androgeni 1101m011i, llirzutizain
REFERENCES
1. Azziz R. Mrcnal androgcns and rcpi.oduc- tivc agcing in fc~naics. Scinin Repinduct Endocrinol 1991:9:246-GO.
2. Pnbtsott RJ. Mcasurcmcnts of 3a, l7b- audrosta~~cdiol gluculunidc in serum and uri~ic and thc correlatiou with skin 5a-re- ductasc activity Fertility & Sterility 1986; 46:222-6.
3. Horton R, Hawlo. L), Lobo R. 3a,17b- androsta~~ctliol gluc~~roilidc in plasina, a n~arkcr of androge~~ actiou in idiopathic hirsutism. J Clin 111vcst 1982;69:1203.
4. Horlorl R, Markers of pcripl~cral androgcu production. In Scrio A4, cd. Sexual DiCCer- cntiation: Basic and Cliuical Aspect. Ncw I'ork: Ravcn Prcss; 1984:261-73.
5. Horlon R, Lobo R.Pcriphcra1 androgcns and thc role of a~ldrostancdiol glucurouide. Clin Endocrinol Mctab 1986; 1.5293.
G. Ckttumls JD, Edergscn VL. Metabolism and co~ljugation of "c-plagcstcronc by bovine livcr and adipose tissuc, in vitro. Steroids 1982;40:287-306.
7. Ferty~on D, Gallruoy JD. Clinical asscssmcl*t of body hair growth ill wvomcn. J Cliu Endocrinol Mctab 1962;21: 1440-7.
Osredkro J, fihovec I, Jmnoz~ec N, KocijanEiE A, Preblj J. Salivary free testosterone in hirsutis~n. Aun Clin Biocl~e~n 1986;26:522- 6. Smlojlik E, Kirsch,rwr MA, Silbcr D, Schneidet- C , Ertel H. Elevated production and meta- bolic clea~ance ratcs of androgens in 11101-dibly obcse women. J Clin Endocrinol Mctab 1984;59:949.
@fir M , P,-iplj J, KoczjanLiE A. The correla- ti011 bctwccn clinical and horinonal param- ctcrs in andi~gcnisation. Acta Europ Fert 1989;20:31-3.
Mijaurki M , 'liikaysu S, Karnhawa T Activ- ity of testosterone 5a-reductase in hair fol- licks of wvo~ncn with polycystic ovaries. J Endocrinol 1978;78:445.
Scanlon MJ, Whorwood CB, Frankr S, Reed MJ, lames VHT. Serum androstanediol elucu- " ronide co~lccntratioi~s in normal and hirsute wvomcli and patients with thyroid dysfunc- tion. Clin Endocr 1988;29:529-38.
13. Gornjxl A, Wright Ktiltenn Mau~mi~Jorz~is tl Contribution of plasma androstcndion to 5a-a~~drosta~~ctliol glucu~onide in women with idiopathic Ilirsutis~n. J Clin Er~docrinol Mctab l986;62:441-9.
BIOCHEMIA MEDICAgod. 5, b~ 2-3, 1995.
