Clinical Endocrinology (1992) 37, 203-206
High serum sex hormone-binding globulin (SHBG) in premature thelarche A. Belgorosky, E. Chaler and M. A. Rivaroia Laboratorio de Investigaciones, Hospital de Ped!atria Garrahan, Buenos Aires, Argentina (Received 2 December 1991; returned for revision 30 January 1992; finally revised 8 April 1992; accepted 29 April 1992)
Summary OBJECTIVE We determined serum sex hormone-binding globulin (SHBG), serum dehydroeplandrosterone sulphate, serum oestradiol and serum testosterone and Its fractions In girls wlth premature thelarche. DESIGN Blood was drawn from girls with recently dlagnosed (3-12 weeks) premature thelarche. Serum was kept frozen for at least one year before hormonal determlnatlon to exclude precocious puberty by cltnlcat evaluation. PATIENTS Seventeen girls with premature thelarche aged 043-7.16 years were studied, and compared with a group of 22 normal prepubertal girls. MEASUREMENTS SHBG was measured by saturatlon analysls and serum dehydroeplandrosterone sulphate, serum total oestradiol and serum total testosterone were determined by radloimmunoassay. Non-SHBG-bound testosterone and free testosterone were calculated from an equation derived from the law of mass actlon. RESULTS Median serum SHBG In premature thelarche was 137 nmolll (range 64-221), signiflcantly higher than In normal controls, 93.7 (32-172) (P < 0.05) non-parametric test of medians. Serum SHBG decreasedslgnlflcantly wlth age in controls but not in premature thelarche. No difference was found in serum dehydraepiandrasterone sulphate. Median serum total testosterone (0.34 nmolll, 0.17-0.97), median serum non-SHBG-boundtestosterone (0.04 nmolll, 0.02-040) and median free testosterone (2.2 pmolll, 1G 4 . 5 ) were slgniflcantly lower In premature thelarche than In control (P < 0.001). CONCLUSIONS Serum SHBG Is high and bioavailable T is low In girls wlth premature thelarche. This mlght alter the oestrogenlandrogen ratlo in the breast.
Premature thelarche, defined as breast development in girls before the age of 8 years, is a common and benign condition (Capraro et al., 1971, Root & Shulma, 1986). Mills et al. Correspondence: Dra A. Belgorosky, Hospital de Pediatria Garrahan, Pozos I88 1, Buenos Aires, Argentina (1245).
(1 98 1) reported that in premature thelarche breast develop-
ment may regress (30%), persist unchanged in size (50%) or increase in size (10%). The pathophysiological mechanism that causes premature thelarche is still unknown. Histological evidence indicates that ovaries are not totally quiescent during prepubertal years suggesting that some oestrogen secretion occurs (Peters et al., 1978). A transient partial activation of the hypothalamic pituitary ovarian axis with excessive FSH secretion has been proposed (Llicki et al., 1984; Pasquino et al., 1980). However, most girls do not show signs of breast tissue proliferation before puberty. One possibility is that, in the presence of similarly low ovarian secretion of oestrogens, breast stimulation occurs in some girls because of differences in target cells responsiveness to oestrogens (Llicki et al., 1984; Pescovitz et al., 1988). It has been proposed that sex hormone-binding globulin (SHBG) (Pardridge, 1981; Anderson, 1974) regulates the oestrogen/ androgen balance in target tissues. Since testosterone (T) binds more tightly to serum SHBG than oestradiol or oestrone (El), it would be expected that changes in the concentration of SHBG would differentially affect the free concentration of these steroids. In order to analyse the roles of serum SHBG and sex hormones in premature thelarche, we experimentally determined serum SHBG, total T, total E2 and dehydroepiandrosterone sulphate (DS), and calculated values for non-SHBG-bound T and free T in 17 girls with premature thelarche.
Materials and methods Clinical material
Seventeen girls with premature thelarche, age 0.83-7.16 years, were studied. Mean age was 3.28 f2.3 ( fSD) years. Breast enlargement had been noticed recently in all subjects (from 3 to 12 weeks). Precocious puberty was excluded by clinical evaluation during one-year follow-up. The mean height SD score was 0.35 f 1.2 ( fSD) and mean chronological age - bone age difference was +0.28 f0.56 ( fSD) years. In all these patients, pelvic sonography was normal and no ovarian cysts were found. Studies on a control group of 22 normal prepubertal girls (Tanner stage I) were reported previously (Belgorosky & Rivarola, 1988). Mean age was 4.5 2.34 years. No significant difference was found between mean ages in control and premature thelarche girls. Samples for endocrine studies were obtained at the time of diagnosis.
+
203
204
Clinical Endocrinology (1992) 37
A. Belgorosky et a/.
The study was approved by Research Committee of the Hospital de Pediatria Garrahan, acting as scientific as well as ethical committee.
Table 1 Median (range) serum SHBG, DS, total (TT) non-
SHBG-bound (nmolil) and free T (pmol/l) and total E2 (TE2), non-SHBG-bound and free E2 (pmol/l) concentration in 17 girls with premature thelarche (PT)and in 22 control prepubertal girls (C)
Methods
Skeletal age was estimated by the method of Greulich and Pyle (1959). Serum total T (TT) was measured by RIA (Campo et al., 1981). The imprecision profile (the withinassay coefficients of variation at different doses of the doseresponse curve) was below 10% for TT concentrations between 0.25 and 17.3 nmol/l. In premature thelarche girls, serum E2was measured by RIA using reagents provided by Serono. Assay sensitivity was 30 pmol/l. The imprecision profile ranged from 9.3% at 36.7 pmol/l to 3.9% at 400 pmol/ 1. In normal girls, total serum E2 was measured by RIA using reagents provided by International CIS (SB-ESTR Kit). Assay sensitivity was I5 pmol/l. The imprecision profile range from 25% at 20 pmol/l to 5% at 400 pmol/l. Serum DS was determined by RIA using reagents provided by Diagnostic Products Corp. (TKDS 2 Kit, Los Angeles, CA). The sensitivity of the assay was 36 nmol/l. The imprecision profile ranged from 20% at 140 nmol/l to 5% at 5400 nmol/l. All steroid determinations were carried out in duplicate. Serum SHBG was determined by saturation analysis as previously described (Belgorosky & Rivarola, 1981). Non-SHBGbound and free T were calculated using an equation derived from the law of mass action as described previously (Belgorosky & Rivarola, 1986, 1988). The equations of the regression lines, correlation coefficients, analyses of variance of linear regression, SEs of regression coefficients and nonparametric tests of medians were determined using the General Statistic Pac for a Hewlett Packard HP-86B computer (Palo Alto, CA). Results
As shown in Table 1, in the 17girls with premature thelarche, median serum SHBG was significantly higher while TT, nonSHBG-bound T and free T were significantly lower than in the 22 normal girls: 137 nmol/l (P
High serum sex hormone-binding globulin (SHBG) in premature thelarche A. Belgorosky, E. Chaler and M. A. Rivaroia Laboratorio de Investigaciones, Hospital de Ped!atria Garrahan, Buenos Aires, Argentina (Received 2 December 1991; returned for revision 30 January 1992; finally revised 8 April 1992; accepted 29 April 1992)
Summary OBJECTIVE We determined serum sex hormone-binding globulin (SHBG), serum dehydroeplandrosterone sulphate, serum oestradiol and serum testosterone and Its fractions In girls wlth premature thelarche. DESIGN Blood was drawn from girls with recently dlagnosed (3-12 weeks) premature thelarche. Serum was kept frozen for at least one year before hormonal determlnatlon to exclude precocious puberty by cltnlcat evaluation. PATIENTS Seventeen girls with premature thelarche aged 043-7.16 years were studied, and compared with a group of 22 normal prepubertal girls. MEASUREMENTS SHBG was measured by saturatlon analysls and serum dehydroeplandrosterone sulphate, serum total oestradiol and serum total testosterone were determined by radloimmunoassay. Non-SHBG-bound testosterone and free testosterone were calculated from an equation derived from the law of mass actlon. RESULTS Median serum SHBG In premature thelarche was 137 nmolll (range 64-221), signiflcantly higher than In normal controls, 93.7 (32-172) (P < 0.05) non-parametric test of medians. Serum SHBG decreasedslgnlflcantly wlth age in controls but not in premature thelarche. No difference was found in serum dehydraepiandrasterone sulphate. Median serum total testosterone (0.34 nmolll, 0.17-0.97), median serum non-SHBG-boundtestosterone (0.04 nmolll, 0.02-040) and median free testosterone (2.2 pmolll, 1G 4 . 5 ) were slgniflcantly lower In premature thelarche than In control (P < 0.001). CONCLUSIONS Serum SHBG Is high and bioavailable T is low In girls wlth premature thelarche. This mlght alter the oestrogenlandrogen ratlo in the breast.
Premature thelarche, defined as breast development in girls before the age of 8 years, is a common and benign condition (Capraro et al., 1971, Root & Shulma, 1986). Mills et al. Correspondence: Dra A. Belgorosky, Hospital de Pediatria Garrahan, Pozos I88 1, Buenos Aires, Argentina (1245).
(1 98 1) reported that in premature thelarche breast develop-
ment may regress (30%), persist unchanged in size (50%) or increase in size (10%). The pathophysiological mechanism that causes premature thelarche is still unknown. Histological evidence indicates that ovaries are not totally quiescent during prepubertal years suggesting that some oestrogen secretion occurs (Peters et al., 1978). A transient partial activation of the hypothalamic pituitary ovarian axis with excessive FSH secretion has been proposed (Llicki et al., 1984; Pasquino et al., 1980). However, most girls do not show signs of breast tissue proliferation before puberty. One possibility is that, in the presence of similarly low ovarian secretion of oestrogens, breast stimulation occurs in some girls because of differences in target cells responsiveness to oestrogens (Llicki et al., 1984; Pescovitz et al., 1988). It has been proposed that sex hormone-binding globulin (SHBG) (Pardridge, 1981; Anderson, 1974) regulates the oestrogen/ androgen balance in target tissues. Since testosterone (T) binds more tightly to serum SHBG than oestradiol or oestrone (El), it would be expected that changes in the concentration of SHBG would differentially affect the free concentration of these steroids. In order to analyse the roles of serum SHBG and sex hormones in premature thelarche, we experimentally determined serum SHBG, total T, total E2 and dehydroepiandrosterone sulphate (DS), and calculated values for non-SHBG-bound T and free T in 17 girls with premature thelarche.
Materials and methods Clinical material
Seventeen girls with premature thelarche, age 0.83-7.16 years, were studied. Mean age was 3.28 f2.3 ( fSD) years. Breast enlargement had been noticed recently in all subjects (from 3 to 12 weeks). Precocious puberty was excluded by clinical evaluation during one-year follow-up. The mean height SD score was 0.35 f 1.2 ( fSD) and mean chronological age - bone age difference was +0.28 f0.56 ( fSD) years. In all these patients, pelvic sonography was normal and no ovarian cysts were found. Studies on a control group of 22 normal prepubertal girls (Tanner stage I) were reported previously (Belgorosky & Rivarola, 1988). Mean age was 4.5 2.34 years. No significant difference was found between mean ages in control and premature thelarche girls. Samples for endocrine studies were obtained at the time of diagnosis.
+
203
204
Clinical Endocrinology (1992) 37
A. Belgorosky et a/.
The study was approved by Research Committee of the Hospital de Pediatria Garrahan, acting as scientific as well as ethical committee.
Table 1 Median (range) serum SHBG, DS, total (TT) non-
SHBG-bound (nmolil) and free T (pmol/l) and total E2 (TE2), non-SHBG-bound and free E2 (pmol/l) concentration in 17 girls with premature thelarche (PT)and in 22 control prepubertal girls (C)
Methods
Skeletal age was estimated by the method of Greulich and Pyle (1959). Serum total T (TT) was measured by RIA (Campo et al., 1981). The imprecision profile (the withinassay coefficients of variation at different doses of the doseresponse curve) was below 10% for TT concentrations between 0.25 and 17.3 nmol/l. In premature thelarche girls, serum E2was measured by RIA using reagents provided by Serono. Assay sensitivity was 30 pmol/l. The imprecision profile ranged from 9.3% at 36.7 pmol/l to 3.9% at 400 pmol/ 1. In normal girls, total serum E2 was measured by RIA using reagents provided by International CIS (SB-ESTR Kit). Assay sensitivity was I5 pmol/l. The imprecision profile range from 25% at 20 pmol/l to 5% at 400 pmol/l. Serum DS was determined by RIA using reagents provided by Diagnostic Products Corp. (TKDS 2 Kit, Los Angeles, CA). The sensitivity of the assay was 36 nmol/l. The imprecision profile ranged from 20% at 140 nmol/l to 5% at 5400 nmol/l. All steroid determinations were carried out in duplicate. Serum SHBG was determined by saturation analysis as previously described (Belgorosky & Rivarola, 1981). Non-SHBGbound and free T were calculated using an equation derived from the law of mass action as described previously (Belgorosky & Rivarola, 1986, 1988). The equations of the regression lines, correlation coefficients, analyses of variance of linear regression, SEs of regression coefficients and nonparametric tests of medians were determined using the General Statistic Pac for a Hewlett Packard HP-86B computer (Palo Alto, CA). Results
As shown in Table 1, in the 17girls with premature thelarche, median serum SHBG was significantly higher while TT, nonSHBG-bound T and free T were significantly lower than in the 22 normal girls: 137 nmol/l (P
