Serum Concentrations of Estrone, Androstenedione, Testosterone and Sex-hormone-binding Globulin in Postmenopausal Women with Breast Cancer and in Age-matched Controls

.


8-792856
According to the first of them a decreased urinary excretion of etiocholabolone and androsterone was associated with an increased risk of breast cancer and in the manifest disease with a poor prognosis and a poor response to endocrine ablative surgery (cf. 19). Contradictory results have however been reported (cf. 62) and the androgen excretion has also been shown to be decreased as a nonspecific consequence of illness (61).
Secondly, the "estriol hypothesis" was derived from the fact that carcinogenic potential has been demonstrated for estrone (El) and estradial (E2) in experimental research but not for estriol (E3) and that E3 has the capacity to impede certain estrogenic activities of El and E 2 (33). The relative amount of E3 in relation to El and E 2 was therefore suggested as an important determinant for breast cancer risk. This hypothesis was supported when population studies revealed a parallelism between the urinary estriol ratio and a high, intermediate and low breast cancer risk, respectively (13,41). The difference was most marked in young women, a finding combined with the previous finding of a reduced breast cancer risk due to an early age at first birth (40).
Recent research has revealed several lines of data which has seriously brought into challenge the concepts that the pattern of excretion of the urinary androgens and estrogens is causally related to the etiology of breast cancer (cf. 32, 6 2 ) . Results derived from determinations of the steroid hormones in serum are very few and have not resulted in any convincing alternative hypothesis (32).
The design of the present investigation was not primarily aimed at analysis of steroid hormones and their binding globuline. The availability of serum within the frame of an ongoing epidemiologic breast cancer study which included an assessment of thyroid function and an access to recently developed radioimmunoassays did, however, enable the study. The aim was therefore to compare the serum concentrations of El, androstenedione (A), testosterone (T) and sexhormone-binding globulin (SHBG) in an unselected series of postmenopausal women with a newly diagnosed breast cancer with those in a group of age-matched, non-hospitalized, postmenopausal women without breast cancer.

Patients
This study was based on 149 postmenopausal women included in a series of 179 women with breast cancer registered consecutively during five months in four Swedish counties. The population within this area was uniform as to race and nationality. Only two of the breast cancer patients who were diagnosed during the observation period (2/181) refused therapy and participation in the study.
Hospital records were written on a special form and all patients answered a comprehensive questionnaire concerning their reproductive history, gynecologic d i s e a s e s , h e i g h t , w e i g h t , drug consumption and o t h e r f a c t o r s of epidemiologic i n t e r e s t . A l l p a t i e n t s were c l a s s i f i e d according t o t h e TNM-classification (23) ( Table 1 ) . Two blood samples were drawn f o r a n a l y s i s of SHBG and hormones i n t h e serum.
The s e r a were d i r e c t l y s e n t t o Uppsala and a r r i v e d t h e same evening or t h e n e x t morning and were f r o z e n and s t o r e d a t -9OOC u n t i l analyzed. t h e f i r s t one. A s a r e s u l t each b r e a s t cancer p a t i e n t had one age-matched cont r o l with n o t more t h a n 3 days a g e d i f f e r e n c e t o t h e corresponding p a t i e n t .  c a n t l y h i g h e r (p (Table 2 ) . There v a l u e s exceeding

RESULTS
The mean v a l u e i n t h e p a t i e n t group (132 pmol/l) w a s between the groups was very small albeit significant (p < 0.05)( Table 2). The skewness of the distributions (Fig. 4 ) was less pronounced than for El (Fig. 2) and A (Fig. 3 ) and the arithmetic means -1.65 nmol/l ( 0 . 4 8 ng/ml) and 1.52 nmol/l (0.44 ng/ml)close to the logarithmic.
Sex-hormone-binding globulin (SHBG). In contrast to E l , A and T this distribution was shifted towards lower values in the patient group (Fig, 5 ) . The difference is highly significant (p < 0.01). Due to some very high values in the patient group (Fig. 5 ) the mean values -4 0 . 2 nmol/l and 47.3 nmol/ldiffer less than the median values -4 0 . 1 and 5 0 . 9 , respectively ( Table 2). The arithmetic means were 4 5 . 2 and 5 0 . 8 nmol in the patient and control groups.
Relations to age, weight and stage of disease. Subgrouping of the material according to age revealed no obvious trend in the differences between patients and controls for any of the variables. Thus in all decades from 4 5 to 7 5 years or more the patients, on the average, had a slightly higher A and T concentration in the serum. E l concentration did not differ between patients and controls aged 45-54 years but the patients had higher mean values in all subsequent agegroups. The patients had a lower mean SHBG value in all groups except in that including women over 7 5 years of age. Not any of the variables was significantly correlated to age, either in the control or in the patient group ( Table 3 ) .   were, as reported elsewhere (4), not statistically significant in either weight, height or the Quetelet's index.

2507
When weight and the index for overweight (Quetelet's index) were correlated to the steroid hormone levels the correlation coefficients were very low (Table   3). Except for a low degree of significance (p < 0.05) for the correlation between El and Quetelet's index in the control group (Table 3) they were all insignificant (p > 0.05). The same was true in the patient group in the correlation to the stage of the disease according to the TNM classification (Table 3).
A slight but significant negative correlation was found between Quetelet's index and the SHBG, in both groups (Table 3).
Interrelations between El, A and T were finally analyzed with calculation of correlation coefficients and with a multiple regression analysis using El as a dependent variable and A and T as independent variables in addition to age and the weight index.
In the control group there was a highly significant correlation between El and A (r=0.48) (Fig. 6) and El and T (r=0.45) (Fig. 7). The situation was quite different in the patient group where no significant correlation (r=0.10) was found between El and A (Fig. 8) and a low correlation (~0 . 2 5 ) between El and T ( Fig. 9).
According to the multiple regression analysis the variances in the serum level of El could be explained to 14 % and 58 % in the patient and control group, respectively, by variations primarily in A and T with very little contribution to the determination coefficient from age and the weight index. SHBG was not significantly correlated to El, A or T in any of the groups (r=0.17).

DISCUSSION
The estrogen and androgen metabolism and the serum concentration after the menopaus differ in many aspects from the premenopausal state. The serum concentrations of estradiol and estriol were thus generally found to be low ( (16) normal values have been reported. Our mean value for androstenedione in the control group when corrected for the blank, is in the range found by some other investigators using the same antibody for the radioimmuno-a s s a y ( 2 , 3 , 4 2 ) b u t somewhat lower t h a n t h a t r e p o r t e d by o t h e r s who u s u a l l y found v a l u e s i n t h e i n t e r v a l 0.75-1.09 ng/ml ( 1 6 , 1 7 , 25, 5 8 ) .

The u s e of non-hospitalized c o n t r o l s made a c o m p a r a b i l i t y with t h e p a t i e n t group with r e g a r d t o a r e c e n t o p e r a t i o n impossible. S u r g i c a l trauma h a s been
shown t o induce a small and t r a n s i t o r y f a l l of A and T b u t no change of E l i n normal women ( 2 8 ) . The f a c t t h a t t h e s e changes were normalized w i t h i n one w e e k  The conversion rate of A to E l has repeatedly been found to be 1. In conclusion, serum concentrations of E l and SHBG have, to our knowledge, not been studied before in breast cancer patients and thus the increased E l and decreased SHBG never observed, whereas the increased A and T levels have some support in a few recent studies. Although an increased production of estrogen precursors seems to be the most reasonable explanation for these findings, they are difficult to fit into a definite hypothesis and uniform support for this view was not perceivable in earlier studies. The significance of these findings needs therefore further conformation, primarily by studies of steroid hormone kinetics including production, conversion and metabolic clearance in breast cancer patients and comparable controls.