Sarah Pearce, Mitchell Dowsett and J. Alan McKinna
Abstract
Several studies have recently reported the percentage of non-protein-bound (NPB) oestradiol (E2) to be higher in patients with breast cancer than in normal controls. Using postmenopausal volunteers, we have examined the fractional binding of E2 and testosterone (T), as well as total E2 and T, sex-hormone binding globulin (SHBG), iuteinising hormone (LH) and follicle stimulating hormone (FSH), in normal women, those at risk of developing breast cancer and women with breast cancer at first diagnosis and first recurrence. No significant differences were observed in either the concentration or in the percentage of NPB E2 or T, or in any of the other hormones measured. The validity of our observations were confirmed by expected relationships between E2, T, SHBG and body mass. EurJ Cancer, Vol. 27, No. 3, pp. 259-263, 1991.
Introduction
A LARGE AMOUNT of epidemiological and experimental evidence implicates oestrogens in the aetiology of breast cancer. However, despite many varied studies the precise role of these hormones remains unknown. Total blood oestradiol (E2) levels have frequently been measured in breast cancer patients, but this may be inappropriate for assessing biological activity as oestradiol circulates extensively bound to sex-hormone binding globulin (SHBG) and albumin, leaving only a small percentage ( <2%) of the steroid in the non-protein-bound (NPB) form [1]. Recently there has been much active debate as to the biological availability of these three fractions to the tissue.
In 1981, Siiteri compared the percentage NPB E2 in breast cancer patients with that of controls and found that the level Was higher in the cancer patients [2]. Since then five studies [3-7] have all reported a higher percentage of NPB E2 in breast cancer patients than in controls. Two of these studies [5, 6] also showed an increase in the albumin-bound E2 fraction in the breast cancer patients. Two other studies [8, 9], found no significant difference between breast cancer patients and controls. Moore and his colleagues also reported a lower proportion of E2 bound to SHBG in the blood of women who went on to develop breast cancer than in those who did not [10]. Based on this evidence coupled with the observation that Japanese women (who have a lower incidence of breast cancer than western women) also have a higher proportion of E2 bound to SHBG [10], they proposed that this parameter may be a marker for breast cancer risk.
For a given change in SHBG binding capacity, the change in percentage NPB testosterone (T) is greater than that in the percentage NPB E2 [1]. Any difference observed in percentage of NPB E2 would therefore be expected to manifest itsel# in a greater change in percent NPB T. This parameter would therefore be expected to be a more sensitive marker of breast cancer risk than NPB E2. In addition if this effect were reflected in an increased concentration of NPB T, it may be important in, increasing the availability of T to the aromatase enzyme, which is responsible for converting androgens into oestrogens.
In this study we have measured total, albumin-bound and NPB E2 and T as well as SHBG, lu remising hormone (LH) and follicle stimulating hormone (FSH) in groups of postmenopausal women: normals with and without a family history of breast cancer; women with diffuse breast disease, with and without a family history of breast cancer; localised benign breast disease patients, with and without a family history of breast cancer; breast cancer patients at first diagnosis and at first recurrence of the disease.
SUBJECTS and methods
Subjects
Clotted blood samples were obtained from the Early Diagnostic Unit and the Breast Clinic of the Royal Marsden Hospital, London. The study was approved by the local ethics committee and volunteers signed a consent form. The following data was recorded at the time the sample was taken: height, weight, age, menopausal status, parity, dietary habits, concurrent medical complaints, medication and family history. All volunteers who had undergone menopause less than two years previously were excluded, as were any women on hormone replacement therapy or who had received any other endocrine treatment within the previous three months. After blood collection, serum samples were stored at -20°C in 1 ml aliquots until analysed.
The volunteers were divided into eight groups for analysis, based on clinical examination: (1) normal; (2) normal with a family history; (3) diffuse benign; (4) diffuse benign with a family history; (5) localised benign; (6) localised benign with a family history; (7) first diagnosis breast cancer prior to surgery; and (8) first recurrence breast cancer prior to therapy. For the purposes of this study family history was defined as at least one first degree relative with breast cancer.
Methods
SHBG binding capacity was measured using the two tier column technique of Iqbal and Johnson [11]. Total E2 was measured by radioimmunoassay using a highly specific antiserum [12]. The sensitivity of the assay was 2.8 pmol/1. Total testosterone was also measured by radioimmunoassay (STRIA, St Thomas’ Hospital, London). The sensitivity of this assay was 0.14 nmol/l.
The percentage of NPB steroid was measured by centrifugal- ultrafiltration dialysis (CUFD) [13] with our previously published modifications [14]. The percentage of albumin-bound steroid was measured by ammonium sulphate precipitation [15] also using our previously reported modifications [14].
The percentage of steroid bound to SHBG was calculated by subtraction of the percentage of NPB and albumin-bound steroid. The concentration of the NPB, albumin-bound and SHBG-bound steroid was calculated by multiplication of the measured total steroid concentration and the percentage of the steroid in each fraction. LH and FSH were measured by j radioimmunoassay (Chelsea RIA kit). The sensitivity of the assays were 0.7 and 0.3 IU/1 respectively.
Statistics
Statistics were performed using SPSS-PC. Pearson correlations, linear regression, analysis of variance, range and Student-Newman-Keuls tests and analysis of covariance were used as appropriate.
Results
Subjects
Initially controls and patients were matched for age, weight and height. Quetelet’s index (QI; wt[kg]/ht[m2]) was used as a measure of body mass. As can be seen from Table 1, first diagnosis and first recurrence breast cancer patients were significantly older than the other groups. The first recurrence group also had a significantly higher QI value than all but the localised benign groups. There were also some differences between the cancer patients and the other groups in the number of years past the menopause, but this is probably a reflection of the age difference in these groups. Throughout the sample population, we observed a significant inverse relationship between QI and SHBG (see Table 2) and a significant correlation between QI and total E2 levels and total T levels. We also observed a correlation between FSH and total E2 levels and QI. The fractional binding of both E2 and T also correlated with SHBG binding capacity.
Table 1. Factor comparison of the 8 menopausal groups: mean (S.E.)
Table 2. Correlations and linear regressions between variables
Storage time
The majority of the samples were collected over a two year period and we investigated whether the length of time a sample had been stored had any influence on any of the endocrine parameters measured. No significant effect of length of storage was observed.
Hormone levels
The hormone levels for each of the groups studied are compared in Figs 1-3. There were no significant differences between any of the groups in the levels of LH, FSH, SHBG (Fig. 1), total oestradiol (Fig. 2a) or total testosterone (Fig. 3a). There were, however, a number of significant differences in some of the subfractions of oestradiol and testosterone.
Figure 1. SHBG binding capacity in (1) normal; (2) normal with family history; (3) diffuse benign; (4) diffuse benign with a family history; (5) localised benign; (6) localised benign with a family history; (7) 1st diagnosis breast cancer; (8) 1st recurrence breast cancer. Bars = mean(S.D.),
In patients with diffuse breast disease there was a significantly higher percentage of albumin-bound oestradiol (Fig, 2c) than in patients at first diagnosis of breast cancer. There were1, however, no significant differences between any of the groups of normal subjects in this parameter. No significant differences were found between the groups in the percentage (Fig. 2b) or concentration of NPB oestradiol (Fig. 2d), or in the concentration pf albumin- bound oestradiol (Fig. 2e).
Figure 2. . Fractional binding of oestradiol; (a) total E2; (b) % NPB E2; (c) % albumin-bound E2; (d) concentration of NPB E2; (e) concentration of albumin-bound E2. Groups 1-8 are as for Fig. 1. Bars = mean (S.D.).
A significant difference was observed in the percentage of albumin-bound T between the control group and those with diffuse or localised benign breast disease and a family history of breast cancer (Fig. 3c). A significant difference was also observed between patients at first diagnosis of breast cancer and all the increased risk groups (benign, diffuse, and control, diffuse and benign with a family history), but not with the control group. The first diagnosis patients had a lower percentage of albumin- bound T than the others.
Women newly diagnosed as having breast cancer had a significantly lower percentage of NPB T than diffuse patients and control women with a family history of breast cancer (Fig. 3b). Despite all the differences in percentage of albumin-bound 1 between the groups, the concentration of albumin-bound T (Rg- 3e), was significantly different only between the first diagnosis breast cancer patients and those diagnosed as having localised benign breast disease and a family history of breast cancer. No significant difference was observed in the seijum concentration of NPB T (Fig. 3d).
Figure 3. Fractional binding of testosterone (a) total T; (b) % NPB T; (c) % albumin-bound T; (d) concentration of NPB T; (e) concentration of albumin-bound T. Groups 1-8 are as for Fig. 1. Bars = means (S.D.).
Law fat diet
No significant difference was observed in any of the parameters when the women who categorised themselves as being on a low fat diet were compared to those on a “normal” diet.
Discussion
Since the original observation by Siiteri that women with breast cancer had a higher percentage of NPB E2 than normal controls [2], a number of studies have examined this further. We have confined the current study to the investigation of postmenopausal women. Five studies [3-7] all supported the finding of a higher percentage of NPB E2 in patients with breast cancer in this group. Two other studies however [8, 9] could find no difference between the cancer patients and normal controls. Even in the four studies which did show a difference there is a marked discordance in the quantification of the percent NPB E2 and in the relationship of the other binding fractions with breast cancer incidence. Moore et al [3] reported a difference in both the proportion and concentration of non-protein-bound E2 between the two groups while Reed et al. [4] only observed the difference in the proportion unbound. Bruning et al. [8] reported no difference in percentage NPB E2 but with higher total E2 levels found a higher calculated concentration of NPB E2. Differences also exist in the percentage of protein-bound E2 which were measured in the studies of Reed et al. [4] and , Langley et al. [5]. There are thus a series of discordant findings in these studies many of which are probably due to the differences in techniques employed.
In this study we have examined plasma levels of E2 and T and the differential binding of them in not only breast cancer patients and normal controls but also those women thought to be at risk of developing breast cancer, in an attempt to discover whether these parameters have any bearing on breast cancer, the disease and its development. Measurements of E2 have been reported but the fractional binding of T, which because of its binding relationship with SHBG would be expected to show greater changes than that observed with E2, has not been reported previously. The concentration of both non-protein and protein-bound E2 and T were also calculated in this study. No significant differences were found in either the concentration or in the percentage of NPB E2 or T between any of the groups and the controls.
The results in this study support the widely reported observations that body mass index (Quetelet’s index) shows a significant inverse correlation writh SHBG [16] and a positive correlation with total E2 levels. There was also a significant relation between SHBG binding capacity and the percentage albumin- bound T and E2. These findings are what would be expected from the literature concerning the relation between E2, T, SHBG and body mass. We also observed significant relations between the fractional binding of the steroids and SHBG, which could be predicted from their reported associations. The confirmation of these relations in our data support the validit r of our observations.
Moore et al. [10] suggested that women who went on to develop breast cancer had lower SHBG binding capacities than normal women. We have been unable to lend support to this as we have observed no difference in SHBG levels between the groups. Although Langley et al [5] found that length of storage at -20°C increased the dissociation rate of steroids from SHBG we found no evidence of this. Storage at -20°C had no effect on any of the parameters measured. The samples were all stored in 1 ml aliquots and were not thawed and refrozen.
In the data presented here we have found no evidence that there is an association between E2, its differential binding in breast cancer patients or those considered to be at risk of developing breast cancer compared to controls. We also found no reason to support the theory that SHBG may prove to be a good marker for women at risk of breast cancer. The discrepancies between our finding and those of other workers may be partly explained by our careful collection of samples and use of well- validated techniques: potentially confounding variables have not been controlled to the same degree in other studies which are all of similar or smaller size. The size of the effect was so small that power function analysis failed to produce an optimum group size. It is notable also that more recent work of Siiteri’s group in a much larger study indicated no significant differences in plasma levels of oestrogens, the differential binding of E2 or SHBG [9].
Acknowledgements
Acknowledgements—This work was funded by the Breast Cancer Research Trust.
References
Anderson DC. Sex Hormone binding globulin. Clin Endocniol 1974,3, 69-96.
Siiteri PK. Extraglandular oestrogen formation and serum binding of oestradiol: relationship to cancer. J Endocrinol 1981,89,119.
Moore JW, Clark GMG, Bullbrook RD, et al. Serum concentrations of total and non-protein-bound oestradiol in patients with breast cancer and normal controls. IntJ Cancer 1982, 29,17-2).
Reed MJ, Cheung RW, Noel CT, Dudley HAF, James VHT. Plasma levels of estrone, estrone sulfate and estradiol, and the j percentage of unbound estradiol in postmenopausal women with and without breast disease. Cancer Res 1983,43, 3940-3943.
Langley MS, Hammond GL, Bardsley A, Sellwood RA, Anderson DC. Serum steroid binding proteins and the bioavailability of oestradiol in relation to breast disease. J Natl Cancer Inst 1985,75, 823-829.
Ota DM, Jones LA, Jackson GL, Jackson PM, Kemp K, Bauman D. Obesity, non-protein-bound estradiol levels and distribution of j estradiol in the sera of breast cancer patients. Cancer 1986, 57, 558-562.
Ingram DM, Nottage EM, Willcox DL, Roberts A. Oestrogen binding and risk factors for breast cancer. Br J Cancer 1990, 61, 303-307.
Bruning PF, Bonfrér JMG, Hart AAM. Non-protein bound oestradiol, sex hormone binding globulin, breast cancer and breast cancer y risk. BrJ Cancer 1985,51,479-484.
Ernster VL, Wrensch MR, Petrakis NL, et al. Benign and malignant breast disease: initial study results of serum and breast fluid analyses of endogenous estrogens, J Natl Cancer Inst 1987,79,949-960.
Moore JW, Clark GMG, Hoare SA, et al. Binding of oestradiol to blood proteins and aetiology of breast cancer. IntJ Cancer 19a , 38,625-30.
Iqbal MJ, Johnson MW. Study of steroid binding by a novel "two tier" column employing Cibachrome Blue F3 G-A-Sepharose Sex-hormone-binding-globulin. J SteroidBiochem 1977,8,977-9
Dowsett M, Goss P,Powles TJ, et al. Use of the aromatase inhibitor 4-hydroxyandrostenedione in postmenopausal breast cancer: optimization of therapeutic dose and route. Cancer Res 1987, 47; 1957-1961.
Hammond GL, Nisker JA, Jones LA, Siiteri PK. Estimation of the percentage of free steroid in undiluted serum by centrifigal ultrafiltration dialysis. J BiolChem 1980,255, 5023-5026.
Pearce S, Dowsett M, Jeffcoate SL. Three methods compared for estimating the fraction of testosterone and estradiol not bound to sex-hormone-binding globulin. Clin Chem 1989,35, 632-635
O’Connor S, Baker HWG, Dulmanis A, Hudson B. The measurement of sex steroid binding globulin by differential ammonium sulphate precipitation. J Steroid Biochem 1973,4, 331-339.
16. Moore JW, Key TJA, Bulbrook RD, et al. Sex hormone binding globulin concentrations in a population of normal women who had never used exogenous hormones. BrJ Cancer 1987, 56,661-666.