Thyroid and Sex Hormones Ratios in Predicting Breast Cancer Risk

Purpose: The study intended to analyze the incidence of thyroid related diseases and to assess thyroid to sex hormones ratios of breast cancer (BC) patients Methods: TSH, T 3 , T 4 , Estrogen, progesterone and testosterone levels of newly diagnosed BC patients (n=155) aged 30 to 75 years and age-matched normal controls (n=75) were analyzed. Data on history of thyroid related diseases were collected from an interviewer administered questionnaire. Thyroid/sex hormone ratios were analyzed and compared against healthy women. Results: History of thyroid related diseases was signicantly higher (p<0.05) in BC patients compared to controls. Patients (10%) with history of thyroid related diseases were excluded from the study. Subclinical hyperthyroidism was identied among 14% of the remaining BC patients and was the only dysfunction (7%) among healthy women. Signicantly higher (p<0.05) mean T 3 and T 4 values, lower TSH levels and non-signicant levels of estrogen and progesterone were observed in patients with BC when compared to healthy. Serum Testosterone of BC patients were signicantly low (p<0.05). Considering the thyroid to sex hormones ratios among postmenopausal women, T 3 /testosterone, T 4 /testosterone, T 3 /estrogen, T 4 / estrogen, ratios were signicantly different in the two groups and the highest signicance was found with T 3 /testosterone. Cutoff values studied from receiver operative characteristic curves indicated that a woman having T 3 /testosterone above 7.47 showed 12.5 times odds (p=0.000) of being diagnosed with BC. Conclusion: Incidence of thyroid related diseases are higher among BC patients and elevation of T 3 /testosterone ratio indicated a signicant risk of BC. However, a study involving a larger number of participants could conrm the above.


Introduction
The impact of hyper and hypothyroidism on breast cancer (BC) is researched with inconclusive results.
Some studies disclose profound effects of hyperthyroidism on BC cell proliferation [1] some portray association between hypothyroidism and BC [2]. The thyroid disease incidence is higher among BC patients when compared to apparently healthy individuals. Signi cantly high mean T 3 and T 4 and low TSH values in postmenopausal BC patients when compared to controls implicate an association of hyperthyroidism and BC [3]. Free T 3 and T 4 concentrations were higher in BC patients when compared to controls and benign breast tumors [4]. A dose-response positive association of T 3 with the risk of BC exists no such association between TSH and BC in postmenopausal women [5]. In addition, T 3 levels positively associate with invasive BC [6]. In contrast, hypothyroidism and low-normal T 4 are related with an increased risk of BC in post-menopausal women [2]. In contrast to both above observations, some studies report unaltered thyroid pro les in BC women [7]. Thyroid disorders such as hypothyroidism, hyperthyroidism or autoimmune thyroiditis did not have a higher incidence in BC patients or patients with benign breast tumors [4]. A negative correlation between TSH and T 3 is seen in early BC but not in advanced BC [8]. Thus the exact impact of thyroid hormones in BC development and progression is not recognized [5].
Substantial changes in the expression of thyroid hormone receptors suggest a possible deregulation that could trigger BC development [9]. Estrogen like effects of thyroid hormones is suspected to be impacting BC development [10]. T 3 is believed to promote BC cell proliferation and increase the effect of estrogen on cell proliferation in some BC cell lines indicating the role of T 3 in BC development and progression [11].
Similarly postmenopausal BC patients have signi cantly increased thyroid hormone/ estrogen ratios suggesting a possible tumor growth promoting effect due to the misbalance of the hormones [3].
However, data on distribution of thyroid to sex hormones of BC patients is not reported.
Previous study conducted on the same study sample reported signi cant low levels of serum testosterone concentrations among BC patients irrespective of the menopausal status. However, serum estrogen or progesterone concentrations of post-menopausal BC patients were not signi cantly different when compared with postmenopausal healthy women [12].
Thus this study was designed to analyze the incidence of thyroid related diseases and to analyze the thyroid pro les (TSH, T 3 and T 4 ) of BC patients and compare with apparently healthy females. Attempts will be made to assess any signi cant associations with thyroid hormone / sex hormone levels in developing BC among Sri Lankan BC patients.

Study sample
Newly diagnosed female BC patients (n = 155) who have not had any treatment for breast cancer (surgery, chemotherapy, radiotherapy) were identi ed from Apeksha Hospital (National Cancer Institute, Maharagama). Age matched apparently healthy females (n = 75) were selected for the comparative study. Informed written consent was obtained from all participants before engaging in the study. Data on history of thyroid related diseases, menopausal status, hormonal contraceptive usage and hormone replacement therapies for any clinical condition were collected using an interviewer administered questionnaire.

Sex Hormones And Ratios Of Hormones
Serum estrogen, progesterone and testosterone levels of the same study sample were measured using MINI VIDAS immune analyzer (Biomerix, France) [12] and thyroid/sex hormone ratios were calculated.

Statistical Analyses
Statistical data analysis was carried out using SPSS version 16.0 (2007, SPSS for Windows, SPSS Inc., Chicago, IL, USA) package. The quantitative data with skewed distribution were presented as median (Inter quartile range). The qualitative data were expressed by calculating the frequency and percentage. P value of less than 0.05 (p < 0.05) was considered to be signi cant. Non-parametric signi cances were analysed by Mann-whitney U test. Correlations of parametric and non-parametric data were analysed by Pearson and Spearman test respectively. Receiver operative characteristic (ROC) curve was plotted for determination of cut off values of some selected biochemical parameters.

Ethical Approval and Informed Consent
All procedures performed in the study involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Ethical clearance for the study was obtained from Ethics Review Committee, Faculty of Medical Sciences, University of Sri Jayewardenepura, Sri Lanka (652/12, 13/48). The approval for registering patients and accessing histopathology data were obtained from the Director, National Cancer Institute, Maharagama, Sri Lanka. Informed written consent was obtained prior to enrolling participants.

Incidence of Thyroid related diseases
Among breast cancer patients majority were (63%) postmenopausal women with an average age of 63 ± 7 years at the diagnosis of the carcinoma. Among the patients 10% (n = 16) reported a past history of thyroid related diseases and 6 were on medication for different thyroid related disorders. BC patients with known thyroid dysfunctions before diagnosis of BC were excluded from the study. Serum TSH, T 3 and T 4 levels of remaining patients and of apparently healthy age matched females were analyzed (Table 1).
Subclinical hypothyroidism was observed to be 14% among the remaining BC patients, and only 7% of females categorized as apparently healthy had subclinical hypothyroidism. When compared with apparently healthy females a woman with thyroid disorders had a relative risk of 1.3 (CI 1.04-1.13) of having BC.

Sex Hormone Concentrations Of Bc Patients
Among the study sample 37% of BC patients (n = 57) were premenopausal. Serum estrogen and progesterone concentrations at each phase among premenopausal BC patients were not signi cantly different (p > 0.05) when compared with age matched controls. Serum testosterone concentrations of premenopausal BC patients were signi cantly lower (p = 0.001) than apparently healthy females [12]. However, since the number of premenopausal BC patients in each menstrual phase is comparatively less, and the hormone concentrations signi cantly varied according to each phase, the comparative statistical analysis was conducted with the sex hormone levels of postmenopausal BC patients and apparently healthy age matched postmenopausal women.
Serum estrogen and progesterone concentrations of postmenopausal BC patients were not signi cantly different to that of apparently healthy women. However, serum estrogen of these BC patients were noticeably lower. When compared with premenopausal women, postmenopausal women had signi cantly lower (p = 0.000) serum estrogen and progesterone concentrations. Median (Inter quartile range) testosterone concentrations of postmenopausal BC and healthy women were 0.16(0.18) ng/mL and 0.21(0.22) ng/mL respectively. Serum testosterone concentrations of BC patients were signi cantly low (p = 0.001) irrespective of menopausal status when compared with healthy women [12].

Thyroid Hormones To/ Sex Hormone Ratios
Considering the thyroid pro le of the studied BC patients, even though the mean concentrations of thyroid hormones were within the normal reference range, signi cantly elevated levels of T 3 and T 4 concentrations were observed among BC patients when compared to apparently healthy. Among the studied sex hormones, serum testosterone was signi cantly low (p < 0.05) and a considerable differences in the estrogen concentrations were was observed though not signi cant (p > 0.05). Thus in order to study the possible risk associations with respect to thyroid and sex hormones, the ratio of thyroid hormones to sex hormones were studied and compared with apparently healthy women ( Table 2). Signi cant differences in T 3 /estrogen, T 4 / estrogen were found among the two groups and T 3 /testosterone and T 4 / testosterone ratios indicated a high signi cance (p = 0.000). Thus ROC curves were used (Fig. 1) to identify a possible predictor of BC risk and to nd a cutoff value with a higher sensitivity and speci city. According to the Fig. 01, among all studied thyroid/sex hormone ratios, T 3 / testosterone ratio showed the highest sensitivity and speci city with highest area under the curve being 93% indicating the possibility of using it as a predictor of risk compared to other studied parameters in BC diagnosis. According to the ROC curve the cutoff value was calculated as 7.47. Thus T 3 / testosterone value above 7.47 was identi ed as a predictive indicator of identifying BC risk. T 3 / progesterone or T 4 / progesterone ratios were not signi cant in the study sample.

Discussion
Thyroid and sex hormones have been implicated in mammary tumorigenesis and development.
Effects of estrogen represent an increase in biological activities and therefore, in conjunction with T 3 can act directly on mammary tissue by promoting differentiation [13]. Due to these multiple hormonal interactions as well as the ubiquitous role that thyroid hormones play in the body's overall metabolism, the role that thyroid hormones may play in establishing and maintaining BC is exactly not known. Studies have established a direct action of thyroid hormones on the development of the normal mammary gland. But whether an alteration in thyroid status affects mammary tumor risk as well as development and growth is not entirely clear and needs to be studied further.
Among the BC patients in the present study sample, a considerable number of BC patients (n = 16) reported a past history of thyroid related diseases and among the remaining BC patients, the incidence of subclinical hypothyroidism was twice as high as among apparently healthy individuals. Studies reveal increased risk of BC in women with hyperthyroidism [14]. Indicating an association between level of thyroid function and BC risk and the present study con rms the same for the rst time in Sri Lanka.
Among the BC patients even though serum TSH levels were noticeably lower, serum T 3 and T 4 levels were signi cantly elevated indicating a possible impact of these on tumor development or progression. Cell line studies reveal that T 3 can promote BC cell proliferation and increase the effect of estrogen on cell proliferation. Thus T 3 may play a role in BC development and progression [11].
Circulating estrogens and androgens are found to be positively associated with the risk for BC in premenopausal women [15]. However, previous ndings indicate non-signi cant difference in serum estrogen and progesterone levels in BC patients and signi cant low levels of testosterone [12]. The higher bioavailability of testosterone counteract the proliferative effects of estrogen on mammary tissue and thereby exert a protective role to the breast, inhibiting cancer development and/or tumour growth [16] which might be a considerable stakeholder in the study group. Also majority (75%) of the BC patients in the present study were either obese or overweight [17] and thus the impact of adiposity related secretions of androgens on BC cannot be undermined [18]. A study reveals a synergistic response between T 3 and high carbohydrate meals [19] whereas the BC patients in the present study were not regularly consuming balanced meals but were on frequent carbohydrate rich meals (unpublished observations). Thus the diet, the sedentary lifestyle and being either overweight or obese might have contributed to the present observations.
Lipid-soluble hormones in the blood are bound to hormone-speci c transport proteins, while a smaller portion is bound to serum albumin. Testosterone-estrogen-binding globulin (SHBG) is a sex hormone-binding globulin that binds to testosterone and estradiol in the blood. Other known steroid-binding globulins are transcortin, primarily associated with progesterone and thyroxine-binding globulin (TBG), for transporting T 4 . Increased estrogen concentrations increase TBG concentrations. The rise in TBG is paralleled by a T 4 increase to maintain a physiological concentration of free T 4 . Besides the effects on TBG concentrations, sex hormones also affect deiodinase activity which might together contribute in BC development [20].
In vitro studies reveal direct stimulatory effects of T 3 on basal production of testosterone and estradiol [21] and according to the present study T 3 /testosterone above 7.47 indicated the highest risk. In other words, while elevated T 3 contribute in BC cell proliferation, lower testosterone concentrations might have reduced the anti-proliferative and pro-apoptotic effect of testosterone on BC. Thus the present study identi es that T 3 / testosterone ratio can predict BC with higher odds when compared with other studied thyroid hormones/ sex hormone ratios in identifying BC risk. The imbalance of thyroid hormones causes the dysfunction of the reproductive system [22] which might also impact on the concentrations of sex hormones.
Interestingly testosterone sometimes functions via conversion to estradiol [23] and lower testosterone in females might impact on obesity and poor glucose control. Considering the HbA1c levels 20% of BC patients showed values above 7% after excluding 13% of BC patients who were already on glycemic control drugs at the time of enrolment to the study.
However further studies are needed to con rm the exact impact of lower serum testosterone and elevated T 3 on developing BC as research on molecular mechanisms involving androgenic pathways in BC is still in their infancy.

Conclusion
Thyroid related diseases are signi cantly higher among BC patients and BC patients showed signi cantly elevated serum T 3 and T 4 levels than controls indicating the possible impact of hyperthyroidism in BC.
Considering the thyroid hormones/sex hormone levels signi cantly increased serum T 3 / estrogen, T 3 /testosterone ratios among postmenopausal BC women implies the impact on hormone imbalance on BC development. Considering the Thyroid hormones/ sex hormone ratios, serum T 3 /testosterone above 7 was identi ed as a potent marker in identifying BC risk among the study sample.

Declarations
Con ict of Interest Statement Authors have not con icts of interest to declare.