sign in sign up
<<
Home , Thyroid

European Review for Medical and Pharmacological Sciences 2017; 21: 346-360 Thyroid disorders in polycystic syndrome

K. KOWALCZYK1, G. FRANIK1, D. KOWALCZYK2, D. PLUTA1, Ł. BLUKACZ1, P. MADEJ1

1Department of Gynaecological , Medical Faculty in Katowice, Medical University of Silesia, Katowice, Poland 2Department of Gynaecology and Obstetrics, Centre of Gynaecology, Obstetrics and Neonatology in Opole, Opole, Poland

Abstract. – OBJECTIVE: Thyroid disorders, Introduction especially Hashimoto’s (HT), are ob- served significantly more often in patients with polycystic ovary syndrome (PCOS) than in the Polycystic ovary syndrome (PCOS) is the most general population – approximately 27% and common female endocrinopathy in women of re- 8%, respectively. This is extremely important in productive age. It is characterized by a wide range young women, because both disorders are con- of symptoms, which can occur in different com- nected with fertility problems. As HT and PCOS binations and with different intensity. First of all, occur together, fertility problems may become a they are: ovarian dysfunction, which can cause serious clinical issue in these patients. rare menstruation or absence of it, hyperandroge- MATERIALS AND METHODS: A systematic literature review in PubMed of PCOS- and HT-re- nism and/or hyperandrogenemia, polycystic ova- 1 lated articles in English, published until Decem- ry/ on gynaecological ultrasound and, in ber 2015 was conducted. many cases, disorders such as resistance, RESULTS: The reasons for prevalence , obesity (especially the central still remain unclear. Genetic and autoimmune type of obesity), hypertension, nonalcoholic ste- backgrounds are recognized to be possible atohepatitis (NASH) and, finally, fully developed common etiological factors. Three genetic poly- morphisms have been described to play a role in metabolic syndrome. PCOS as well as in HT. They are polymorphism The prevalence of PCOS is noted in 8-12% of the gene for fibrillin 3 (FBN3) regulating the of women of reproductive age1; however, in the activity of transforming growth factor-b (TGF-b) most current Danish research2 from 2014, it was and regulatory levels, -re- set at up to 16.6%. These numbers vary because leasing receptor (GnRHR) polymor- of the different diagnostic criteria. Nowadays, the phism and CYP1B1 polymorphism standing for most widely accepted criteria are the Rotterdam hydroxylation. High -to-pro- 1 gesterone ratios owing to anovulatory cycles, as criteria from 2003 . According to them, the diag- well as high estrogen levels during prenatal life, nosis is set if any two out of three criteria are met, disrupt development of the and its func- in the absence of other entities that might cause tion in maintaining immune tolerance, and are these findings: suspected to enhance autoimmune response 1) Oligoovulation and/or . in PCOS. deficiency could be also in- 2) Signs of androgen excess (clinical or biochem- volved in the pathogenesis of HT and PCOS. CONCLUSIONS: The above-mentioned com- ical). mon etiological factors associated with fertili- 3) Polycystic ovaries (one or two) on gynecolog- ty problems in HT and PCOS require further re- ical ultrasound. search. Following these criteria, four phenotypes of PCOS have been marked out: phenotype A, when Key Words: all three criteria are fulfilled and phenotypes with , Hashimoto’s disease, Polycystic ovary syndrome, Fertility disorders. two out of three fulfilled criteria: menstrual dis- orders and hyperandrogenism, hyperandrogenism

346 Corresponding Author: Karolina Kowalczyk, MD; e-mail: [email protected] Thyroid disorders in polycystic ovary syndrome and polycystic ovaries on ultrasound and finally companied by menstrual disorders, may suggest menstrual disorders and polycystic ovaries on ul- the symptoms of PCOS. trasound. The heterogeneity of these phenotypes The presence of receptors for signalling path- is the main cause of difficulties in determining ways (thyrotropin-releasing hormone – TRH, clear pathogenesis, as well as causal treatment TSH) and thyroid has been demon- of PCOS. strated in animal studies in monkey uterus. More- Thyroid disorders may occur with thyroid hor- over, the impact of long-term estrogen and estro- mone deficiency (hypothyroidism), thyroid - hor gen- treatment on their expression mone excess () or with thyroid has been evidenced5. The expression of TSH, hormones within the normal range (euthyreosis). TRα1 and TRβ1 receptors has been stated in the Hashimoto’s thyroiditis (HT) also called chronic human endometrium as well6. The highest TRα1 lymphocytic thyroiditis is the most frequent cause and TRβ1 receptor expression were found in of hypothyroidism and also the most frequent receptive endometrium before ovulation. Thy- autoimmune disorder. The diagnosis is made by roid hormones adjusted endometrial synthesis of detecting elevated levels of anti-thyroid peroxi- mRNA, among others leukemia inhibi- dase antibodies (anti-TPO) and/or antithyroglob- tory factor (LIF), which is important during the ulin (anti-Tg) in the serum and a characteristic implantation process, and transporter 1 7 thyroid image on ultrasound – hypoechoic or het- (GLUT1) . Free (fT3) has been erogeneous echotexture. The gold standard in the detected in follicular fluid8. TSH and its receptors’ diagnosis of HT is lymphocytic infiltration and transcripts have been revealed not only in ovarian fibrosis found on histological examination; how- structures such as oocytes, cumulus oophorus ever, nowadays it is not performed routinely. HT cells, granulosa cells and ovarian , but may occur with clinical hypothyroidism (most also in syncytiotrophoblast villi8. Moreover, in the often), euthyreosis or hyperthyroidism. there are present membrane transport-

According to the National Health and Nutrition ers for triiodothyronine (T3) and thyroxine (T4), Survey (NHANES) – the biggest epidemiological type 2 and 3, which regulate research conducted in the USA on a group of thyroid hormone activity in the placenta, and tran- 17 000 subjects – the frequency of hypothyroid- sthyretin – a protein binding ism with a thyroid-stimulating hormone (TSH) and retinol. In vitro research on bovine oocytes level above 4.5 mIU/L in women of reproductive and embryos suggests that thyroid hormones play age is approximately 4%3. In research on the a role in the implantation process. Oocytes and incidence of thyroid disorders, conducted in Col- embryos mentioned above, cultured in medium orado on 25 000 individuals, the percentage of enriched with thyroid hormones, presented better hypothyroidism increased with age4. Therefore, outcomes regarding blastocyst formation, implan- the most probable thyroid disorder among young tation capacity, decreased rate and vital- women is HT with normal thyroid function and ity after cryopreservation9. elevated levels of anti-TPO and/or anti-Tg. The Overt hypothyroidism is accompanied with prevalence of anti-TPO and anti-Tg in the age TRH increase, which causes hyperprolactinemia, group of 20-29 years is 11.3% and 9.2% respec- (LH) pulsatile dis- tively, while in the age group of 30-39 years it is ruption, sex hormone-binding globulin (SHBG) 14.2% and 14.5%, respectively. synthesis decrease, estrogen peripheral metabo- NHANES revealed that the prevalence of hy- lism disruption and increase of ovarian androgen perthyroidism in the eastern population is low3. production. In research on female pigs, hypothy- TSH below 0.4 mIU/L was observed in 3% of roidism resulted in increased gonadotropin re- subjects of reproductive age. ceptor sensitivity in the ovary, which finally pro- Hypothyroidism has an important impact on moted ovarian hypertrophy, as well as multiple women’s reproductive health. In the prepuber- ovarian formation10. Hypothyroidism may tal period, it can lead to delayed puberty. In contribute to heavy, irregular menstrual bleeding, adult women hypothyroidism, as well as hy- spotting during the , insufficient perthyroidism, is considered to be the cause of endometrial thickness, ovulation disorders and, menstrual disorders and decreased fertility as a in consequence, endometrial proliferative phase consequence. Women with undiagnosed PCOS disorders. are often referred for thyroid examination. On the Sex hormone imbalance has likewise been other hand, overt hypo- or hyperthyroidism, ac- described in hyperthyroidism. It covers: SHBG

347 K. Kowalczyk, G. Franik, D. Kowalczyk, D. Pluta, Ł. Blukacz, P. Madej synthesis increase, estrogen disrup- control group15. It is really interesting, consider- tion, increased estrogen to androgen conversion, ing the fact that analogical research conducted higher basal gonadotropin-releasing hormone on patients with autoimmune diseases such as (GnRH) concentration and stronger pituitary to rheumatoid arthritis, spondyloarthritis or Behçet GnRH response. Menstrual disorders have been disease revealed a decreased ovarian reserve16. also observed, while in women in euthyreosis Significantly, a lower fT3, but normal TSH, the ovulation remains regular. level was marked in infertile women with HT

Taking into consideration the high prevalence treated with LT4 compared to fertile women with of both PCOS and HT in women of reproductive HT during treatment17. The researchers elevated age, common possible etiological and clinical the LT4 dose in the study group in order to obtain associations between them are discussed in the a similar fT3 level in both groups. All 21 women following article. A systematic literature review from the study group conceived and delivered in PubMed of PCOS- and HT-related articles at term. This study indicated that the role of fT3 in English, published until June 2016 was con- in requires further investigation and, ducted. what is more, that TSH within normal limits does not always express the thyroid hormone PCOS, HT and Female Fertility Disorders state adequately. Decreased fertility is associated with HT and In some researches18,19, autoimmune thyroid PCOS, though it seems that in patients with these diseases were connected with a three- to fivefold two diseases, fertility disorders appear more fre- increased miscarriage rate. On the other hand, quently and are be more pronounced. no difference in anti-TPO prevalence was found Antithyroid antibodies are observed in 5-10% among women with unexplained recurrent preg- of women of reproductive age3. The presence of nancy loss and the general population. Anti-TPO antibodies alone does not automatically influence did not have any predictive value concerning thyroid function. Although overt hypothyroidism the course of the next pregnancy20. In another 21 may manifest with serious reproductive disorders, study , preventive LT4 treatment in women with they should retreat after introducing an appro- anti-TPO and normal thyroid function decreased priate treatment. However, patients with HT in the miscarriage rate from 13.8% to 3.5% and pre- euthyreosis may still have difficulties conceiving term delivery rate from 22.4% to 7%. or delivering at term11,12. The rate among infertile women Antithyroid antibodies are the most frequent with and without HT who underwent treatment autoimmune findings in infertile couples with using assisted reproductive technologies (ART) at least two consecutive failed in vitro fertili- was comparable; however, the miscarriage rate zation (IVF) attempts13. The only research esti- was significantly higher among women with HT mating ovarian reserve in HT published so far – 53% vs. 23%22. In the Italian research, the preg- involved women in the premenopausal period nancy rate in women who had undergone ART treated with (LT4). No differences procedures, the group with TSH ≤ 2.5 µIU/mL between study and control groups with regard achieved 22.3%, while in the group with TSH to antral follicle count (AFC), follicle-stimulat- > 2.5 µIU/mL it was 8.9%21. On the other hand, ing hormone (FSH) level and menstrual cycle factors other than autoimmune length were noted14. Anti-mullerian hormone seemed to be more important in delivering at (AMH) was another parameter analyzed in this term after IVF in women older than 38 years, as research. AMH is a protein produced by gran- the prevalence of thyroid disorders in patients ulosa cells in the ovarian follicle during the who delivered and miscarried did not differ sig- reproductive period. The established ranges of nificantly23. Similarly, no correlation was found AMH allow prediction of the ovarian reserve. between anti-TPO presence before pregnancy and Levels below the normal range indicate a de- miscarriage in a small group of patients under- creased ovarian reserve; on the contrary, levels going IVF. Anti-TPO presence did not decrease exceeding the norm are characteristic for PCOS. the chance of conceiving24. The meta-analysis In the research mentioned above, the AMH by Thangartinam et al25 revealed that antithyroid level was, surprisingly, significantly higher in antibodies increase the miscarriage risk three the Hashimoto’s group treated with LT4 than in times and preterm delivery risk twice. Finally, 25 controls, whereas in patients with HT in euthy- LT 4 treatment can reduce these risks by 50% . reosis without treatment it was lower than in the The precise mechanism which explains how an-

348 Thyroid disorders in polycystic ovary syndrome tithyroid antibodies impact fertility has not been ity (defined as lack of conception after one year elucidated yet. Few hypotheses for antibody pres- of regular unprotected sexual intercourse). De- ence impact on miscarriage have been suggested. creased fertility is a consequence of ovarian dys- Possibly, women with increased antithyroid anti- function, which is influenced by obesity, metabol- bodies, despite euthyreosis in laboratory findings, ic disorders, chronic and hormonal may suffer from mild hypothyreosis, or these imbalance28. Excessive androgen concentration antibodies are a secondary manifestation of a in the ovary may contribute to premature lutein- general immune imbalance. Another explanation ization of granulosa cells; further discrepancy in could be the fact that women with HT get preg- paracrine secretion of growth factors in the ovary nant at an older age, which is an independent risk may inhibit oocyte maturation28. An abnormally factor13. Lately, Plowden et al26 have presented high LH concentration during the whole menstru- very interesting and strong evidence in the dis- al cycle has a very negative impact. It is supposed cussion. In a well-designed, prospective cohort to cause incomplete ovarian follicle maturation study conducted on 1228 healthy women with a during the first phase of the cycle and poor endo- history of pregnancy loss, no association between metrial receptiveness in the luteal phase29. Meta-

TSH levels ≥ 2.5 mIU/L (free T4, fT4, within bolic disorders at a cellular level during oocyte normal parameters) or the presence of antithyroid maturation are also underlined as a possible cause antibodies with fecundity, pregnancy loss or live contributing to anovulation. In vitro research re- birth was shown. vealed that granulosa cells of PCOS patients In 2012, the Endocrine Society published without ovulation showed poor glucose uptake, Guidelines for Thyroid Dysfunction in Pregnan- which can aggravate their energetic supply and cy27. According to them, universal screening of worsen oocyte quality in consequence. healthy women for thyroid dysfunction, as well as Women with PCOS undergo treatment using the presence of anti-TPO antibodies, before and ART techniques more frequently compared to during pregnancy is not recommended. In the case the general population (13.7% vs. 1.5%)30. Most of screening for thyroid disorders in newly preg- of the research on miscarriage risk and recurrent nant women, the committee is incoherent. Some pregnancy loss in PCOS was conducted before members recommend screening of all pregnant setting the Rotterdam criteria in 2003. The data women by the ninth week or at the time of their consistent with actual criteria are scarce – ac- first visit. Others recommend targeted screening, cording to the only research, PCOS patients con- among others, in women: over age of 30 years, stitute 10% of women with recurrent pregnancy with thyroid antibodies, with other autoimmune loss31. On the contrary, a Japanese team32 stated diseases, with a prior history of miscarriage or that PCOS is not a predictive factor of subse- preterm delivery and, finally, with infertility. If quent miscarriage in cases of recurrent pregnancy serum TSH is greater than 2.5 mIU/L in the first loss. Factors promoting miscarriages in PCOS trimester or above 3 mIU/L in the second and are responsible for an increase of prothrombot- third trimesters, LT4 should be instituted. ic activity: insulin resistance, hyperinsulinemia, If the hypothyroidism has been diagnosed before hyperandrogenemia, hyperhomocysteinemia and pregnancy, adjustment of LT4 dose to reach serum genetic polymorphisms: plasminogen activator TSH not higher than 2.5 mIU/L before pregnan- inhibitor 1 (PAI-1) gene locus 4G/5G polymor- cy is recommended. However, in the case of phism, responsible for higher PAI-1 concentration nothing except antithyroid antibody presence in in serum, and -converting 33 pregnancy, LT4 treatment is not advised. During (ACE D/I) polymorphism . A significantly lower pregnancy and breast feeding, suggested miscarriage rate has been observed in a group of intake should reach 250 μg per day27. 120 pregnant women receiving metformin during Fertility disorders are a characteristic feature the whole pregnancy (11.6% compared to 36.3% of PCOS. Ovulation disorders are diagnosed in in the control group)34. A lower miscarriage rate 75-85% of PCOS patients. It is estimated that was also marked in a small group of pregnant only 30% of cycles are ovulatory in this group PCOS patients with thrombotic risks, who re- of patients26. PCOS constitutes the most frequent ceived low-molecular-weight heparin and met- cause of female infertility due to the lack of ovu- formin35. Nonetheless, it is essential to conclude lation. In women with PCOS time to conceive is that it is not a recommended treatment according significantly longer than in the general population to current management in PCOS36. Women with and 40-70% of them are diagnosed with infertil- PCOS present a higher risk of developing ges-

349 K. Kowalczyk, G. Franik, D. Kowalczyk, D. Pluta, Ł. Blukacz, P. Madej tational diabetes, pregnancy-induced hyperten- on PCOS phenotype. Women with PCOS with sion and preterm delivery, both in spontaneous regular cycles present notably better metabolic and after ART technique use37. Fetal parameters (including BMI, glucose fasting level, growth abnormalities: small for HOMA-IR index) than women with oligo- and (SGA) and large for gestational age (LGA), most amenorrhoea46. likely due to insulin resistance, are also diag- The latest published47 prevalence of metabol- nosed in these patients more frequently30. The risk ic syndrome in PCOS was estimated at 11.9%. of preterm delivery is strongly correlated with In the Swedish study48, in a group of 43-year-old the presence of pregnancy-induced hypertension. women with PCOS with average BMI 28.3 kg/ Obesity extends the average time to conceive as m2, during 14 years of follow-up, only 23% of well as preterm delivery risk38. The impact of hy- them developed metabolic syndrome. Possibly perandrogenism on the course of pregnancy was this is associated with a usually lower BMI in studied in two Danish researches. A higher risk the female population in Sweden. TSH concen- of preterm delivery and preeclampsia in PCOS tration in PCOS patients correlates with BMI, patients with hyperandrogenism was revealed in waist circumference, diastolic blood pressure, the first one39, whereas no difference between LDL and triglycerides level cor- studied and control groups was observed in the responding with the general population49. The second study40. differences concerning BMI and HOMA-IR Despite the fact that both endocrinopathies were recognized in PCOS patients with TSH discussed in this review contribute to reproduc- between 2 and 4.5 mIU/L and below 2 mIU/L. tive health disorders, their combined influence on HOMA-IR in both groups was independent of fertility has not been investigated so far. The only BMI and age50. case-control study revealed that patients with HT and PCOS coexistence may exacerbate PCOS and increased anti-TPO level are at risk metabolic changes in comparison to disturbances of clomifene citrate resistance with an odds ratio present in the single disease. In fact, girls suffer- equal to 7.741. ing from both PCOS and HT had higher values of: BMI, fasting glucose level, HOMA-IR index Metabolic Risks and cholesterol compared to girls with HT only Metabolic changes observed in HT and PCOS or a control group51. PCOS features correlated 51 are heterogeneous; however, they are often re- negatively with fT4 quartiles . Also, women with lated. They generally cover higher BMI, glucose PCOS and subclinical hypothyreosis had higher and lipid abnormalities. levels of triglycerides, fasting insulin and HO- Metabolic disorders in Hashimoto’s disease MA-IR than women with PCOS alone. Both are more pronounced in patients with overt or groups did not differ with regard to total choles- subclinical hypothyreosis than in euthyroid pa- terol, HDL and LDL cholesterol levels52. 42 tients on T4 replacement therapy . It seems then An interesting report concerning the use of that they are primarily connected with thyroid metformin in obese patients with PCOS and dysfunction and improve with normal hormone hypothyreosis has appeared. These patients were levels42. It is confirmed that total cholesterol lev- prescribed 1500 mg of metformin daily or placebo el, triglycerides, LDL and non-LDL cholesterol for 6 months53. A significant decrease in TSH lev- rise with TSH concentrations, within TSH nor- els was observed in the metformin group, but not mal ranges as well43. Moreover, a cross-sectional in the placebo group, after 6 months of metformin study unveiled that BMI correlated positively treatment. No significant change in3 fT and fT4 with TSH serum concentration and negatively was observed throughout the study in any group53. with fT4 level. Weight gain in a 5-year period, The authors speculate that metformin could have which corresponded with a BMI increase of 1%, a possible central mechanism of action. also correlated with thyroid function. The impact of thyroid function on BMI was equal to smoking Joint Prevalence and physical activity44. The first systematic, prospective study54 on Metabolic changes in PCOS are well known the prevalence of autoimmune thyroid diseases and widely reported45. In HT metabolic chang- in PCOS, published in 2004, involved 175 pa- es result from thyroid dysfunction degree, in tients with PCOS and 168 healthy women, with the same way, insulin resistance degree and the an average age of 28 years; 26.9% of patients presence of other cardiovascular risks depend with PCOS and 8.3% of women in the control

350 Thyroid disorders in polycystic ovary syndrome group had elevated concentrations of anti-TPO Genetic Susceptibility and/or anti-Tg. The patients in the PCOS group Studies on the families and siblings of pa- presented significantly higher antibody and TSH tients with HT revealed strong hereditary sus- levels as well. The characteristic features of HT ceptibility to the disease. The HT risk in children on ultrasound were observed in 42.3% of PCOS and siblings of a patient with HT increases 32 patients and only in 6.5% of subjects in the con- and 21 times, respectively, with predominance trol group. Finally, HT, meeting both diagnostic in women58. Several genes related to develop- criteria, was three times more likely to occur ment, progression and symptom aggravation have in PCOS women (in 20.6% and 6.5%, respec- been found. They are: human leukocyte antigens tively)54. In the most recent study55, the higher (HLA-DR), cytotoxic T--associated prevalence of HT in PCOS was confirmed – HT protein 4 (CTLA4), CD40, protein phos- was diagnosed in 27% of PCOS patients (with an phatase 22 (PTPN22), interleukin 2 receptor al- average age of 24 years) compared to 8% of con- pha (IL2RA) and vitamin D receptor (VDR)59,60. trols. An interesting aspect of HT risk in PCOS The gene for Tg is the only thyroid-specific gene patients was underlined in the Brazilian study. among them59. The prevalence of autoimmune thyroid disease Familial occurrence of PCOS is widely recog- in 65 women with PCOS and 65 without PCOS nized. In first-degree relatives, a more frequent was estimated as 43.1% and 26.2%. However, presence of PCOS features has been document- after excluding the impact of weight and insulin ed61. Among genes suspected to influence PCOS resistance, no greater risk in the PCOS group was development are: fibrillin 3 (FBN3), insulin, in- found. Thus, weight and insulin resistance have sulin receptor, insulin receptor substrate 1, tran- been recognized to be independent risk factors for scription factor 7-like 2, calpain 10, the fat mass- developing HT56. Eventually, the authors of the and obesity-associated protein SHBG and VDR61. meta-analysis from 2013, on the relationship be- However, the results of research on most of them tween thyroiditis and PCOS, which involved six are inconclusive. Lately, in genome-wide asso- studies, stated that the prevalence of AIT, serum ciation studies (GWAS), the DENND1A gene, TSH, anti-TPO and anti-Tg positive rate in PCOS coding a protein participating in endosomal mem- patients were all significantly higher than those in brane transport, was regarded as a susceptibility control groups57. gene for PCOS for the Asian population61. The issue of PCOS prevalence in HT patients has been considered in only one research so far. The role of Genetic Polymorphisms The PCOS incidence in girls of 13-18 years of The FBN3 polymorphisms, by affecting TG- age with HT was significantly higher than in Fß activity, are taken into consideration in both a control group without anti-TPO (46.8% and PCOS and HT pathogenesis. TGFß is secreted 4.3%, respectively)51. by numerous cells, including macrophages, and influences cell recognition, proliferation and ap- Joint Etiology and Pathogenesis optosis62. FBN1, FBN2 and FBN3 are genes cod- The etiology of HT and PCOS is complex and ing for fibrillins – components of the extracellular not yet fully clarified. In HT etiology the role matrix which enable TGF-b binding63. As long as of genetic, immune and environmental factors TGF-b is bound by fibrillins, it remains inactive, as well as iodine insufficiency/excess is under- until its release allows proper TGF-b activity. lined. It is believed that PCOS has a polygenic Despite the fact that, according to GWAS, none of background. Genetic factors are responsible for the TGF-b signalling pathway are crucial susceptibility to the disease and subsequent en- in PCOS, the role of TGF-b activity is considered vironmental influence, mainly high-calorific diet in PCOS pathogenesis, especially in prenatal pro- and scarce physical activity, stimulate the devel- gramming of PCOS62. High FBN3 expression was opment of the disease. Regarding PCOS, the role observed in fetal tissues, comprising ovarian stro- of autoimmune factors is also discussed. ma as well62. After the first trimester of pregnancy The possible common genetic and autoimmune FBN3 expression in ovarian stroma disappears. factors, including the influence of genetic poly- In this manner, FBN3 regulates the activity of morphisms and other factors like tumor growth TGF-b, which takes part in ovarian development factor-b (TGF-b), regulatory T cells (Tregs), the and function during prenatal life64,65. thymus, vitamin D deficiency and sex hormone Higher TGF-b1 levels were shown in women disorders are discussed in the following section. with PCOS, apart from women with PCOS carry-

351 K. Kowalczyk, G. Franik, D. Kowalczyk, D. Pluta, Ł. Blukacz, P. Madej

70 ing allele 8 (A8) of D19S884 in the FBN3 gene. ated with T4, fT3 and fT4 , which may represent The TGFß1 level in these women did not differ a third link between genetic polymorphisms in from findings in healthy women without the A8 PCOS and thyroid function. mutation63. Thus, according to previous findings the FBN3 polymorphism determines whether The Role of the Thymus women with PCOS present normal or increased Self-tolerance, which maintains and counter- TGF-b1 levels. Additionally, A8 carriers with acts autoimmunity, is enabled by two mech- PCOS had higher concentrations anisms. The first one is based on the central compared with A8-negative women with PCOS63. immune tolerance and involves the destruction of As the TGF family influences cell proliferation autoreactive T cells in the thymus during prenatal and apoptosis, a higher TGFß1 concentration dur- life. The second one is connected with peripheral ing fetal life would explain collagen accumulation immune tolerance, in which Tregs play the key and fibrosis in the ovarian stroma of a polycystic role71. Tregs originate from the thymus and naïve ovary. TGF-b is one of the molecular mediators peripheral T cells. Their role is to suppress the im- playing a role in vessel remodelling, leading to mune system and to prevent an excessive immune atherosclerosis and -angiotensin-aldosterone response71. An experimental mouse model of au- system activation contributing to hypertension. toimmune thyroid disease revealed that CD4+ Both disorders are observed in the course of CD25+ FOXP3+ Tregs take part in breaking the PCOS66. The results of vitamin D3 supplemen- excessive autoimmune reaction72. TGFß induces tation on TGF-b1 bioavailability in PCOS have FOXP3 expression, which afterwards stimulates been shown. In the group supplemented with vi- Treg formation72. According to the TGFß1 role tamin D, TGF-b1 bioavailability decreased and so discussed above, lower Treg levels should be ex- did triglycerides, the Ferriman-Gallwey score and pected in HT than in healthy subjects. menstrual cycle length67. This research unveils Animal models have shown the influence of the possibility of a new therapeutic approach for on immune factors taking part in PCOS PCOS patients with increased TGF-b1 concen- pathogenesis. Estrogen injections in female mice trations (A8-negative). They could benefit from before the tenth day of gestation, a period when anti-TGF drug implementation, which certainly the thymus matures, resulted in anovulation and will be the issue of coming studies. polycystic ovaries73. In mice which were deprived In HT lower TGFß1 serum concentration has of the thymus before the injection and given ma- been found than in controls. Moreover, there was ture thymocytes afterwards, no polycystic ovary 74 no increase after LT4 treatment, which indicates a structure was observed . When high doses of rather closer relationship with HT than with hy- estrogens did not affect the maturing thymus, pothyreosis per se64,68. TGFß takes part in immune and developed thymocytes were supplied later tolerance development – it enforces FOXP3 pro- ovulation, as well as proper ovary structure, was tein expression and subsequent Treg formation58. preserved. After estrogen injection, mice with an Therefore, TGF-b is supposed to be involved in intact thymus had significantly lower thymocyte the onset autoimmune diseases, for example, HT. levels than controls74. It is believed that estrogens A hypothesis that women with PCOS carrying A8 affecting the maturing thymus lead to cessation of D19S884 in the FBN3 gene have lower TGFß1 of Treg production, which among other things levels, and thereby are more prone to HT devel- results in the formation of numerous ovarian fol- opment than women with A8-negative PCOS, de- licles. This thesis supports the theory of autoim- serves consideration and further investigation64. mune factors in the etiology of PCOS64,74. A relationship between the 3’-UTR variant A second important conclusion from the study rs1038426 of gonadotropin-releasing hormone mentioned above is that estrogens must act during receptor (GnRHR) and insulin secretion, insulin a proper time window in order to cause irrevers- sensitivity, insulin and TSH serum concentrations ible PCOS symptoms75. At the time, they induce in PCOS has been evidenced. Following this hy- steroidogenic enzyme activity in the ovary76. In pothesis insulin secretion, insulin resistance and the following research, rats were given a prena- thyroid function in PCOS could be regulated by tal estrogen, or dihydrotestosterone GnRHR polymorphisms69. injection. Interestingly, polycystic ovaries and The CYP1B1 gene is related to PCOS and the similar hypothalamic neurotransmitter secretion enzyme which oxidizes estradiol to 4-hydroxye- were observed after both estrogen and testoster- stradiol. CYP1B1 L432V (rs1056836) is associ- one injections, but not after dihydrotestosterone

352 Thyroid disorders in polycystic ovary syndrome ones77. Whether it can be simply attributed to the phase88. Likewise, an additional estrogen supply aromatization reaction still remains a question. correlated negatively with anti-TPO presence42. In women who were prescribed diethylstilbes- The symptoms of both multiple sclerosis and trol (DES) in the USA in the years 1940-1970, the rheumatoid arthritis tend to be relieved during highest infertility rate was observed in women ex- pregnancy, especially in the third trimester, when posed between the ninth and twelfth week of pre- estrogen and levels are the highest89. natal life78, the period of the most intense thymus Apart from T suppressor cells, Th1 response development. Higher prevalence of autoimmune and activation of CD8+, androgens reduce the diseases was also noted79. activity of the remaining components of the im- Considering the issue of fetal PCOS program- mune system90. Progesterone inhibits macrophage ming, a different hypothesis – hyperandroge- proliferation, IL6 synthesis and peripheral an- nization – cannot be omitted. In patients with tibody production90. It is supposed that proges- congenital adrenal hyperplasia, polycystic ovaries terone fluctuations during the menstrual cycle are observed more frequently80. In sheep and correspond to reversible functional changes in the rhesus monkeys submitted to androgenization in immune system91. prenatal life, disorders characteristic for PCOS In the course of the menstrual cycle, higher were recognized in adult life81. Moreover, rats levels of estrogens during the luteal phase and exposed to androgens prenatally had higher val- menstruation and lower levels during the follicu- ues of insulin, and fasting glucose level lar phase lead to a shift of the immunity mediated as well as glucose tolerance test82. The impact by T helper cells – from Th1 to Th2 response92. of prenatal androgenization on DNA epigenetic In young women during the menstrual cycle the modifications was assessed in a primate model83. IL6 level (related to Th2) correlated negatively The observed changes were, among others, in the with progesterone level – it was highest in the TGFß signalling pathway. follicular phase and lowest in the luteal phase93. It has been evidenced that IL6 inhibits FOXP3 The Role of Sex Hormones induction and, as a consequence, Treg formation. Sex hormone participation in the pathogenesis The authors concluded that estrogens inhibit and of autoimmunity seems to be reliable, as women progesterone promotes Treg production93. How- suffer from autoimmune diseases significantly ever, other researchers have come to different more often than men. Five percent of the general conclusions – according to them, estrogens pro- population has an and 78% mote Tregs, because their number rose in the late of them are females. The onset of autoimmune follicular phase and declined in the luteal phase92. diseases in women is observed earlier, and it often According to most studies dealing with the influ- correlates with sex hormone rise84. The female to ence of female sex hormones on Tregs in vitro, male ratio among subjects with autoimmune thy- in turn, both estrogens and progesterone promote roiditis is significantly lower in the pre-puberty Treg synthesis and enhance their function94. Addi- period in comparison with puberty and adulthood tionally, in a mouse model progesterone induced (1.6; 6.7 and 10.3, respectively)85. Tregs characterized by increased stability and In animal models, estrogens contribute to an greater efficiency in suppressing inflammation increase of activity and to a decline of T than Tregs induced by TGFß195. It turned out that cell activity86. The production of androgens also stimulate Treg differentiation, by turned out to be greater in females than males87. modulating the expression of FOXP389. A statis- Estrogens were described to decrease T suppres- tically significant correlation occurred around the sor cell activity, increase B cell activity, enhance ovulation period. secretion of interleukin 6 (IL6) – a cytokine Pregnancy is connected with a few immunity related to Th2 – and, finally, to shift the immune regulations in order to ensure immune tolerance response to Th2 and antibody secretion84. Women to fetal antigens, but mainly with the Th1 to present a higher CD4+/CD8+ ratio compared Th2 shift96. It is probably a result of the estro- to men84. On the other hand, supraphysiological gen-induced production of Tregs, with both Th1 concentrations of estrogens (for example, during and Th2. However, Th2 seem to be less prone pregnancy or oral contraception therapy) turn out to this suppression and tend to prevail in preg- to cause immunosuppression. Oral contraception nancy97. After delivery, the level of Tregs turns therapy prevented multiple sclerosis exacerba- down and shifts the immune response to the Th1 tions before menstruation and in the follicular type, which often results in autoimmune disease

353 K. Kowalczyk, G. Franik, D. Kowalczyk, D. Pluta, Ł. Blukacz, P. Madej aggravation97. A few large retrospective studies form – are related to HT102. The vitamin D level revealed the association between parity and the itself affects polymorphism of the VDR gene and risk of autoimmune thyroid disease98. other genes related to vitamin D103. Genetic vari- Women with PCOS usually have similar estrogen ants connected with a lower vitamin D range may levels, higher testosterone and lower progesterone play a role in autoimmune disease onset. Some levels compared to healthy women54. In PCOS research has displayed that vitamin D deficiency patients with oligomenorrhoea or anovulatory cy- is connected with autoimmune thyroid disease cles, periods of low progesterone concentration and risk and thyroid dysfunction severity as well as high estrogen-to-progesterone ratio occur frequent- with anti-TPO presence104. The degree of vitamin ly. Considering the previously discussed impact of D deficiency is correlated with hypothyreosis estrogens and progesterone on autoimmunity, it is severity. The probable explanation would be vi- justified that this situation may lead to increased sus- tamin D’s effect on VDR expression in dendritic ceptibility to autoimmune disorders54,65. Androgens, cells. VDR stimulation induces self-tolerance, on the contrary, suppress activity; acting through CD4+ CD25+ FOXP3+ Treg dif- however, it appears that they are not potent enough ferentiation, which is responsible for the periph- to overcome estrogen-induced action. Therefore, the eral action. In fact, vitamin D supplementation in imbalance between estrogens, progesterone and an- healthy subjects significantly increased the Treg drogens in PCOS may constitute a predisposing fac- percentage in relation to starting levels105. None- tor to HT development. Referring this hypothesis to theless, the following researchers did not confirm the different PCOS phenotypes, the highest risk of the association with thyroid dysfunction severi- autoimmune disease would be expected in patients ty106. In a recent study no differences regarding with anovulation and without hyperandrogenism. vitamin D level in patients with HT, Graves’ Patients presenting the classic phenotype with oli- disease and control groups were observed107. A gomenorrhoea and hyperandrogenism would have statistically significant correlation between an- an intermediate risk, and the lowest susceptibility ti-TPO and vitamin D deficiency in women, but would be expected in patients with ovulatory cy- only in a premenopausal group, has been found in cles and hyperandrogenism65. A study comparing a large Korean population study. It may indicate PCOS phenotypes in the context of HT has not a joint estrogen and vitamin D contribution to been conducted yet, although it has been evidenced thyroid autoimmunity108. that the estrogen-to-progesterone ratio correlates Vitamin D deficiency is also believed to play a with the anti-TPO concentration in women with role in women with PCOS and could be the miss- PCOS. No dependence regarding androgen levels ing link in the explanation of metabolic disorders. has been shown99. The results of a study comparing The relationship between low vitamin D level and infertile women with PCOS and infertile women the components of metabolic syndrome has been with regular cycles are also closely related. Sub- shown109. VDR gene polymorphism and genes clinical hypothyroidism with TSH > 2.5 µIU/mL related to vitamin D correlated with hormonal and has been found significantly more often in PCOS metabolic parameters in PCOS; however, their patients than in regularly menstruating patients. No direct effect on predisposition to develop PCOS correlation between TSH and androgens has been has not been confirmed105. Only in the Iranian unveiled, yet TSH level correlated positively with study were VDR ApaI polymorphisms found to insulin resistance and BMI100. predispose to PCOS. Vitamin D range correlates negatively with The (Controversial) Role of Vitamin D insulin resistance, hyperinsulinemia and hyper- The positive impact of vitamin D on the im- androgenism109. However, it is worth mentioning mune system has been shown and it is believed that this correlation was not confirmed in lean that it can prevent autoimmunity development101. women with PCOS110. Vitamin D supplementa- Inadequate vitamin D intake may be connected tion is supposed to adjust menstrual regularity with a higher incidence of autoimmune diseas- and normalize the AMH level111. es. Vitamin D binds with VDR, which has been The effect of vitamin D supply was also stud- found in many tissues including , ied with regard to factors lately associated with monocytes and dendritic cells101. PCOS: advanced glycation end products (AGEs), Two polymorphisms connected with vitamin D markers of inflammation and oxidative stress. – VDR gene and CYP27B1, hydroxylase convert- Supplementation of vitamin D increased the lev- ing hydroxyvitamin D3 (25(OH)D) into an active el of soluble receptors of AGEs (which binds

354 Thyroid disorders in polycystic ovary syndrome and deactivates them)112. In the Iranian research level is not able to reduce the stimulating impact vitamin D and calcium supplementation led to of estrogens on autoimmunity. a decrease of inflammation and oxidative stress Vitamin D deficiency is related to the patho- markers113. genesis of autoimmune thyroid disorders as well However, in systematic reviews on the role of as PCOS. However, from the research available vitamin D in metabolic and hormonal disorders in at present, we cannot draw many certain conclu- PCOS, published in 2013 and 2015, it was point- sions. In relation to the latest publications, we ed out that the results of the research included can also discuss whether vitamin D deficiency is were inconclusive, mainly due to methodolog- a consequence of PCOS or just coexists with it. ical reasons. The decision considering routine To elucidate this issue and the differences caused screening of vitamin D level in PCOS patients by seasons of the year, various definitions of de- and eventual supplementation can be made after ficiency and diverse analytical methods in a ran- conducting double-blind randomized trials114,115. domized population study should be performed. The mechanism of more pronounced metabolic disorders in cases of coexistence of both disorders Conclusions compared to single occurrence requires further explanation. Thyroid disorders, especially HT, in the course To sum up, it seems that HT and PCOS may of PCOS, are observed significantly more fre- be closely related, which reflects frequent com- quently than in the general population. However, mon prevalence. The data concerning fertility their concomitance in the context of fertility dis- issues, joint immune background, the role of sex orders has not been investigated extensively. hormones, vitamin D and the cause of metabol- The data concerning joint etiology, pathogen- ic disorder exacerbation still remain ambiguous. esis and clinical consequences are scarce. This Further, detailed research is needed to be carried is probably caused by complex etiology of both out in these fields. diseases and changes of PCOS diagnostic criteria over time. Genetic susceptibility contributes to the de- Conflict of Interest velopment of both diseases; however, a com- The authors declare no conflicts of interest. mon genetic background has not been discovered. Three genetic polymorphisms: FBN3, GnRHR References and CYP1B1, have been described to play a role in PCOS as well as in HT. Polymorphism of the 1) Rotterdam ESHRE/ASRM-Sponsored PCOS consensus FBN3 gene seems to be most suitable, due to the workshop group. Revised 2003 consensus on dia- relationship with TGFß activity, and Treg level gnostic criteria and long-term health risks related in consequence. The role of TGFß in the patho- to polycystic ovary syndrome (PCOS). Human genesis of both disorders has been understood. Reprod 2004; 19: 41-47. Among patients with PCOS only A8 D19S884 2) Lauritsen MP, Bentzen JG, Pinborg A, Loft A, Forman JL, Thuesen LL, Cohen A, Hougaard DM, Nyboe Ander- carriers had a normal TGFß1 level; others have sen A. The prevalence of polycystic ovary syndrome been found to have increased values compared in a normal population according to the Rotterdam with the general population. The first ones would criteria versus revised criteria including anti-Mulle- be more prone to develop autoimmune diseases. rian hormone. Hum Reprod 2014; 29: 791-801. High estrogen levels during prenatal life may 3) Hollowell JG, Staehling NW, Flanders WD, Han- disrupt development of the thymus and its role non WH, Gunter EW, Spencer CA, Braverman LE. Serum TSH, T4, and thyroid antibodies in the in creating immune tolerance – by that means it United States population (1988 to 1994): Natio- could contribute to both HT and PCOS devel- nal Health and Nutrition Examination Survey opment. (NHANES III). J Clin Endocrinol Metab 2002; 87: Women with PCOS, compared to healthy 489-499. women, may be more susceptible to the devel- 4) Canaris GJ, Manowitz NR, Mayor G, Ridgway EC. opment of HT because of imbalance between The Colorado thyroid disease prevalence study. estrogens, progesterone and androgens. Probably, Arch Intern Med 2000; 160: 526-534. Hulchiy M, Zhang H, Cline JM, Hirschberg AL, Sahlin the estrogen level is not opposed adequately by 5) L. Receptors for thyrotropin-releasing hormones, progesterone, which can be low due to anovula- thyroid-stimulating hormones, and thyroid hormo- tion. Apparently, even a relatively high androgen nes in the macaque uterus: effects of long-term

355 K. Kowalczyk, G. Franik, D. Kowalczyk, D. Pluta, Ł. Blukacz, P. Madej

sex hormones treatment. 2012; 19: 19) Stagnaro-Green A, Glinoer D. Thyroid autoimmuni- 1253-1259. ty and the risk of miscarriage. Best Pract Res Clin 6) Catalano RD, Critchley HO, Heikinheimo O, Baird DT, Endocrinol Metab 2004; 18: 167-181. Hapangama D, Sherwin JR, Charnock-Jones DS, Smith 20) Yan J, Sripada S, Saravelos SH, Chen ZJ, Egner W, SK, Sharkey AM. Mifepristone induced progestero- Li TC. antibody in women ne withdrawal reveals novel regulatory pathways with unexplained : prevalen- in human endometrium. Mol Hum Reprod 2007; ce, prognostic value, and response to empirical 13: 641-654. thyroxine therapy. Fertil Steril 2012; 98: 378-382. 7) Aghajanova L, Stavreus-Evers A, Lindeberg M, Land- 21) Negro R, Formoso G, Mangieri T, Pezzarossa A, Dazzi gren BM, Sparre LS, Hovatta O. Thyroid-stimulating D, Hassan H. Levothyroxine treatment in euthyroid hormone receptor and thyroid hormone receptors pregnant women with autoimmune thyroid dise- are involved in human endometrial physiology. ase: effects on obstetrical complications. J Clin Fertil Steril 2011; 95: 230-237. Endocrinol Metab 2006; 91: 2587-2591. 8) Zhang SS, Carrillo AJ, Darling DS. Expression 22) Poppe K, Glinoer D, Tournaye H, Devroey P, van Steir- of multiple mRNAs in teghem A, Kaufman L, Velkeniers B. Assisted repro- human oocytes, cumulus cells and granulosa duction and thyroid autoimmunity: an unfortunate cells. Mol Hum Reprod 1997; 3: 555-562. combination? J Clin Endocrinol Metab 2003; 88: 9) Costa NN, Cordeiro MS, Silva TV, Sastre D, Santana 4149-4152. PP, Sá A.L, Sampaio RV, Santos SS, Adona PR, Miran- 23) Reh A, Chaudhry S, Mendelsohn F, Im S, Rolnitzky L, da MS, Ohashi OM. Effect of triiodothyronine on Amarosa A, Levitz M, Srinivasa S, Krey L, Berkeley AS, developmental competence of bovine oocytes. Grifo JA, Danoff A. Effect of autoimmune thyroid Theriogenology 2013; 9: 295-301. disease in older euthyroid infertile woman during 10) Fitko R, Kucharski J, Szlezyngier B. The importance of the first 35 days of an IVF cycle. Fertil Steril 2001; thyroid hormone in experimental ovarian cyst for- 95: 1178-1181. mation in gilts. Anim Reprod Sci 1995; 39: 159-168. 24) Muller AF, Verhoeff A, Mantel MJ, Berghout A. Poppe K, Velkeniers B, Glinoer D. 11) Thyroid disease Thyroid autoimmunity and abortion: a prospective and female reproduction. Clin Endocrinol 2007; study in women undergoing in vitro fertilization. 66: 309-321. Fertil Steril 1999; 71: 30-34. 12) Medenica S, Nedeljkovic O, Radojevic N, Stojkovic 25) Thangaratinam S, Tan A, Knox E, Kilby MD, Franklyn M, Trbojevic B, Pajovic B. Thyroid dysfunction J, Coomarasamy A. Association between thyroid and thyroid autoimmunity in euthyroid women in achieving fertility. Eur Rev Med Pharmacol Sci autoantibodies and miscarriage and preterm bir- 2015; 6: 977-987. th: meta-analysis of evidence. Br Med J 2011; 5: 9-342. 13) Lambert M, Hocké C, Jimenez C, Frantz S, Papaxan- Plowden TC, Schisterman EF, Sjaarda LA, Zarek thos A, Creux H. Repeated in vitro fertilization 26) SM, Perkins NJ, Silver R, Galai N, DeCherney AH, failure: Abnormalities identified in the diagno- Mumford SL. stic assessment. Gynecol Obstet Fertil 2016; 9: Subclinical hypothyroidism and 30214-30224. thyroid autoimmunity are not associated with 14) Saglam F, Onal ED, Ersoy R, Koca C, Ergin M, Erel fecundity, pregnancy loss, or live birth. J Clin O, Cakir B. Anti-Müllerian hormone as a marker Endocrinol Metab 2016; 6: 2358-2365. of premature ovarian aging in autoimmune thyroid 27) De Groot L, Abalovich M, Alexander EK, Amino N, disease. Gynecol Endocrinol 2015; 2: 165-168. Barbour L, Cobin RH, Eastman CJ, Lazarus JH, Luton D, Mandel SJ, Mestman J, Rovet J, Sullivan S. 15) Tuten A, Hatipoglu E, Oncul M, Imamoglu M, Acik- Mana- goz AS, Yilmaz N, Ozcil MD, Kaya B, Misirlioglu AM, gement of thyroid dysfunction during pregnancy Sahmay S. Evaluation of ovarian reserve in Hashi- and postpartum: an Endocrine Society clinical moto’s thyroiditis. Gynecol Endocrinol, 2014; 10: practice guideline. J Clin Endocrinol Metab 2012; 708-711. 8: 2543-2565. 16) Henes M, Froeschlin J, Taran FA, Brucker S, Rall KK, 28) Fauser BC, Tarlatzis BC, Rebar RW, Legro RS, Balen Xenitidis T, Igney-Oertel A, Lawrenz B, Henes JC. Ova- AH, Lobo R, Carmina E, Chang J, Yildiz BO, Laven rian reserve alterations in premenopausal women JS, Boivin J, Petraglia F, Wijeyeratne CN, Norman RJ, with chronic inflammatory rheumatic diseases: Dunaif A, Franks S, Wild RA, Dumesic D, Barnhart K. impact of rheumatoid arthritis, Behçet’s disease Consensus on women’s health aspects of polycy- and spondyloarthritis on anti-Müllerian hormone stic ovary syndrome (PCOS): the Amsterdam levels. Rheumatology 2015; 9: 1709-1712. ESHRE/ASRM-Sponsored 3rd PCOS Consensus 17) Sowinski J, Sawicka-Gutaj N, Gutaj P, Ruchała M. Workshop Group. Fertil Steril 2012; 1: 28-38. The role of free triiodothyronine in pathogenesis 29) Katulski K, Czyzyk A, Podfigurna-Stopa A, Genaz- of infertility in levothyroxine-treated women with zani AR, Meczekalski B. Pregnancy complications thyroid autoimmunity--a preliminary observatio- in polycystic ovary syndrome patients. Gynecol nal study. Gynecol Endocrinol 2015; 31: 116-118. Endocrinol 2015; 2: 87-91. 18) Prummel MF, Wiersinga WM. Thyroid autoimmuni- 30) Roos N, Kieler H, Sahlin L, Ekman-Ordeberg G, ty and miscarriage. Eur J Endocrinol 2004; 150: Falconer H, Stephansson O. Risk of adverse pre- 751-755. gnancy outcomes in women with polycystic ovary

356 Thyroid disorders in polycystic ovary syndrome

syndrome: population based cohort study. BMJ 44) Knudsen N, Laurberg P, Rasmussen LB, Bulow I, Per- 2011; 8: 343. rild H, Ovesen L, Jørgensen T. Small differences in 31) Cocksedge KA, Saravelos SH, Metwally M, Li TC. thyroid function may be important for body mass How common is polycystic ovary syndrome in index and the occurrence of obesity in the popula- recurrent miscarriage? Reprod Biomed Online tion. J Clin Endocrinol Metab 2005; 90: 4019-4024. 2009; 19: 572-576. 45) Dunaif A. Insulin resistance and the polycystic 32) Sugiura-Ogasawara M, Sato T, Suzumori N, Kitaori T, ovary syndrome: mechanism and implications for Kumagai K, Ozaki Y. The polycystic ovary syndrome pathogenesis. Endocr Rev 1997; 18: 774-800. does not predict further miscarriage in Japanese 46) Strowitzki T, Capp E, von Eye Corleta H. The degree couples experiencing recurrent miscarriages. Am of cycle irregularity correlates with the grade J Reprod Immunol 2009; 61: 62-67. of endocrine and metabolic disorders in PCOS 33) Chakraborty P, Goswami SK, Rajani S, Sharma S, patients. Eur J Obstet Gynecol Reprod Biol 2010; Kabir SN, Chakravarty B, Jana K. Recurrent pre- 149: 178-181. gnancy loss in polycystic ovary syndrome: role 47) Lerchbaum E, Schwetz V, Giuliani A, Obermayer-Piet- of hyperhomocysteinemia and insulin resistance. sch B. Hypertriglyceridemic waist is associated PLoS One 2013; 8: 64446. with impaired glucose tolerance in polycystic 34) khattab S, Mohsen IA, Foutouh IA, Ramadan A, ovary syndrome. Nutr Metab Cardiovasc Dis Moaz M, Al-Inany H. Metformin reduces abortion 2013; 23: 15-16. in pregnant women with polycystic ovary syndro- 48) Hudecova M, Holte J, Olovsson M, Larsson A, Berne me. Gynecol Endocrinol 2006; 22: 680-684. C, Sundstrom- Poromaa I. Prevalence of the meta- 35) Ramidi G, Khan N, Glueck CJ, Wang P, Goldenberg bolic syndrome in women with a previous diagno- N. Enoxaparin-metformin and enoxaparin alone sis of polycystic ovary syndrome: long-term follow may safely reduce pregnancy loss. Transl Res up. Fertil and Steril 2011; 96: 1271-1274. 2009; 153: 33-43. 49) Anaforoglu I, Topbas M, Algun E. Relative associa- 36) Mathur R, Alexander CJ, Yano J, Trivax B, Azziz R. tions of polycystic ovarian syndrome vs metabolic Use of metformin in polycystic ovary syndrome. syndrome with thyroid function, volume, nodula- Am J Obstet Gynecol 2008; 199: 596-609. rity and autoimmunity. J Endocrinol Invest 2011; 34: 295-264. 37) Qin JZ, Pang LH, Li MJ, Fan XJ, Huang RD, Chen HY. Mueller A, Schofl C, Dittrich R, Cupisti S, Oppelt PG, Obstetric complications in women with polycystic 50) Schild RL, Beckmann MW, Haberle L. ovary syndrome: a systematic review and meta-a- Thyroid-stimu- nalysis. Reprod Biol Endocrinol 2013; 11: 56-70. lating hormone is associated with insulin resistan- ce independently of body mass index and age 38) De Frène V, Vansteelandt S, T’Sjoen G, Gerris J, in women with polycystic ovary syndrome. Hum Somers S, Vercruysse L, De Sutter P. A retrospective Reprod 2009; 24: 2924-2930. study of the pregnancy, delivery and neona- tal outcome in overweight versus normal weight 51) Ganie MA, Marwaha RK, Aggarwal R, Singh S. women with polycystic ovary syndrome. Hum High prevalence of polycystic ovary syndrome Reprod 2014; 29: 2333-2338. characteristics in girls with euthyroid chronic lym- phocytic thyroiditis: a case-control study. Eur J 39) Naver KV, Grinsted J, Larsen SO, Hedley PL, Jørg- Endocrinol 2010; 162: 1117-1122. ensen FS, Christiansen M, Nilas L. Increased risk of preterm delivery and pre-eclampsia in women 52) Celik C, Abali R, Tasdemir N, Guzel S, Yuksel A, Aksu with polycystic ovary syndrome and hyperandro- E, Yilmaz M. Is subclinical hypothyroidism con- genaemia. BJOG 2014; 121: 575-581. tributing dyslipidemia and insulin resistance in 40) Mumm H, Jensen DM, Sørensen JA, Andersen LL, women with polycystic ovary syndrome? Gynecol Ravn P, Andersen M, Glintborg D. Hyperandrogeni- Endocrinol 2012; 28: 615-618. sm and phenotypes of polycystic ovary syndrome 53) Taghavi SM, Rokni H, Fatemi S. Metformin decrea- are not associated with differences in obstetric ses thyrotropin in overweight women with polycy- outcomes. Acta Obstet Gynecol Scand 2015; 94: stic ovarian syndrome and hypothyroidism. Diab 204-211. Vasc Dis Res 2011; 8: 47-48. 41) Ott J, Aust S, Kurz C, Nouri K, Wirth S, Huber 54) Janssen OE, Mehlmauer N, Hahn S, Offner AH, Gar- JC, Mayerhofer K. Elevated antithyroid peroxidase tner R. High prevalence of antibodies indicating Hashimoto’s thyroiditis are in patients with polycystic ovary syndrome. Eur J associated with the treatment response in inferti- Endocrinol 2004; 150: 363-369. le women with polycystic ovary syndrome. Fertil 55) Garelli S, Masiero S, Plebani M, Chen S, Furmaniak J, Steril 2010; 94: 2895-2897. Armanini D, Betterle C. High prevalence of chronic 42) Cooper DS, Biondi B. Subclinical thyroid disease. thyroiditis in patients with polycystic ovary syn- Lancet 2012; 379: 1142-1154. drome. Eur J Obstet Gynecol Reprod Biol 2013; 43) Asvold BO, Vatten LJ, Nilsen TI, Bjøro T. The asso- 169: 248-251. ciation between TSH within reference range and 56) Novais Jde S, Benetti-Pinto CL, Garmes HM, Jales serum lipid concentrations in a population-based RM, Juliato CR. Polycystic ovary syndrome and study. The HUNT study. Eur J Endocrinol 2007; chronic autoimmune thyroiditis. Gynecol Endocri- 156: 181-186. nol 2015; 31: 48-51.

357 K. Kowalczyk, G. Franik, D. Kowalczyk, D. Pluta, Ł. Blukacz, P. Madej

57) Du D, Xuelian L. The relationship between thyroi- 71) Sakaguchi S, Yamaguchi T, Nomura T, Ono M. Regu- ditis and polycystic ovary syndrome: a meta-a- latory T cells and immune tolerance. Cell 2008; nalysis. Int J Clin Exp Med 2013; 6: 880-889. 133: 775 -787. 58) Dittmar M, Libich C, Brenzel T, Kahaly GJ. Increased 72) Chen W, Jin W, Hardegen N, Lei KJ, Li L, Marinos familial clustering of autoimmune thyroid disea- N, McGrady G, Wahl SM. Conversion of periphe- ses. Horm Metab Res 2011; 43: 200-204. ral CD4CCD25K naive T cell to CD4CCD25C regulatory T cells by TGF-b induction of tran- 59) Zaletel K, Gaberscek S. Hashimoto’s thyroiditis: scription factor Foxp3. J Expl Med 2003; 198: from genes to the disease. Curr Genomics 2011; 1875-1886. 12: 576-588. 73) Chapman JC, Kunaporn S, Shah S, Kaiki-Astara A, Dong YH, Fu DG. 60) Autoimmune thyroid disease: Michael SD. The differential effect of injecting mechanism, genetics and current knowledge. Eur estradiol-17b, testosterone, and hydrocortisone Rev Med Pharmacol Sci 2014; 23: 3611-3618. during the immune adaptive period on the fertility 61) Chen ZJ, Zhao H, He L, Shi Y, Qin Y, Shi Y, Li Z, You of female mice. Am J Reprod Immunol 2001; 46: L, Zhao J, Liu J. et al. Genome-wide association 288-297. study identifies susceptibility loci for polycystic 74) Chapman JC, Min SH, Freeh SM, Michael SD. The ovary syndrome on chromosome 2p16.3, 2p21 estrogen injected female mouse: new insight into and 9q33.3. Nat Genet 2011; 43: 55-59. the etiology of PCOS. Reprod Biol Endocrinol 62) Raja-Khan N, Urbanek M, Rodgers RJ, Legro RS. 2009; 7: 47-57. The role of TGF-b in polycystic ovary syndrome. 75) Cruz G, Barra R, González D, Sotomayor-Zárate R, Reprod Sci 2014; 21: 20-31. Lara HE. Temporal window in which exposure to 63) Raja-Khan N, Kunselman AR, Demers LM, Ewens KG, estradiol permanently modifies ovarian function Spielman RS, Legro RS. A variant in the fibrillin-3 causing polycysticovary morphology in rats. Fertil gene is associated with TFG-b and inhibin B Steril 2012; 5: 1283-1290 levels in womenwith polycystic ovary syndrome. 76) Sotomayor-Zárate R, Dorfman M, Paredes A, Lara Fertil Steril 2010; 94: 2916-2919. HE. Neonatal exposure to estradiol valerate pro- 64) Gaberscek S, Zaletel K, Schwetz V, Pieber T, Ober- grams ovarian sympathetic innervation and folli- mayer-Pietsch B, Lerchbaum E. Mechanisms in Endo- cular development in the adult rat. Biol Reprod crinology. Thyroid and polycystic ovary syndro- 2008; 4: 673-680. me. Eur J Endocrinol 2015; 172: 9-21. 77) Sotomayor-Zárate R, Tiszavari M, Cruz G, Lara HE. 65) Kosova G, Urbanek M. Genetics of the polycystic Neonatal exposure to single doses of estradiol or ovary syndrome. Mol Cell Endocrinol 2013; 373: testosterone programs ovarian follicular develop- 29-38. ment-modified hypothalamic neurotransmitters 66) Bobik A. Transforming growth factor-betas and and causes polycystic ovary during adulthood in vascular disorders. Arterioscler Thromb Vasc Biol the rat. Fertil Steril 2011; 6: 1490-1496. 2006; 26: 1712-1720. 78) Palmer JR, Hatch EE, Rao RS, Kaufman RH, Herbst AL, Noller KL, Titus-Ernstoff L, Hoover RN. 67) Irani M, Seifer DB, Grazi RV, Julka N, Bhatt D, Kalgi Inferti- B, Irani S, Tal O, Lambert-Messerlian G, Tal R. Vita- lity among women exposed prenatally to diethyl- min D supplementation decreases TGF-β1 bioa- stilbestrol. Am J Epidemiol 2001; 154: 316-321. vailability in PCOS: a randomized placebo-con- 79) Noller KL, Blair PB, O’Brien PC, Melton LJ III, Offord trolled trial. J Clin Endocrinol Metab 2015; 100: JR, Kaufman RH, Cotton T. Increased occurrence 1307-1404. of autoimmune disease among women exposed 68) Akinci B, Comlekci A, Yener S, Bayraktar F, Demir in utero to diethylstilbestrol. Fertil Steril 1988; 49: T, Ali Ozcan M, Yuksel F, Yesil S. Hashimoto’s 1080-1082. thyroiditis, but not treatment of hypothyroidism, is 80) Hague WM, Adams J, Rodda C, Brook CG, de Bruyn associated with altered TGF-b1 levels. Arch Med R, Grant DB, Jacobs HS. The prevalence of polycy- Res 2008; 39: 397-401. stic ovaries in patients with congenital adrenal 69) Li Q. Yang G, Wang Y, Zhang X, Sang Q, Wang hyperplasia and their close relatives. Clin Endo- H, Zhao X, Xing Q, He L, Wang L. Common crinol 1990; 4: 501-510. genetic variation in the 30-untranslated region of 81) Veiga-Lopez A, Ye W, Phillips DJ, Herkimer C, gonadotropin-releasing hormone receptor regu- Knight PG, Padmanabhan V. Developmental pro- lates gene expression in cella and is associated gramming: deficits in reproductive hormone with thyroid function, insulin secretion as well as dynamics and ovulatory outcomes in prenatal, insulin sensitivity in polycystic ovary syndrome testosterone-treated sheep. Biol Reprod 2008; patients. Hum Genet 2011; 129: 553-561. 4: 636 - 647. 70) Zou S, Sang Q, Wang H, Zhang X, Sang Q, Wang H, 82) Yan X, Dai X, Wang J, Zhao N, Cui Y, Liu J. Prenatal Zhao X, Xing Q, He L, Wang L. Common genetic androgen excess programs metabolic derange- variation in CYP1B1 is associated with concentra- ments in pubertal female rats. J Endocrinol 2013; tions of T4, FT3 and fT4 in the sera of polycystic 1: 119-129. ovary syndrome patients. Mol Biol Rep 2013; 40: 83) Xu N, Kwon S, Abbott DH, Geller DH, Dumesic DA, 3315-3320. Azziz R, Guo X, Goodarzi MO. Epigenetic mecha-

358 Thyroid disorders in polycystic ovary syndrome

nism underlying the development of polycystic 99) Arduc A, Aycicek Dogan B, Bilmez S, Imga Nasiroglu ovary syndrome (PCOS) like phenotypes in pre- N, Tuna MM, Isik S, Berker D, Guler S. High preva- natally androgenized rhesus monkeys. PLoS One lence of Hashimoto’s thyroiditis in patients with 2011; 11: e27286. polycystic ovary syndrome: does the imbalance 84) Quintero OL, Amador-Patarroyo MJ, Montoya-Ortiz between estradiol and progesterone play a role? G, Rojas- Villarraga A, Anaya JM. Autoimmune Endocr Res 2015; 4: 204-210. disease and gender: plausible mechanism for the 100) Morgante G, Musacchio MC, Orvieto R, Massaro female predominance of autoimmunity. J Autoim- MG, De Leo V. Alterations in thyroid function mun 2011; 38: 109-119. among the different polycystic ovary syndrome phenotypes. Gynecol Endocrinol 2013; 11: 967- 85) Mariotti S, Prinzis A, Ghiani M, Cambuli VM, Pilia S, 969. Marras V, Carta D, Loche S. Puberty is associated Pludowski P, Holick MF, Pilz S, Wagner CL, Hollis with a marked increase of the female sex predo- 101) BW, Grant WB, Shoenfeld Y, Lerchbaum E, Llewel- minance in chronic autoimmune thyroiditis. Horm lyn DJ, Kienreisch K, Soni M. Res 2009; 72: 52-56. Vitamin D effects on musculoskeletal health, immunity, autoim- Ahmed SA, Hissong BD, Verthelyi D, Dooner K, Becker K, 86) munity, cardiovascular disease, cancer, fertility, Karpozuglu-Sahin E . Gender and risk of autoimmune pregnancy, dementia and mortality – a review diseases: possible role of estrogenic compounds. of recent evidence. Autoimmun Rev 2013; 12: Environ Health Perspect 1999; 107: 681-686. 976-989. 87) Tornwall J, Carey AB, Fox RI, Fox HS. Estrogen in 102) Lopez ER, Zwermann O, Segni M, Meyer G, Reincke autoimmunity: expression of estrogen receptors M, Seissler J, Herwig J, Usadel KH, Badenhoop K. A in thymic and autoimmune T cells. J Gend Specif promoter polymorphism of the CYP27B1 gene is Med 1999; 2: 33-40. associated with Addison’s disease, Hashimoto’s 88) Zorgdrager A, De Keyser J. Menstrually related thyroiditis, Graves’ disease and worsening of symptoms in multiple sclerosis. J mellitus in Germans. Eur J Endocrinol 2004; Neurol Sci 1997; 149: 95-97. 151: 193-197. Lin MW, Tsai SJ, Chou PY, Huang MF, Sun HS, Wu Confavreux C, Hutchinson M, Hours MM, Cortino- 103) 89) MH. vis-Tourniaire P, Moreau T. Rate of pregnancy-related Vitamin D receptor 1a promotor K1521 G/C and K1012 A/G polymorphisms in polycystic relapse in multiple sclerosis. Pregnancy in multiple ovary syndrome. Taiwan J Obstet Gynecol 2012; sclerosis group. N Engl J Med 1998; 339: 285-291. 51: 565-571. Seli E, Arici A. 90) Sex steroids and the immune system. 104) Shin DY, Kim KJ, Kim D, Hwang S, Lee EJ. Low Immunol Allergy Clin North Am 2002; 22: 407-433. serum vitamin D is associated with anti-thyroid 91) Hughes GC. Progesterone and autoimmune dise- peroxidase antibody in autoimmune thyroiditis. ase. Autoimmun Rev 2012; 11: 502-514. Yonsei Med J 2014; 55: 476-481. 92) Pennell LM, Galligan CL, Fish EN. Sex affects 105) Prietl B, Pilz S, Wolf M, Tomaschitz A, Ober- immunity. J Autoimmun 2012; 38: 282-291. mayer-Pietsch B, GraningerW, Pieber TR. Vitamin D 93) Angstwurm MW, Gartner R, Ziegler-Heitbrock HW. supplementation and regulatory T cells in appa- Cyclic plasma IL-6 levels during normal menstrual rently healthy subjects: vitamin D treatment for cycle. Cytokine 1997; 9: 370-374. autoimmune diseases? Isr Med Assoc J 2010; 12: 136-139. 94) Lee JH, Ulrich B, Cho J, Park J, Kim CH. Progestero- 106) Tamer G, Arik S, Tamer I, Coksert D. Relative vita- ne promotes differentiation of human cord blood min D insufficiency in Hashimoto’s thyroiditis. fetal T cells into T regulatory cells but suppresses Thyroid 2011; 21: 891-896. their differentiation into Th17 cells. J Immunol D’Aurizio F, Villalta D, Metus P, Doretto P, Toz- 2011; 187: 1778-1787. 107) zoli R. Is vitamin D a player or not in the Lee JH, Lydon JP, Kim CH 95) . Progesterone suppres- pathophysiology of autoimmune thyroid disea- ses the mTOR pathway and promotes generation ses? Autoimmun Rev 2015; 5: 363-369. of induced regulatory T cells with increased sta- 108) Choi YM, Kim WG, Kim TY, Bae SJ, Kim HK, Jang bility. Eur J Immunol 2012; 10: 2683-2696. EK, Jeon MJ, Han JM, Lee SH, Baek JH, Shong 96) Wegmann TG, Lin H, Guilbert L, Mosmann TR. Bidi- YK, Kim WB. Low levels of serum vitamin D3 are rectional cytokine interactions in the maternal-fe- associated with autoimmune thyroid disease in tal relationship: is successful pregnancy a TH2 pre-menopausal women. Thyroid 2014; 24: 655- phenomenon? Immunol Today 1993; 14: 353-356. 661. 97) Gaberscek S, Zaletel K. Thyroid physiology and 109) Wehr E, Pilz S, Schweighofer N, Giuliani A, Kopera autoimmunity in pregnancy and after delivery. D, Pieber TR, Obermayer-Pietsch B. Association of Exp Rev Clin Immunol 2011; 7: 697-707. hypovitaminosis D with metabolic disturbance 98) Jørgensen KT, Pedersen BV, Nielsen NM, Jacobsen in polycystic ovary syndrome. Eur J Endocrinol F, Frisch M. and risk of female pre- 2009; 161: 575-582. dominant and other autoimmune diseases in a 110) Sahin S, Eroglu M, Selcuk S, Turkgeldi L, Kozali S, population-based Danish cohort. J Autoimmun Davutoglu S, Muhcu M. Intrinsic factors rather 2012; 38: 81-87. than vitamin D deficiency are related to insulin

359 K. Kowalczyk, G. Franik, D. Kowalczyk, D. Pluta, Ł. Blukacz, P. Madej

resistance in lean women with polycystic ovary influences biomarkers of inflammation and oxi- syndrome. Eur Rev Med Pharmacol Sci 2014; dative stress in overweight and vitamin D-de- 19: 2851-2856. ficient women with polycystic ovary syndrome: 111) Rashidi B, Haghollahi F, Shariat M, Zayerii F. The a randomized double-blind placebo-controlled effects of calcium-vitamin D and metformin on clinical trial. Clin Endocrinol 2015; 83: 888-894. polycystic ovary syndrome: a pilot study. Taiwan 114) He C, Lin Z, Robb SW, Ezeamama AE. Serum vita- J Obstet Gynecol 2009; 2: 142-147. min D levels and polycystic ovary syndrome: a 112) Irani M, Minkoff H, Seifer DB, Merhi Z. Vitamin D systematic review and meta-analysis. Nutrients Increases Serum Levels of the Soluble Receptor for 2015; 6: 4555-4577. Advanced Glycation End Products in Women With 115) Krul-Poel YH, Snackey C, Louwers Y, P, Lambalk PCOS J Clin Endocrinol Metab 2014; 99: 886-890. CB, Laven JS, Simsek S. The role of vitamin D in 113) Foroozanfard F, Jamilian M, Bahmani F, Talaee R, metabolic disturbances in polycystic ovary syn- Talaee N, Hashemi T, Nasri K, Asemi Z, Esmaillza- drome: a systematic review. Eur J Endocrinol deh A. Calcium plus vitamin D supplementation 2013; 6: 853-865.

360