全文

1. Introduction

Uterine leiomyomas are benign tumors of the uterine smooth muscle cells. They are clinically apparent in up to 25% of reproductive-age women. Heavy or abnormal uterine bleeding, pelvic pain, infertility, and recurrent pregnancy loss are generally associated with leiomyomas [1].

 

Although leiomyomas are the most common gynecologic tumors in premenopausal women, their etiology is still under investigation. To date, genetic and epigenetic factors, sex steroids, growth factors, cytokines, chemokines, and extracellular matrix components are important factors thought to be involved in the pathogenesis of leiomyomas. The fact that fibroids occur during the reproductive period and regress after menopause indicates a dependency on the influence of ovarian hormones [2].

 

In addition to uterine leiomyomas, thyroid gland dysfunction is commonly observed within the subfertile population. Thyroid function is suspected to be crucially involved in reproductive processes and dysfunction. Primarily, hypothyroidism has been proposed to be associated with several etiologies of infertility, including impaired ovulation, fertilization, and implantation, as well as a risk factor for miscarriage and late pregnancy complications [3]. The most prevalent cause of subclinical or overt hypothyroidism and one of the most common autoimmune endocrine disorders in the female population is Hashimoto's thyroiditis, characterized by the presence of antibodies against thyroid peroxidase (TPO-Ab) and/or antibodies against thyroglobulin (TG-Ab) [4].

 

A possible correlation between uterine leiomyoma and thyroid disease was reported decades ago. In 1989, Lange and Meinen focused on thyroid dysfunction in an examination of 79 women who had undergone hysterectomy. Women with leiomyomas had pathological TRH-stimulation test results and thyroid antibodies more often than women without leiomyomas [5]. Recent reports on this topic are scarce. A possible relationship between lipoleiomyoma, an unusual uterine fatty tumor, and concomitant metabolic disorders, e.g. diabetes mellitus and hypothyreosis, has also been described [6] and [7].

 

Given the high prevalence and clinical relevance of both diseases and the indications in the literature of a possible link, we intended to systematically investigate the possible association between uterine leiomyomas and thyroid function, including thyroid antibodies, in a population of subfertile women.

 

2. Materials and methods

2.1. Study population and design

In a retrospective study, all women with primary or secondary sterility who underwent reproductive surgery by hysteroscopy and laparoscopy/laparotomy at the Medical University of Vienna, Department of Gynecology and Obstetrics, Vienna, Austria, from January 2007 to January 2011, were included. The study aims were to evaluate possible associations between the presence of uterine leiomyomas and (i) the presence of overt hypothyroidism (primary objective), (ii) the level of anti-thyroid peroxidase antibodies (TPO-Ab) and thyroglobulin antibodies (TG-Ab), and (iii) thyroid-stimulating hormone (TSH) levels, in women who had not been pre-treated with thyroid hormone supplementation (secondary objectives). At the initial visit, overt hypothyroidism was defined as follows: (i) if TSH and levo-thyroxin (T4) were >5.0 μIU/ml and <64.5 nmol/l, respectively, or (ii) if the woman was under treatment with thyroid hormone supplementation due to overt hypothyroidism as defined above. If women were pre-treated, medical records providing TSH and T4 levels were obtained in order to verify the diagnosis of overt hypotyhroidism.

 

Data were retrieved by retrospective chart review. In accordance with a recent comprehensive review [8], we included the following available parameters as risk factors for uterine leiomyomas in addition to thyroid parameters (TSH, TPO-Ab, TG-Ab): age, African heritage, age at menarche, parity and body mass index (BMI). In addition, we noted the type of sterility (primary/secondary).

 

No patient had received hormonal therapy for myomas before the operation. All women were premenopausal. The Institutional Review Board of the Medical University of Vienna approved the study.

 

2.2. Diagnosis of uterine leimomyomas

All women underwent preoperative transvaginal ultrasound examination, performed by highly experienced operators. Gynecological transvaginal ultrasound examinations are subject to internal quality assurance. All myomas suspected on gynecologic ultrasound could be verified during surgery. The exact size of the leiomyomas—the secondary outcome parameter—was retrieved either from histologic reports if the tumor had been removed, or from preoperative ultrasound.

 

Submucosal uterine leiomyomas were considered relevant for infertility [9].

 

2.3. Laboratory analyses

Hormonal parameters were assessed within the three months prior to surgery in a morning-fasted state. Samples were obtained during the first six days of the menstrual cycle. All serum parameters were determined using commercially available assays. At our department, the normal ranges are 0.44–3.77 μU/ml for TSH, 0–34 IU/ml for TPO-Ab, and 0–33 IU/ml for TG-Ab.

 

2.4. Statistical analysis

Nominal variables are reported as numbers and frequencies, and continuous variables as median and range. Statistical analyses were performed with the SPSS software package, version 19 (SPSS, Chicago). A logistic regression model with Wald's tests was used to test the statistical significance of all coefficients. Adjusted odds ratios (OR) are given, including the 95 percent confidence interval (95% CI). To evaluate factors predicting leiomyoma size, a generalized linear model with Poisson link function was used. Differences between groups were tested with the Wilcoxon rank sum test with continuity correction. Differences were considered statistically significant if p was <0.05.

 

3. Results

A total of 215 women were included. Forty-six patients (21.4%) had missing values which affected only the parameters “age at menarche” and “body mass index”. Table 1 shows general patient characteristics. Eleven of the 16 women (68.8%) with overt hypothyroidism had been pre-treated with thyroid hormone supplementation for a median of six months (range, 2–17) at the time of the initial visit. One or more uterine leiomyomas were found in 51 cases (23.7%). The localization of the leiomyomas was as follows: there were 20 submucosal (39.2%) and 30 subserosal leiomyomas (58.8%) as well as one leiomyoma in the broad ligament (2.0%). Intra-abdominal abnormalities associated with infertility that were diagnosed/confirmed during surgery are listed in Table 2. In the group of patients with uterne leiomyoma, 31 (60.8%) underwent surgery for diagnostic reasons, which was comparable to the group of patients without uterine leiomyomas (n = 80, 48.8%, p = 0.134).

 

Details on the analysis of parameters associated with the presence of uterine leiomyoma are provided in Table 3. After multivariate analysis, three parameters remained significant, with African heritage the most important (OR 27.80), followed by overt hypothyroidism (OR 3.10) and increasing age (OR 1.23).

 

 

We then focused on women with uterine leiomyoma. When comparing women without overt hypothyroidism (n = 37) to those with overt hypothyroidism (n = 7), the latter revealed significantly larger myomas (median 70 mm, range 5–88; vs. median 30 mm, range 2–93, respectively; p = 0.007). The three parameters that were found to be significantly associated with the presence of uterine leiomyoma, namely age, African heritage, and overt hypothyroidism, were included in a generalized linear model to predict leiomyoma size. Overt hypothyroidism and increasing age were significantly associated with a larger leiomyoma size (p < 0.001).

 

4. Comment

In this retrospective study, after multivariate analysis overt hypothyroidism was found to be associated with the presence of uterine leiomyoma. Moreover, women with overt hypothyroidism revealed larger myomas than those without.

 

This is in accordance with a study that included women who underwent hysterectomy, published in 1989 [5]. The authors reported that, in women with leiomyomas, abnormal results of the TRH stimulation test were found more often than in controls (23% vs. 3.7%). Moreover, myoma patients were more frequently positive for microsomal antibodies and/or TG-Ab. The latter result was significant (p = 0.0019). This is somewhat in accordance with our data. In a univariate analysis, TPO-Ab and TG-Ab were more frequently elevated in women with a leiomyoma, but this difference did not remain significant after multivariate analysis, whereas overt hypothyroidism proved to be of influence. This suggests hypothyroidism as the major thyroid factor for the development of uterine leiomyoma. Thus, autoimmune thyroiditis, the most common cause of hypothyroidism [4], might only be a co-contributor.

 

In the 1940s, a link between thyroid disorders and uterine leiomyomas was often discussed in the literature. Specifically, an association between nodular goiter and the presence of myomas [10], as well as a high prevalence of hyperthyroidism, had been described [11]. We cannot comment on the latter, since there were no cases of hyperthyroidism, either pre-treated or untreated, in our study population. In 1948, a case series of 100 women with uterine leiomyoma was published, which demonstrated a high prevalence of thyroid disorders in this population, namely nodular goiter (30%), hyperthyroidism (29%), and hypothyroidism (12%). The study was limited by the lack of a control group. The article by Lange and Meinen, published in 1989, was the last study dealing with thyroid function and its association with uterine leiomyoma. Only a few case reports have been published since then [6] and [7].

 

Age is a known risk factor for hypothyroidism. In our analysis it was also a significant risk factor for the presence of leiomyomas. Since both parameters remained significant after multivariate analysis, both have to be seen as independent risk factors.

 

The present study explores possible explanations for the association between hypothyroidism and uterine leiomyoma. As extensively reviewed [12], leiomyomas express alpha and beta estrogen receptors and are thought to be estrogen-dependent. This mitogenic effect of estrogen on leiomyoma tissue is mediated through the up-regulation of several growth factors [13] and by the induction of progesterone receptors [14]. Although various growth factors have been suggested as possible promoters of leiomyoma development and growth [12], only tumor necrosis factor beta has been shown to be overexpressed in leiomyoma tissue compared to normal myometrium [15]. To date, however, there are no studies that have evaluated the expression of TSH or thyroid hormone receptors in leimyoma tissue. Thus, the impact of thyroid dysfunction on growth factors in leiomyoma tissue can only be hypothesized. Notably, a direct effect of elevated TSH levels on myoma growth seems questionable, since overt hypothyroidism was associated with only a moderate risk for uterine leiomyoma in our study (OR 1.30).

 

Since there are, however, significantly lower mean levels of estrogen and progesterone receptors in large myomas, as opposed to small myomas [16], it has been suggested that large nodes are likely not as dependent on the activity from the high levels of these receptors as they are on other factors [12]. This is supported by the observation that women with overt hypothyroidism revealed significantly larger leiomyomas in our study population.

 

It has also been hypothesized that leiomyomas could be caused, in part, by a systemic immune milieu that is chronically inflammatory with a predominance of T-helper cytokines [17]. Evidence suggests that several proinflammatory molecules are expressed during states of hypothyroidism [18]. This might also link thyroid dysfunction with the presence of uterine leiomyoma.

 

It is evident that thyroid metabolism has effects on the reproductive system. Hypothyroidism might lead to hypogonadotropic hypogonadism, but is also associated with polycystic ovary syndrome [19]. This suggests that thyroid disorders have some influence on the hypothalamic pituitary ovarian axis. Moreover, hypothyroidism can lead to an increase in prolactin levels. An increase in prolactin receptors in leimyoma tissue has been reported [20].

 

As mentioned by Lange and Meinen [5], one cannot rule out the possibility that fluctuations in the hypothalamo–pituitary–ovarian axis that lead to the development of uterine leiomyoma might also alter thyroid function. It is well known that the application of exogenous estrogen triggers the production of thyroxine-binding globulin and an increase in its binding avidity, which is accompanied by a marked inhibition of the peripheral turnover of thyroxine [21]. In subjects who are prone to thyroid disorders and might not be capable of compensating for these alterations, thyroid dysfunction could, hypothetically, be induced by sex-hormone dependent mechanisms.

 

One has to consider several possible weaknesses when interpreting the results of this study. First, one might argue that the vast majority of patients with overt hypothyroidism were already under thyroid hormone supplementation, which might have biased the results. However, both hypothyroidism and uterine leiomyoma develop over a long period, and, thus, likely existed before they had been diagnosed for the first time. We therefore, consider the association between the two diseases reasonable even though many women in our study had already been pre-treated for hypothyroidism. Second, one could be concerned about undetected leiomyomas. However, in all patients with leiomyomas that had been suspected on ultrasound, leiomyomas could be verified during hysteroscopy and laparoscopy/laparotomy. Thus, such issues should not have affected the results on the association between leiomyomas and hypothyroidism in this infertile patient population. Moreover, Austria is a region with a high percentage of benign euthyroid goiters, as in the majority of German-speaking countries. The lack of thyroid ultrasound examinations in our data set has to be seen as a study limitation, since we cannot comment on a hypothetical association between goiter and leiomyomas.

 

One could also argue that the study might be biased by the fact that only infertile women were included. However, patients were referred to our center for primary diagnostic evaluation of infertility and, thus, were not pre-selected for uterine leiomyomas, polycystic ovary syndrome or others. One could also be concerned about pre-selection of women who finally underwent surgery. On the one hand, thyroid dysfunction alone was the likely cause for infertility in some cases and these women became pregnant with thyroid hormone supplementation before surgery would have been recommended. This, however, would apply to patients both with and without clinically relevant leiomyomas and should not have introduced any kind of bias. On the other hand, women with leiomyomas were probably more likely to undergo surgery than other infertile patients. However, the rate of primary diagnostic laparoscopy/laparotomy did not differ between patients with and without uterine leiomyomas. This is supported by the fact that in >60% of cases uterine leiomyomas were not submucosal and, thus, considered of only minor relevance for infertility (Table 2). In consequence, we consider the inclusion of only infertile women only as a minor study limitation. From a scientific point of view, however, we can conclude that our data are applicable only to an infertile patient population, though it is likely that hypothyroidism would exert a general effect on leiomyoma growth in all women regardless of whether or not they suffer from infertility. Last not least, the retrospective design is a study limitation when interpreting our results.

 

In conclusion, overt hypothyroidism, but not autoantibodies against the thyroid gland that are indicative of Hashimoto's thyroiditis, were associated with the presence of uterine leiomyoma in our study. We hope these results will encourage researchers to investigate this topic. Further evidence is needed to clarify this issue, since the results likely are of clinical impact. If hypothyroidism promotes leiomyoma development, women with leiomyomas might be advised to undergo screening for thyroid dysfunction in order to treat hypothyroid patients early and, thereby, inhibit leiomyoma growth.

 

Conflict of interest statement

All authors declare that they have no commercial interest, financial interest, and/or another relationship with manufacturers of pharmaceuticals, laboratory supplies, and/or medical devices or with commercial providers of medically related services.

 

Acknowledgements

None.