Sex Hormones in Males and Females with Active Central Serous Chorioretinopathy Open Access

Central serous chorioretinopathy (CSC), the fourth most common wet maculopathy, is a chorioretinal disease characterized by subretinal fluid (SRF) accumulation [1, 2]. The SRF most often occurs in the central part of the retina, and persistent SRF can lead to atrophy of the photoreceptors causing an irreversible decrease in vision-related quality of life [1, 3]. CSC is primarily caused by abnormalities in the choroid; the vascular layer of the eye located underneath the retina. These choroidal abnormalities include hyperpermeability of the capillaries and congestion of blood flow with strongly dilated veins [2, 4‒6]. Even if SRF accumulation occurs unilaterally, choroidal abnormalities are frequently observed in the other eye, indicating that systemic factors may play a role in the pathophysiology of CSC [4]. Strikingly, the male sex is the strongest risk factor for CSC; approximately 80% of the cases are males [7]. Another distinct risk factor for CSC is exposure to corticosteroids, which entails both synthetic glucocorticoids and endogenous hypercortisolism (Cushing’s syndrome) [8‒10]. Furthermore, pregnancy has been reported as a risk factor for CSC, with spontaneous recovery of the acute CSC episode after termination of the pregnancy [11]. Collectively, these risk factors suggest a role for steroid hormones in the etiology of CSC, although the underlying mechanisms are yet to be unraveled [2].

Next to corticosteroids, sex hormones resemble the other major class of steroid hormones. Sex hormones are categorized into androgens and estrogens, of which the main ones are testosterone and estradiol (E2), respectively. Interestingly, androgens have been described to modulate the downstream effects of corticosteroids [12]. Therefore, based on the high incidence of CSC in males [7], it can be hypothesized that the levels of sex hormones may pose a vulnerability to develop CSC. Testosterone levels have been studied in CSC patients, but the available small studies show conflicting results [13‒18]. Of note, not only total levels but also freely circulating hormone concentrations and the ratio to E2 are of clinical value. In a study by Van Oosterhout and co-workers on migraine patients, a disease that occurs more frequently in females compared to males, male patients with migraine had been found to have higher serum levels of E2 compared to healthy male controls [19]. As a result of the elevated E2 levels, the free testosterone (fT)/E2 ratio was disturbed, leading to a relative androgen insufficiency in males with migraine [19].

In the current study, serum levels of testosterone and E2, as well as the fT/E2 ratio, are evaluated in a large cohort of males and females with active CSC. The goal of this study was to provide better insight into the levels of sex hormones in CSC patients of both sexes and to reveal potential imbalances in the fT/E2 ratio, which may provide novel insights into the pathophysiology of CSC.

A total of 206 chronic CSC patients (183 males, 23 females), diagnosed at the Department of Ophthalmology at the Leiden University Medical Center (Leiden, The Netherlands) between July 2014 and January 2020, were included in this study. The diagnosis of chronic CSC was established based on previously published criteria [1]. All of the following characteristics had to be present: serous SRF on optical coherence tomography, area of multifocal diffuse leakage or irregular retinal pigment epithelium window defects on fluorescein angiography, and corresponding hy­-perfluorescence on indocyanine green angiography [1]. Patients were excluded if any other posterior segment pathology was present, such as any form of choroidal neovascularization or age-related macular degeneration. In order to only obtain the results from CSC patients with active disease, patients were only eligible if SRF was detected on optical coherence tomography within 7 days of the day of blood sampling. Healthy controls (29 males, 30 females) without known ophthalmological medical history were recruited at the Leiden University Medical Center. The healthy controls were either employed at the Leiden University Medical Center or visited the hospital for voluntary blood donation. None of the healthy controls visited the hospital due to suspicion of any disease. Due to limitations in the selection of healthy controls, it was not possible to perform an age-matched study. Of note, the use of contraceptives and the menstrual cycle status were unknown across all female study participants. This information could not be obtained from the control group due to local privacy regulations. All patients and controls signed a written informed consent for blood sampling and study participation. This study was reviewed and approved by the local medical Ethical Committee of the Leiden University Medical Center, approval number B14.003/SH/sh.

Blood sampling was performed according to the same protocol for all participants of the study. Blood was collected in ethylenediaminetetraacetic acid tubes and stored on room temperature. Within 1 h after sampling, samples were centrifuged, and aliquots of serum were frozen at −80°C until batch analysis. Isolation and storage of samples occurred between June 2015 and January 2020. Albumin (g/L) was determined with a colorimetric assay using a Roche COBAS 8000 Modular (Roche Diagnostics, Basel, Switzerland). Sex hormone-binding globulin (SHBG, nmol/L), testosterone (nmol/L), and E2 (pmol/L) were determined using the ECLIA immune assay on the e602 module of the Roche COBAS 8000 Modular (Roche Diagnostics). fT levels were calculated according to a previously published method [20].

A power calculation was based on the fT/E2 ratio [19]. With α set at 0.05 and β at 0.10, we estimated that a control group of n = 30 and a patient group of n = 29 would be required to detect a 10% change with a power of 80%. For statistical analysis, an unpaired Student’s t test and a linear regression analysis were performed using Prism v9.0.1 (GraphPad Software Inc., San Diego, CA, USA). A pvalue <0.05 was considered statistically significant. Adjusted p values are shown after applying a Bonferroni correction for multiple testing. The data of males and females were analyzed separately.

A total of 183 male patients (mean age 52, median 52, range 30–88 years) and 23 female patients (mean age 59, median 58, range 45–79 years) with active chronic CSC participated in the study. The control groups consisted of 29 healthy males (mean age 48, median 47, range 31–69 years) and 30 healthy females (mean age 47, median 46, range 34–67 years). An unpaired t test showed that there was no statistically significant difference between the male control and male patient group (p = 0.132), but there was a statistically significant difference in age between the female controls and patients (p < 0.001).

For the males, no statistically significant differences were observed between the controls and patients in serum levels of SHBG (controls 38.7 nmol/L, patients 38.9 nmol/L, adj. p = 0.999), total testosterone (controls 15.5 nmol/L, patients 15.1 nmol/L, adj. p > 0.999), and E2 (controls 96.0 nmol/L, patients 98.9 nmol/L, adj. p > 0.999) (Fig. 1). Albumin (controls 47.8 g/L, patients 46.0 g/L, adj. p = 0.006) was found to be statistically significant lower in the male patient group compared to the controls. The calculated levels of fT (controls 279.8 pmol/L, patients 280.3 pmol/L, adj. p > 0.999) and the fT/E2 ratio (controls 3.2, patients 3.0, adj. p > 0.999) showed no statistically significant difference.

In the female patients, SHBG (controls 94.2 nmol/L, patients 50.4 nmol/L, adj. p = 0.001) levels were found to be statistically significantly lower compared to controls (Fig. 2). Moreover, total testosterone (controls 0.48 nmol/L, patients 0.58 nmol/L, adj. p > 0.999) and albumin (controls 43.3 g/L, patients 44.9 g/L, adj. p = 0.277) showed a tendency toward higher levels in patients. E2 (controls 300 nmol/L, patients 143 nmol/L, adj. p = 0.922) levels showed a tendency toward lower levels in patients compared to controls, reflecting a significantly higher fT/E2 ratio (controls 0.06, patients 0.18, adj. p = 0.018) in female CSC patients. Because of the fact that some studies described an age-related trend for SHBG levels in women [21, 22], we performed a linear regression analysis between SHBG and age, which showed that there was no interaction between these variables (Fig. 3). In addition, we have analyzed the dataset with the age range in the male group up to a maximum of 70 years old (online suppl. Fig. 1; for all online suppl. material, see www.karger.com/doi/10.1159/000526052), as well as the female control group limited to controls >45 years old (online suppl. Fig. 2). Importantly, the use of contraceptives, which is known to substantially elevate SHBG levels, was not known for any of the participants [23, 24].

In the present study, we found that total testosterone and E2 levels were not altered in both male and female CSC patients. In males, albumin levels were found to be significantly lower in CSC patients compared to controls. Moreover, only in females, SHBG levels were substantially lower in the patient group compared to the control group, together with a significantly higher fT/E2 ratio in female CSC patients.

Our findings could imply that concentrations of free steroid hormones may be altered in CSC patients as levels of the steroid hormone transportation proteins albumin and SHBG differed between patients and controls. Albumin is an abundant serum protein that binds a range of biologically active steroids, albeit with low specificity and affinity [25]. Minor differences in albumin levels were observed between controls and patients, which were decreased levels in male patients and a tendency toward increased levels in female patients. However, given the relatively low affinity of albumin to androgens and estrogens, it is unlikely that these small differences play a significant role in the etiology of CSC. In contrast, SHBG binds steroids more specifically and with much higher affinity [25]. SHBG was found to be lower in female CSC patients than in healthy female controls, and this could be essential in the pathogenesis of CSC. The relatively low SHBG levels in female CSC patients underlie the high levels of free (i.e., bioactive) testosterone and the high fT/E2 ratio in this group as SHBG is a major determinant in the calculation of these values. Intriguingly, none of the female CSC patients showed relatively high SHBG levels (e.g., above 100 nmol/L), which is in contrast to the control group, in which 40% of the females showed SHBG levels above 100 nmol/L. Therefore, it may be speculated that high SHBG levels are protective for the development of CSC as this protein binds excess androgens and prevents potential exacerbations in the fraction of free (i.e., bioavailable) testosterone. A potential causal role for testosterone in CSC is further supported by case reports on both males and females in whom exogenous testosterone administration had been described to induce CSC, with resolution of SRF after cessation of the supplementation in some cases [13‒16]. Collectively, this may indicate that testosterone may trigger CSC in both sexes, with SHBG being a limiting factor only in females.

Moreover, in patients with polycystic ovary syndrome, a common endocrinological condition characterized by excess androgens, a reduction in SHBG is used as a biomarker of disease severity [26], and several cases have been described that linked the occurrence of CSC to concomitant polycystic ovary syndrome [27]. Collectively, these data warrant further studies into the sex hormone homeostasis of females with CSC, especially into SHBG levels, preferably with sufficient clinical information on the use of contraceptives and menstrual cycle. Other clinical measures that have been associated with decreased SHBG levels include body mass index and fasting insulin levels [21].

Sex hormone levels in male CSC patients were found to be within the physiological range. Therefore, it may be the case that physiological levels of sex androgens in males are in itself “permissive” for CSC to develop or that other steroid species play a role in the strong male preponderance to develop CSC. Previously, a study by Schellevis et al. [28] reported increased levels of the steroid hormones androsterone, estrone, etiocholanolone, and androstenedione in male CSC patients, which are mainly metabolites or precursors of testosterone and E2, and are believed to have rather limited biological effects. Of note, the calculation of the fT/E2 ratio, which requires analysis of SHBG and albumin levels, was not performed in that study. Nevertheless, the results indicate imbalances in the steroid hormone homeostasis in males with CSC [28], and therefore, it is of interest to perform a similar analysis in a cohort of female CSC patients. Alternatively, it may be speculated that actions of sex hormones within the physiological range, such as testosterone, have an effect on ocular tissue that promotes the development of CSC. In principle, this would then require ocular expression of the main testosterone receptor, the androgen receptor. However, the expression of this receptor has been described to be nearly absent in single-cell RNA sequencing studies on human ocular tissue and immunohistochemistry studies from our group on androgen receptor expression in the posterior eye [29, 30]. However, androgen receptor expression has not been evaluated in the sclera, the rigid and fibrous tissue covering the exterior eye. This may be of particular interest since scleral thickness has recently been hypothesized to be involved in the regulation of venous choroidal blood outflow at the point where choroidal veins pass through the sclera and exit the eye [31].

A particular strength of our study is the number of patients that were included (183 males, 23 females), which is the largest cohort of CSC patients analyzed for sex hormone levels to date. Moreover, we only included patients with typical CSC that were considered to have an “active” CSC episode, based on the presence of SRF within 7 days of blood sampling. A shortcoming of our study is the lack of clinical information on the female cohort, which was also still relatively small. A potential explanation for the observed difference in SHBG levels in the female cohort could also be a difference in baseline characteristics between the control and patient group, such as more frequent use of oral contraceptives, which is known to be strongly associated with markedly increased levels of SHBG [23, 24]. Unfortunately, we were unable to retrieve reliable information on the use of oral contraceptives in this study due to privacy regulations. There was a significant difference in age between the female control group (mean age 47) and patient group (mean age 59). However, we did not find an association between age and SHBG in our study (Fig. 3), although a decline in SHBG has been described due to the transition to menopause [22]. Therefore, we cannot exclude that the difference in age in this study played a role in the observation of low SHBG levels in the female patient group. Nonetheless, our findings merit further and more thorough evaluation of sex hormones in females with CSC, and these studies preferably take potential confounding factors into account such as age, menstrual cycle status, contraceptives, and body mass index [22‒24].

In conclusion, our study revealed that sex hormone levels are not disturbed in male CSC patients and that females with CSC show an increased fT/E2 ratio, which may be explained by decreased SHBG levels compared to healthy females. Future studies could focus on an in-depth analysis of steroid hormones in females to confirm these observations in the female group. Additionally, in vitro assays using steroid hormones may shed a light on the effects of these hormones on choroidal endothelial cells, which are believed to be a key player in the pathophysiology of CSC by the manifestation of vascular hyperpermeability [2, 32]. This may include the administration of “excess” androgens which has been suggested to trigger CSC, the steroid metabolites and precursors identified by Schellevis et al. [28], or the use of serum from CSC patients with an active episode in cell culture assays. Finally, the integration of patient-specific research using induced pluripotent stem cells can be of particular value as it may reveal intrinsic cellular signaling pathways in CSC (e.g., a “hypersensitivity” to steroids) that help to explain the complex pathophysiology of this disease.

This study protocol was reviewed and approved by the local medical Ethical Committee of the Leiden University Medical Center (approval number B14.003/SH/sh) and followed the tenets of the Declaration of Helsinki. Written informed consent was obtained from all participants.

The authors have no conflicts of interest to declare.

Supported by the following foundations: Oogfonds, Stichting Leids Oogheelkundig Ondersteuningsfonds, ZonMw “Meer Kennis Minder Dieren” (MKMD; Grant No. 2018/01831/ZONMW), and the International Foundation for Ethical Research (IFER) Graduate Fellowship Program. The following organizations supported this study via UitZicht (Grant No. UZ2018-7): Stichting Macula Fonds, Retina Nederland Onderzoek Fonds, Stichting BlindenPenning, Algemene Nederlandse Vereniging ter Voorkoming van Blindheid, and Landelijke Stichting voor Blinden en Slechtzienden. The funding organizations had no role in the design or conduct of this research. They provided unrestricted grants.

Joost Brinks contributed to the planning, conduct, data analysis, and writing of the article. Elon H.C. van Dijk contributed to the planning, writing, editing, and reviewing of the manuscript. Roula Tsonaka assisted in the experimental design, statistical analysis, and contributed to editing and reviewing of the manuscript. Onno C. Meijer and Camiel J.F. Boon supervised the project and contributed to the experimental design, planning, and review of the manuscript.

All data generated or analyzed in this study are included in this article. Further inquiries can be directed to the corresponding author.