Choroid circulatory changes like hyper-permeability, stasis, and ischemia cause the leakage of retinal pigment epithelium (RPE) and the formation of subretinal fluid (SRF). This phenomenon results in neurosensory retinal detachment most often localized to the macula with or without focal serous pigment epithelial detachment and is known as central serous chorioretinopathy (CSCR). Gass (1967) [1] coined the term “central serous chorioretinopathy” for this condition using the knowledge that the choroid and RPE were the primary tissues involved.

CSCR is the fourth-most common condition among non-surgical retinal diseases according to incidence rate [2]. In most patients, symptoms and SRF occurring in CSCR disappear spontaneously within 3-4 months [3]. In their work, Daruich et al. [4] defined chronic CSCR as CSCR with an anatomic change, such as diffuse retinal epitheliopathy. However, in the present article, acute and chronic CSCR are differentiated according to disease duration. A consensus has not yet been made on the generally accepted duration of distinction between acute and chronic CSCR. Since many authors define chronic CSCR as that where symptoms last > 3 months, we also defined chronic CSCR as that with a ≥ 3-month duration [3].

A variety of treatment methods exist for CSCR, including intravitreal injection of anti-vascular endothelial growth factor (anti-VEGF) agents, full-dose or half-dose verteporfin photodynamic therapy (vPDT), argon or micropulse laser treatment, beta-blocker therapy, and mineralocorticoid receptor antagonist (MRA) therapies such as eplerenone or spironolactone [5]. There have been many randomized controlled trials (RCTs) in which half-dose vPDT has resulted in a significantly greater proportion of patients with complete resolution of SRF and improved function in both patients with acute and chronic CSCR [6,7]. However, vPDT is difficult to deploy in many hospitals due to the cost of verteporfin and accessibility to the equipment to generate the specialized lasers.

Research has shown that MRA reduces retinal vasculopathy in an oxygen-induced retinopathy rat model [8]. Later, Bousquet et al. [9] conducted a prospective non-comparative study using eplerenone in CSCR patients and Herold et al. [10] published a prospective interventional case series using spironolactone in CSCR patients. These articles reported a best-corrected visual acuity (BCVA) improvement and complete resolution of SRF [9,10]. Spironolactone induces severe adverse effects, such as gynecomastia and vaginal bleeding, but eplerenone has relatively few adverse effects due to selectivity for the aldosterone receptor [11].

Two meta-analyses of the results of the studies using MRA in CSCR patients have been published, but contradicting conclusions have been drawn regarding the effectiveness of MRA [12,13].

In this article, a meta-analysis was performed for BCVA, SRF, choroidal thickness (CT), and central macular thickness (CMT) at each follow-up point from RCT results using eplerenone in CSCR patients.

Search strategy and study selection

A systematic search was performed from database inception to May 2023 in the Medline, EMBASE, and Cochrane literature databases to find studies that have applied oral eplerenone therapy to CSCR patients. To briefly explain the search strategy, Medline’s medical subject headings (MeSH), Embase subject headings (Emtree), and Cochrane medical terms for “CSCR” and “eplerenone” were used as search expressions. The selection criteria for studies were as follows: 1) RCTs, 2) patients with CSCR, 3) oral eplerenone therapy, and 4) control protocol of placebo therapy or observation. Meanwhile, the exclusion criteria were as follows: 1) non-RCT trials, 2) other diseases, 3) interventions other than eplerenone.

Data extraction

Excluding those with data duplication, 223 articles were searched. Two investigators independently selected 107 potentially eligible studies by reviewing the titles and abstracts of the retrieved papers. Afterward, a full-text review of each paper was conducted. In the case of disagreement in the selection of literature, divergences were resolved by discussing with each other. Finally, five articles were selected (Fig. 1).

Fig. 1. Flow diagram of study inclusion in this review. RCT = randomized controlled trial.

Information extracted from each article included the country in which the study was conducted, the CSCR duration, the number of patients, the characteristics of study participants, and the follow-up period. As outcomes, mean difference values and standard deviations between baseline and follow-up points were extracted for BCVA, SRF, CT, and CMT. In addition, the number of patients with complications during each study was extracted. In the case of a crossover study using eplerenone and spironolactone, if a complication was not clearly attributed to spironolactone, it was assumed that all complications occurred in the eplerenone group to conservatively judge the risk of oral eplerenone therapy [14]. Complications defined in the selected five articles were divided into the following two categories: 1) systemic adverse events, including hyperkalemia and decreases in renal function, and 2) ocular adverse events, including new choroidal neovascularization and severe decreases in visual acuity. If data could not be extracted from the paper, the authors of the article were contacted by email to obtain the data [14-16]. In the crossover study using eplerenone and spironolactone, only data for up to 1 month before starting crossover therapy were used [14]. The term “end of the study follow-up point” used in this article refers to the end of each trial or its last follow-up point.

Quality assessment

For the five RCTs finally selected, quality assessment was performed using the Cochrane Collaboration tool for assessing the risk of bias (ROB 2.0, 2019 version; Cochrane). Two investigators analyzed independently and resolved divergences through consensus.

Statistical analyses

All procedures were performed according to the Preferred Reporting Items for Systematic reviews and Meta-analyses (PRISMA) guideline; the PRISMA 2009 checklist is shown in Supplementary Table 1. Outcome data were analyzed using RevMan version 5.3 (Cochrane). When the unit of BCVA was the chart score of the Early Treatment of Diabetic Retinopathy Study, it was converted to a logarithmic minimum angle of resolution (logMAR) value using the formula “logMAR = 1.7 - 0.02 × chart score” [17]. When data were presented only as a figure, the Web Plot Digitizer version 4.4 (Automeris) program was used to extract exact numerical values of the data from the image [15]. When only the p-value was available (i.e., there was no information on the standard deviation), the standard deviation was extracted by calculating the t value and standard error of the t test using the p-value [14,15]. When the standard deviation couldn’t be extracted from the values presented in the article and there was no response to an email requesting data, a standard deviation of the other follow-up time in the same group was imputed or the calculated pooled variance was used [14,15]. Under the assumption that the different studies were estimating different, yet related, intervention effects, a random-effects model and the Mantel-Haenszel method were used in this meta-analysis [18]. The disease duration of articles was chronic or acute. Also, the control group underwent placebo treatment or observation. The end of the study follow-up time varied between studies. As possible heterogeneity was suspected, the I² statistic was used to evaluate the degree of heterogeneity. Moderate heterogeneity was defined by I² > 25%, while severe heterogeneity was indicated by I² > 75%.

Finally, five RCTs were selected; among them, there were four studies covering chronic CSCR and one study covering acute CSCR (Table 1). A total of 252 patients received either oral eplerenone therapy or placebo therapy or observation; overall, there were 134 patients in the eplerenone group and 118 patients in the control group without eplerenone therapy. The smallest sample size was 19 patients, while the largest sample size was 114 patients, and the median sample size was 40 patients. The follow-up periods of three studies were all < 3 months, that of one study was 6 months, and that of one study was 12 months, respectively. All results with a follow-up period of < 3 months favored the eplerenone group. Results collected during the 6-month and 12-month follow-up periods, respectively, were not statistically significant.

Table 1

Characteristics of included trials

First Author	Published Country	Disease Duration (mo)	EPL/CTL (n/n)	EPL Rx	CTL Design	F/U (mo)	Conclusion of RCT	Quality Assessment
Lotery (2020)	UK	Chronic (> 4)	57/57	25 mg 1 wk + 50 mg 12 mo	PCB	12	NS	Low ROB
Pichi (2017)	USA	Chronic (> 3)	20/20	25 mg 1 wk + 50 mg 3 wk	PCB	1	Favor EPL	Some concerns
Rahimy (2018)	USA	Chronic (> 3)	15/6	25 mg 1 wk + 50 mg 8 wk	PCB	2	Favor EPL	Low ROB
Schwartz (2017)	Israel	Chronic (> 4)	13/6	25 mg 1 wk + 50 mg 11 wk	PCB	6	NS	Some concerns
Venkatesh (2020)	India	Acute (< 3)	29/29	50 mg 1-3 mo	OBS	3	Favor EPL	High ROB

Mo = months; EPL = eplerenone; CTL = control; n = number of eyes; Rx = prescription; F/U = follow-up; RCT = randomized controlled trial; wk = weeks; PCB = placebo; NS = not significant; ROB = risk of bias; OBS = observation.

Quality assessment

The quality of included RCTs was assessed using ROB 2.0 and summarized in Table 1 and Fig. 2. If, during the randomization process, blindness was maintained by the patient and the evaluator but potentially broken by the pharmacist, the domains of “Randomization process” and “Deviations from the intended interventions” were evaluated instead as “Some concerns” [15].

Fig. 2. Risk-of-bias summary: judgments about each risk-of-bias item for each included study.

During the crossover study using spironolactone and eplerenone, different pill identifiers were written on the surfaces of spironolactone tablets and eplerenone tablets; therefore, it is possible that the patients were aware of their allocations. The domains of “Deviations from the intended interventions” and “Measurement of the outcome” were therefore evaluated as “Some concerns” [14].

In the study in which the intervention group took eplerenone tablets and the control group underwent observation, it was very likely that the patients knew their treatment allocation, so the domain of “Deviations from the intended interventions” was evaluated as “Some concerns” and the domain of “Measurement of the outcome” was evaluated as “high risk of bias” [16].

BCVA

The comparison of the mean differences in BCVA from baseline between the eplerenone group and the control group was different for each follow-up period (Fig. 3). At the end of the study follow-up period, the eplerenone group showed a statistically significant improvement in BCVA compared to the control group (95% confidence interval [CI], -0.08 to -0.02; p = 0.001). At 2 and 3 months after receiving oral eplerenone therapy, the eplerenone group had significantly improved BCVAs compared to the control group (2 months: 95% CI, -0.11 to -0.06; p < 0.00001) (3 months: 95% C I, -0.03 to -0.01; p < 0.00001). However, after 6 months, the mean difference in BCVAs in the eplerenone group did not show a statistically significant difference from that in the control group (95% CI, -0.25 to 0.11; p = 0.44).

Fig. 3. Forest plots of effect sizes in best-corrected visual acuity (A) at 1 month, (B) at 2 months, (C) at 3 months, (D) at 6 months, and (E) at the end of study follow-up. SD = standard deviation; CI = confidence interval.

SRF

At the end of study follow-up, the comparison of the mean difference in SRF from baseline between the eplerenone group and the control group did not show a statistically significant difference (95% CI, -111.67 to 47.47; p = 0.43) (Fig. 4). The same result was noted at all follow-up points after receiving oral eplerenone therapy.

Fig. 4. Forest plots of effect sizes in subretinal fluid (μm) (A) at 1 month, (B) at 2 months, (C) at 3 months, and (D) at the end of study follow-up. SD = standard deviation; CI = confidence interval.

CT

At the end of study follow-up, the comparison of the mean differences in CT from baseline between the eplerenone group and the control group did not show a statistically significant difference (95% CI, -63.99 to 40.96; p = 0.67) (Fig. 5). However, at 1 month after receiving oral eplerenone therapy, the eplerenone group had significantly reduced CT values compared to the control group (95% CI, -64.83 to -23.29; p < 0.0001).

Fig. 5. Forest plots of effect sizes in choroidal thickness (μm) (A) at 1 month, (B) at 2 months, (C) at 3 months, and (D) at the end of study follow- up. SD = standard deviation; CI = confidence interval.

CMT

At the end of study follow-up, the comparison of the mean Figure 4. differences in CMT from baseline between the eplerenone group and the control group was not statistically significant (95% CI, -113.76 to 69.94; p = 0.64) (Fig. 6).

Fig. 6. Forest plots of effect sizes in central macular thickness (μm) at the end of study follow-up. SD = standard deviation; CI = confidence interval.

Complications

At the end of study follow-up, there was no statistically significant difference in the risk of complications occurring between the eplerenone group and the control group (95% CI, -0.04 to 0.33; p = 0.12) (Fig. 7).

Fig. 7. Forest plots of risk differences in complications at the end of study follow-up. M-H = Mantel-Haenszel method; CI = confidence interval.

At the end of study follow-up, oral eplerenone therapy for CSCR patients appeared to improve BCVA statistically significantly compared to the control protocol. However, when meta-analysis was performed at each follow-up point, the BCVA improved statistically at 2 and 3 months after the start of oral eplerenone therapy, but there was no statistically significant change at 6 months compared to the control group (Fig. 3). Wang et al. [12] performed meta-analysis using five RCTs covering MRA therapies such as spironolactone or eplerenone in CSCR patients. As a result, it was also reported that there is a modest benefit of improved BCVA achievable with MRA therapy for CSCR patients at 1 and 2 months [12]. Importantly, the improvement in BCVA after the use of oral eplerenone therapy in the present study is modest. According to our results, the mean difference in BCVA compared to the control group was 0.08 at 2 months and 0.02 at 3 months, respectively. Although this cannot be considered clinically very significant, it is a large enough improvement that patients could read about one more line on the logMAR chart in the second month. Kitzmann et al. [19] reported that 74 patients diagnosed with CSCR from 1980-2002 had a mean onset age of 41 years (range, 29-56 years), suggesting onset is common in young and working-age individuals. Although it may only bring modest BCVA improvement, shortterm use of eplerenone may reduce the discomfort of daily life in patients with CSCR.

According to a retrospective study of 21 acute CSCR patients treated with anti-VEGF therapy (bevacizumab or ranibizumab), the BCVA (logMAR) was 0.35 at baseline, 0.18 at 1 month, and 0.19 at 3 months after injection, respectively, confirming peak effectiveness at 1 month [20]. In addition, Chan et al. [6] reported that the central foveal thickness was 456.0 μm at baseline, 203.9 μm at 1 month, and 166.0 μm at 3 months after treatment in 43 eyes treated with half-dose vPDT. The efficacy of anti-VEGF therapy and half-dose vPDT was highest at 1 month and decreased thereafter. In this meta-analysis, the effect of eplerenone showed a similar trend. In the eplerenone group, the average changes in BCVA (logMAR) compared to baseline at 1, 2, and 3 month(s) were -0.11, -0.19, and -0.15, respectively. During the first month of taking eplerenone, the BCVA improvement was 0.11, confirming peak effectiveness, while, over the next 1-2 months, the BCVA improvement was 0.08, suggesting its therapeutic effect gradually decreases (Fig. 3).

Of the five RCTs included in this meta-analysis, three had a follow-up period of < 3 months. All three of these RCTs reported that the eplerenone therapy group had statistically significant improvements in BCVA compared to that in the control group. On the other hand, two RCTs with follow-up periods of ≥ 6 months did not report significant BCVA improvements in CSCR patients compared to the control group (Table 1). The authors believe that the effect of oral eplerenone therapy is short-lived, with an effect that lasts only up to 3 months.

On the other hand, no statistically significant changes in anatomic outcomes such as SRF, CT, and CMT were found among CSCR patients compared to the control group (Figs. 4-6). Loo et al. [21] reported that even atrophic RPE tracts inferior to the macula were not significantly associated with an adverse visual acuity outcome. It has been reported that there is no difference in final visual acuity between cases where intravitreal anti-VEGF injection was not performed when the SRF was < 200 μm and cases where additional injections were performed whenever SRF was detected in patients with neovascular age-related macular degeneration [22]. In addition, Daugirdas et al. [23] pointed out that, in an RCT study published in 2020 [17], only the SRF of one cross-section was measured and compared, and the results may be different when comparing SRF volumes. What should be noted here is the fact that SRF decreased significantly compared to baseline in the eplerenone group. In fact, the studies of Lotery et al. (2020) [17] and Schwartz et al. (2017) [15], which concluded that eplerenone was not superior to placebo, also reported a significant decrease in SRF compared to baseline in the eplerenone group. However, even in the placebo or observation group, SRF decreased statistically significantly compared to baseline. In other words, the decrease in SRF in the eplerenone group was not significantly different compared to that in the control group. Nevertheless, it is difficult to clearly explain why the use of eplerenone can lead to an improvement in BCVA without improvements in anatomic parameters such as SRF compared to the control group. Further research on this subject is needed in the future.

In this meta-analysis, the complication risk associated with eplerenone was assessed strictly and conservatively. In a crossover study using eplerenone and spironolactone [14], all patients with complications except for one patient with gynecomastia caused by spironolactone were assumed to be attributable to eplerenone. Eplerenone reduction management due to high serum potassium levels [24] was also included as a complication of eplerenone. One case of suspected rhabdomyolysis caused by performing strenuous exercise at the gym [15] was also included as a complication due to eplerenone. Four studies reported no complications in the control group, while one study [17] documented 31 events. The authors believe that this happened for the following two reasons: 1) in this study [17], complications included common symptoms like headache and nausea and 2) not only severe adverse events but also mild adverse events were considered as complications in this study. There may also be other explanations for the complications; for example, hyperkalemia is a side effect of eplerenone but can also be caused by chronic kidney disease or diabetes. Even with this conservative assessment, however, there was no statistically significant difference in the occurrence of complications between the eplerenone and control groups. In addition, there were no severe adverse effects in any of the five RCTs included in this meta-analysis. Eplerenone is an oral medication that is relatively safe compared to more disruptive therapies such as micropulse focal laser therapy or half-dose vPDT [16].

BCVA showed low heterogeneity in this meta-analysis (I² = 17%). However, severe heterogeneity existed i n t he meta-analysis of SRF, CT, CMT, and complications (I² = 94%, 98%, 96%, and 79%), respectively. Reasons for severe heterogeneity may include the following: 1) participants of the RCTs included in this meta-analysis had different disease durations, e.g., chronic CSCR versus acute CSCR, 2) the control group treatments also differed from each other, e.g., placebo therapy versus observation, 3) there were cases where photo-dynamic therapy was performed as usual care [17], 4) the RCTs included in this meta-analysis adopted various follow-up periods, and 5) the dosage of eplerenone and its duration of use were different between studies (Table 1).

The limitations of this paper are as follows. Four RCTs used a placebo group as a control group, whereas the Venkatesh (2020) [16] study used an observation group. Therefore, the effects of the Venkatesh (2020) [16] study include a placebo effect, which may weaken the conclusions of this paper.

In addition, a meta-analysis was performed by mixing acute and chronic CSCR cases, which may also undermine the results of this paper. In acute cases, spontaneous remission is expected, and follow-up is performed without treatment; conversely, in chronic cases, additional treatment is often considered. However, patients often discover symptoms late, and the classification of CSCR based on disease duration has certain inherent flaws. Extrafoveal leaks with no ocular symptoms may go unnoticed on many occasions [25]. Also, if patients have good vision in one eye, symptoms in the other eye can be masked. Therefore, the authors believe that short-term use of eplerenone could be considered regardless of acute or chronic CSCR status. In the future, if more RCTs are performed for acute CSCR, a separate meta-analysis can be performed only for acute CSCR.

In conclusion, BCVA seems to improve for up to 3 months when oral eplerenone therapy is administered for CSCR. However, the effect of eplerenone therapy did not appear to persist when observed for a long time. The complications of eplerenone are tolerable. Clinically, early and short-term eplerenone treatment up to 3 months in CSCR patients can improve BCVA and alleviate the discomfort of patients.

The authors declare no conflicts of interest relevant to this article.

Conception (G.H., J.A.S.); Design (G.H., J.Y.L., W.K.C., J.A.S.); Data acquisition (G.H., D.S.K., D.I.K.); Analysis (G.H., D.S.K., D.I.K.); Interpretation (G.H., J.Y.L., W.K.C., J.A.S.); writing (G.H., D.S.K., D.I.K.); Review (All authors); Final approval of the article (All authors)