Topical Application of Bunazosin Hydrochloride Suppresses Myopia Progression With an Increase in Choroidal Blood Perfusion

Abstract

Purpose: The incidence of myopia has rapidly increased in recent decades, making it a growing public health concern worldwide. Interventions to suppress the progression of myopia are needed; one suggested strategy is the prevention of choroidal thinning, which can improve choroidal blood perfusion (ChBP). Bunazosin hydrochloride (BH) is an alpha1-adrenergic blocker and commercialized glaucoma eye drop that increases in blood circulation in the eye. In this study, we evaluated the efficacy of BH in suppressing the progression of myopia in a lens-induced murine model.

Methods: Lens-induced myopia was induced in 3-week-old C57BL/6 J mice with −30 diopter (D) lenses for three weeks. Refractive error, axial length, and choroidal thickness were evaluated at three and six weeks of age using an infrared photorefractor and a spectral domain optical coherence tomography (OCT) system. Moreover, ChBP and scleral thickness were evaluated using swept-source OCT and histological analysis.

Results: Compared with the controls, the administration of BH eye drops suppressed the myopic shift of refractive error (mean difference ± standard error in the eye with −30 D lens, −13.65 ± 5.69 D vs. 2.55 ± 4.30 D; P < 0.001), axial elongation (0.226 ± 0.013 mm vs. 0.183 ± 0.023 mm; P < 0.05), choroidal thinning (−2.01 ± 1.80 µm vs. 1.88 ± 1.27 µm; P < 0.001), and scleral thinning (11.41 ± 3.91 µm vs. 19.72 ± 4.01 µm; P < 0.01) with myopia progression and increased ChBP (52.0% ± 4.1% vs. 59.5% ± 6.3%; P < 0.05). The suppressive effect of BH eye drops was dose-dependent and higher than that of other glaucoma eye drops and alpha1 blockers.

Conclusions: These results demonstrate the potential of BH eye drops in the treatment of myopia and support further investigation of their efficacy in humans. Further studies are needed to determine the mechanism of action and long-term safety of this treatment.

Over the past few decades, the prevalence of myopia has increased substantially worldwide. Approximately 50 years ago, the prevalence of myopia in East Asian countries was approximately 20%, but has now risen to more than 80%.¹ Also, a dramatic increase has been seen in Europe and the United States.¹ It has been reported that the coronavirus disease 2019–related quarantine has further accelerated the growth of the myopia population.^2,3 High myopia is associated with complications such as macular disease, retinal detachment, and glaucoma, which can lead to blindness.⁴ Additionally, children with mild myopia at ages six or seven are more likely to develop high myopia in upper elementary school.⁵ 

Currently, pharmacological and optical approaches, such as 1% atropine eye drops, spectacles, multifocal contact lenses, and orthokeratology, are effective in suppressing the progression of myopia.^6–9 Moreover, considering the increasing medical costs associated with myopia-related diseases and the high prevalence of myopia, interventions to suppress myopia progression and to investigate the underlying mechanisms of myopia development have recently been investigated. We previously demonstrated several treatments that halted myopia progression in a murine lens–induced myopia (LIM) model^10–12 and suggested that the prevention of choroidal thinning along with myopia progression may be critical. Choroidal thinning has been observed in myopia-induced eyes of chicks.¹³ In humans, negative refraction is associated with an increase in axial length and a decrease in choroidal thickness (ChT).^14,15 Furthermore, previous studies have reported decreased choroidal blood flow in animal models of myopia^10,16 and in patients with high myopia.^17,18 This decrease in choroidal blood flow may serve as potential predictive markers for the onset and progression of myopia. Reduced choroidal blood perfusion (ChBP), which might be caused by thinning, can cause a hypoxic response in the sclera, resulting in scleral remodeling in the axial elongation direction.¹⁹ Myopia progression induced by form-deprivation myopia in guinea pigs is correlated with changes in ChT and ChBP.²⁰ Moreover, during the recovery phase after myopia induction in animal models, an increase in ChT^21,22 and ChBP²³ has been observed. Therefore maintaining ChT and ChBP levels is believed to be a strategy for preventing the development and progression of myopia. 

Based on this idea, there have been attempts to use glaucoma agents, which have the effect of lowering intraocular pressure, to increase blood circulation in the eye for myopia suppression as a drug repositioning strategy.^24–27 In previous studies, commercialized glaucoma eye drops such as timolol, one of the β-blockers, and latanoprost, a prostaglandin, have been applied to the treatment of myopia.^28–30 Although they have been shown to be effective in suppressing axial elongation to some extent, the results have not been consistent. Consequently, most of these studies have not yet reached clinical application. 

In this study, we aimed to reposition an existing eye drop to suppress the progression of myopia by maintaining ChT. Among the many current commercialized eye drops, we focused on bunazosin hydrochloride (BH), an alpha-1-adrenergic blocker, which selectively inhibits α1-adrenergic receptors in vascular smooth muscle cells and alleviates vasoconstriction.^27,31 Moreover, BH is clinically safe because it has been used as an ocular hypotensive agent for glaucoma and benign prostatic hyperplasia treatments. We examined the myopia-suppressive effect of BH by intraperitoneal injection (IP) or eye drops and evaluated changes in refractive error, axial elongation, and ChT using the LIM mouse model we previously developed. We also evaluated the increase in ChBP and the maintenance of scleral thickness after BH administration and compared the effects of other glaucoma drugs and alpha-1-adrenergic blockers. These results could lead to drug repositioning of therapeutic agents for myopia and could be used as an effective countermeasure in the rapidly increasing myopia population. 

Material and Methods

LIM Mouse Model

Animal experiments in this study were approved by the Ethics Committee on Animal Research of the Keio University School of Medicine and adhered to the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research, Institutional Guidelines on Animal Experimentation at Keio University, and the Animal Research: Reporting of In Vivo Experimental Guidelines for the Use of Animals in Research. Three-week-old male C57BL/6J mice weighing 9 to 11 g were purchased from CLEA Japan Inc and fed normal chow and tap water. Fewer than five mice were kept in one cage under an approximately 50 lx fluorescent light with a 12-hour diurnal period at 23°C ± 3°C. 

A mouse model of LIM was established, as previously described.^32,33 Mice were anesthetized with a combination of midazolam (40 µg/100 µL; Sandoz K.K., Tokyo, Japan), medetomidine (7.5 µg/100 µL; Domitor, Nippon Zenyaku Kogyo Co., Ltd., Fukushima, Japan), and butorphanol tartrate (50 µg/100 µL; Meiji Seika Pharma Co., Ltd., Tokyo, Japan). After anesthesia was initiated, the scalp was cut to expose an approximately 1 cm² area of the skull, and the periosteum was removed with etching fluid. The frames of the eyeglasses were prepared specifically for the mice using a three-dimensional printer and 0 diopter (D) and −30 D lenses (Rainbow Optical Laboratory Co., Ltd., Tokyo, Japan) were attached to the frames. A pair of eyeglasses was attached to the mouse head using a self-curing dental adhesive resin cement (Super-Bond, Sun Medical Co., Ltd., Shiga, Japan). As an internal control, all left sides of the eyeglasses contained 0 D lenses and the right sides contained −30 D lenses. The eyeglasses were cleaned at least twice a week. 

Drug Administration

BH (0.01%; B689585; Toronto Research Chemicals Inc., Toronto, Ontario, Canada), tafluprost (0.0015%; T004820; Toronto Research Chemicals Inc.), unoprostone isopropyl ester (UIE) (0.12%; 16681; Cayman Chemical, Ann Arbor, MI, USA), prazosin hydrochloride (0.01%; P7791; Sigma-Aldrich), and urapidil hydrochloride (0.01%; U100; Sigma-Aldrich Corp., St. Louis, MO, USA) were dissolved in PBS. Ten microliters of drug eye drops were administrated to both eyes daily, unless otherwise mentioned. BH solution (10 mL/kg) was administered to the mice daily by IP. The control group was treated with PBS. 

Experiment 1. Comparison of BH Administration by IP and Eye Drop

Three-week-old LIM mice (n = 24) were randomly assigned to three groups. The mice received daily IP of either PBS (10 µL/g), 0.01% BH (10 µL/g), or 0.01% BH eye drops for a duration of three weeks (Fig. 1A). The treatments were administered at a consistent time each afternoon. Before LIM surgery and after three weeks, refractive error, axial length, and ChT were measured (Supplementary Fig. S1, Supplementary Table S1). ChBP was assessed using OCT-A in both eyes of the PBS IP group and the BH eye drop group (Supplementary Table S2). Additionally, the enucleated eyes were used to measure scleral thickness (Supplementary Table S3). 

Figure 1.

 

BH suppresses LIM in mice. The mice received daily IP injection of either PBS (10 µL/g), 0.01% BH (10 µL/g), or 0.01% BH eye drops for a duration of three weeks (A). Changes in refractive error (B) and axial length (C) between zero and three weeks after LIM. The average values at baseline (before LIM) and three weeks after LIM are shown in Supplementary Figures S1A and S1B. n = 8. *P < 0.05, **P < 0.01, ***P < 0.001, two-way ANOVA with a Bonferroni post hoc test. Bars represent mean ± standard deviation.

Experiment 2. Effects of BH Dose Dependency

Three-week-old LIM mice (n = 32) were randomly assigned to four groups. The mice received daily administration of PBS for control or BH eye drops in varying concentration (0.0001%, 0.001%, 0.01%) for a period of three weeks (Fig. 4A). The treatments were consistently administered at the same time each afternoon. Before LIM surgery and after three weeks, refractive error, axial length, and ChT were measured (Supplementary Fig. S2, Supplementary Table S4). 

Experiment 3. Effects of BH Administration Number

Three-week-old LIM mice (n = 40) were randomly assigned to five groups. The mice received daily administration of PBS eye drops for control or 0.01% BH eye drops once for three weeks, once a week, twice a week, and daily for a duration of three weeks (Fig. 5A). The treatments were consistently administered at the same time each afternoon. Before LIM surgery and after three weeks, refractive error, axial length, and ChT were assessed (Supplementary Fig. S3, Supplementary Table S5). 

Experiment 4. Comparison of Glaucoma Drugs and α1 Blockers

Three-week-old LIM mice (n = 48) were randomly assigned to six groups. The mice received daily administration of PBS for control, 0.01% BH, 0.0015% tafluprost, 0.12% UIE, 0.01% prazosin hydrochloride, or 0.01% urapidil hydrochloride eye drops for a duration of three weeks (Fig. 6A). The treatments were consistently administered at the same time each afternoon. Before LIM surgery and after three weeks, refractive error, axial length, and ChT were assessed (Supplementary Fig. S4, Supplementary Table S7). 

Refractive Error and Axial Length Measurement

As described in previous studies,^32–34 mouse refraction and axial length were measured at three and six weeks of age using an infrared photorefractor (Steinbeis Transfer Center, Graz, Austria) and a spectral domain-OCT system (SD-OCT, Envisu R4310; Leica, Wetzlar, Germany) tuned for mice. Before the measurements, mice were administered mydriasis with 0.5% tropicamide and 0.5% phenylephrine eye drops (Santen Pharmaceutical Co., Ltd., Osaka, Japan). After assuring mydriasis, the mice were anesthetized by IP injection with 0.01 mL/g of midazolam, medetomidine, and butorphanol tartrate. The axial length was determined as the distance between the vertex of the cornea and the retinal pigment epithelium (RPE) layer near the optic nerve. 

ChT Measurement

ChT was measured at three and six weeks of age using SD-OCT after mydriasis and anesthesia as previously described.¹¹ Briefly, from the OCT image of the annular retinochoroidal section at a point 500 µm from the optic nerve, the encircled area from the border of the RPE to the posterior surface of the choroid was quantified using ImageJ (NIH) (Supplementary Fig. S5). Considering choroidal thickness variations by location, the average ChT was determined by the encircled area by dividing its circumference. To validate the measurement method of ChT, we performed the interexaminer and intraexaminer repeatability tests (Supplementary Fig. S6). The intraclass correlation coefficients for both tests were found to be high (>0.9), indicating a high level of measurement stability and reproducibility. 

ChBP Measurement

With the mice under anesthesia, ChBP was measured after three weeks of myopia induction using swept-source OCT (SS-OCT, XEPHILIO OCT-S1; Canon Medical Systems, Tokyo, Japan). SS-OCT assesses in vivo blood flow by analyzing changes in intensity and phase information arising from the motion of red blood cells, as detected through repeated OCT scans at the same location, resulting in the generation of red signals.^35,36 B-scan OCT angiography images, which show the retinochoroidal section passing through the center of the optic nerve from the en face OCT images, were selected to quantify ChBP signals in the choroid (Supplementary Figure S7). In OCT angiography, the ChBP are reversed beyond the RPE layer because of the presence of abundant pigment particles in the RPE layer.^38,38 This results in the formation of areas devoid of blood perfusion, represented as red noise points, as known as flow voids. Previous studies have found that a negative correlation between ChBP and the presence of flow voids in the choriocapillaris.^17,39 Therefore ChBP was determined as the area without red signals in the choroidal layer and measured the proportion of the region without red signals relative to the total area of the choroid using ImageJ according to our previous study.¹⁰ 

Scleral Thickness Measurements

After three weeks of myopia induction, both eyes were enucleated from eight mice in each group, fixed with 4% paraformaldehyde overnight, and embedded in paraffin with the superior aspect oriented upwards using an automated tissue processor (Tissue-Tek VIP 5 Jr; Sakura Finetek Japan Co., Ltd., Tokyo, Japan). These sample blocks were sliced from the inferior side, yielding three sections per sample around the optic nerve head into 5 µm sections by a microtome (REM-710; Yamato Kohki, Saitama, Japan). The sections were stained with hematoxylin and eosin (H&E) and visualized under a microscope (BX53; Olympus, Tokyo, Japan). Images of the sections were captured using a 20× objective (Uplansapo20×; Olympus). Scleral thickness was measured by averaging three measurements taken 500 µm from the optic nerve head toward the nasal region. 

Statistical Analyses

All data are presented as mean ± standard deviation. Two-way analysis of variance with the Bonferroni post hoc test was used to analyze the statistical significance of the differences using IBM SPSS Statistics version 27 (IBM Corp, Armonk, NY, USA). Statistical significance was set at P values of <0.05. The effect size and power for each measure were calculated using G*power 3.1 (Supplementary Tables S7–S12).^40,41 

Results

Suppressive Effect of BH on LIM in Mice

To examine the suppressive effect of BH on myopia progression, we administered IP or eye drops of BH to mice (as described in the Material and Methods). Eyes treated with −30 D lenses showed a significant myopic shift compared with those treated with 0 D lenses in the control group (P < 0.001) (Fig. 1B). Mice treated with IP injection or eye drops of BH showed a smaller refractive change with −30 D lenses than the control group with −30 D lenses (P = 0.002 and P < 0.001). The change in the axial length of eyes with −30 D lenses in the control group was significantly larger than that in the AL of eyes with 0 D lenses (P = 0.006) (Fig. 1C). In the BH-administered group, there was no significant difference between the −30 D and 0 D lenses (IP injection, P = 0.813; eye drop, P = 0.890). 

BH Suppressed Choroidal Thinning and Increased ChBP in LIM Mice

Previous studies demonstrated a negative correlation between ChT and AL during myopia progression.^42–44 In the control group, the ChT of eyes with a −30 D lens was lower than that of eyes with a 0 D lens (P = 0.001) (Fig. 2A). However, in BH-treated eyes, choroidal thinning was significantly suppressed (0 D lens vs. −30 D lens: IP injection, P > 0.999 and eye drop, P > 0.999). To evaluate ChBP in LIM mice with or without BH eye drop administration, blood perfusion signals were measured using SS-OCT (Fig. 2B). In the control group, the blood signals of the choroidal region (which represents ChBP) in eyes with a −30 D lens was lower than that of eyes with a 0 D lens (P = 0.038) (Fig. 2C). In the BH administered group, the ChBP signals in both eyes were higher than those in the control group, respectively (0 D lens, P = 0.023; −30 D lens, P = 0.045), indicating that BH can increase ChBP. 

Figure 2.

 

BH maintains ChT and increases ChBP in LIM mice. (A) Changes in ChT between zero and three weeks after LIM. The average values at baseline (before LIM) and three weeks after LIM are shown in Supplementary Figure S1C. n = 8. **P < 0.01, ***P < 0.001, two-way ANOVA with a Bonferroni post hoc test. (B) En face OCT images (upper) and OCT angiography images of the retinochoroidal section (below) of the eyes with a 0 D or −30 D lens administered PBS IP or BH eye drop after three weeks of LIM. (C) The proportion of ChBP signals in the eyes with a 0 D or −30 D lens administered PBS IP or BH eye drop after three weeks of LIM. n = 8. *P < 0.05, Two-way ANOVA with a Bonferroni post hoc test. Bars represent mean ± standard deviation.

BH Suppressed Scleral Thinning in Myopia Progression

We previously reported that scleral thickness decreased with the progression of axial myopia in an LIM murine model.⁴⁵ To assess the scleral thickness with or without BH eye drops, we performed H&E staining of cross-sectioned enucleated mouse eyes after LIM (Fig.3A). Scleral thickness was thinner in eyes with −30 D lenses than in those with 0 D lenses (P = 0.0014) (Fig. 3B). However, in both eyes treated with BH, the thickness was maintained, suppressing scleral thinning caused by the progression of myopia. 

Figure 3.

 

BH suppresses scleral thinning in LIM mice. (A) Hematoxylin and eosin-stained images of the cross-sectioned eyes with a 0 D or −30 D lens administered PBS or BH eye drop after three weeks LIM. Scale bar: 50 µm. (B) Scleral thickness measured at the point of 500 µm from the optic nerve head. n = 8. *P < 0.05, **P < 0.01, two-way ANOVA with a Bonferroni post hoc test for comparison of the eyes. Bars represent mean ± standard deviation.

Myopia Suppressive Effects of BH That Depend on the Dose and Number of Eye Drops Administered

The commercial dose for the glaucoma ophthalmic solution (Detantol; Santen Pharmaceutical Co.) is 0.01% BH solution. To evaluate the efficacy of BH eye drops at a concentration of less than 0.01% in suppressing myopia, three different concentrations, 0.0001%, 0.001%, and 0.01% (normal dose), were administered to LIM mice, along with a control group that received PBS eye drops. Eyes treated with −30 D lenses showed a significant myopic shift compared with those treated with 0 D lenses in the PBS- (P < 0.001), 0.0001% (P < 0.001), and 0.001% (P < 0.001) BH-treated groups (Fig. 4B). The 0.01% BH-treated group showed a dramatically smaller change in refractive error with a −30 D lens than the PBS group at −30 D (P = 0.001). Similar to the results of refractive error, the eyes with −30 D lenses showed greater axial eye growth than those with 0 D lenses in the PBS- (P = 0.026) and 0.0001% (P = 0.039) BH-administered groups (Fig. 4C). There was no significant difference between the eyes with 0 D and −30 D lenses in the 0.001% (P = 0.650) and 0.01% (P = 0.976) groups. The ChT of the eyes with the −30 D lens decreased in the PBS group (P < 0.001) (Fig. 4D). Although the deviations were large, the results showed a trend that choroidal thinning was suppressed as the administration dose increased. 

Figure 4.

 

We found that 0.001% and 0.01% BH suppressed LIM in mice. The mice received daily administration of PBS for control or BH eye drops in varying concentration (0.0001%, 0.001%, 0.01%) for a period of three weeks (A). Changes in the refractive error (B), axial length (C), and choroidal thickness (D) between zero and three weeks after LIM. The average values at baseline (before LIM) and three weeks after LIM are shown in Supplementary Figure S2. n = 8. *P < 0.05, ***P < 0.001, two-way ANOVA with a Bonferroni post hoc test for comparison of the eyes with 0 D and −30 D of each group. #P < 0.05, ##P < 0.01, ###P < 0.001, two-way ANOVA with a Bonferroni post hoc test for comparison of eyes with a −30 D lens in each group. Bars represent the mean ± standard deviation.

Next, to evaluate the effective number of administrations to suppress myopia progression, BH eye drops were administered to LIM mice once for three weeks, once a week (every Monday), twice a week (every Monday and Thursday), and every day. The eyes treated with −30 D lenses showed significantly larger refractive changes than those treated with 0 D lenses in PBS (P < 0.001), once a week (P < 0.001), once for three weeks, and twice a week groups (Fig. 5B). The eyes with −30 D lenses in once a week (P = 0.015), twice a week (P = 0.001), and every day (P < 0.001) BH-administered groups showed smaller myopic shifts than those in the PBS group. Regarding changes in axial length, eyes treated with −30 D lenses showed larger axial elongation than those treated with 0 D lenses in the groups with PBS (P = 0.014) or one for three weeks (P = 0.029) (Fig. 5C). Axial elongation was suppressed in other BH groups. Changes in ChT also showed that the administration of BH suppressed thinning and was more effective as the number of administrations increased (Fig. 5D). 

Figure 5.

 

Daily administration of BH suppresses LIM in mice. A BH eye drop was administered to LIM mice once for three weeks, once a week, twice a week, and every day (A). Changes in refractive error (B), axial length (C), and choroidal thickness (D) between zero and three weeks after LIM. The average values at baseline (before LIM) and three weeks after LIM are shown in Supplementary Figure S3. n = 8. *P < 0.05, **P < 0.01, ***P < 0.001, two-way ANOVA with a Bonferroni post hoc test for comparison of the eyes with 0 D and −30 D lens of each group. #P < 0.05, ###P < 0.001, two-way ANOVA with a Bonferroni post hoc test for comparison of the eyes with a −30 D lens in each group. Bars represent mean ± standard deviation.

Comparison of Glaucoma Drugs and α1 Blockers in Terms of Myopia-Suppressive Effects

To compare the myopia-suppressive effect of BH with other existing glaucoma drugs and α1 blockers, eye drops of tafluprost, UIE, prazosin hydrochloride, or urapidil hydrochloride were administered to LIM mice. Tafluprost (0.0015%) and UIE (0.12%) are commercial glaucoma eye drops. They have been reported to not only decrease intraocular pressure but also improve retinal and optic nerve head circulation.^46,47 The change in refractive error with tafluprost eye drops showed a myopic shift in eyes with the −30D lens (P < 0.001), whereas that with UIE eye drops showed a myopic shift in both eyes with the 0 D (−10.16 ± 5.27 D) and −30 D (−8.34 ± 6.74 D) lenses (Fig. 6B). Tafluprost and UIE eye drops suppressed axial elongation with no significant difference between the eyes with 0 D and −30 D lenses (Fig. 6C). Tafluprost eye drops decreased ChT in the eye with −30 D lenses compared to that of BH group (P = 0.016) (Fig. 6D). 

Figure 6.

 

Alpha1 blockers, BH and prazosin hydrochloride, suppress LIM in mice. 0.01% BH, 0.0015% tafluprost, 0.12% UIE, 0.01% prazosin hydrochloride, or 0.01% urapidil hydrochloride eye drops were administered to LIM mice daily (A). Changes in refractive error (B), axial length (C), and choroidal thickness (D) between zero and three weeks after LIM. The average values at baseline (before LIM) and three weeks after LIM are shown in Supplementary Figure S4. n = 8. *P < 0.05, ***P < 0.001, two-way ANOVA with a Bonferroni post hoc test for comparison of the eyes with 0 D and −30 D of each group. #P < 0.05, ##P < 0.01, ###P < 0.001, two-way ANOVA with a Bonferroni post hoc test for comparison of the eyes with a −30 D lens in each administered group. Bars represent mean ± standard deviation.

Regarding α1 blockers, the refractive error changes with both prazosin hydrochloride and urapidil hydrochloride eye drops showed no significant difference between the eyes with 0 D and −30 D lenses (Fig. 6B). However, the axial length of the eye with −30 D lens in the urapidil group was higher than that in BH (P = 0.008) group (Fig. 6C). Changes in ChT of the eye with −30 D lens increased in the administration of BH (P = 0.015) and prazosin (P = 0.001) compared to PBS group (Fig. 6D). 

Discussion

We observed that BH, a selective α1-adrenoceptor antagonist, can suppress the thinning of the sclera and choroid and increase ChBP in LIM mice. Therefore, we propose that the topical application of BH through eye drops may be a viable approach for suppressing myopia progression. BH was originally developed as a treatment for benign prostatic hyperplasia and has minimal side effects when applied via eye drops or other methods. Furthermore, BH acts locally in the eye and promotes the outflow of aqueous humor from the uveoscleral outflow tract, which helps lower intraocular pressure. Previous research has shown that ChBP is significantly increased and intraocular pressure is decreased by 0.005% BH solution in normal rabbits.⁴⁸ Our results confirmed that ChBP decreased in myopia-induced eyes compared with that in control eyes, but BH administration led to higher ChBP in myopia-induced eyes than in those without administration. 

A correlation between ChT and axial length in patients with myopia has been reported, and consistent results have been shown in many myopia-induced experimental models.^{21,30,44,49–51} In our previous studies using a lens-induced mouse model, we demonstrated choroidal thinning with the progression of myopia.^11,52,53 Therefore studies have been conducted to prevent the progression and development of myopia by maintaining ChT and increasing ChBP levels.^20,54,55 

Zhou et al.⁵⁴ injected prazosin, one of the α1 blockers used in our current study, into the inferior peribulbar space in a form-deprivation myopia guinea pig model. By evaluating ChT and ChBP, prazosin increased both thickness and ChBP in the injected eye compared with those in the myopia-induced eye and suppressed myopia and axial elongation. The effects of maintaining the choroid and suppressing myopia were consistent with the results of the eye drops in the current study. 

Conversely, myopia was enhanced in the same model by injecting quinpirole, a dopamine D2 receptor agonist, to reduce ChBP levels.¹⁶ This reduction in ChBP may induce a hypoxic condition in the sclera, subsequently triggering remodeling of the scleral extracellular matrix and elongation of the axial length, resulting in myopia development.¹⁹ 

In Figure 2, it may appear that the changes in choroidal thickness and blood perfusion do not align. The thickness represents the change before and after myopia induction, while the blood perfusion reflects the value after myopia induction, so that it may cause a discrepancy. However, a comparison of choroidal thickness and blood perfusion values at 6 weeks of age revealed a correlation (Supplementary Fig. S8; R = 0.42, P = 0.018). It is thought that the increase in blood perfusion caused by the vasodilator action of BH may have prevented the decrease in choroidal thickness during myopia progression. Further investigation into the causal relationship between changes in thickness and blood perfusion is needed. 

Although BH, prazosin, and urapidil are classified as alpha-1-adrenergic blockers, which exert their effects by binding to alpha-1 adrenergic receptors in vascular smooth muscle and inducing vasodilation,^27,31 urapidil also possesses central agonist activity at the serotonin 5 HT1A receptor.⁵⁶ In this study, both BH and prazosin showed similar suppressive effects on choroidal thinning. Conversely, urapidil did not exhibit such effects and resulted in an axial elongation in the myopia-induced eye. A previous study using human hypertrophied prostatic adenomas showed that the selective α1 inhibitory effect was much higher in BH and prazosin than in urapidil.⁵⁷ Consistently, we showed that the effect of urapidil in suppressing myopia by maintaining ChT was lower than that of BH or prazosin. 

Under the hypothesis that myopia development may occur due to axial elongation caused by an increase of intraocular pressure,⁵⁸ several studies have been conducted on the drug repositioning of glaucoma treatments to inhibit the progression of myopia.^{8,28,29,59–62} It has been reported that some beta-adrenergic receptor antagonists^28,60,61 and prostaglandin analogs^29,62 have inhibitory effects on myopia progression, but they have not yet been widely used in clinical practice. For example, in a randomized clinical trial of 0.25% timolol in 150 Danish children over two years of age, there was no significant difference in myopia progression between the timolol and spectacle group.³⁰ Moreover, tafluprost and UIE are prostaglandin analogs that selectively bind to prostaglandin E and F receptors in the ciliary muscle, leading to relaxation of the ciliary muscle and increased outflow of aqueous humor.^63–65 Although eye drops of tafluprost or UIE, which are prostaglandin analogs, have been shown to improve blood circulation in the retina⁴⁶ or optic nerve head⁴⁷ in glaucoma patients, they did not show a preventive effect on myopic shift and choroidal thinning in this study. In this study, we used a concentration of 0.0015% tafluprost, which corresponds to the concentration found in commercially available glaucoma eye drops. Dong et al.⁶⁶ reported dose-dependent dilation of arteries in rabbits with tafluprost. Compared to their results, the concentration used in our study reduces the vasodilatory effect by more than half. Therefore it is speculated that a higher concentration of eye drops may be necessary to achieve choroidal vasodilation for myopia suppression. 

UIE has been shown to exhibit a vasodilatory effect on constricted arteries and vessels induced by endothelin-1.^67,68 Polska et al.⁶⁷ reported that four drops of UIE increased choroidal blood flow, suggesting that the single drop used in our study may not be sufficient to increase choroidal blood perfusion for myopia suppression. Furthermore, the underlying mechanism for the occurrence of myopia even in the eyes with a 0 D lens that were not myopia-induced by UIE eye drops remains unclear, it should be needed for further investigation into this matter. 

The findings from our experimental mouse model have not been proven by clinical examinations, which is a limitation of this study. Experimental studies using animal models have shown that reduced choroidal blood flow and thickness can lead to scleral hypoxia and subsequent scleral remodeling, contributing to myopia development.^16,19,20 On the other hand, the observed choroidal thinning may be a biophysical consequence of the stretching of the eye during myopic progression, as reported in previous studies on patients with high myopia.^69–71 Further research is needed to elucidate the mechanism underlying the suppression of myopia progression by BH and the relationship between choroidal changes and axial changes. Although the clinical safety of BH as an eye drop has been confirmed, further experiments and clinical trials based on the current results are needed. However, it is important to examine the safety of BH in children to prevent myopia at an early stage. This study showed that BH eye drops suppressed myopia progression with maintaining ChT and that ChBP could be used as a countermeasure against the rapidly increasing myopic population. 

Acknowledgments

Supported by Grants-in-Aid for Scientific Research (KAKENHI) from the Ministry of Education, Culture, Sports, Science and Technology to Heonuk Jeong (20K18357 and 23K15937) and by AMED-CREST under Grant Number JP22gm1510007. Also supported by the Grant for Myopia Research from Tsubota Laboratory, Inc. (Tokyo, Japan). 

Disclosure: H. Jeong, (P); D. Lee, None; X. Jiang, (P); K. Negishi, Seed (F), i.com medical GmbH (F); K. Tsubota, (P), Santen (F), Kowa (F), Rohto Pharmaceutical (F), SBI Holdings (F), Tsubota Laboratory (E, F, I, R), TissueTech (I), Cellusion (I), Restore Vision (I); T. Kurihara, (P), Fuji Xerox (F), Kowa (F), Rohto Pharmaceutical (F), Santen (F), Seed (F), Tsubota Laboratory (F, I), Waskasa Seikatsu (F), Restore Vision (I)