Preoperative Use of the Rho Kinase Inhibitor Ripasudil Protects the Corneal Endothelium from Trabeculectomy-Induced Damage

Introduction

Rho kinase (ROCK) regulates cell shape and size through cytoskeleton modulation, contributing to cell contraction regulation in a calcium-independent manner.¹ In the field of ophthalmology, ROCK inhibitors have emerged as treatment options for glaucoma^2,3 and corneal endothelial disorders.^4,5

The ROCK inhibitor Y-27632 promotes adhesion, inhibits apoptosis, and enhances the proliferation of corneal endothelial cells in cynomolgus monkeys.⁶ Similarly, in humans, corneal endothelial cell injection with Y-27632 supplementation can restore cornea clarity, maintain normal morphology, and achieve remission of corneal edema.⁷ Y-27632 has demonstrated promising results in treating acute surgical trauma, preventing the progression of bullous keratopathy, and managing various corneal diseases.^8,9

Other ROCK inhibitors that have been approved for clinical use in glaucoma, including ripasudil and netarsudil, warrant further investigation for the treatment of corneal endothelial disorders.¹⁰ Ripasudil eye drops, with an established safety profile since their introduction in Japan in 2014,^11,12 have demonstrated potential in protecting the corneal endothelium post-surgical trauma.¹³

Trabeculectomy (TRAB), a surgical treatment for glaucoma, can result in iatrogenic corneal endothelial damage.¹⁴ As a standard intraoperative strategy, ophthalmic viscosurgical devices (OVDs), including dispersive agents, are used during cataract surgery to protect the corneal endothelium by coating the endothelial surface and minimizing mechanical and ultrasound-related damage. However, endothelial cell loss can still occur despite these protective measures. This damage disrupts the endothelial pump function, leading to corneal stromal edema, a form of irreversible bullous keratopathy that impairs vision.¹⁵ TRAB-induced surgical trauma is a significant contributor to corneal transplantation.¹⁶

Although the effectiveness of postoperative ripasudil use for corneal endothelium protection has been reported,¹³ few studies focused on preoperative use. In this study, we aimed to investigate the efficacy of preoperative ripasudil administration in protecting the corneal endothelium of patients who underwent TRAB for glaucoma. We compared the outcomes between those who received and did not receive preoperative ripasudil.

Materials and Methods

Patient Population

In this retrospective observational study, we evaluated a total of 139 consecutive primary open-angle glaucoma (POAG) eyes from 139 patients who underwent TRAB (TRAB alone or combined with phacoemulsification and intraocular lens implantation [TRAB+PHACO+IOL]) at the Hyogo Medical University. The study was approved by the Institutional Review Board of the Hyogo Medical University (approval number: 4709) and adhered to the tenets of the Declaration of Helsinki. Informed consent was obtained from all participants. This study did not involve patient or public participation in its design, recruitment, or conduct.

POAG diagnosis was based on the following: the open anterior chamber angles on gonioscopy, the presence of localized retinal nerve fiber layer defects on red-free examination or optic nerve head damage (cup-to-disc ratio asymmetry >0.2 between fellow eyes, rim thinning, notching, or excavation), and the glaucomatous visual field damage. Glaucomatous visual field damage was defined as beyond the normal limits based on a glaucoma hemifield test or three abnormal points with a p < 5% probability of being normal, one with p < 1% by pattern deviation, or pattern standard deviation of p < 5% on two consecutive, reliable tests (≤33% fixation losses and false negatives, ≤15% false positives). Cases of pseudoexfoliation, neovascular, secondary, and developmental glaucoma were excluded.

The inclusion criteria were as follows: (1) patients with POAG who underwent TRAB alone or TRAB+PHACO+IOL, (2) specular microscope images obtained within 12 months before and after the operation, and (3) patients without other anterior segment disorders (eg bullous keratopathy, epithelial erosions, infections, conjunctivitis, and nasolacrimal duct obstruction). The exclusion criteria were as follows: (1) patients with a history of postoperative ripasudil use, (2) absence or poor quality of specular microscopy images, and (3) a preoperative specular microscope image obtained within 6 hours after ripasudil instillation (in the ripasudil group) to avoid endothelial cell density (ECD) underestimation attributed to the morphological change of the endothelium induced by ripasudil.¹⁷

Following exclusions, 68 POAG eyes that underwent TRAB alone and 71 that underwent TRAB+PHACO+IOL were included. Among the total of 139 eyes, 67 (48.2%) were treated with preoperative ripasudil (Kowa Company, Ltd., Tokyo, Japan, twice a day) for > 3 months until the day before surgery, including 38 (55.9%) of the 68 eyes in the TRAB alone group and 29 (40.8%) of the 71 (40.8%) in the TRAB+PHACO+IOL group. No glaucoma eye drops, including ripasudil, were administered postoperatively in any group until the postoperative specular microscopy examination.

TRAB was performed under local anesthesia. A fornix-based conjunctival flap and a partial-thickness scleral flap were created in the superior quadrant. In the TRAB+PHACO+IOL group, standard phacoemulsification and IOL implantation were performed. Thereafter, a trabecular block was excised to create a sclerostomy, and peripheral iridectomy was performed. The scleral flap was closed with adjustable tension using sutures, and the conjunctiva was sutured watertight. In all cases, mitomycin C was generally used in the sub-Tenon pocket at a concentration of 0.4 mg/mL for 3 min. No incidents of surgical complications, such as anterior chamber hemorrhage, capsule rupture, or lens drop, were noted. Moreover, no incidents of bullous keratopathy attributable to surgery were noted.

Preoperative and Postoperative Examinations

Images of corneal endothelial cells (0.55 mm high, 0.25 mm wide) were taken within 12 months preoperatively and within 12 months postoperatively using a non-contact specular microscope (Specular Microscope 8; Konan Co. Ltd., Tokyo, Japan). ECD, percentage of hexagonal cells (HEX), central corneal thickness (CCT), and coefficient of variation (CV) were calculated. In the ripasudil group, the preoperative examinations were performed >6 hours after ripasudil instillation to avoid ECD underestimation attributed to the morphological change of the endothelium induced by ripasudil.¹⁷ Patient fixation determined the imaging point, wherein locations were based on the patient’s primary line of sight rather than the normal vertex. The central area was confirmed using a monitoring camera. Intraocular pressure (IOP) was measured using the applanation tonometer. The glaucoma drug score was calculated by assigning one point for each glaucoma eye drop and two points for the combined eye drops and oral carbonic anhydrase inhibitors.¹⁸

Statistical Analysis

The Wilcoxon signed-rank test was used to compare ECD, HEX, and CV pre- and postoperatively in both the ripasudil and without-ripasudil groups. The Mann–Whitney U-test was used to compare ECD, ECD loss (%), HEX, CV, CCT, logMAR visual acuity, IOP, glaucoma drug score, duration of surgery, age, and sex between the groups. To adjust for potential confounding, a multivariable linear regression analysis was conducted to evaluate the association between the ECD loss rate and the treatments with and without preoperative ripasudil. The regression model was adjusted for age, sex, surgeon, preoperative glaucoma drug score, preoperative ECD, and follow-up duration. Sex and surgeon were treated as categorical variables. Regression coefficients and two-sided p values were calculated using the ordinary least squares method. A p value of less than 0.05 was considered statistically significant. All the statistical analyses were performed using the SPSS version 25.0 software (IBM Corp., released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp. USA).

Results

Patient Demographics and Baseline Characteristics

This study included 139 eyes from 139 patients, with a mean age of 69.0 ± 10.0 years. No statistically significant differences were observed between the ripasudil and without-ripasudil groups (Table 1) in terms of age, male/female ratio, ECD, HEX, CV, CCT, logMAR visual acuity, IOP, and duration of surgery. The glaucoma drug score in the ripasudil group was naturally higher than that in the without-ripasudil group owing to the administration of ripasudil eye drops. Two surgeons performed the surgery; 50 of the 67 eyes that received ripasudil were operated on by surgeon 1, and the remaining 17 by surgeon 2. Among the 72 eyes that did not receive ripasudil, 45 were operated on by surgeon 1. The distribution exhibited no significant differences (p = 0.12, Pearson’s chi-squared test). Specular microscope images were obtained preoperatively (40.7 ± 57.7 days) and postoperatively (71.6 ± 66.0 days). The postoperative observation period did not vary significantly between the ripasudil and without-ripasudil groups (82.3 ± 69.7 days and 61.5 ± 61.1 days, respectively; p = 0.10).

Table 1 Comparison of the Preoperative Background Characteristics Between the Ripasudil and Without-Ripasudil Groups for Trabeculectomy

Postoperative ECD Changes

For the total 139 eyes, the ripasudil group demonstrated minimal change in the mean ECD following surgery (preoperative: 2303.4 ± 463.7 cells/mm² vs postoperative: 2292.7 ± 455.3 cells/mm², p = 0.68). Conversely, the without-ripasudil group exhibited a significant decrease in the mean ECD after surgery (preoperative: 2431.6 ± 320.8 cells/mm² vs postoperative: 2265.3 ± 387.6 cells/mm², p = 1.3×10⁻⁵). ECD loss, calculated with the formula [−(ECD after surgery − ECD before surgery) / ECD before surgery × 100%], was significantly lower in the ripasudil group (−0.16 ± 11.02%) than in the without-ripasudil group (6.66 ± 11.45%) (p = 0.0036, statistical power = 0.93, Figure 1).

Figure 1 ECD loss postoperatively in the ripasudil and without-ripasudil groups for trabeculectomy. ECD loss is calculated as follows: −(ECD postoperatively − ECD preoperatively)/ECD preoperatively × 100%. The horizontal lines in the box and whisker plots represent the median values. The bottom and top of the boxes represent the lower and upper quartiles, respectively. The bars represent the minimum and maximum values within 1.5 times the lower and upper quartiles. Abbreviation: ECD, endothelial cell density.

The multivariable linear regression analysis revealed that preoperative ripasudil treatment was independently associated with a significantly lower ECD loss after adjustment for age, sex, surgeon, preoperative glaucoma drug score, preoperative ECD, and postoperative observation period (p = 0.0024).

Sub-Analysis

Among the eyes that underwent TRAB alone, the ripasudil group (38 eyes) demonstrated minimal change in the mean ECD following surgery (preoperative: 2212.6 ± 465.8 cells/mm² vs postoperative: 2225.0 ± 474.0 cells/mm², p = 0.87), whereas the without-ripasudil group (30 eyes) exhibited a significant decrease (preoperative: 2294.6 ± 377.0 cells/mm² vs postoperative: 2170.4 ± 409.7 cells/mm², p = 0.010). However, ECD loss was not significantly lower in the ripasudil group (−1.03 ± 11.42%) than in the without-ripasudil group (5.20 ± 10.17%) (p = 0.09, statistical power = 0.66, Figure 2).

Figure 2 ECD loss postoperatively in the ripasudil and without-ripasudil groups for trabeculectomy alone. ECD loss is calculated as follows: −(ECD postoperatively − ECD preoperatively)/ECD preoperatively × 100%. The horizontal lines in the box and whisker plots represent the median values. The bottom and top of the boxes represent the lower and upper quartiles, respectively. The bars represent the minimum and maximum values within 1.5 times the lower and upper quartiles. Abbreviation: ECD, endothelial cell density.

Similarly, among the eyes that underwent TRAB+PHACO+IOL, the ripasudil group (29 eyes) exhibited minimal change in the mean ECD after surgery (preoperative: 2422.5 ± 440.8 cells/mm² vs postoperative: 2381.4 ± 421.2 cells/mm², p = 0.56), whereas the without-ripasudil group (42 eyes) demonstrated a significant decrease (preoperative: 2529.5 ± 232.7 cells/mm² vs postoperative: 2333.0 ± 360.7 cells/mm², p = 2.4×10⁻⁴). ECD loss was significantly lower in the ripasudil group (0.97 ± 10.56%) than in the without-ripasudil group (7.70 ± 12.29%) (p = 0.040, statistical power = 0.79, Figure 3). Representative preoperative and postoperative images of the same eye from each of the ripasudil and without-ripasudil groups are shown in Figures 4 and 5, respectively.

Figure 3 ECD loss postoperatively in the ripasudil and without-ripasudil groups for trabeculectomy, combined with phacoemulsification and intraocular lens implantation. ECD loss is calculated as follows: −(ECD postoperatively − ECD preoperatively)/ECD preoperatively × 100%. The horizontal lines in the box and whisker plots represent the median values. The bottom and top of the boxes represent the lower and upper quartiles, respectively. The bars represent the minimum and maximum values within 1.5 times the lower and upper quartiles. Abbreviation: ECD, endothelial cell density.

Figure 4 Representative image of specular microscopy from the ripasudil group for trabeculectomy alone. The endothelial cell density was preserved. Abbreviations: CD, endothelial cell density; AVE, average of the cell area (μm2); SD, standard deviation of the cell area; MAX, maximum of the cell area; CV, coefficient of variation; MIN, minimum of the cell area; 6A, percentage of hexagonal cells; NUM, analyzed number of the cells; PACHY, central corneal thickness.

Figure 5 Representative image of specular microscopy from the without-ripasudil group for trabeculectomy combined with phacoemulsification and intraocular lens implantation. The endothelial cell density was decreased. Abbreviations: CD, endothelial cell density; AVE, average of the cell area (μm2); SD, standard deviation of the cell area; MAX, maximum of the cell area; CV, coefficient of variation; MIN, minimum of the cell area; 6A, percentage of hexagonal cells; NUM, analyzed number of the cells; PACHY, central corneal thickness.

Other Outcomes

For both TRAB and TRAB+PHACO+IOL, HEX and CV did not change significantly after surgery in the ripasudil and without-ripasudil groups (Supplementary Table 1).

Discussion

The findings of this study suggest that preoperative ripasudil administration offers protection against corneal endothelial cell loss in patients with glaucoma who underwent TRAB. Using ripasudil preoperatively suppresses the endothelial loss rate to almost 0%. The protective effect of ripasudil is positive for not only eyes that undergo TRAB but also those that undergo TRAB+PHACO+IOL. The protective effect is positive in TRAB group, but only the ECD loss was not significant. The statistical power for ECD loss was 66% suggesting that the study may have been underpowered to detect a difference in this outcome.

The combination of TRAB, phacoemulsification, and IOL implantation causes more endothelial cell damage than TRAB alone.¹⁸ The endothelial protective effects of ripasudil were satisfactory for TRAB alone and even sufficient for TRAB+PHACO+IOL, despite the higher surgical stress.¹⁹ TRAB, especially when combined with mitomycin C use, causes corneal endothelial damage.^20,21 In our study, all patients received mitomycin C, and TRAB-induced endothelial damage was observed in the without-ripasudil group.

Our findings are consistent with those of a previous study demonstrating the effectiveness of postoperative ripasudil use in protecting the corneal endothelium of patients with low ECD during cataract surgery.¹³ The protective effect of ripasudil is partially mediated by the ability of ROCK inhibitors to promote endothelial cell migration.^6,22,23 Particularly, Okumura et al¹⁰ reported that ripasudil promoted in vitro human endothelial cell proliferation and rabbit corneal endothelial wound healing; however, rabbit endothelial cells proliferate under physiological conditions, whereas human endothelial cells do not.²⁴ Nevertheless, ripasudil may enhance the function of human endothelial cells by improving cell survival, as reported in our study.

ROCK inhibitors can promote corneal endothelial cell proliferation in cultured cells.^6,7 In the case presented in Figure 4, the postoperative ECD exceeded the preoperative value. While further investigation is required to identify the extent to which topical ROCK inhibitor treatment can exert a similar effect under physiological conditions, the possibility that ROCK inhibition contributed to endothelial cell proliferation in vivo cannot be completely ruled out.

Although postoperative ripasudil administration has been discussed, the mechanism underlying the protective effect of preoperative ripasudil use remains unclear. One possibility is that ripasudil may have a persistent effect despite its decaying concentration postoperatively.²⁵ This could allow ripasudil to influence endothelial cell proliferation, triggered by surgical trauma.¹⁰ However, this explanation might be contradicted by studies demonstrating the disappearance of ripasudil-induced morphological changes in the human endothelium within 6 hours of instillation.²⁶ Alternatively, preoperative ripasudil use might strengthen the endothelium, allowing it to better withstand surgical stress. This possibility is supported by reports demonstrating that ripasudil increases the expression of actin, N-cadherin (a marker of adherence junctions), and Na⁺-K⁺-ATPase (a marker of pump function) in rabbit corneal endothelium.^10,27

Ripasudil causes transient morphological change of the endothelium.^17,26 Owing to the morphological changes, the non-contact specular microscope produced unstable data for determining ECD within 1 h of ripasudil instillation. As the corneal endothelium morphology gradually recovered to normal within 6 h, in the ripasudil group, our examinations were performed after >6 hours of ripasudil instillation to avoid ECD underestimation attributed to the morphological change of the endothelium.

Whether preoperative or postoperative ripasudil use offers better endothelial protection remains unclear. A previous study reported on postoperative ripasudil use in patients with low ECD during cataract surgery,¹³ and the present study on preoperative use reports almost complete inhibition of ECD loss induced by TRAB alone or TRAB+PHACO+IOL. In addition, the desirable preoperative instillation period remains unclear. In this study, the instillation period was >3 months, and ECD loss was almost absent in the TRAB group; thus, 3 months seems to be a sufficient period. However, further examination is warranted to fully understand these issues.

The mean glaucoma drug score was 4.5 in the ripasudil group and 3.5 in the without-ripasudil group, exhibiting a difference of 1 point. This difference corresponds to the presence or absence of ripasudil itself as one component of the glaucoma drug score. Although the possibility that it might reflect a difference in baseline characteristics cannot be excluded, it is also attributable to the inclusion of ripasudil in the treatment regimen. In addition, possible synergistic or antagonistic effects of other topical medications on the corneal endothelium should be considered. However, because most patients in the present cohort were treated with multiple concurrent topical agents, it was difficult to analyze the independent or interactive effects of individual medications. This issue should be addressed in future studies.

Ripasudil use has also been explored in the treatment of Fuchs’ endothelial dystrophy. In this approach, Descemet’s membrane and endothelium are removed (up to 4 mm in diameter) from the central cornea, followed by ripasudil administration.^28,29 This procedure eliminates the damaged endothelium, allowing the healthy endothelium from the periphery to mitigate and compensate for the eliminated area. Collectively, available data suggest that ripasudil may be useful in protecting the corneal endothelium during intraocular surgeries, such as TRAB.

The use of ripasudil eye drops, a formulation approved for clinical use, corroborates our findings and the potential application of the drug. Our study supports the idea of developing ROCK inhibitor formulations as eye drops for both preoperative and postoperative treatments of surgery-induced acute corneal endothelial damage, including that by TRAB. Additionally, these findings suggest that ripasudil might be not only beneficial for the treatment of TRAB but also broadly useful for protection of the corneal endothelium after intraocular surgery.

Regarding the effect of mitomycin C and TRAB, ECD loss is typically observed during or immediately after surgery, with no further cell loss between 3–12 months postoperatively.³⁰ Although our postoperative observation period of approximately 2–3 months was relatively short, the observed ECD loss in the without-ripasudil group was consistent with that observed previously (5–10%),³¹ suggesting that our study captured most of the ECD loss.

A limitation of this study was its retrospective design. The period of the preoperative ripasudil use was not uniform. Future studies should prospectively evaluate our findings using a randomized, blinded design. The data presented here highlight the importance of examining a larger cohort of patients with and without preoperative ripasudil administration to confirm its endothelial protective effect. A relatively small sample was evaluated in this study; thus, a larger sample size could contribute to different results.

Conclusion

This study demonstrated that preoperative ripasudil administration effectively protects against TRAB-induced endothelial damage. Given its established safety profile in patients with glaucoma, these findings suggest the potential protective efficacy of preoperative ripasudil application in a broader range of conditions affecting the corneal endothelium during intraocular surgery.