IOP Reduction in Nonhuman Primates by Microneedle Injection of Drug-Free Hydrogel to Expand the Suprachoroidal Space

Abstract

Purpose: Expansion of the suprachoroidal space (SCS) by a hydrogel injection has been shown to reduce intraocular pressure (IOP) in rabbits as a potential treatment for ocular hypertension in glaucoma. Here, we evaluate the safety and efficacy of this approach in hypertensive and normotensive eyes in nonhuman primates.

Methods: A microneedle was used to inject a hyaluronic acid–based hydrogel or saline solution (control) into the SCS of cynomolgus monkey eyes that were either normotensive (n = 7 experimental; n = 2 control eyes) or had induced ocular hypertension (n = 6 experimental; n = 3 control eyes). IOP and the degree of SCS expansion were monitored over time by tonometry and ultrasound biomicroscopy, respectively. Safety was evaluated through slit lamp, fundus, and histology examinations.

Results: In hypertensive eyes, SCS injection with hydrogel initially reduced IOP by 47.5 ± 16.7%, and IOP returned to baseline in 38 days. In normotensive eyes, hydrogel injection initially reduced IOP by 38.8 ± 8.1% and IOP gradually returned to baseline also in 39 days. Sham injections resulted in mild IOP reduction in hypertensive eyes and normotensive eyes. The hydrogel injections were well tolerated by clinical assessments.

Conclusions: IOP was reduced in nonhuman primates for over one month by sustained SCS expansion. This procedure was safe and simple to perform. These data confirm the translational potential of this treatment method. Further optimization of the hydrogel may provide longer durations of IOP reduction.

Translational Relevance: A microneedle injection of hydrogel into the suprachoroidal space may provide a non-surgical, non-pharmacologic treatment for ocular hypertension in glaucoma patients.

Introduction

Glaucoma, a progressive optic neuropathy,¹ is the leading cause of irreversible blindness, affecting 80 million individuals worldwide² and projected to increase to more than 111 million by 2040.³ Although the complex pathophysiology of glaucomatous optic neuropathy remains to be fully determined, it is known that significant, sustained intraocular pressure (IOP) reduction can prevent further vision loss.^4,5 Thus, all current treatments attempt to lower IOP, preferably indefinitely. 

Existing therapies for glaucoma, including pharmacologic methods, laser therapy, and surgery, have notable limitations. For example, anatomic and physiological barriers lead to poor ocular bioavailability of medications delivered by eyedrops, often requiring drops to be administered multiple times daily.^6,7 This increases the burden on patients, decreasing treatment adherence,^8,9 and is associated with local and systemic side effects from chronic therapy.^10,11 Laser therapy is comparable in efficacy to initial pharmacologic therapy; however, patients may still progress to requiring medical or surgical intervention.^12,13 Surgical treatments are associated with serious sight-threatening complications, such as endophthalmitis and suprachoroidal hemorrhage; are costly; and often require revisions because of device migration and exposure.^14–16 Thus, there is a pressing need for alternative, novel methods to effectively lower IOP in glaucoma, ideally avoiding the complications of drugs and surgery. 

IOP is determined by a balance of aqueous humor production rate, drainage resistance through the conventional and unconventional (uveoscleral) pathways, and outflow backpressure (episcleral venous pressure). Aqueous humor drains from the eye via the trabecular meshwork (conventional pathway) or into the posterior segment for clearance across the sclera or via the choroidal vasculature (unconventional pathway).¹⁷ Either reduced production or increased drainage of aqueous humor decreases IOP. The suprachoroidal space (SCS), a potential space between the sclera and choroid, is involved in the unconventional outflow pathway; in this pathway, aqueous humor passes through the ciliary muscle into the SCS before permeating across the sclera or entering the choroid.¹⁷ Thus, expansion of the SCS is a potential mechanism for IOP reduction. 

Reduction of IOP via modification of the SCS to facilitate greater aqueous humor outflow has been explored previously, but with limited success. Surgical incision into the SCS is not currently used because of high rates of postoperative complications.¹⁸ Devices that use the SCS for IOP reduction initially appeared promising in clinical trials; unfortunately, the only approved device was later recalled because of long-term damage to the corneal endothelium.^19–22 Other SCS devices in development have demonstrated sustained IOP reduction but require surgical implantation of the device.^23,24 

In contrast to surgical SCS access or device implantation, injections of a viscous agent into the SCS have shown IOP reduction without the use of pharmacologic agents.²⁵ Our research previously demonstrated a sustained (four-month) lowering of IOP in normotensive rabbits after a single SCS injection of a hydrogel.²⁶ The in situ–forming hydrogel caused SCS expansion and was hypothesized to lower IOP by facilitating greater unconventional outflow, with the degree of SCS expansion correlating with the degree of IOP reduction. Furthermore, these injections were very well tolerated as judged by clinical and histological examination. 

In this approach, microneedles less than 1 mm-long were used to deliver the hydrogel to the SCS by crossing the conjunctiva and sclera witout penetrating the choroid. This approach is advantageous, offering minimally invasive and targeted SCS access²⁷ and thus avoiding the risks of incisional surgery. Furthermore, microneedle injections can be performed clinically as an in-office procedure.^28,29 The safety and reliability of this technology have been demonstrated by a treatment recently approved by the Food and Drug Administration for macular edema by SCS microneedle injection of a long-acting steroid, triamcinolone acetonide.³⁰ 

In this study, we asked whether hydrogel injection into the SCS would produce sustained IOP reduction in nonhuman primates, like it did in rabbits. To answer this question, we evaluated the effect on IOP caused by SCS expansion via hydrogel injection by microneedles in cynomolgus monkeys, in both hypertensive and normotensive eyes. 

Methods

Microneedle and Hydrogel Preparation

Microneedles were prepared from either 27-gauge or 30-gauge hypodermic needles (Air-Tite, Tochigi, Japan), as previously described (Fig. 1).²⁶ Briefly, the needles were ground to a length of 900 to 1000 µm and a bevel angle of 30° to 40° using a cordless rotary tool (Dremel 800; Robert Bosch, Gerlingen, Germany), and the appropriate length and bevel were confirmed using a stereomicroscope (Olympus SZX12; Evident, Tokyo, Japan). The needle was then sterilized with ethylene oxide (Anprolene AN74j sterilizer, Andersen Products, Haw River, NC, USA). 

Figure 1.

 

Representative images of a microneedle and hydrogel used for injections. (A) Microneedle (arrow) connected to a 1 mL syringe. (B) Microneedles were 900 to 1000 µm in length. (C) A representative volume of 50 µL hydrogel after gelation in vitro.

An in situ–forming, crosslinked hydrogel was prepared as previously described using thiol-modified hyaluronic acid (HA-SH) (Glycosil; ESI Bio, Alameda, CA, USA) and poly(ethylene glycol) diacrylate (PEGDA, 3500 Da; JenKem Technology, Beijing, China).²⁶ Aliquots of these materials were prepared in a conventional biosafety cabinet, and sterile Hanks’ balanced salt solution (HBSS; Gibco, Grand Island, NY, USA) was used to dissolve the HA-SH and PEGDA separately. The two solutions were then combined to yield a final concentration of 1.5% (w/v) HA-SH and 4.5% (w/v) PEGDA. The mixture was loaded into a syringe (1 mL Luer-lock plastic syringe; BD Bioscience, San Jose, CA, USA) and injected after 15 minutes at room temperature (20°–25°C) using the microneedle. 

Animals

Eleven adult female cynomolgus monkeys (weight 2–6 kg; 15–22 years old on study initiation) were used in this study. All animal studies were carried out in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Visual Research, and all experimental procedures were approved by the Georgia Institute of Technology and University of Nebraska Medical Center Institutional Animal Care and Use Committees. 

Each monkey had previously undergone unilateral (left eye) argon laser photocoagulation (Ophthalas 532 EyeLite Laser; Alcon Laboratories, Fort Worth, TX, USA) of the trabecular meshwork under isoflurane anesthesia to induce ocular hypertension.³¹ Fifty to 100 laser photocoagulation burns with a spot size of 50 µm, a power of 1000 mW, and an exposure time of 0.5 s were made to the trabecular meshwork in an arc subtending 270° to 340°. Treatments were repeated one or two times with one month recovery between treatments within three months of the first treatment until IOP was persistently elevated in treated eyes. The remaining trabecular meshwork (i.e., an arc subtending 20° to 90°) was left untreated. This procedure to generate hypertension was performed on the animals in this study 10 to 19 years before this study was conducted. Information on the monkeys is summarized in the Table. 

Table.

 

Characteristics of Animals Included in Hypertensive and Normotensive Studies

 

Animals were grouped as either experimental (receiving an injection of hydrogel) or control (receiving an injection of sterile HBSS). We conducted two studies. In the first, hypertensive eyes were injected. Six experimental animals received a 50 µL injection of the hydrogel in their hypertensive eye, whereas three control animals received a 50 µL injection of HBSS in their hypertensive eye. An injection volume of 50 µL was used because this was the same volume used in our previous study in rabbits.²⁶ 

In the second study, normotensive eyes were injected. Given the larger diameter and axial length of monkey eyes³² compared to rabbit eyes,³³ it was also of interest to use a larger volume of 100 µL, which has also been previously studied in rabbits for SCS injection^34,35 and is the volume administered clinically to the suprachoroidal space for treatment of macular edema.³⁶ Thus, seven experimental animals received a 100 µL injection of the hydrogel in their normotensive eye, whereas two control animals received a 100 µL injection of HBSS in their normotensive eye. 

Suprachoroidal Space Injections

General anesthesia was induced with ketamine (intramuscular injection of 5–15 mg/kg) and maintained with inhaled isofluorane (2%–3%) in oxygen gas. Before SCS injection, eyes were irrigated with sterile saline solution and sterilized with povidone-iodine 5% solution (Betadine; Alcon Laboratories). Topical ophthalmic anesthetic (proparacaine 0.5%; Akron, Lake Forest, IL, USA) was then applied, and a lid speculum was placed to separate the eyelids. Either hydrogel (experimental eyes) or sham (HBSS; control eyes) was injected into the SCS using a microneedle 3 mm posterior to the limbus in the inferotemporal or superotemporal quadrant of the eye. SCS injection was confirmed by ultrasound biomicroscopy, described below. Ophthalmic antibacterial ointment (neomycin, polymyxin B sulfate and bacitracin zinc ophthalmic ointment USP; Bausch & Lomb, Rochester, NY, USA) was applied immediately after injection and the following day. 

Tonometry and Clinical Evaluations

IOP was measured using a pneumatonometer (Reichert Model 30; Reichert Technologies, Depew, NY, USA) in awake animals between 8 AM and 10 AM, and the average of two or three consecutive measurements was recorded. Baseline IOP values were taken as an average over three measurements (measured at three separate timepoints over the course of approximately one month before injection for hypertensive eyes and four days before injection for normotensive eyes) in each injected eye. Baseline IOPs for each injected eye of every animal are shown in Supplementary Figure S1 in the Supporting Information. 

Clinical evaluations included examinations of the fundus, anterior segment, and SCS, and were performed under general anesthesia (ketamine, 5–15 mg/kg) by a board-certified ophthalmologist. Anterior segment examinations were performed using a Kowa SL-15 handheld slit lamp (Kowa, Tokyo, Japan). Fundus examinations were performed with a binocular indirect ophthalmoscope with a handheld 20 D ophthalmoscopic lens after animals received tropicamide 1% (Bausch & Lomb) and 2.5% phenylephrine dilating eye drops (Justice Ophthalmics, Cordova, TN, USA). Fundus imaging was performed using a smartphone camera (IPhone 12; Apple, Cupertino, CA, USA) and a conventional 20 D lens. 

The extent of SCS expansion was determined by analysis of images of the limbal region, acquired by ultrasound biomicroscopy (UBM Plus; Accutome, Malvern, PA, USA) with a sterile probe cover (ClearScan; Eye-Surgical-Instruments, Plymouth, MN, USA). From these images, the maximum eye wall thickness (EWT) was measured, as previously described.²⁶ Briefly, EWT was defined as the distance from the external conjunctival surface to the internal limiting membrane of the retina using the scale provided by the software on the ultrasound system (Accutome connect 8.02.02) and ImageJ (Version 1.53k; NIH, Bethesda, MD, USA). Baseline values of EWT were measured immediately before injection at the injection site, and changes in EWT (defined as “Delta EWT”) were defined as the difference between EWT measurements at each time point and the eye's baseline EWT value. 

The approximate timepoints for tonometry and clinical evaluations for both studies were as follows (see Supplementary Table S1 for exact timepoints for each study). The hypertensive study was completed first and helped to inform the timepoints for the normotensive study. 

• • Baseline IOPs were measured as described above.
• • In the hypertensive eye study:  
  • ○ Awake IOP was measured daily for one week after injection and then once per week until IOP returned to baseline. A penlight examination was performed at each IOP check and followed up with a slit lamp examination if any concerns about infection or inflammation, such as hyperemia or lid swelling, were noted.
  • ○ Sedated fundus imaging and slit lamp examinations were performed on the day of injection and then once a month until IOP returned to baseline.
  • ○ Sedated UBM images were acquired on the day of injection and then once every two weeks.
• • In the normotensive eye study:  
  • ○ Awake IOP was measured daily for two weeks after injection and then twice weekly until IOP returned to baseline. Animals were monitored for signs of infection or inflammation as described above.
  • ○ Sedated fundus imaging, slit lamp examinations, and UBM imaging were done twice per month until IOP returned to baseline.

Histology Tissue Preparation and Morphologic Analysis

Three months (normotensive eyes) and eight months (hypertensive eyes) after injection, monkeys (total 10 eyes; including non-injected eyes: n = 2, HBSS-injected eyes: n = 2, hydrogel-injected eyes: n = 6; Table) were anesthetized with ketamine (10 mg/kg), and their eyes were fixed in situ with 4% paraformaldehyde for 30 minutes. After euthanasia, eyes were enucleated and further immersion-fixed with 4% paraformaldehyde for 48 hours. Subsequently, eyes were stored in phosphate-buffered saline solution before further tissue processing. 

Radial dissection of the eyes was performed at the injection site, and corresponding sections were prepared from the noninjected eye to serve as a negative control. The sections underwent dehydration using a series of ethanol concentrations, gradually increasing from 70% to 100%, and xylene. Dehydrated sections were then paraffin-embedded, and 8- to 10-µm sections were cut and stained using the Hematoxylin and Eosin Stain Kit (Vector Laboratories, Newark, CA, USA), following the manufacturer's suggested protocol after gradient alcohol dewaxing. Images at magnification ×4, ×20, and ×40 were captured using a light microscope (Olympus America, Center Valley, PA, USA). 

Statistical Analysis

As previously described, baseline IOP values were computed as the average of three individual IOP measurements per eye taken prior to injection (Supplementary Fig. S1) and are presented as mean ± standard deviation (SD; Table). For each eye, we then defined “Delta” measures as the changes with respect to baseline (i.e., Delta IOP at a given time was the IOP at that time minus the mean baseline IOP for that eye). For graphing purposes, we then pooled Delta IOP values over all eyes within each treatment group at each time point. 

We observed that, after an initial IOP drop, IOP slowly returned to baseline (i.e., Delta IOP slowly returned to zero). This return was reasonably well fit by a straight line (see below), and because we were interested in the duration of IOP lowering, we used least squares linear regression to model Delta IOP versus time after injection (GraphPad Prism software, version 10.0.1 (170) for macOS; GraphPad Software, San Diego, CA, USA). The duration of IOP lowering efficacy was then determined as the time at which the upper 95% confidence band of the treatment group Delta IOP, as predicted by the fitted regression line,²⁶ was equal to zero. This corresponds approximately to a P value = 0.05 associated with the null hypothesis that Delta IOP was not different from zero, as previously described.^26,37 For experimental hypertensive eyes, the regression was fit between days 5 to 45 after injection, while for experimental normotensive eyes the regression was fit from days 5 to 39 post-injection. We chose these time ranges for statistical analysis because Delta IOP data from later timepoints and from days 1–4 both demonstrated plateau regions (i.e., by visual inspection) and thus were not suited to fitting by a model describing changes with time. For normotensive control eyes, all timepoints were included in the regression model. Descriptive statistics for each model are found in Supplementary Table S3. 

An exception to the above was the hypertensive control eyes, for which the regression fit did not result in a slope significantly different from zero (Supplementary Table S3). We thus defined a “Delta baseline IOP” value, computed as individual baseline IOP measurements minus the mean of those measurements for each eye. This quantity has a mean value of zero, by definition, but has variance associated with it which is required for statistical testing. To obtain this measure of variance, we pooled all individual Delta baseline IOP values (three per eye) over all eyes within each treatment group and computed a standard deviation from these pooled data (Supplementary Table S2). These standard deviations were used to generate confidence intervals for each treatment cohort. Finally, for this cohort of eyes, Delta IOP was compared to zero (the mean Delta baseline IOP) at all timepoints with a paired t-test. 

Correlations of Delta IOP on Delta EWT for hypertensive and normotensive studies were calculated by least squares linear regression (GraphPad Prism software); measurements from all timepoints were included. 

Results

Hydrogel Injection Into the SCS Lowered IOP in Both Ocular Hypertensive and Normotensive Monkey Eyes

We tested the hypothesis that SCS expansion caused by a hydrogel injection would reduce IOP in nonhuman primates. In experimental animals with ocular hypertension, we unilaterally injected a single 50 µL volume of an in situ–forming hyaluronic acid-based hydrogel into hypertensive eyes and measured the change in IOP from baseline IOP over time; control animals were injected with the same volume of HBSS. For hydrogel injections, the initial IOP decrease was 15.1 ± 7.2 mm Hg one day after injection (a 47.5% ± 16.7% decrease), and statistical analysis demonstrated that IOP remained significantly lower than baseline for 38 days (P = 0.05; Fig. 2A). Sham injections in hypertensive eyes reduced IOP by −4.78 ± 5.12 mm Hg (P = 0.0094; Fig. 2C). 

Figure 2.

 

Effect of hydrogel injection on intraocular pressure and SCS expansion. Hypertensive and normotensive monkey eyes were injected with the crosslinked hydrogel (A and B, respectively) or a sham injection of Hanks’ Balanced Salt Solution (C and D, respectively). Change in IOP compared to pre-injection baseline values (Delta IOP) is plotted. Delta baseline IOP means (blue solid lines, equal to zero by definition) and confidence intervals (blue dotted lines; see text) are shown in each panel. In panels A, B, and D, the regression fit (solid black line) and confidence band (gray area) are shown; the intersection of the confidence band and the mean Delta baseline IOP is indicated by the black box, which indicates the time point at which the Delta IOP was no longer significantly different from zero. In panel C, the mean Delta IOP (solid black line) and confidence intervals (gray area) are shown in black. Plotted data are the average ± SD of six (A, hydrogel), three (C, sham), seven (B, hydrogel) or two (D, sham) replicates.

We also tested our hypothesis in normotensive eyes. Experimental animals received a single 100 µL hydrogel injection, while control animals received a HBSS sham injection of the same volume. Hydrogel injection reduced IOP by 8.3 ± 2.2 mm Hg one day after injection (a 38.8% ± 8.1% decrease), and the decrease in IOP from baseline remained statistically significant 39 days after injection (P = 0.05; Fig. 2B). The sham injections in normotensive eyes initially reduced IOP by 3.8 ± 4.0 mm Hg (a 15.9% ± 14.0% decrease), and IOP reduction remained significant for seven days after injection (P = 0.05; Fig. 2D). In summary, hydrogel injection into the SCS was associated with a sustained decrease in IOP in both hypertensive and normotensive eyes. 

SCS Expansion Correlated With IOP Decrease

We also hypothesized that the decrease in IOP from its baseline level would correlate with the degree of SCS expansion, as we previously observed in rabbits.²⁶ We therefore monitored SCS expansion by measuring EWT by ultrasound biomicroscopy, with Delta EWT being used as a measure of SCS expansion, because adjacent ocular tissues, such as the sclera and choroid, were not expected to change in thickness after injection (Fig. 3). 

Figure 3.

 

Representative ultrasound biomicroscopy images of the SCS after hydrogel injection. (A) Hypertensive and (B) normotensive eyes are shown after injection with crosslinked hydrogel or sham injection of Hanks’ Balanced Salt Solution. All images are in the same orientation (C, cornea; I, iris; CB, ciliary body; Ch, choroid; H, hydrogel; S, sclera). The dotted yellow line approximately outlines the hydrogel in the SCS; these images were used to determine EWT. Each injection type (sham or hydrogel) across time points is from the same eye. Images are representative of three sham eyes and six hydrogel eyes for hypertensive animals and two sham eyes and seven hydrogel eyes for normotensive animals.

In hypertensive eyes, EWT increased by 1.74 ± 0.50 mm immediately after hydrogel injection, corresponding to the initial 15.1 mm Hg decrease in IOP and then slowly returned to baseline for the duration of decreased IOP (Fig. 4A). Delta EWT and Delta IOP demonstrated moderate correlations that reached statistical significance (linear regression, R² = 0.26; Pearson r = −0.51, P = 0.0005), with each 500 µm increase in EWT corresponding to an approximately 2.7 mm Hg IOP decrease (Fig. 4C). 

Figure 4.

 

Correlation of Delta IOP versus Delta EWT. Delta IOP (circles; right axis) with linear regression fit and confidence band (black line and gray area; right axis) and Delta EWT (square; left axis) are shown for (A) hypertensive and (B) normotensive eyes. A cross-plot of Delta EWT versus Delta IOP (data from graphs A and B) are shown for (C) hypertensive and (D) normotensive hydrogel-injected eyes. Both hypertensive (linear regression R² = 0.26, Pearson r = −0.51, P = 0.0005) and normotensive (R² = 0.60, Pearson r = −0.78, P < 0.0001) eyes demonstrated that Delta IOP was negatively correlated with Delta EWT. In panels C and D, the solid lines represent the linear regression fits, with the gray-shaded areas representing the 95% confidence bands of the linear regressions.

In normotensive eyes, EWT initially increased by 2.4 ± 0.20 mm, corresponding to the initial 8.3 mm Hg decrease in IOP, and, similar to the situation in hypertensive eyes, gradually returned to baseline as IOP returned to its baseline (Fig. 4B). Delta EWT and Delta IOP were significantly correlated (linear regression, R² = 0.60; Pearson r = −0.78, P < 0.0001), with each 500 µm increase in EWT corresponding to an approximately 1.5 mm Hg IOP decrease in ocular normotensive monkey eyes (Fig. 4D). The initial changes in EWT were significantly different (P < 0.0088) for hypertensive versus normotensive eyes (1.74 ± 0.50 mm vs. 2.4 ± 0.20 mm, respectively). 

SCS Hydrogel Injections Were Well Tolerated

We evaluated the safety of SCS injections of the crosslinked hyaluronic-acid hydrogel. Sham injections of saline solution in both hypertensive and normotensive eyes did not result in any notable changes in ocular surface appearance, and eyes appeared normal by slit lamp and fundus examinations immediately after injection and at all subsequent timepoints (Fig. 5). 

Figure 5.

 

Representative fundoscopic images of eyes after hydrogel injection. (A) Hypertensive and (B) normotensive eyes are shown after injection with crosslinked hydrogel or sham injection of Hanks’ balanced salt solution. Images are from the same eye across time points for each injection type. Images are representative of three sham eyes and six hydrogel eyes for hypertensive animals and two sham eyes and seven hydrogel eyes for normotensive animals. No changes in retinal morphology due to injection were identified at any timepoint. N, nasal; T, temporal.

In both hypertensive and normotensive eyes, hydrogel injections also did not change ocular surface appearance immediately after injection, and no irritation was evident in normotensive eyes over time after injection. However, in hypertensive hydrogel injected eyes (n = 6), we observed mild ptosis and eyelid edema one day after injection, at which point topical antibiotics were applied once. This irritation may have resulted from animal handling rather than hydrogel injection, as discussed below. The irritation lessened by the following day (Fig. 6) and resolved within a week. The mild generalized hyperemia was limited to the ocular surface, and no anterior chamber cells or flare were noted on slit lamp examinations. Slit lamp examinations at all other timepoints also showed no anterior chamber reaction. Furthermore, fundoscopy did not show any changes to the posterior eye, including no evidence of retinal detachment or tears, choroidal effusions, hemorrhage, or hydrogel leakage, at any time point. No significant structural abnormalities, such as cyclodialysis cleft or ciliary body detachment, were observed by ultrasound biomicroscopy. 

Figure 6.

 

Representative photographs of the external eye after hydrogel injection. (A) The injected eye immediately before and after injection and (B) both injected and control eyes on day 1 and day 2 after injection. Hydrogel was injected in the inferotemporal quadrant of the left eye, and the injection site was visible immediately after injection (arrow). Mild edema occurred one day after injection and was nearly resolved by the second day. N, nasal; S, superior.

Ocular tissues were excised from both normotensive and hypertensive eyes at least three months after hydrogel or HBSS injection and examined histologically after hematoxylin and eosin staining by a trained morphologist (H.G.). No active inflammation or fibrosis was observed at the site of injection (Fig. 7). The degree of pigmentation seen in histologic images was in the normal range for monkeys and did not appear to be affected by the injection. Evidence of SCS expansion was seen in the images, but was believed this to be an artefact of histological preparation because UBM imaging had shown complete closure of the SCS (Figs. 3 and 4). While no active inflammation or fibrosis was observed, gel residue and multinuclear giant cells were present in the SCS of eyes injected with 100 µL (2/3 eyes; tissue obtained three months after injection; Figs. 8A, 8B). In contrast, these features were not observed in the SCS of noninjected eyes (0/2 eyes) or eyes injected with saline solution (0/2 eyes; Figs. 8C, 8D) or 50 µL of gel (0/3 eyes; tissue obtained eight months after injection). 

Figure 7.

 

Low- and medium-magnification representative histologic images of ocular tissue collected at least three months after SCS injection. (A, B) A noninjected ocular hypertensive eye (negative control). (C, D) A saline-injected normotensive eye at the injection area. (E, F) A hydrogel-injected normotensive eye at the injection area. Images A, C, and E were captured under magnification ×4, whereas images B, D, and F show further magnification of the boxed areas 3 mm from the limbus (approximate location of injection) captured under magnification ×20. Expansion of the SCS is likely an artefact of histology preparation, with no evidence of active fibrosis or inflammation at the injection site. Variations in pigment could be observed but were within the normal range in monkeys.

Figure 8.

 

High-magnification representative histologic images of ocular tissue after injection. (A, B) A normotensive eye injected with 100 µL hydrogel. Images are at the injection area three months after injection. (C, D) A hypertensive eye injected with 50 µL of hydrogel. Images are at the injection area eight months after injection. Images A and C were captured under magnification ×4, whereas images B and D show magnification ×40 of the boxed areas 3 mm from the limbus (approximate location of injection). Blue arrows indicate gel residue, and the yellow arrow points to the giant cells.

Discussion

Microneedles were used to inject hydrogel into the SCS, thereby expanding this potential space and lowering IOP in hypertensive and normotensive eyes of cynomolgus monkeys. A single injection of hydrogel decreased IOP without significant safety concerns in both hypertensive and normotensive eyes. This result is consistent with prior findings in rabbits,²⁶ and reinforces the translational potential of this method for IOP lowering in the treatment of glaucoma. 

Effect of SCS Expansion by Hydrogel Injection on IOP

A significant IOP decrease was observed for over one month in hypertensive and normotensive eyes receiving hydrogel (Figs. 2A, 2C). Although the exact mechanism of IOP decrease caused by hydrogel has yet to be definitively determined, SCS expansion by the injected hydrogel likely increases aqueous humor drainage via the unconventional pathway,²⁶ similar to the increase in unconventional outflow evident in cyclodialysis.^38–40 For example, in monkeys, experimentally induced ciliochoroidal detachment by surgical injection of a solution into the SCS resulted in transient IOP reduction.⁴¹ 

Under normal physiological conditions, unconventional outflow results from aqueous humor passing through the interstitial spaces of the ciliary muscle and into the supraciliary and suprachoroidal spaces, and then further through the sclera (uveoscleral), choroid and vortex veins (uveovortex), or lymphatic vessels (uveolymphatic) pathways.^17,42 The ciliary body is an important part of the unconventional outflow pathway and changes in ciliary muscle structure alter unconventional outflow rates.^43–45 The hydrogel placement (as shown in Fig. 3) was adjacent to the ciliary muscle, and although no ciliary body detachment was noted on UBM imaging, the ciliary muscle structure may have been affected, reducing resistance and thus increasing flow through the unconventional outflow route. 

We also observed a mild IOP decrease in sham-injected eyes (Figs. 2B, 2D), which is consistent with transient IOP reductions seen with SCS saline solution injections in rabbits and may be associated with the injection itself. In normotensive eyes receiving saline solution, this IOP decrease was transient and of short duration, whereas in hypertensive eyes receiving saline solution it was of longer duration; however, both treatment groups had very small sample sizes (two and three eyes, respectively), and, considering our previous studies in rabbits,²⁶ we expect that a larger study would show that the IOP-lowering effects of saline solution alone are minimal.⁴⁶ 

Previously, we demonstrated that injection of 50 µL of crosslinked hyaluronic-acid hydrogel into the SCS in normotensive rabbit eyes decreased IOP for four months.²⁶ In this study, the same volume of hydrogel was injected into hypertensive monkey eyes, whereas twice that volume was injected into normotensive monkey eyes. In both experiments, the duration of the resulting IOP decrease was considerably shorter than in rabbits, as was the corresponding duration of SCS expansion. In other words, the injected hydrogel appeared to be cleared more rapidly in monkeys than in rabbits, which was associated with a faster return to baseline IOP. 

The cause of the different dynamics in monkeys versus rabbits is uncertain, but several possible explanations may be considered. First, there may be a higher concentration of hyaluronidase in monkey ocular tissues than in rabbit eyes.^47–49 The hydrogel used in the current study is hyaluronic acid based and thus can be degraded by endogenous hyaluronidase, similar to hyaluronic acid–based hydrogels used in other applications.^50,51 Alternatively, anatomical and functional differences of the ciliary muscle may contribute to faster hydrogel clearance. Specifically, monkeys accommodate more than rabbits and have a more developed ciliary muscle,^17,52 and it is possible that ciliary muscle-induced mechanical deformation contributes to hydrogel degradation and/or clearance.⁵³ Finally, monkeys have greater baseline unconventional outflow rates than rabbits,¹⁷ and fluid flow impacts polymer degradation profiles.^54,55 The role of these possible effects in human eyes will be important to study when assessing future translational potential. 

It is of interest to consider the effect of injected hydrogel volume. As noted above, hypertensive eyes received a 50 µL hydrogel injection, whereas normotensive eyes received 100 µL, leading to a significantly greater initial change in EWT in normotensive versus hypertensive eyes. Despite this initial difference, both groups of eyes demonstrated similar durations of IOP reduction and EWT expansion. In previous studies, fluid injection into the SCS increased SCS expansion by a certain amount, after which injection of additional volume resulted in lateral flow of the fluid in the SCS, rather than a further increase in SCS width.²⁷ However, injection of more viscous formulations, such as a carboxymethylcellulose, resulted in a greater maximum SCS width due to the solution's significant resistance to lateral flow,⁵⁶ consistent with our hydrogel injections. Taken together, these data suggest that there is a volume “threshold” to enable increased unconventional outflow and IOP lowering, after which point further volume injection at the same site may not improve efficacy. Delta EWT data, where the SCS remained expanded ∼1 mm on day 25 for both hypertensive and normotensive eyes. Thus, there may be no therapeutic benefit to increasing injection volume beyond 50 µL of hydrogel; however, the optimal injection volume in non-human primates needs to be further investigated. It also remains to be determined whether injections at multiple locations improve efficacy, although previous experiments in rabbits suggested this was not the case.²⁶ 

Correlation Between SCS Expansion and IOP Reduction

Changes in eye wall thickness, a measure of SCS expansion, were correlated with changes in IOP in both hypertensive and normotensive eyes, similar to previous observations in normotensive rabbit eyes. This correlation was weaker in hypertensive eyes, possibly due in part to the greater variability in hypertensive eye IOP (Table), as previously reported in this laser-induced ocular hypertensive monkey model.^31,57,58 Furthermore, it has been suggested that there may be long-term changes in unconventional outflow in monkey eyes after laser treatment, such as small artificial clefts created from laser burns during the glaucoma-induction procedure.³¹ Thus it may have been that certain monkeys in this study had greater-than-normal unconventional outflow rates at the start of our experiments, so that the hydrogel injection was less efficacious. In humans, unconventional outflow decreases with age⁴³ and in glaucoma,⁵⁹ which may offer a greater therapeutic effect of this treatment. 

Safety Profile of Hydrogel Injection Into the SCS

Hydrogel injections into the SCS were well tolerated, with no significant abnormalities being noted by slit lamp examinations, ultrasound biomicroscopy, or fundus imaging at any timepoint. The 900 to 1000 µm microneedle length resulted in a successful SCS administration without evidence of choroidal hemorrhage or intravitreal injection in all monkeys in the study. Our microneedle length was similar to that of 900–1100 µm microneedles that are currently used in humans for SCS delivery of triamcinolone acetonide as approved by the Food and Drug Administration.³⁶ 

On histologic analysis, eyes with hydrogel injection had gel residue present three months after injection but not eight months after injection (Fig. 8); this difference may be due to both the postinjection duration and volume of injected gel (100 µL vs. 50 µL), and the immune response should be considered when determining the optimal injection volume. It is likely trace amounts of gel may remain even after SCS expansion and IOP return to baseline and eventually degrade. Multinuclear giant cells present in hydrogel-injected eyes (Fig. 8) suggest some foreign body response to the hydrogel. However, no active inflammation or fibrosis was seen in eyes with hydrogel injection at three months and eight months after injection (Fig. 7), suggesting that local immune response may resolve with degradation of the gel. This is somewhat different than our study in rabbits, where eyes with hydrogel exhibited minimal to moderate fibrotic formation²⁶ and may be due to species differences in ocular immune response.^60,61 Optimization of the hydrogel formulation may reduce the inflammatory response, and additional studies are needed to assess the clinical relevance of these results. 

Although there was mild edema and ptosis on external examination one day after hydrogel injection in hypertensive eyes (Fig. 6), this was nearly fully resolved by the second day and fully resolved within one week. This edema was seen only in eyes receiving hydrogel injections, and affected the periocular tissues, rather than being localized to the injection site or involving the anterior chamber. Given this spatial distribution, it is likely that the edema did not result from the hydrogel material itself but rather from the greater physical manipulation of the eye in anesthetized animals during the hydrogel injection; for example, by stabilizing the eye in the orbit for the duration of the hydrogel injection (15–30 seconds) versus sham injections HBSS (<5 seconds). Furthermore, for the normotensive study, injection technique was improved to minimize ocular manipulations, and ptosis and edema were correspondingly not observed for hydrogel-injected eyes. 

Translational Potential of SCS Hydrogel Injections for Glaucoma Therapy

SCS hydrogel injections can provide a minimally invasive method to reduce IOP for extended periods of time without the use of medication or surgery. Expansion of the SCS demonstrated significant initial IOP decreases of 47.5% and 38.8% for hypertensive and normotensive eyes, respectively, which corresponds to similar or even greater IOP reduction than current topical medications.⁶² Furthermore, the use of microneedles allowed for a precise injection targeted to the SCS without the need for any surgical procedure, and this injection has the potential to be performed in the clinic setting, similar to current intravitreal and SCS injections. 

Both 27-gauge and 30-gauge microneedles, which are the same gauges commonly used in intravitreal injections,⁶³ were initially tested in this study, and the 27-gauge microneedle was judged by the surgeon to be easier to use for injecting the fairly viscous hydrogel because of its reduced flow resistance compared to the smaller 30-gauge microneedle. Because the hydrogel injection does not involve any pharmacologic agents to reduce IOP, it may be especially advantageous for patients who cannot tolerate certain topical medications or do not experience sufficient IOP reduction from eye drops. 

A notable finding of this study is the shorter duration of IOP reduction in the monkey model versus previous studies in rabbits.²⁶ This observation motivates optimization of the hydrogel to decrease clearance rate, thereby prolonging the treatment duration. As previously mentioned, the hydrogel used in this study consisted of hyaluronic-acid polymer with PEGDA as a crosslinker; although the addition of crosslinker slows degradation, the hydrogel can still undergo degradation by hyaluronidase.²⁶ Thus, new formulations that slow down hydrogel degradation could be beneficial. Developing a longer-lasting hydrogel would be expected to lower IOP for longer durations, decreasing the required frequency of repeat injections. Determination of the exact mechanism of IOP reduction from hydrogel SCS expansion may also better inform various aspects of this treatment method, such as the optimal volume and placement of hydrogel for maximum IOP reduction. 

Limitations

There were several limitations to this study. First, we used a laser-induced ocular hypertension model that is known to produce variable hypertensive IOP levels and does not fully mimic glaucoma.³¹ The laser procedure was performed over a decade ago, further increasing variability of IOP levels, as shown in the baseline IOP measurements in the hypertensive eyes (Table). This may have in turn resulted in variable increases in unconventional outflow, thereby reducing the statistical power of this study. 

Second, the control group (sham injection) sample sizes for both hypertensive and normotensive studies were small (n = 3 and n = 2 eyes, respectively). As a result, control group IOP data also had high variability. Furthermore, although use of a nonhuman primate model demonstrates greater translational potential of this treatment method compared to non-primate models, anatomical and physiological differences between monkeys and humans should be considered; in particular, cynomolgus monkeys have higher absolute and relative rates of unconventional outflow versus humans.¹⁷ 

Third, we were also limited in our evaluation of this treatment method's safety profile. Although no overt safety concerns were present during both experiments, further safety studies including examination of visual acuity, visual fields, and retinal function should be conducted. Additionally, previous SCS devices implanted ab interno have shown decreased corneal endothelial cell density;⁶⁴ although our approach uses a transscleral approach, central corneal thickness should be measured in future studies. We also did not perform additional histologic evaluation for fibrosis, such as by staining for α-smooth muscle actin, in the eyes involved in this study. In our prior study, histologic examination of SCS injections of the crosslinked hydrogel in rabbits demonstrated a mild, localized inflammatory response,²⁶ but the clinical significance of these findings is unknown; further histopathological analysis would be valuable in future studies using a nonhuman primate model. Finally, all animals used in this study were females, so future preclinical work should consider animals of both sexes. 

Conclusions

We demonstrated IOP reduction for over one month in normotensive and hypertensive eyes of cynomolgus monkeys, with concomitant expansion of the SCS, after injection of a hydrogel with a minimally invasive microneedle. When combined with prior data in rabbits, this study establishes greater translational potential of this technology as a means to reduce IOP over extended durations. With further optimization, SCS expansion by hydrogel injection may provide a nonsurgical, nonpharmacological treatment option for patients with glaucoma. 

Acknowledgments

Supported by the Georgia Research Alliance, BrightFocus Foundation (G 2022013S) and the Rifkin Family Glaucoma Research Fund. 

Disclosure: Y.G. Chung, None; S. Fan, None; V. Gulati, None; H.-L. Li, None; H. Gong, None; C.B. Toris, None; M.R. Prausnitz, Clearside Biomedical, Inc. (I, O), US20230201110A1 (P); C.R. Ethier, Equinox Ophthalmic (C), Injectsense (I), US20230201110A1 (P)