Suprachoroidal Space Alterations Following Delivery of Triamcinolone Acetonide: Post-Hoc Analysis of the Phase 1/2 HULK Study of Patients With Diabetic Macular Edema

Introduction

Pharmaceuticals delivered by intravitreal injection have revolutionized the management of the most common causes of blindness in developed countries, including diabetic macular edema (DME), neovascular age-related macular degeneration, and retinal venous occlusive disease (RVO).¹ Current medications encompass two classes: anti-vascular endothelial growth factor (VEGF)-A and corticosteroid agents. Although intravitreal corticosteroid pharmaceuticals are highly effective for the management of exudative retinal diseases, they are often relegated to second-line interventions due to their well-recognized tendency to accelerate cataract formation and cause elevated intraocular pressure (IOP).² Delivery of corticosteroids into the suprachoroidal space (SCS) may allow treatment of posterior pole pathologies while minimizing levels in the anterior parts of the eye, theoretically decreasing their impact on the crystalline lens and trabecular meshwork.³ Suprachoroidal injection of CLS-TA, a preservative-free, terminally sterilized, aqueous ophthalmic suspension of triamcinolone acetonide (TA) (Clearside Biomedical, Alpharetta, GA) has demonstrated preliminary safety and efficacy for the management of noninfectious uveitis,⁴ RVO,⁵ and DME,⁶ with multiple ongoing phase 2 and 3 randomized, controlled trials. Although animal models have been used to study SCS alterations following injection of TA,^7–10 studies examining changes in the anterior SCS in humans following suprachoroidal delivery of CLS-TA are lacking. The current study considers the impact of suprachoroidal CLS-TA on SCS anatomy among eyes of subjects with DME in the HULK (Open-Label Study of the Safety and Efficacy of Suprachoroidal CLS-TA Alone or in Combination with Intravitreal Aflibercept for the Treatment of Diabetic Macular Edema) trial.

Patients and Methods

Following institutional review board approval and confirmation of written, informed consent of the current sub-study, subjects within the ongoing prospective HULK trial (IND 115683; NCT02949024) receiving suprachoroidal CLS-TA (0.1 cc, 4 mg) for DME were systematically imaged with anterior segment spectral-domain optical coherence tomography (AS-OCT) (Spectralis; Heidelberg Engineering, Heidelberg, Germany). At each visit, line (15°, 768 A-scans) scans of eight predefined quadrants, four standard and four oblique, were captured of the study and fellow eye (Figure 1). HULK patients who had completed the HULK trial prior to the current sub-study were invited back for a single imaging session. Among the limited number of eyes still enrolled within the HULK trial, AS-OCT images were obtained before and 30 minutes following suprachoroidal injection of CLS-TA at each consecutive injection visit. If no suprachoroidal CLS-TA was administered, a total of 16 images were captured during the session, whereas 24 images — 16 of the study eye and eight of the fellow eye — were captured if the patient received CLS-TA treatment.⁶ All study procedures adhered to the tenets set forth in the Declaration of Helsinki and the Health Insurance Portability and Accountability Act.

Figure 1. Visual representation of the different anterior-segment spectral-domain optical coherence tomography (AS-OCT) scan quadrants labelled in green (A). Starting from the 12-o'clock position on the right eye and proceeding clockwise: superior (I), superior-nasal (II), nasal (III), inferior-nasal (IV), inferior (V), inferior-temporal (VI), temporal (VII), and superior-temporal (VIII). Below, a superior AS-OCT scan of a study eye prior to suprachoroidal CLS-TA injection (B). The calipers used to analyze the thicknesses of the choroid (orange), sclera (blue), and choroid plus suprachoroidal space plus sclera (green) are displayed along with 1,000 μm incremental measurements posterior from Schlemm's canal (red).

Suprachoroidal width was measured using calipers native to Heidelberg Eye Explorer software (Heidelberg Engineering, Heidelberg, Germany) (Figure 1). A 1:1 pixel-to-micron ratio was applied. Images without a clearly visible choroid were excluded from analysis (Figure 2). Measurements were taken at 2 mm, 3 mm, 4 mm, and 5 mm posterior to Schlemm's canal. Thickness of the SCS was calculated by subtracting the thickness of the choroid and the sclera from the thickness of the choroid plus SCS plus sclera (Figure 1). Data were analyzed by pooling and averaging all measurements observed at a given position in each of the eight quadrants. The resulting means were statistically analyzed using two-tailed Student's t-tests for between-group comparisons and paired t-tests for within-group comparisons.

Figure 2. Anterior segment optical coherence tomography (AS-OCT) scan of the superior quadrant of a study patient after receiving a suprachoroidal injection of CLS-TA (A). From left to right: Schlemm's canal (I), scleral spur (II) episcleral vasculature (III), sclera (IV), choroid (V), suprachoroidal space (VI). Example of an AS-OCT scan excluded from analysis due to insufficient definition of the choroid (B).

A subgroup analysis of eyes that were imaged before and after a suprachoroidal CLS-TA injection was conducted by comparing the average change of each of the eight quadrants. Comparisons between quadrants were performed by averaging measurements from the four positions within each quadrant. The mean suprachoroidal width of each of the eight quadrants were statistically compared to one another using one-way analysis of variance and Tukey's honest significant difference test. All statistical analyses were conducted using R version 3.3.1 (R Project for Statistical Computing, www.r-project.org). P values less than .05 were considered statistically significant.

Results

Suprachoroidal Imaging Among Eyes Imaged Before and 30 Minutes After Suprachoroidal CLS-TA Injection Within the HULK Trial

Three imaging sessions involving two patients were performed at the time of suprachoroidal CLS-TA injection. A total of 48 AS-OCT images were captured. Twenty (42%) were included, whereas 28 (58%) were excluded due to an inability to adequately visualize the choroid. Prior to CLS-TA injection in these eyes, mean suprachoroidal width was 9.9 μm (± 11.1 μm). Among the six adequately visualized quadrants, mean suprachoroidal width increased to 75.1 μm (± 74.6 μm) (P < .001) 30 minutes following suprachoroidal CLS-TA administration. Among eyes imaged before and after a suprachoroidal CLS-TA injection, mean suprachoroidal width increased to 17.2 μm (± 14.0 μm), 90.7 μm (± 86.9 μm), 112.3 μm (± 78.1 μm), and 56.8 μm (± 44.7 μm) at 2 mm, 3 mm, 4 mm, and 5 mm posterior to Schlemm's canal, respectively, when all six adequately imaged quadrants were averaged. A comparison of preinjection and postinjection values found significant expansion at 3 mm (P = .019), 4 mm (P = .004), and 5 mm (P = .047); no significant increase in suprachoroidal width was observed 2 mm or less posterior to Schlemm's canal (P = .13) (Figure 3) 30 minutes following injection.

Figure 3. The anatomic changes induced before (orange) and 30 minutes after (blue) a suprachoroidal CLS-TA injection at 2 mm (A), 3 mm (B), 4 mm (C), and 5 mm (D) posterior to Schlemm's canal, averaged over all quadrants. Standard error of the mean is represented by the error bars.

The eight quadrants of the two eyes receiving suprachoroidal CLS-TA injection were compared. Data from the 2 mm, 3 mm, 4 mm, and 5 mm positions were averaged to find the mean suprachoroidal width in each quadrant after an injection. A total of three injections, two (67%) superior-temporally and one (33%) superior-nasally, were imaged. Using data averaged across three injection visits, the superior-temporal quadrant was statistically significantly larger 30 minutes after an injection than the superior, nasal, inferior-nasal, and inferior-temporal quadrants (P < .05), but not statistically significantly larger than the temporal, superior-nasal, and inferior quadrants. At each of the three injection visits, the three regions with the greatest expansion were the superior-temporal region at two visits and the superior-nasal region at one visit. The greatest mean expansion of the superior-temporal regions were 298.5 μm (± 41.7 μm) and 94.0 μm (± 88.4 μm), whereas the superior-nasal region expanded to a mean of 52.0 μm (± 8.7 μm). When aggregated across visits, the superior-temporal quadrant increased by a mean of 175.8 μm (± 130.0 μm). Mean scleral width also increased slightly from 468.5 μm (± 46.6 μm) to 492.5 μm (± 103.1 μm), possibly due to postinjection irritation and subsequent swelling; however, this increase was nonsignificant (P = .217). Mean suprachoroidal width returned to 14.9 μm (± 15.0 μm) by 1 month after the last suprachoroidal CLS-TA injections (P = .221). The suprachoroidal width of the five areas that experienced the greatest expansion after an injection were compared; there was no significant difference between the mean suprachoroidal width before (4.2 μm ± 3.3 μm) and after (5.4 μm ± 3.9 μm) injection at the 1-month follow-up period (P = .625).

Discussion

The changes induced within the SCS following suprachoroidal pharmaceutical injections are incompletely understood. The current sub-study within the larger HULK prospective trial found that the SCS width among patients treated with CLS-TA increased significantly 30-minutes following an injection, returned to preinjection levels within 1 month, and appeared stable after a mean follow-up of 4.8 months following the last CLS-TA injection into the SCS with no apparent measurable impact on SCS anatomy.

In comparison to the current study, Willoughby et al. reported an analysis utilizing enhanced-depth OCT imaging targeting the posterior-pole in patients with macular edema secondary to RVO, in which the posterior SCS widened measurably following the first suprachoroidal injection in nine (24%) of 38 patients; Willoughby et al. only observed this outcome in the posterior pole of eyes with a visible SCS at baseline.¹¹ Prior research may provide insight into the initially apparent conflict between these two data sets. A study involving 18 primate eyes reported that the posterior SCS experiences a mean of 2.9 mm Hg less pressure than the anterior SCS and has a 3.7 mm Hg lower pressure than the mean IOP.¹² This reduced hydrostatic pressure experienced by the posterior SCS may allow slight separation of the choroid in that region from the adjacent sclera, creating a visible SCS at baseline in some patients. It is possible the anterior SCS did not show sustained expansion in the current study due to greater hydrostatic forces relative to the posterior SCS. In the current study, the increased hydrostatic pressure experienced by the anterior SCS may minimize the visibility of the SCS, making this anterior extension of the SCS more of a potential space rather than a real space in most patients. Furthermore, ex vivo analyses of rabbit eyes have highlighted the reduced adhesive capacity of the lamellae connecting the choroid to the sclera after suprachoroidal injections.¹² If these findings in animal eyes are applicable to human eyes, the higher pressure experienced by the anterior SCS may offset the reduced adhesion of the lamellae, preventing a consistently visible SCS after CLS-TA has been cleared from the SCS.

Within the current work, expansion of the SCS appeared to begin beyond 2 mm posterior to Schlemm's canal among eyes imaged before and after a suprachoroidal CLS-TA injection. One potential explanation for this observation is a contact force exerted by the ciliary muscle on the choroid.¹³ During imaging, patients focused on an adjustable light in specific positions to expose the desired scleral quadrants. Contraction of the ciliary muscle while viewing the fixation point presumably caused the muscle belly to expand, pushing on the choroid and inhibiting CLS-TA from expanding into the underlying SCS. Resistance to filling would be highest near the origin of the ciliary muscle, the scleral spur (Figure 2). Therefore, expansion of the SCS would be increasingly limited toward the scleral spur.

Among eyes imaged before and 30 minutes following CLS-TA injection, expansion of the SCS was observed to be most prominent proximal to the injection site, as the regions with the greatest expansion coincided with the injection site, with notable variability in extent of expansion across different quadrants and across different patients. Many factors likely impact the timing and extent of diffusion of CLS-TA within the SCS, including variable force exerted during the injection procedure itself, variable intrinsic rigidity of the adjacent sclera that may impact SCS expansion, variable globe curvature, and variable density of material on either the inside (eg, vitreous) or outside (eg, subconjunctival hemorrhage) of the SCS that may restrict SCS expansion.

Pharmacokinetic analyses using animal models of various suprachoroidally administered fluids have found a relationship between drug formulation, lipophiliticy, concentration, and distribution time.^7,10,16 A comparative study between saline, indocyanine green (ICG), and a TA suspension of similar concentration to CLS-TA found TA had a slower distribution rate; this was attributed to the suspension formulation of TA.¹⁰ In comparison, ICG and saline are both solutions that are rapidly cleared from the SCS through the surrounding vasculature. Additionally, Chiang et al.⁷ found that highly viscous fluids continued to spread throughout the SCS of rabbit eyes up to 2 days after injection. In the current sub-study, the 30-minute lapse between SCS injection and repeat SCS imaging may not have allowed sufficient time for CLS-TA to diffuse completely, resulting in SCS changes artificially appearing to be localized around the site of initial delivery, and potentially explaining the large standard deviations reported in the current study. To analyze the variability of the current data set, the index of dispersion, a measurement of variability of a data set, was calculated by dividing the sample variance by the mean. Values less than one indicate a clustered data set, whereas values greater than one are a sign of high variability. This calculation returned a value of 84.7, indicating the data was over dispersed with substantial variance, consistent with the clinical finding that CLS-TA was unequally distributed throughout the quadrants, leading to varying degrees of SCS expansion.

To the authors' knowledge, this the first report of anatomic changes induced in the anterior SCS immediately following suprachoroidal injection in humans. The current results suggest that AS-OCT before and after suprachoroidal injection may allow confirmation of delivery of the pharmaceutical by assessing SCS changes. The correlation of changes in the SCS following suprachoroidal injection with clinical outcomes is yet to be explored.

The current sub-study carries many limitations. First, no baseline SCS imaging was performed. Instead, comparisons were made between the study eye and fellow eye to gauge changes from baseline. Second, the small sample size may have affected the observed data. Third, 51% of all images captured were ungradable due to inability to visualize the SCS. Future studies should consider using swept-source OCT (SS-OCT) rather than spectral-domain OCT due to the higher penetration of SS-OCT wavelengths into intraocular structures.¹⁷ Fourth, repeated SCS imaging at fixed time-points following injection may allow more complete characterization of SCS expansion and contraction. Fifth, SCS injection technique may substantively impact observed results; for example, CLS-TA reflux out of the SCS may partially explain the differences in SCS expansion observed 30 minutes following injection. Rabbit models have attributed injection volume and speed of needle withdrawal can affect material reflux.^9,10 In the current study, prospective injection protocols were followed to minimize reflux; 100 μL were injected and clinicians held the syringe in place briefly after completing the delivery of CLS-TA into the SCS. Recording and reviewing future injections may help determine if substantial CLS-TA reflux occurs and if this correlates with subsequent changes in the SCS anatomy.

The current sub-study within the prospective HULK trial found that suprachoroidal CLS-TA injections induced a measurable expansion of the SCS, which returned to baseline within 1 month, with study eyes indistinguishable from fellow eyes at that time with no lasting impact on SCS anatomy.