Short-term Assessment of Subfoveal Injection of Adeno-Associated Virus-Mediated hCHM Gene Augmentation in Choroideremia Using Adaptive Optics Ophthalmoscopy

This case series uses adaptive optics scanning light ophthalmoscopy to assess the short-term integrity of the cone mosaic following subretinal injections of adeno-associated virus-mediated hCHM gene augmentation in patients with choroideremia.

Key Points

Question

To what extent is cone photoreceptor mosaic integrity and density maintained following subretinal injection of adeno-associated virus vector designed to deliver a functional version of the CHM gene (AAV2-hCHM) augmentation for choroideremia?

Findings

In this case series of 9 patients, cone mosaic integrity and cone photoreceptor density were maintained following subretinal injection of AAV2-hCHM for choroideremia, although mild shortening of the photoreceptor outer segments was detected.

Meaning

These findings provide safety information at the cellular level for both the subretinal injection surgical technique and the AAV2-hCHM study agent.

Abstract

Importance

Subretinal injection for gene augmentation in retinal degenerations forcefully detaches the neural retina from the retinal pigment epithelium, potentially damaging photoreceptors and/or retinal pigment epithelium cells.

Objective

To use adaptive optics scanning light ophthalmoscopy (AOSLO) to assess the short-term integrity of the cone mosaic following subretinal injections of adeno-associated virus vector designed to deliver a functional version of the CHM gene (AAV2-hCHM) in patients with choroideremia.

Design, Setting, and Participants

This longitudinal case series study enrolled adult patients with choroideremia from February 2015 to January 2016 in the US. To be included in the study, study participants must have received uniocular subfoveal injections of low-dose (5 × 10¹⁰ vector genome per eye) or high-dose (1 × 10¹¹ vector genome per eye) AAV2-hCHM. Analysis began February 2015.

Main Outcomes and Measures

The macular regions of both eyes were imaged before and 1 month after injection using a custom-built multimodal AOSLO. Postinjection cone inner segment mosaics were compared with preinjection mosaics at multiple regions of interest. Colocalized spectral-domain optical coherence tomography and dark-adapted cone sensitivity was also acquired at each time point.

Results

Nine study participants ranged in age from 26 to 50 years at the time of enrollment, and all were White men. Postinjection AOSLO images showed preservation of the cone mosaic in all 9 AAV2-hCHM–injected eyes. Mosaics appeared intact and contiguous 1 month postinjection, with the exception of foveal disruption in 1 patient. Optical coherence tomography showed foveal cone outer segment shortening postinjection. Cone-mediated sensitivities were unchanged in 8 of 9 injected and 9 of 9 uninjected eyes. One participant showed acute loss of foveal optical coherence tomography cone outer segment–related signals along with cone sensitivity loss that colocalized with disruption of the mosaic on AOSLO.

Conclusions and Relevance

Integrity of the cone mosaic is maintained following subretinal delivery of AAV2-hCHM, providing strong evidence in support of the safety of the injections. Minor foveal thinning observed following surgery corresponds with short-term cone outer segment shortening rather than cone cell loss. Foveal cone loss in 1 participant raises the possibility of individual vulnerability to the subretinal injection.

Introduction

The advent of gene therapy for inherited retinal degenerations has revolutionized the field of ophthalmic care.¹ The US Food and Drug Administration’s recent approval of Luxturna (Spark Therapeutics) has provided the first clinically available treatment option for RPE65-associated retinal degeneration² and given hope to numerous other patients who experience genetic blinding disease.³ Indeed, numerous clinical trials testing gene augmentation to treat other inherited retinal degenerations are being conceived or are in progress.^4,5,6

Choroideremia is one such inherited retinal degeneration where gene augmentation is being tested in multi-institutional gene therapy clinical trials.^{7,8,9,10,11,12,13,14,15,16,17,18} Choroideremia is an X-linked degeneration caused by variants in the CHM gene, which encodes Rab escort protein 1 (REP1), a protein thought to be involved in membrane trafficking.^19,20 Variants in CHM lead to progressive degeneration of the photoreceptors, retinal pigment epithelium (RPE), and choroid.^21,22,23,24 Patients with choroideremia typically present in their youth with nyctalopia and visual field defects. The earliest clinically detectable abnormalities include RPE demelanization, disruption of the photoreceptor outer segments, and severe rod photoreceptor dysfunction starting in the near midperipheral retina.^25,26,27 Cone dysfunction as well as centrifugal and centripetal movement of the degenerative process^21,27 causes progressive constriction of the visual field and eventual involvement of the foveal center, which results in degraded visual acuity typically in the fifth decade of life.^{21,25,27,28,29}

Retinal imaging in choroideremia has demonstrated retained central islands of neural retina, with sharp borders demarcating the transition to severely degenerated areas.^29,30 Cross-sectional imaging with optical coherence tomography (OCT) has revealed that, independent of the cellular origin of the primary mechanism of disease, shortening or loss of the photoreceptor outer segments are the earliest clinically detectable abnormalities, preceding overt structural loss within the RPE and choroid.^{21,24,27,29,31}

At the cellular level, imaging with adaptive optics scanning laser ophthalmoscopy (AOSLO) has enabled visualization of the photoreceptor mosaic both in health and disease.³² This technique involves measuring and compensating for the optical aberrations of the living eye to obtain diffraction-limited imaging through the pupil.³³ In choroideremia, AOSLO imaging has revealed the photoreceptor mosaic remains contiguous up to the edge of sharp borders between relatively preserved and atrophic retina.^29,34,35 Within the retained central islands, local regions of the photoreceptor mosaic can exhibit either normal or reduced cone densities with dim and mottled wave-guided reflectance profiles. AOSLO imaging combined with microperimetry has confirmed the existence of sharp transitions between functioning retina and severe sensitivity losses that collocate with the rapid transitions in retinal structure.³⁰ The results suggest the RPE, in addition to rods, is an autonomous site of degeneration in choroideremia. Cones ultimately are lost as well, either by mechanisms that occur in parallel or as a consequence of severe RPE and/or choroidal changes.^{21,27,29,30,34,36,37,38,39}

Current gene augmentation strategies for managing choroideremia use a subretinal injection of adeno-associated virus vector to deliver normal copies of the CHM gene (AAV2-hCHM) to the residual islands of viable retina hoping to restore REP1 function and thereby maintain or improve levels of central vision.^7,17,21 The procedure is not without risk in that it involves forcefully delivering a fluid under the neural retina, thus separating the photoreceptors within the residual island from the underlying supportive RPE. This process of intentionally detaching the photoreceptors from the RPE has the potential to damage the structural integrity of the entire retina, particularly the RPE and photoreceptors. In the present study, we use AOSLO to gain insight into the short-term changes of the cone mosaic following macular subretinal injections of adeno-associated virus-mediated (AAV2) hCHM in choroideremia. We evaluate the cone mosaic structure in conjunction with foveal measures of cone outer segment length and vision prior to and 1 month after the subfoveal injections.

Methods

Study Participants and General Procedures

This research adhered to the Declaration of Helsinki⁴⁰ and was approved by the institutional review boards at the University of Pennsylvania and the Children’s Hospital of Philadelphia. All participants provided written and verbal informed consent before voluntarily enrolling (February 2015 to January 2016) in the study and were compensated $10 per hour for participating. The study participants also provided informed consent and were enrolled in a dose escalation phase 1/2 clinical trial testing the safety of the subretinal delivery of AAV2-hCHM in choroideremia (NCT02341807).

Study participants underwent a complete ophthalmic examination before and 1 month after the subretinal delivery of AAV2-hCHM, including dark-adapted foveal sensitivity testing, OCT imaging, and AOSLO imaging. Cone sensitivity was measured at the fovea using a modified Humphrey Field Analyzer (HFA II-I; Carl Zeiss Meditec) using a 1.75° diameter, 650-nm stimuli presented at the fovea following 30 minutes of dark adaptation.²¹ Axial lengths were recorded using an IOL-Master (Carl Zeiss Meditec). AOSLO images were proportionally scaled by axial length. Within 87 days from the baseline imaging session (mean [SD], 27 [25] days; range, 4-87 days), study participants received unilateral subretinal injection of low-dose (up to 5 × 10¹⁰ vector genome per eye) or high-dose (up to 1 × 10¹¹ vector genome per eye) AAV2-hCHM per the phase 1/2 clinical trial protocol. As previously reported, the injection blebs covered the entire extent of the residual central islands including the foveal center.¹⁶ For quantitative analyses of the OCT cross-sections, images from postoperative visits were coregistered to baseline and resampled at 10-fold the original resolution. Longitudinal reflectivity profiles from the foveal center (juxtafovea in jpatient 07 to avoid retinal tracks and ellipsoid zone [EZ] discontinuation) were generated using ImageJ.^21,41 Longitudinal reflectivity profiles aligned by the main RPE/Bruch membrane (BrM) signal peak were used to determine the distance between the EZ signal peak and the peak at the base of the RPE/BrM. Finally, changes in dark-adapted cone sensitivity and foveal EZ-to-RPE/BrM distance were compared with cone mosaic morphology as determined by AOSLO imaging.

AOSLO Imaging

The custom-built, multimodal AOSLO system used in this study has been previously described.^42,43 Three photomultiplier tubes (Hamamatsu Corporation) were configured to record confocal and nonconfocal split-detection near-infrared reflectance image sequences at 18 Hz simultaneously. AOSLO image sequences were acquired over the central 3° surrounding fixation and along each meridian until reaching the atrophic border or reaching approximately 15° eccentricity.

AOSLO image sequences from each imaging session were desinusioded, registered, and semiautomatically montaged as previously described.^44,45,46 Montages from each time point were then manually aligned using Adobe Photoshop. Macroscopic image features, such as blood vessels and the contours of the central island, were used for an initial alignment; the longitudinal alignment was then refined for cone-by-cone accuracy over regions of interest (ROIs). When necessary, 1-month images were scaled to baseline images.

Cone Density Measurements

Four ROIs from each eye were selected for measurement of cone densities. ROIs were manually cropped from the baseline and 1 month postinjection split detection AOSLO montages for both injected and control eyes of all study participants. One grader (J.I.W.M.) manually identified cones in all ROIs using custom software.⁴⁷ The grader was masked to injected vs control eye and time point for each study participant and was able to adjust the brightness and contrast of the image while selecting cones. Cone centers were used to determine Voronoi boundaries and bound cone density was calculated for each ROI.⁴⁸ Using the generalized estimating equations, cone densities, cone densities change from baseline, and percent change in cone densities from baseline were compared between uninjected and injected eyes, while accounting for intereye correlation and repeated measure correlation for cone densities at 4 locations.⁴⁹ Similar analysis were performed for comparing cone densities between baseline and 1-month time points in uninjected and injected eyes. Analysis began February 2015. Two sided P < .05 was considered to be statistically significant.

Results

Nine patients with molecularly confirmed choroideremia participated in the study. Study participant characteristics are shown in the eTable in the Supplement. All patients self-reported as White men and ranged in age from 26 to 50 years at the time of enrollment. As previously reported, surgeries were uneventful.¹⁶ Foveal cone sensitivity was unchanged at 1 month postinjection for 8 of 9 injected eyes and all 9 control eyes. One participant, patient 11, showed a loss in foveal cone sensitivity in the injected eye (eTable in the Supplement). As previously reported,¹⁶ this same participant demonstrated a loss in visual acuity (3 lines) in the injected eye while all other visual acuities remained unchanged.

Shortening of Cone Outer Segments After Subretinal Gene Therapy

Overall, at 1 month postinjection, OCTs show that compared with baseline images the laminar architecture of the retina is qualitatively unchanged in both the injected and uninjected eyes and that the subretinal bleb containing AAV2-hCHM resolved in injected eyes (eFigure 1 in the Supplement). Quantitative analyses showed a normal foveal EZ-to-RPE/BrM distance (mean [2SD] normal, 52 [15] μm) at baseline in all study participants except patients 05 and 07 with the most severe foveal abnormalities. At the 1 month postinjection time point, this distance was shorter in the AAV2-hCHM–injected eyes, suggestive of foveal cone outer segment shortening (eFigure 1C in the Supplement). The differences between the measures at 1 month and baseline exceeded the variability of the measurements (±4.42 μm) in 4 study participants, was borderline relative to the variability in another 4 participants and unchanged in 1 participant. The EZ-to-RPE/BrM distance in uninjected eyes remained unchanged at 1 month compared with baseline.

Photoreceptor Mosaic Integrity and Cone Density

Nonconfocal split detection AOSLO at baseline revealed the photoreceptor mosaic within the central island of remaining retina was intact and contiguous out to the border of atrophy, at which point, the photoreceptor mosaic exhibited a sharp transition to atrophic retinal regions. AOSLO at 1 month postinjection also revealed a contiguous mosaic in 8 of 9 injected eyes and all 9 uninjected eyes (Figure 1 and eFigures 2-18 in the Supplement). Global features within the montages could be aligned longitudinally. However, local distortions in adjacent images both within and between time points precluded cone-by-cone alignment across the full montage in both injected and uninjected eyes. Thus, ROIs within the montages were selected for cone-by-cone alignment across time points using rigid transforms (translation, rotation, scale) only (eFigures 19-27 in the Supplement). Cone-by-cone alignment was attained at multiple retinal locations within the montage in all eyes. Qualitatively, this manual alignment was easier to perform in uninjected eyes and ROIs in uninjected eyes showed accurate cellular alignments over a larger distance than ROIs in injected eyes.

Figure 1. Nonconfocal Split Detection Adaptive Optics Scanning Laser Ophthalmoscopy Montage of the Photoreceptor Inner Segment Mosaic at Baseline and 1 Month Postinjection in the Injected Eye of Patient 06.

The same retinal features are observed longitudinally in both montages and the photoreceptor mosaic remains intact following the subretinal injection of adeno-associated virus-mediated hCHM. Asterisks denote the foveal location in the mosaic. Arrowheads highlight example features of the cone mosaic that are observed at both time points. Yellow boxes show an enlarged region of the cone mosaic within the montage; this location was one of the regions of interest used for cone density measurement. Blue and magenta boxes denote the locations of the regions of interest used for cone density measurements shown in Figure 2.

Cones could be manually identified and bound cone density determined for longitudinally aligned ROIs in all eyes (Figure 2). Cone densities were similar across time points for all ROIs in both injected and uninjected eyes (eFigure 28 in the Supplement). In injected eyes, cone density (mean [SE]) was 24 027 (1991) cones/mm² at 1 month postinjection compared with 24 401 (2361) cones/mm² at baseline (mean [SE] cones/mm², 24 027 [1991] vs 24 401 [2361]; P = .43). Mean (SD) cone densities in uninjected eyes were 24 284 (3051) cones/mm² at 1 month compared with 24 491 (3022) cones/mm² at baseline (mean [SE] cones/mm², 24 284 [3051] vs 24 491 [3022]; P = .41). Summarizing across all ROIs, there was no statistical difference observed between cone densities measured in injected and uninjected eyes at baseline (mean [SE] cones/mm², 24 401 [2361] vs 24 491 [3022]; P = .97; Table) or at 1 month (mean [SE] cones/mm², 24 027 [1991] vs 24 284 [3051]; P = .91; Table). There was no difference between injected and uninjected eyes in cone density change from baseline (mean [SE] cones/mm², 374 [472] vs 207 [249]; P = .80; Table) nor in percent change from baseline in cone density at 1 month (mean [SE] %, −0.56 [2.56] vs 1.2 [1.05]; P = .60; Table).

Figure 2. Adaptive Optics Regions of Interest Aligned Between Time Points From the Injected Eye of Patient 06.

An intact cone mosaic is visible before and after the subretinal injection of adeno-associated virus-mediated hCHM. Cones were manually identified (blue dots) and bound cone density was calculated for each region of interest. The adaptive optics montages in Figure 1 show the retinal locations of the 2 regions of interest, based on the corresponding blue and magenta colored squares.

Table. Comparison of Cone Density (Cones/mm²) Between Injected and Uninjected Eyes .

Variable	Eyes, mean (SE)a		Difference (95% CI)b	P valueb
	Uninjected	Injected
Baseline	24 491 (3022)	24 401 (2361)	90 (−4505 to 4685)	.97
1 mo Postinjection	24 284 (3051)	24 027 (1991)	257 (−4119 to 4632)	.91
Difference (baseline – 1 mo postinjection)	207 (249)	374 (472)	−167 (−1447 to 1113)	.80
% Difference	1.20 (1.05)	−0.56 (2.56)	1.76 (−4.79 to 8.32)	.60

^a Eighteen eyes were analyzed.

^b From the generalized estimating equations⁴⁹ that account for the correlations from repeated measures at 4 locations and intereye correlation.

One-month postinjection AOSLO images of participant patient 09 revealed a local loss of cones at the fovea (Figure 3). This area of cone loss was colocated both with the retinal region that revealed a loss of foveal sensitivity and with a region showing disruption of the EZ band on OCT at the 1-month postinjection time point (eFigure 1 in the Supplement). Despite the foveal cone loss, cones were visible in the surrounding parafoveal regions at the same time point. Similar to the other participants, the photoreceptor mosaic was intact at regions outside of the fovea for patient 09, longitudinal alignments of the cone mosaic were possible, and cone densities were unchanged in ROIs selected for cone density quantifications.

Figure 3. Adaptive Optics Nonconfocal Split Detection Montage in the Injected Eye of Patient 09 at 1 mo Postinjection.

The black box shows an area of loss of cone inner segments surrounded by intact cone inner segments at the fovea; yellow boxes show the intact cone inner segment mosaic in parafoveal regions.

Discussion

Managing choroideremia by gene augmentation represents a substantial departure from the treatment scenario of the earlier experience in RPE65-inherited retinal degenerations that culminated with the first US Food and Drug Administration–approved gene therapy product for use in the clinic.^3,18 In mid- to end-stage choroideremia, there is no alternative but to treat small fragile central islands of relatively preserved retina that sustain limited fields of vision that often support reading levels that are above the legal limit of blindness. The scenario is quite different from the treatment of functionally blind retinas that are less fragile in diseases within the spectrum of RPE65–Leber congenital amaurosis.^3,18 In fact, the resulting shift in the benefit-to-risk ratio is to be expected for the larger group of non–Leber congenital amaurosis inherited retinal degenerations at the disease stages that are typically considered for initial clinical trials, making choroideremia a model for an entire group of genetic retinal diseases that await treatment solutions. Although the safety profile of the subretinal injections is now better understood, quantitative approaches, particularly at the cellular level, are needed to understand the mechanisms that lead to unwanted outcomes, particularly when the remaining vision is threatened by an invasive procedure, such as a therapeutic retinal detachment. In the current study, we used OCT and AOSLO imaging to document possible short-term changes of the photoreceptor mosaic following subfoveal injections of AAV2-hCHM. We demonstrate that at 1 month postinjection, the cone mosaic resettled on the RPE following resolution of the subretinal bleb. The cone mosaic remained intact in 8 of 9 study participants (except patient 09) and outside of the foveal center in all participants. Quantification of cone densities revealed no measurable difference between the injected and uninjected eyes and no measurable changes in cone densities between baseline and 1 month postinjection. These results show there was no widespread cone loss across the retained area of central retina targeted by the retinal detachment. Thus, we make 2 important conclusions regarding the safety of the AAV2-hCHM experimental therapy. First, cone photoreceptors did not drop out as a consequence of mechanical or acute inflammatory changes in response to the presence of AAV2-hCHM in the subretinal space. Second, the therapeutic retinal detachment, as performed by us, did not result in detectable short-term changes of the density of the photoreceptor mosaic, although mild shortening of the photoreceptor outer segments was detected. Altogether, our results provide safety information at the cellular level for both the surgical technique and the AAV2-hCHM study agent and confirm cellular-level safety signals that up to now were only available through histopathologic studies in normal nonhuman primates.^{50,51,52,53,54}

However, the subretinal injection is not without risk. Foveal thinning and the occurrence of full thickness loss of retinal tissue (macular holes) are known complications of the procedure.³ The loss of the photoreceptors at the fovea in 1 study participant (patient 09) raises the possibility of individual vulnerability to the subfoveal injection, an issue reported in at least 1 participant in each of the CHM gene therapy clinical trial reports.¹⁸ Patient 09’s surgery was considered uneventful. However, it is unclear why the photoreceptors at the fovea of patient 09 did not withstand the subfoveal injection because the parafoveal cones did survive the intervention (Figure 3); we suggest that the cone loss may have resulted from mechanical factors of the surgery rather than toxicity to the study agent. All study participants were at a similar stage of their central retinal disease with remodeled foveas that were within normal limits of thickness (except patient 05) or even thicker than normal.^21,27 In fact, 2 patients (patients 07 and 08) showed proximity of the transitional zones of structural disorganization to the foveal center, a factor known to predict the decline in visual acuity as part of the natural history of the disease,²¹ yet they did not have an unfavorable outcome.¹⁶ Inspection of patient 09’s OCT cross-sections reveals faint EZ and interdigitation zone signal that may be indicative of fragile or more abnormal photoreceptors, as has been described in certain forms of cone photoreceptor inherited degenerations.^55,56 Perhaps these may be structural signs that dictate modified surgical approaches.^57,58 Further studies are needed to address predisposing factors in patients with similar outcomes.

Our OCT results at 1 month postinjection showed a decrease in outer segment length in comparison with baseline. Preliminary results reported from an ongoing randomized clinical trial (NCT02341807) showed that the foveal thickness slowly recovers over 6 months postinjection.¹⁶ This, taken together with the AOSLO imaging results, leads to the conclusion that the measured decrease in thickness is caused by short-term outer segment shortening as opposed to cone loss. We hypothesize the forced detachment between the outer segment tips and the RPE from the surgical injection rather than the vector itself causes this short-term shortening of the outer segments. For unknown reasons, the recovery of the outer segment length after retinal detachments follows a time course that differs from the normal renewal rate of the outer segment and seems to be independent of the cause of the retinal detachment, whether short-lived and intentional, such as during the delivery of gene therapy products by subretinal injections, or after spontaneous primary macula-off detachments.^59,60,61,62 Beyond purely mechanical factors affecting the outer retina, complex interactions are known to occur after retinal detachments, such as the response of the inner retina and the underlying RPE to the therapeutic detachment, topics in need of investigation.^63,64,65,66

Although there have been reports documenting the safety of subretinal injections targeting the macula in choroideremia and other inherited retinal degenerations, this work is, to our knowledge, the first study to apply AOSLO to investigate gene therapy intervention for a blinding disease. Gene therapy aims to prevent cellular death and/or restore function to cells that are still surviving and AOSLO enables noninvasive visualization of individual cells. As a practical matter, the design and economics of clinical trials puts a high value on accurately measuring outcomes reasonably soon after the experimental interventions. The Spark Therapeutics–funded clinical trial for choroideremia did not include AOSLO imaging as an outcome measure, but the results from this ancillary study suggest that AOSLO might be suitable as a precise anatomic outcome measure in future trials involving subretinal injections. Further, techniques such as optoretinography^{67,68,69,70,71} and AOSLO microperimetry^30,72,73,74 have complemented AOSLO imaging by allowing the direct or indirect evaluation of photoreceptor function at the cellular level. The emergence of these tools may prove impactful for assessing the short- and long-term safety and efficacy of gene therapies for blinding diseases.

Limitations

Longitudinal cone density measurements from AOSLO images are limited by manual image processing techniques including cell-by-cell alignment and cone identification over small regions of interest. Techniques such as automated longitudinal montaging⁴⁴ and automatic quantification of cone density over the full montage⁷⁵ will be beneficial, especially for studies that enroll large numbers of participants.

Conclusions

In conclusion, our data support the short-term safety of subretinal injections of AAV2-hCHM. Additional follow-up will be required to assess the long-term safety and efficacy of subretinal injection and delivery of AAV2-hCHM for preventing or restoring vision loss caused by choroideremia.

Supplement.

eTable. Study Participant Characteristics

eFigure 1. Cone outer segment shortening after subretinal gene therapy

eFigure 2. AOSLO montages, PN-01 Injected Eye

eFigure 3. AOSLO montages, PN-01 Uninjected Eye

eFigure 4. AOSLO montages, PN-03 Injected Eye

eFigure 5. AOSLO montages, PN-03 Uninjected Eye

eFigure 6. AOSLO montages, PN-04 Injected Eye

eFigure 7. AOSLO montages, PN-04 Uninjected Eye

eFigure 8. AOSLO montages, PN-05 Injected Eye

eFigure 9. AOSLO montages, PN-05 Uninjected Eye

eFigure 10. AOSLO montages, PN-06 Uninjected Eye

eFigure 11. AOSLO montages, PN-07 Injected Eye

eFigure 12. AOSLO montages, PN-07 Uninjected Eye

eFigure 13. AOSLO montages, PN-08 Injected Eye

eFigure 14. AOSLO montages, PN-08 Uninjected Eye

eFigure 15. AOSLO montages, PN-09 Injected Eye

eFigure 16. AOSLO montages, PN-09 Uninjected Eye

eFigure 17. AOSLO montages, PN-11 Injected Eye

eFigure 18. AOSLO montages, PN-11 Uninjected Eye

eFigure 19. Cone mosaic ROIs, PN-01

eFigure 20. Cone mosaic ROIs, PN-03

eFigure 21. Cone mosaic ROIs, PN-04

eFigure 22. Cone mosaic ROIs, PN-05

eFigure 23. Cone mosaic ROIs, PN-06

eFigure 24. Cone mosaic ROIs, PN-07

eFigure 25. Cone mosaic ROIs, PN-08

eFigure 26. Cone mosaic ROIs, PN-09

eFigure 27. Cone mosaic ROIs, PN-11

eFigure 28. Cone density at one-month post-injection versus baseline

eReferences