A Systematic Review and Meta-Analyses of Interventional Clinical Trial Studies for Gene Therapies for the Inherited Retinal Degenerations (IRDs)

IRDs are one of the leading causes of visual loss in children and young adults. Mutations in over 271 genes lead to retinal dysfunction, degeneration and sight loss. Though no cure exists, gene augmentation therapy has brought hope to the field. This systematic review sought to assess the efficacy of available gene therapy treatments for IRDs. Databases and public resources were searched for randomised controlled trials (RCTs) and non-randomised studies of interventions (NRSIs). Standard methodological procedures were used, including a risk-of-bias assessment. One RCT and five NRSIs were assessed, all for adeno-associated virus two (AAV2)-mediated treatment of RPE-specific 65 kDa (RPE65)-associated LCA (Leber congenital amaurosis). Five outcomes were reported for meta-analyses. Modest improvements in visual acuity, ambulatory navigation/mobility testing or central retinal thickness was observed. There was significant improvement in red and blue light full-field stimulus testing (FST) (red light risk ratio of 1.89, treated v control, p = 0.04; and blue light risk ratio of 2.01, treated v control, p = 0.001). Study design assessment using a ROBIN-I tool (Cochrane Library) showed risk-of-bias judgement to be “low/moderate”, whilst there were “some concerns” for the RCT using a RoB-2 tool (Cochrane Library). Although comparison by meta-analysis is compromised by, amongst other issues, a variable amount of vector delivered in each trial, FST improvements demonstrate a proof-of-principle for treating IRDs with gene therapy.

There is considerable phenotypic variability between IRDs and historically they have been grouped into several different disease patterns, including retinitis pigmentosa (RP), cone dystrophies, cone-rod dystrophies and Leber congenital amaurosis (LCA) [15]. Further, many attempts have been made at genotype-phenotype correlation. However, in reality, there is both considerable inter-allelic disease overlap and marked intra-allelic disease variability. Thus, the progress made in genetic diagnosis of IRDs has been invaluable. removed due to duplication, leaving one hundred and eight articles to be screened. Eighty-seven records were excluded that did not include relevant information. Twentyone articles were accessed for eligibility; fifteen articles were excluded: one was not applicable for meta-analysis (a study on gene therapy for choroideremia), five were follow-up studies and nine articles included duplicate data. This left six final articles (Appendix A.3) for review and meta-analysis:

Outcomes
In total, 23 different assays were reported and analysed across the six studies in Figure  2 and Table 1 (including Figure S1a-c, Appendix A.4 and Table A1). Safety data was not collected on specific AAV2 vectors, having been examined in other independent studies on interventional clinical studies in the retina [37,[59][60][61][62]. Only one assay, visual acuity (VA), was common to all six papers. Of the 23 assays reviewed, only five outcomes were reported for meta-analysis-VA (logMAR), mobility, red light full-field stimulus (FST) testing (log10(cd.s/m 2 )), blue light full-field stimulus (FST) testing (log10(cd.s/m 2 )) and central retinal thickness (CRT).

Outcomes
In total, 23 different assays were reported and analysed across the six studies in Figure 2 and Table 1 (including Figure S1a-c, Appendix A.4 and Table A1). Safety data was not collected on specific AAV2 vectors, having been examined in other independent studies on interventional clinical studies in the retina [37,[59][60][61][62]. Only one assay, visual acuity (VA), was common to all six papers. Of the 23 assays reviewed, only five outcomes were reported for meta-analysis-VA (logMAR), mobility, red light full-field stimulus (FST) testing (log10(cd.s/m 2 )), blue light full-field stimulus (FST) testing (log10(cd.s/m 2 )) and central retinal thickness (CRT). Biomolecules 2021, 11, x 6 of 64 All continuous and dichotomous data was reported. If either continuous and dichotomous data were available, then analysis was used to compare and contrast the models. However, continuous data was preferred from a statistical perspective as some information risked being lost in categorical data. All continuous and dichotomous data was reported. If either continuous and dichotomous data were available, then analysis was used to compare and contrast the models. However, continuous data was preferred from a statistical perspective as some information risked being lost in categorical data. Table 1. Study selection of six (6) research articles for review and meta-analysis outcomes for LCA2-RPE65 gene therapy.  (a) vg-vector genomes; (b) ITT-intention to treat; (c) BCVA-Best corrected visual acuity, (logMAR); (d) RR-risk ratio; (e) 95% CI-95% confidence interval; (f) FST-full-field stimulus testing (red and blue wavelength), log10(cd.s/m 2 ).

Visual Acuity Measured by logMAR
A 0.30 logMAR (3 line) mean post-operative change of VA was accepted as being a "clinically meaningful" improvement [43]. VA results were reported in Figure 3. Overall, outcomes showed a benefit of treatment compared to control eyes, but did not meet statistical significance.

Visual Acuity Measured by logMAR
A 0.30 logMAR (3 line) mean post-operative change of VA was accepted as being a "clinically meaningful" improvement [43]. VA results were reported in Figure 3. Overall, outcomes showed a benefit of treatment compared to control eyes, but did not meet statistical significance. Finally, an analysis of dichotomous data on visual outcomes post treatment (better or worse) was performed. Five studies provided individual patient data to allow this analysis (n = 58 treated eyes; n = 47 untreated eyes). The line of no effect showed an RR of 1.13 (95% CI 0. 83, 1.53), indicating an improvement with treatment that did not reached clinical significance (p = 0.44) ( Figure S2 (Supplemental)).

Mobility
Given the disparity between the four different mobility methods used in the studies in terms of size, light intensity, scoring and reporting, no direct comparison was possible. Instead, a meta-analysis of dichotomous data (better/worse post-treatment) was performed, To do so, four sub-groups were defined, according to light intensity used to illuminate the mobility mazes: (a): mobility under a single light of intensity of 4 lux; (b): mobility under a "low" ambient light level (0.2, 0.6, 1, 2 or 4 lux), broadly scotopic light; (c): mobility under a "high" ambient light level (10,15,50,100    Finally, an analysis of dichotomous data on visual outcomes post treatment (better or worse) was performed. Five studies provided individual patient data to allow this analysis (n = 58 treated eyes; n = 47 untreated eyes). The line of no effect showed an RR of 1.13 (95% CI 0.83, 1.53), indicating an improvement with treatment that did not reached clinical significance (p = 0.44) ( Figure S2 (Supplemental)).

Mobility
Given the disparity between the four different mobility methods used in the studies in terms of size, light intensity, scoring and reporting, no direct comparison was possible. Instead, a meta-analysis of dichotomous data (better/worse post-treatment) was performed, To do so, four sub-groups were defined, according to light intensity used to illuminate the mobility mazes: (a): mobility under a single light of intensity of 4 lux; (b): mobility under a "low" ambient light level (0.2, 0.6, 1, 2 or 4 lux), broadly scotopic light; (c): mobility under a "high" ambient light level (10,15,50,100 Table S3 (Supplemental)).
Under a light intensity of 4 lux ( Figure 4, "Lux 4"), analysis of 4 studies showed an RR of 1.03 (95% CI 0.75, 1.42), indicating an improvement with treatment that did not reach clinical significance (p = 0.84). Under low ambient light ("Low Lux 0.2 to 4"), analysis of 4 studies showed an RR of 1.35 (95% CI 0.78, 2.35), indicating an improvement with treatment that did not reach clinical significance (p = 0.29). Under high ambient light ("High Lux 10 to 100"), analysis of 4 studies showed an RR of 0.42 (95% CI 0.12, 1.50), indicating a worsening with treatment that did not reach clinical significance (p = 0.18). Analysis of all ambient light levels ("All lux levels 0.2 to 100") of 4 studies showed an RR of 1.15 (95% CI 0.84, 1.58), indicating an improvement with treatment that did not reach clinical significance (p = 0.39). Under a light intensity of 4 lux ( Figure 4, "Lux 4"), analysis of 4 studies showed an RR of 1.03 (95% CI 0.75, 1.42), indicating an improvement with treatment that did not reach clinical significance (p = 0.84). Under low ambient light ("Low Lux 0.2 to 4"), analysis of 4 studies showed an RR of 1.35 (95% CI 0.78, 2.35), indicating an improvement with treatment that did not reach clinical significance (p = 0.29). Under high ambient light ("High Lux 10 to 100"), analysis of 4 studies showed an RR of 0.42 (95% CI 0.12, 1.50), indicating a worsening with treatment that did not reach clinical significance (p = 0.18). Analysis of all ambient light levels ("All lux levels 0.2 to 100") of 4 studies showed an RR of 1.15 (95% CI 0.84, 1.58), indicating an improvement with treatment that did not reach clinical significance (p = 0.39).

Full-Field Stimulus (FST) Testing for Red and Blue Wavelength
Only two studies used FST testing that allowed for a meta-analysis. Both continuous and dichotomous (better/worse) data was analysed (Figure 5a-d). . Meta-analysis shows no significant improvement in ambulatory navigation/mobility following treatment. RevMan meta-analysis of dichotomous data showed no significant improvement in performance across all light intensities analysed.

Full-Field Stimulus (FST) Testing for Red and Blue Wavelength
Only two studies used FST testing that allowed for a meta-analysis. Both continuous and dichotomous (better/worse) data was analysed (Figure 5a-

Central Retinal Thickness (CRT)
Three studies reported CRT outcomes as measured by optical coherence tomography (OCT), but only two included quantitative data that allowed for meta-analysis. Analysis of dichotomous data (thinner/thicker) at 1 year post treatment, showed a RR of 1.15 (95% CI 0.45, 3.00), indicating an increase of CRT with treatment that did not reach clinical significance (p = 0.77) (Figure 6a). Analysis of dichotomous data for a long term timepoint (3 Figure 5. Meta-analysis shows significant improvement in full field sensitivity in response to red and blue light (log10(cd.s/m 2 ) following treatment [(a-d)]. RevMan analysis of dichotomous FST data shows significant improvement with red (RR1.89; p = 0.04) and blue (RR 2.01; p = 0.001) stimuli. Continuous data shows improvement with blue (mean difference 1.69, p < 0.00001) but not red (mean difference 0.89, p = 0.07) light.

Central Retinal Thickness (CRT)
Three studies reported CRT outcomes as measured by optical coherence tomography (OCT), but only two included quantitative data that allowed for meta-analysis. Analysis of dichotomous data (thinner/thicker) at 1 year post treatment, showed a RR of 1.15 (95% CI 0.45, 3.00), indicating an increase of CRT with treatment that did not reach clinical significance (p = 0.77) (Figure 6a). Analysis of dichotomous data for a long term timepoint (3 years), showed a RR of 1.29 (95% CI 0.33, 5.10), indicating an increase of CRT with treatment that did not reach clinical significance (p = 0.72) (Figure 6b).
Biomolecules 2021, 11, x 12 of 64 years), showed a RR of 1.29 (95% CI 0.33, 5.10), indicating an increase of CRT with treatment that did not reach clinical significance (p = 0.72) (Figure 6b). RevMan meta-analysis of dichotomous data showed no significant improvement in CRT measurement.

Risk of Bias Tools within Studies
Cochrane risk-of-bias tools were used to assess study reliability; ROBIN-I methods [52], for non-randomised study designs, and RoB-2 methods [63], for randomised clinical trials. Overall, a risk-of-bias judgement was reported "low/moderate", with a predicted direction of bias "towards null/unpredictable" for the 5 NRSIs, and a report with "some concerns" and a predicted direction of bias with "favours experimental" for the RCT (Table 2; Appendix A.5, Table A2. and Appendix A.6, Table A3).  Table 3 provided a summary of 12 meta-analyses reported for each of the outcomes, a PRISMA summary of a structured abstract in Appendix A.7, and a PRISMA checklist in Appendix A.8 (Table A4).

Risk of Bias Tools within Studies
Cochrane risk-of-bias tools were used to assess study reliability; ROBIN-I methods [52], for non-randomised study designs, and RoB-2 methods [63], for randomised clinical trials. Overall, a risk-of-bias judgement was reported "low/moderate", with a predicted direction of bias "towards null/unpredictable" for the 5 NRSIs, and a report with "some concerns" and a predicted direction of bias with "favours experimental" for the RCT ( Table 2; Appendix A.5, Table A2. and Appendix A.6, Table A3).  Table 3 provided a summary of 12 meta-analyses reported for each of the outcomes, a PRISMA summary of a structured abstract in Appendix A.7, and a PRISMA checklist in Appendix A.8 (Table A4).

Discussion
Inherited retinal dystrophies (IRDs) are a leading cause of visual loss in children and adults of working age. Formerly untreatable, the emergence of gene augmentation therapy represents a real paradigm shift in patient care. We thus performed a systematic review and meta-analysis of interventional clinical trials to assess the efficacy of gene therapies for IRDs, thus delivering useful information for both clinicians and patients. The purpose of this systematic review is also based on a "fair test" [64], grounded in evidence-based medicine [65,66] (and Figure S1). To test such new therapies, it is critical to assess how transparent results show clear benefit for the patient. This requires that methodology, study design and outcome measures have to provide a clear and reasonable conclusion for the impact on the patient. A systematic review and meta-analysis of IRD patient outcomes for gene therapy is critical in order to support the field [67].
A search of peer-reviewed literature found that only gene therapies to treat Leber congenital amaurosis (LCA) met the criteria for addressing the original question (Appendix A.1). LCA is a rare disorder and gene therapy is an expensive treatment, which led to studies with small patient numbers. Further, the particularly severe phenotype of the disease, with low visual acuity from birth, led to difficulties in assessing the effect of treatment.
Of the 6 studies analysed, a significant drawback to the meta-analysis performed here is the variability in vector design and concentration of virus injected sub-retinally. All studies analysed used an AAV2 serotype, with most using an AAV2/2 capsid. However, one study used an AAV2/4 capsid. Further, some studies used a hybrid chicken β-actin promoter with a cytomegalovirus enhancer, whilst some used the human RPE65 promoter. Treatment doses ranged from 10 8 to 10 12 vg, in volumes from 0.15 to 1.0 mL. This spans a number of logarithmic steps in each dose, potentially compromising the comparison of the results within the 6 studies. Despite all this, in our view, the similarities in products compared in the meta-analyses outweigh the differences. All contain the same recombinant human RPE65 gene, all are packaged in a similar AAV2 vector and all use a similar sub-retinal surgical procedure for delivery. All were used to treat the same trial population (RPE65-LCA2 patients). Further, there were similar criteria for controls and there was considerable overlap in trial duration and endpoints. Finally, we felt the comparison appropriate as pre-clinical work has shown good photoreceptor transduction and expression efficiency. As such, despite the analyses' obvious limitations, we felt it appropriate in order to increase numbers of this rare disease and thus improve statistical power. Given the differences outlined, it is extremely encouraging to note the significant improvements in full-field stimulus (FST) testing that are seen following meta-analysis.
With gene augmentation in its infancy, it is perhaps unsurprising that there were variabilities in the biomarkers used to determine treatment efficacy. In total, 23 outcome assays were used. Visual acuity was the only outcome used in all six studies analysed. Five studies used Goldmann perimetry, four used ambulatory navigation/mobility and three used electroretinography. A further nineteen assays were used in two or less studies. Many of the assays were not comparable for several reasons. Five studies assessed visual fields using Goldmann perimetry, although different studies presented different isopters with variable follow up time. Further, a lack of quantitative data in some studies meant overall meta-analysis was not possible. Three studies used electroretinography as an assay, but two provided no data. Other assays used in two or more studies were unsuitable for meta-analysis due to a lack of quantitative data or irreconcilable differences in the way data was presented. As the field evolves, it is hoped agreed standards for methods and reporting will be established, allowing for easier meta-analyses of trials.
Visual acuity is the gold standard assay by which retinal disease treatments are assessed. Though our meta-analysis showed only modest improvement with gene therapy (in terms of clinical or statistical significance), the result is perhaps not surprising given the low-vision phenotype of LCA patients. Two studies (Bainbridge and Testa) did not provide raw visual acuity data and instead patient vision was determined from results presented in study graphs. This was undertaken by two independent researchers, with a mean of the two readings being used, but an element of uncertainty remains with the overall result due the unavailability of raw data within the actual papers. It should be noted that an I 2 value 65% indicates substantial statistical heterogeneity within the VA assays ( Figure 3). As such, little weight can be placed upon the outcome of our VA analyses.
The study designs often dictated that the eye with the worse vision was treated, with control eyes having a better baseline vision. Although logMAR vision charts determine a linear improvement in vision with each letter or line gained, if treatment and control arms have different baseline values, bias is introduced and outcomes may be influenced as a result. Without adjustment, it may be unclear what impact arises from the treatment effect, as opposed to the treated eye being worse at baseline. Even adjusted data may not be robust enough to eliminate this confounding factor. Emerging gene therapy trials, where both eyes are treated and compared for one year to a deferred treatment group, should address this issue.
Four studies reported mobility testing as a key outcome. Mobility testing for the MLMT assay (Russell et al.) received criticism by an independent commentator [68] and reviewers in the FDA regarding uneven luminance levels [29]. In addition, we note that the MLMT assay results were indirect. A "passing level" of the assay compared baselines between 1-year timepoints however, the original data for measuring speed, time, accuracy (and further components) for assessment, were not included in the paper or the Biologicals License Application (BLA). Further, the MLMT assay used a logarithmic scale, based on light intensity (lux), which was then subsequently converted to an ordinal scale (ranging from −1 to 6), such that a two-point change in the ordinal scale may have a different interpretation depending on the baseline score (Table S4 (Supplemental)) [68].
Due to disparate methodologies (maze size and design, measurement, quantification and reporting), only analysis of dichotomous data (better/worse post treatment) was possible. This risks overestimating the benefit in certain studies. For example, Russell et al. reasoned that results in their maze required at least 2 levels of improvement on their assessment scale to accept the result as showing therapeutic benefit, whereas we defined even a 1 level gain post-operatively as performing 'better'.
Some studies had datapoints missing, while quantitative data was missing from others, and required interpretation from results presented in study graphs (Jacobson et al.). Though RevMan analysis of dichotomous data suggested overall improvement in mobility, statistical significance was not reached. At present, there is no better test available for assessing the impact of gene therapy on visual function and so, as the field develops, it would be advantageous if some standardisation of the test could be agreed upon, recently supported by other literature [69][70][71].
The use of full-field stimulus (FST) testing (white, red and blue wavelength) is highly relevant because few research tools can quantify changes in visual perception if sight loss is as severe as it is in an RPE65/LCA2 population. Thus, the FST data carries extra significance. FST results presented in the studies was at times confusing. One study (Russell et al.) alternately presented white light FST results in log10(cd.s/m 2 ) units and −log10(cd.s/m 2 ) units, whilst not commenting on their red and blue light FST results (Russell et al.). A further study only described results in terms of "log10" units, which we interpreted as log10(cd.s/m 2 ) units (Jacobson et al.), thus allowing for meta-analysis. Although the FDA, as part of the Biologics Licence Application (BLA) review for Luxturna [29], stated 'the direct clinical benefit of FST is not clear', it is apparent from meta-analysis of these two studies that retinal sensitivity improves with AAV-mediated gene augmentation therapy for RPE65-mediated LCA2. The significance of this cannot be underestimated. It is proof of principle that visual improvement is achievable with this technology and gives us hope that similar benefit could be achieved when other alleles are targeted.
The first attempt to use gene augmentation therapy for retinal disease has led to an FDA and EMA approved product (voretigene neparvovec-rzyl [Luxturna]). Improvements in surgical technique and improved knowledge of treatment technicalities (e.g., virus concentration) could mean subsequent iterations of these therapies show improved efficacy. Further, the more novel outcomes for mobility may drive innovative end-points. Though some concerns were raised by our Cochrane risk of bias analysis, further treatments targeting more common disease-causing genes [37] will mean increased patient numbers in trials and may allow for blinded evaluations, resulting in more robust studies.
As of April 2021, there are > 40 interventional gene therapy trials for IRDs reported at clinicaltrials.gov, from both academic and commercial institutions targeting several different IRD genes [72][73][74][75][76]. This meta-analysis highlights the need for consistency of trial design to allow comparison of gene products, but also shows the potential this technology has for addressing a leading cause of blindness in children and adults of working age.

Conclusions
The objective of this work was to conduct a systematic review of interventional clinical trial studies for IRDs and to assess and compare the effectiveness of available gene therapy treatments. Following the search, review and analysis of the relevant studies, the systematic review concluded that a meta-analysis for AAV-RPE65 gene therapy for LCA2 reported a modest improvement for visual acuity, mobility and full-field stimulus testing (FST), with FST improvements reaching statistical significance. In terms of a recommendation to support the IRD patient communities and researchers, we propose that full and openaccess data is key. If the field is to be progressed and improved, then objective and transparent results need to be shared in order to improve outcomes, analysis, reporting and interpretation.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/biom11050760/s1, Figure S1: (a) A simple hierarchy of evidence; (b) A list of 23 assays from the 6 selected studies; (c) Each of the 23 assays re-grouped and colour coded to clearly distinguish how the specific assays were to be used in the study, Figure S2: Visual acuity logMAR, with a random effects model and summary statistic for dichotomous data showed, Table S1: PICOS results. The PICOS search terms, keywords, MeSH terms, search strings and Boolean operators were used and identified in Materials & Method (using Ovid Database), additionally de-fined Appendices A.1-A.3, Table S2: All mean difference (MD) values for all visual acuity logMAR changes across all six (6) papers. All data was retrieved and analysed by two independent authors, Table S3: All ambulatory navigation/mobility across all six (6) papers. All data was retrieved and analysed by two independent authors, Table S4

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Conflicts of Interest:
The authors declare no conflict of interest.

Appendix A Appendix A.1. Ovid Search Results in MEDLINE and EMBASE
The overall aim of this study was to identify a systematic review and meta-analyses of interventional clinical trial studies for gene therapies for IRDs. The aims of the study were to identify, extract, analyse and critique outcomes for gene therapy treatments from a relevant and specific population, in particular: (i) identify, search and collate the available research data from IRD patients for gene therapy treatment; (ii) extract and assess the relevant data from an IRD population and prepare a metaanalysis of the available outcomes and potential impact of the research reporting; (iii) analyse and critique relevant outcomes from key IRD studies. This systematic review and meta-analyses were performed for interventional clinical trial outcomes for approved gene therapies for IRDs. The systematic review used a structured search approach with a PICOS process (Population, Intervention, Comparison, Outcomes and Study). The specific research question was to search interventional clinical trials (the study design) for published gene therapies (the intervention) for IRD patients (the population), for the purpose of improving the disorder (the outcomes), given there was little or no treatment available (the comparison).
The    (1)  Willingness to adhere to protocol and long-term follow-up as evidenced by written informed consent or parental permission and subject assent (where applicable).

1
Male and female subjects of any ethnic group are eligible for participation in this study, providing they meet the following criteria:

5
Age eight years old or older at the time of administration. 5 Visual acuity not better than 20/60 and not worse than hand motion in both the treated eye and the fellow eye;

6
Ability to comply with research procedures; 6 Subjects must be evaluable on mobility testing (the primary efficacy endpoint) to be eligible for the study. Evaluable is defined as: 1) The ability to perform mobility testing within the luminance range evaluated in the study. Individuals must receive an accuracy score of ≤ 1 during screening mobility testing at 400 lux or less to be eligible; individuals with an accuracy score of > 1 on all screening mobility test runs at 400 lux, or those who refuse to perform mobility testing at screening, will be excluded.
2) The inability to pass mobility testing at 1 lux. Individuals must fail screening mobility testing at 1 lux to be eligible; individuals that pass one or more screening mobility test runs at 1 lux will be excluded.

6
Visual acuity ≤ 20/160 or visual field less than 20 degrees in the eye to be injected.

6
Visible photoreceptor (outer nuclear) layer on a standard optical coherence tomography (OCT) scan;

8
For females of childbearing potential, a negative pregnancy test at screening and at baseline, and agreement to use effective contraception for 12 months after administration of rAAV2-CB-hRPE65, for sexual activity that could lead to pregnancy;

9
For males of reproductive potential, agreement to use effective contraception for 12 months after administration of rAAV2-CB-hRPE65, for sexual activity that could lead to pregnancy Humoral immune deficiency as evidenced by low tetanus toxoid IgG antibody titers;

2
Patients with, within the past 6 months, a clinically significant cardiac disease or known congestive heart failure, cardiac rhytm and conduction abnormalities

2
Any prior participation in a study in which a gene therapy vector was administered.

2
Unable or unwilling to meet requirements of the study.

2
Presence of epiretinal membrane on OCT; 3

Diabetes mellitus 3
Pre-existing eye conditions that would preclude the planned surgery or interfere with the interpretation of study endpoints or surgical complications;

3
Patients with pulmonaty dysfunction 3 Participation in a clinical study with an investigational drug in the past six months.

3
History of immunodeficiency or other medical conditions that might increase the risk of rAAV2-CB-hRPE65 administration; Use of retinoid compounds or precursors that could potentially interact with the biochemical activity of the RPE65 enzyme; individuals who discontinue use of these compounds for 18 months may become eligible.

4
Participation in a clinical study with an investigational drug in the past six months.

4
Use of anticoagulants or anti-platelet agents within 7 days prior to study agent administration;

Renal impairment 5
Use of anti-platelet agents that may alter coagulation within 7 days prior to study agent administration;

5
Prior intraocular surgery within six months. 5 Pre-existing eye conditions that would preclude the planned surgery or interfere with the interpretation of study endpoints (for example, glaucoma, corneal or lenticular opacities).

5
History of allergy or sensitivity to medications planned for use in the peri-operative period; 6 Immunocompromise 6 Use of immunosuppressive medications; 6 Known sensitivity to medications planned for use in the peri-operative period.

6
Lack of sufficient viable retinal cells as determined by non-invasive means, such as optical coherence tomography (OCT) and/or ophthalmoscopy. Specifically, if indirect ophthalmoscopy reveals less than 1 disc area of retina which is not involved by complete retinal degeneration (indicated by geographic atrophy, thinning with tapetal sheen, or confluent intraretinal pigment migration), these eyes will be excluded. In addition, in eyes where optical coherence tomography (OCT) scans of sufficient quality can be obtained, areas of retina with thickness measurements less than 100 um, or absence of neural retina, will not be targeted for delivery of AAV2-hRPE65v2.

6
For females of childbearing potential, a positive pregnancy test at screening or baseline (within 2 days before rAAV2-CB-hRPE65 administration); Complicating systemic diseases would include those in which the disease itself, or the treatment for the disease, can alter ocular function. Examples are malignancies whose treatment could affect central nervous system function (for example: radiation treatment of the orbit; leukemia with CNS/optic nerve involvement). Subjects with diabetes or sickle cell disease would be excluded if they had any manifestation of advanced retinopathy (e.g., macular edema or proliferative changes). Also excluded would be subjects with immunodeficiency (acquired or congenital) as there could be susceptibility to opportunistic infection (such as CMV retinitis).

7
Complicating systemic diseases or clinically significant abnormal baseline laboratory values. Complicating systemic diseases would include those in which the disease itself, or the treatment for the disease, can alter ocular function. Examples are malignancies whose treatment could affect central nervous system function (for example, radiation treatment of the orbit; leukemia with CNS/optic nerve involvement). Also excluded would be subjects with immuno-compromising diseases, as there could be susceptibility to opportunistic infection (such as CMV retinitis). Subjects with diabetes or sickle cell disease would be excluded if they had any manifestation of advanced retinopathy (e.g., macular edema or proliferative changes). Subjects with juvenile rheumatoid arthritis could be excluded due to increased infection risk after surgery due to poor wound healing. Subjects who are positive for hepatitis B, C, and HIV will be excluded.

7
Females who are breast feeding; Individuals of childbearing potential who are pregnant or unwilling to use effective contraception for four months following vector administration.

8
Prior ocular surgery within six months. 8 Use of any investigational agent, or systemic corticosteroids or other immunosuppressive drug(s), within 3 months prior to enrollment;

9
Severe affective disorder) 9 Any condition that would prevent a subject from completing follow-up examinations during the course of the study;

9
Individuals incapable of performing mobility testing (the primary efficacy endpoint) for reason other than poor vision, including physical or attentional limitations.

9
Known sensitivity to medications planned for use in the peri-operative period.

9
Prior receipt of any AAV gene therapy product; 10 Pregnancy or lactation 10 Any condition that makes the subject unsuitable for the study;

10
Any other condition that would not allow the potential subject to complete follow-up examinations during the course of the study or, in the opinion of the investigator, makes the potential subject unsuitable for the study.

10
Individuals of childbearing potential who are pregnant or unwilling to use effective contraception for the duration of the study.

10
Any condition which leads the investigator to believe that the participant cannot comply with the protocol requirements or that may place the participant at an unacceptable risk for participation.

11
Current, or recent participation, in any other research protocol involving investigational agents or therapies;

11
Subjects will not be excluded based on their gender, race, or ethnicity.

11
Any other condition that would not allow the potential subject to complete follow-up examinations during the course of the study and, in the opinion of the investigator, makes the potential subject unsuitable for the study.

12
Recent receipt of an investigational biologic therapeutic agent.

12
Subjects will be excluded if immunological studies show presence of neutralizing antibodies to AAV2 above 1:1000. MLMT was assessed using both eyes at 1 or more of 7 levels of illumination, ranging from 400 lux (a brightly lit office) to 1 lux (a moonless summer night). Each light level was assigned a score code ranging from 0 to 6. A higher score indicated that a subject was able to pass the MLMT at a lower light level. A score of −1 was assigned to those who could not pass MLMT at 400 lux. The MLMT of each subject was videotaped and assessed by independent graders. The MLMT score was determined by the lowest light level at which the subject was able to pass the MLMT. The MLMT score change was defined as the difference between the score at Baseline and the score at Year 1. A positive MLMT score change from Baseline to Year 1 visit indicated that the subject was able to complete the MLMT at a lower light level.             Study design quality and an overall risk-of-bias judgement showed "Low/Moderate" and "Towards to null/Unpredictable" with the ROBIN-I tool (NRSIs), and "Some concerns" and "Favours experimental" with the RoB-2 tool (RCT).

Conclusions
The objective of this work was to conduct a systematic review of interventional clinical trial studies for IRDs and to assess and compare the effectiveness of available gene therapy treatments. Following the search, review and analysis of the relevant studies, the systematic review concluded that a meta-analysis for AAV-RPE65 gene therapy for LCA2 reported a modest improvement for visual acuity, mobility and full-field stimulus testing (FST). However, other than FST, there was no clinically meaningful benefit and no statistical significance from the six collected studies. One RCT found a clinically meaningful benefit for an assessment for a primary endpoint for mobility, a MD of 1.6 (95% CI 0.72-2.41), p = 0.0013.
In terms of a recommendation to support the IRD patient communities and researchers, we propose that full and open-access data is key. If the field is to be progressed and improved, then objective and transparent results need to be shared in order to improve outcomes, analysis, reporting and interpretation.
Appendix A.8. PRISMA List for LCA2 Studies Identify the report as a systematic review, meta-analysis, or both. Page 1

Structured summary 2
Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implication ns of key findings; systematic review registration number.

Rationale 3
Describe the rationale for the review in the context of what is already known.

Page 1-2 (Introduction)
Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, interventions, comparisons, outcomes, and study design (PICOS). Table S1 (Supplemental).

Protocol and registration 5
Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if available, provide registration information including registration number.
Informal proposal/protocol assessed and peer-reviewed at the University of Edinburgh.

Eligibility criteria 6
Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale.

Information sources 7
Describe all information sources (e.g., databases with dates of coverage, contact with study authors to identify additional studies) in the search and date last searched.

Study selection 9
State the process for selecting studies (i.e., screening, eligibility, included in systematic review, and, if applicable, included in the meta-analysis).

Data collection process 10
Describe method of data extraction from reports (e.g., piloted forms, independently, in duplicate) and any processes for obtaining and confirming data from investigators.

Data items 11
List and define all variables for which data were sought (e.g., PICOS, funding sources) and any assumptions and simplifications made.

Risk of bias in individual studies 12
Describe methods used for assessing risk of bias of individual studies (including specification of whether this was done at the study or outcome level), and how this information is to be used in any data synthesis.

Summary measures 13
State the principal summary measures (e.g., risk ratio, difference in means).

Synthesis of results 14
Describe the methods of handling data and combining results of studies, if done, including measures of consistency (e.g., I2) for each meta-analysis.

Risk of bias across studies 15
Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication bias, selective reporting within studies).

Additional analyses 16
Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regression), if done, indicating which were pre-specified.

Study selection 17
Give numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage, ideally with a flow diagram.

Study characteristics 18
For each study, present characteristics for which data were extracted (e.g., study size, PICOS, follow-up period) and provide the citations.

Risk of bias within studies 19
Present data on risk of bias of each study and, if available, any outcome level assessment (see item 12). Appendices A.5 and A.6.

Results of individual studies 20
For all outcomes considered (benefits or harms), present, for each study: (a) simple summary data for each intervention group (b) effect estimates and confidence intervals, ideally with a forest plot.

Risk of bias across studies 22
Present results of any assessment of risk of bias across studies (see Item 15). Table 2, page 11

Summary of evidence 24
Summarize the main findings including the strength of evidence for each main outcome; consider their relevance to key groups (e.g., healthcare providers, users, and policy makers).

Limitations 25
Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., incomplete retrieval of identified research, reporting bias).

26
Provide a general interpretation of the results in the context of other evidence, and implications for future research. Table  S4a,b.

Funding 27
Describe sources of funding for the systematic review and other support (e.g., supply of data); role of funders for the systematic review. None.