Efficacy and Safety of Gene Therapy for RPGR Gene-Associated X-Linked Retinitis Pigmentosa: A Systematic Review and Meta-Analysis

Abstract

Purpose

This study aimed to conduct a systematic review and meta-analysis to synthesize clinical evidence on adeno-associated viral (AAV)-based RPGR gene therapy for X-linked retinitis pigmentosa (XLRP), guiding clinical decisions and future research.

Methods

Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and Prospective Register of Systematic Reviews (PROSPERO; CRD420251008337) registration, we searched PubMed, Embase, and other databases through March 2025, including controlled clinical trials evaluating AAV gene therapy for RPGR-related XLRP. Two reviewers independently screened literature, extracted data, and assessed bias using the Cochrane tool. Meta-analyses were performed in RevMan version 5.4, calculating risk ratios (RRs) for dichotomous outcomes and mean difference /standardized mean difference (MD / SMD) for continuous outcomes, with random/fixed-effects models based on heterogeneity (I² > 50%).

Results

Five articles from three studies (205 patients) were included. Gene therapy significantly improved low-luminance visual acuity (LLVA) ≥10 Early Treatment Diabetic Retinopathy Study (ETDRS) letters at 6 months (RR = 3.79, P = 0.03) and retinal sensitivity at 6 months (MD = 1.06, P = 0.001) and 12 months (MD = 2.47, P < 0.00001), but not best-corrected visual acuity (VA) or LLVA ≥15 letters. Ocular treatment-emergent adverse event (TEAE) risk increased significantly (RR = 5.52, P < 0.00001), with common events including anterior chamber inflammation and intraocular inflammation. A trend toward higher serious adverse events (SAEs) was observed (odds ratio [OR] = 3.36, P = 0.05).

Conclusions

Gene therapy demonstrates short-term efficacy for XLRP, with 6 months as a critical evaluation time point, but requires vigilance for ocular TEAE risks. Heterogeneity and small sample sizes highlight the need for large-scale, long-term multicenter trials to optimize protocols and advance safe clinical application.

Translational Relevance

XLRP, a severe hereditary retinal dystrophy primarily caused by RPGR gene mutations, lacks curative treatments. AAV-mediated RPGR gene therapy shows promise in early trials, but systemic evaluation of its efficacy and safety is absent. This evaluates gene therapy as a novel approach, noting short-term efficacy but ocular AE risks.

Introduction

X-linked retinitis pigmentosa (XLRP) is a rare and severe inherited retinal dystrophy, representing a notable subset of all retinitis pigmentosa cases.¹ XLRP is genetically heterogeneous, with mutations in the RPGR gene responsible for the majority of cases, whereas RP2 and other rare gene mutations also contribute to the disease.² Thus far, clinical trials of gene therapy have only been reported for RPGR-associated XLRP, and these are the studies analyzed in the current report. These genetic defects disrupt photoreceptor function and ciliary transport, leading to progressive photoreceptor degeneration and severe visual impairment, predominantly affecting male patients due to the X-linked inheritance pattern.³ The disease typically manifests in early childhood with night blindness and peripheral vision loss, eventually progressing to legal blindness by the fourth decade of life.⁴ To date, no curative treatment is available, and current management focuses on symptomatic relief and slowing disease progression.⁵

With the advancement of gene therapy, adeno-associated virus (AAV)-mediated delivery of functional RPGR gene sequences has emerged as a promising approach.⁵ Several phase 1/2 and 2/3 clinical trials—including AAV5-RPGR (botaretigene sparoparvovec),⁶ AGTC-501,⁷ and BIIB112 (cotoretigene toliparvovec)⁸—have reported encouraging outcomes in terms of retinal sensitivity and functional vision improvements.

However, the evidence remains fragmented, and no previous meta-analysis has synthesized the efficacy and safety of AAV-based gene therapies for RPGR-associated XLRP. Therefore, we conducted a systematic review and meta-analysis to evaluate the therapeutic benefits and adverse event (AE) profiles of these interventions across existing clinical trials. This work aims to provide a comprehensive and quantitative summary of current clinical evidence to inform future therapeutic strategies and clinical decision making.

Methods

We registered the review prospectively in the international Prospective Register of Systematic Reviews (PROSPERO) (CRD420251008337) and reported it in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines. A protocol was developed in accordance with the PRISMA guidelines.⁹

Search Strategy

We conducted a systematic literature search in PubMed, Embase, Cochrane Library, Clinicaltrials.gov, and Web of Science databases for studies published up to July 2025. The search terms included “X-linked retinitis pigmentosa,” “gene therapy,” and “RPGR,” combined with filters for clinical trials. Full search strategies are provided in Supplementary Material S1. No language restrictions were applied. Reference lists of relevant articles were manually screened for additional eligible studies (Supplementary Material S2).

Inclusion and Exclusion Criteria

We included clinical controlled trials that evaluated AAV-based gene therapy in patients with genetically confirmed RPGR-associated XLRP caused by mutations in the RPGR gene. Eligible studies had to involve male participants aged 5 years or older with RPGR-associated XLRP, and include those with baseline best-corrected visual acuity (BCVA) between 20/50 (≥65 Early Treatment Diabetic Retinopathy Study [ETDRS] letters) and 20/200 or worse in the study eye. The studies must be clinical controlled trials, including parallel dose-comparison groups, and report at least one of the following primary or secondary clinical outcomes: microperimetry (Macular Integrity Assessment [MAIA] or similar devices) to assess retinal sensitivity, BCVA to measure visual function, low-luminance visual acuity (LLVA) for sensitivity under reduced light conditions, and treatment-emergent adverse events (TEAEs) related to therapy, including ocular and systemic safety. We included trials with follow-up periods sufficient to assess both the safety and efficacy of the treatment, with most studies reporting data at 1 to 24 months.

We excluded non-randomized studies, such as observational studies, cohort studies, and single-arm trials, as well as studies that did not use AAV-based gene therapy vectors or used non-standardized gene therapy interventions (refers to gene therapy protocols and practices lacking unified industry technical specifications, operating procedures, and efficacy-safety evaluation criteria). Examples include gene therapies targeting rare RPGR mutations, for which unified protocols cannot be formulated due to the small patient population that hinders large-scale multicenter trials, as well as preliminary in vivo experiments of novel vectors (e.g., non-AAV viral vectors) at the laboratory stage, whose operating procedures remain unstandardized). Studies with insufficient or unclear reporting on key outcomes, such as BCVA, microperimetry, LLVA, or AEs, were also excluded unless these were addressed clearly in the Supplementary Data.

The Judgment Criteria for Evaluation Indicators

Retinal sensitivity is assessed by macular integrity assessment microperimetry. The retinal sensitivity responder criterion is defined as a ≥7 decibel (dB) improvement from baseline at ≥5 of the 16 central loci (not preselected) within the 10–2 grid.

TEAE is defined as any unfavorable medical event occurring during drug treatment, which either did not exist or was not observed prior to treatment initiation, or first appeared/worsened during the treatment period. Serious adverse event (SAE) refers to any adverse medical event meeting any of the following criteria: (1) death: death of the patient resulting from an adverse reaction to the drug. (2) Life-threatening: events directly threatening the patient’s life (not potential future risks), such as anaphylactic shock or severe arrhythmia (with potential to cause sudden death). (3) Hospitalization or prolongation of hospitalization: requirement for hospitalization due to drug-induced adverse reactions (e.g., drug-induced hepatic failure), or prolongation of an existing hospitalization due to adverse reactions. (4) Permanent or significant disability/impairment: drug-induced permanent vision loss, limb paralysis, or other irreversible or severe functional impairments. (5) Congenital anomaly or birth defect: structural malformation in the fetus due to maternal drug administration. (6) Other medically important events: including but not limited to severe allergic reactions, fulminant hepatitis, severe hemorrhage, critical infections, status epilepticus, and other conditions requiring urgent medical intervention.

Study Selection and Data Extraction

Two reviewers independently screened titles and abstracts, followed by full-text reviews. Discrepancies were resolved through discussion or consultation with a third reviewer. Data were extracted into a standardized form, including study design, participants, interventions, follow-up duration, and primary and secondary outcomes.

Risk of Bias Assessment

Risk of bias was assessed using the Cochrane Risk of Bias 2.0 tool in Review Manager (RevMan) version 5.4. Domains evaluated included randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result.

Statistical Analysis

Meta-analyses were performed using RevMan version 5.4. For dichotomous outcomes (e.g., proportion of responders), odds ratio (OR), risk ratios (RRs) with 95% confidence intervals (CIs) were calculated. For continuous outcomes (e.g., change in retinal sensitivity), mean differences (MDs) or standardized mean differences (SMDs) were used. Heterogeneity was assessed using the I² statistic. A random-effects model was applied when significant heterogeneity was detected (I² > 50%); otherwise, a fixed-effects model was used. Sensitivity analyses and subgroup analyses were conducted where appropriate.

Results

Study Selection

A total of 301 articles were identified through database searches. After computerized deduplication using EndNote 21 and manual duplicate removal, 208 records remained. Based on title and abstract screening, animal studies, cell or organoid studies, reviews, conference abstracts, and observational studies solely describing patient characteristics were excluded, leaving 6 articles. One ongoing clinical trial without available results was subsequently removed. Ultimately, five articles from three studies were included in the analysis (Supplementary Material S4).

Characteristics of Included Studies

All included studies were controlled clinical trials focusing on the efficacy and safety of gene therapy for XLRP caused by RPGR gene mutations (Supplementary Material S5). Three papers reported on studies using the AAV8-coRPGR vector: J. Cehajic-Kapetanovic et al., in 2020, reported outcomes for 18 patients treated with AAV8-coRPGR over a 6-month follow-up. The primary outcome was safety, with secondary outcomes including visual acuity, microperimetry, and central retinal thickness.¹⁰ The L. von Krusenstiern et al. 2023 trial compared eyes from part 1 of the BIIB112 (XIRIUS) study (n = 18) with untreated subgroups from the XOLARIS study (n = 103) to assess the durability of retinal sensitivity improvements associated with single subretinal injections of cotoretigene toliparvovec (BIIB112/AAV8-RPGR) over 12 months. Primary endpoints included TEAEs, changes in retinal sensitivity from baseline, the percentage of participants meeting microperimetry responder criteria, and LLVA.¹¹ Finally, the Byron L. Lam et al. 2024 trial evaluated the efficacy and tolerability of subretinal cotoretigene toliparvovec (BIIB112/AAV8-RPGR) in 32 patients. The primary endpoint was the percentage of participants meeting microperimetry responder criteria, with secondary endpoints including changes in retinal sensitivity and LLVA at 12 months, as well as the proportion of eyes achieving ≥15 and ≥10 ETDRS letter improvements in LLVA from baseline.¹² The trial by M. Michaelides et al., in 2024, enrolled 43 participants to evaluate the safety and efficacy of AAV5-hRKp.RPGR gene therapy, with a follow-up period of 12 months. The primary outcome was safety, whereas secondary efficacy outcomes included static perimetry, microperimetry, visually guided mobility, BCVA, and contrast sensitivity. Safety and efficacy outcomes were assessed at 52 weeks for immediate-treatment participants and 26 weeks for control participants.⁶ The P. Yang et al. 2025 trial included 29 patients with RPGR-associated XLRP across four US centers to evaluate the safety and efficacy of subretinal gene therapy using AGTC-501 (rAAV2tYF-GRK1-RPGR), with a 24-month follow-up. Primary endpoints included TEAEs and SAEs, whereas efficacy was evaluated via mean sensitivity using intermediate visual microperimetry.⁵

Risk of Bias in Included Studies

Figure 1 summarizes the assessment of the risk of bias in included studies.

Figure 1.

Risk of bias in included studies.

Allocation

For two studies, the trial reports showed that the randomization was performed properly, that is, an unpredictable sequence of treatment allocations was appropriately hidden from those who recruited participants into the trial. The other three studies did not use random methods to conduct research, and were directly classified according to the patient's condition and the doctor's judgment.

Masking

All five studies did not use a masked method because patients needed to undergo vitrectomy and other operations. The MeiraGTx trial, the AGTC trial, the trial of Lenore von Krusenstiern et al., and the trial of Jasmina Cehajic-Kapetanovic et al., are open-label clinical trials. The participants, investigators, assessors, and sponsors were all aware of whether the patients received treatment, the identity of the study eye, the specific dose administered, and other relevant information. In the trial of Byron L. Lam et al., assessors were masked to the participants’ treatment groups, specific doses administered, and study eyes. In contrast, participants, investigators, and sponsors were only masked to the low or high dose levels; no masking was applied to other aspects (e.g., whether participants received treatment or the identity of the study eye). However, the impact of non-masking on the outcome measures was minimal. Therefore, masking of participants and personnel is classified as unclear risk.

Incomplete Outcome Data

Only two studies were judged to be at low risk of attrition bias, that is, they reported high rates of follow-up that were reasonably equal between treatment groups (J. Cehajic-Kapetanovic et al. 2020 and L. von Krusenstiern et al. 2023). For three studies, follow-up was not reported in enough detail to make a judgement (Byron L. Lam et al. 2024, M. Michaelides et al. 2024, and P. Yang et al. 2025).

Selective Reporting

In these five studies, all can find trial registration information (M. Michaelides et al. 2024 https://clinicaltrials.gov/: NCT03252847; J. Cehajic-Kapetanovic et al. 2020 https://clinicaltrials.gov/: NCT03116113; L. von Krusenstiern et al. 2023 XIRIUS: https://clinicaltrials.gov/: NCT03116113; XOLARIS: https://clinicaltrials.gov/: NCT04926129; Byron L. Lam et al. 2024 https://clinicaltrials.gov/: NCT03116113; and P. Yang et al. 2025 study https://clinicaltrials.gov/: NCT03316560).

Other Potential Sources of Bias

No other potential sources of bias were noted.

Effects of Gene Therapy

LLVA ≥10 ETDRS Letters

A total of 51 participants were enrolled in this section of the study. At the first, third, sixth, and 12th months, the proportion of participants achieving LLVA ≥10 ETDRS letters above baseline in the experimental group was consistently higher than that in the control group. This improvement peaked at 6 months (39.3% vs. 10.0%), demonstrating the efficacy of gene therapy in enhancing LLVA for patients with RPGR-associated XLRP (see the Table 1).

Table 1.

The Proportion of LLVA ≥10 ETDRS Letters and LLVA ≥15 ETDRS Letters in the Experimental Group and the Control Group in the First, Third, Sixth, and Twelfth Months

	First Month	Third Month	Sixth Month	12th Month
LLVA ≥10 ETDRS letters (%)
Experimental	10/31 (32.3)	10/31 (32.3)	11/28 (39.3)	9/28 (32.1)
Control	2/20 (10.0)	3/19 (15.8)	2/20 (10.0)	5/20 (25.0)
LLVA ≥15 ETDRS letters (%)
Experimental	8/31 (25.8)	8/31 (25.8)	6/28 (21.4)	7/28 (25.0)
Control	1/20 (5.0)	1/19 (5.3)	1/20 (5.0)	1/20 (5.0)

ETDRS, Early Treatment Diabetic Retinopathy Study; LLVA, low-luminance visual acuity.

There did exist a statistically significant difference between the experimental group and the control group (RR = 3.79, 95% CI = 1.18–12.16, P = 0.03) only at 6 months (the experimental group demonstrated superiority). In contrast, other time points, including the first month (RR = 3.09, 95% CI = 0.94–10.14, P = 0.06), the third month (RR = 1.63, 95% CI = 0.64–4.13, P = 0.31), and the 12th month (RR = 1.44, 95% CI = 0.61–3.40, P = 0.41), did not exhibit significant differences (Fig. 2). This finding indicates that the effect of gene therapy interventions is time-dependent, demonstrating clinical value at 6 months. Consequently, it provides a critical time point for the short-term evaluation of RPGR-associated XLRP gene therapy in clinical practice. Meanwhile, when clinicians communicate the efficacy to patients, they should emphasize the significance of periodic evaluation. This approach can offer explicit efficacy feedback to patients, enhance treatment confidence, and prevent misinterpretation of early-stage or long-term effects.

Figure 2.

Patients with LLVA ≥ 10 ETDRS letters at the first, third, sixth, and 12th months of follow-up. (a) Patients with LLVA ≥ 10 ETDRS letters at the first month of follow-up. (b) Patients with LLVA ≥ 10 ETDRS letters at the third month of follow-up. (c) Patients with LLVA ≥ 10 ETDRS letters at the sixth month of follow-up. (d) Patients with LLVA ≥ 10 ETDRS letters at the 12th month of follow-up.

LLVA ≥15 ETDRS Letters

A total of 51 participants were enrolled in this section of the study. In the control group, the proportion of participants achieving LLVA ≥15 ETDRS letters above what the eye had at baseline remained ≤5.3% (1/20–1/19) at all time points (the first, third, sixth, and 12th months). In contrast, the experimental group significantly outperformed the control group across all assessments, indicating sustained and stable effects of gene therapy on LLVA improvement (≥15 letters above baseline; see the Table 1). Although gene therapy shows a tendency toward improvement in LLVA of patients, no statistically significant difference has been observed.

At the first month (RR = 3.74, 95% CI = 0.83–16.78, P = 0.09) and the third month (RR = 3.31, 95% CI = 0.68–16.19, P = 0.14), there was no significant statistical difference in RPGR-associated XLRP gene therapy intervention regarding LLVA ≥15 ETDRS letters. This indicates that the short-term effect of gene therapy was unclear, and, in clinical practice, this stage cannot be relied on to evaluate the improvement effect of LLVA ≥15 letters. Further follow-up observation is required to determine the long-term trend. At the sixth month (RR = 3.27, 95% CI = 0.71–15.07, P = 0.13), although the observation time was extended, no significant difference was found. It may be necessary to further investigate the factors affecting the improvement of LLVA (such as individual genetic differences, treatment dosage, etc.) or expand the sample size to improve the detection efficiency. The long-term observation lasting 12 months (RR = 3.87, 95% CI = 0.88–17.01, P = 0.07) showed that gene therapy did not show a clear advantage in LLVA ≥15 letters, and the application of gene therapy in clinical practice may need to be carefully selected. Meanwhile, attempts can be made to enhance the therapeutic effect by optimizing the gene therapy vector, adjusting the treatment timing, or combining with other therapies (Fig. 3).

Figure 3.

Patients with LLVA ≥ 15 ETDRS letters at the first, third, sixth, and 12th months of follow-up. (a) Patients with LLVA ≥ 15 ETDRS letters at the first month of follow-up. (b) Patients with LLVA ≥ 15 ETDRS letters at the third month of follow-up. (c) Patients with LLVA ≥ 15 ETDRS letters at the sixth month of follow-up. (d) Patients with LLVA ≥ 15 ETDRS letters at the 12th month of follow-up.

BCVA ETDRS

A total of 67 participants were enrolled in this section of the study. At the sixth month, there was no significant difference in the improvement of BCVA ETDRS by gene therapy overall (MD = 1.69, 95% CI = −0.40 to 3.77, P = 0.11). However, the study by M. Michaelides et al. (MD = 3.64, 95% CI = 0.88–6.40) suggested that gene therapy might have a positive effect, whereas the study by J. Cehajic-Kapetanovic et al. (MD = −0.90, 95% CI = −4.08 to 2.28) did not support this. High heterogeneity (I² = 78%) weakened the stability of the overall conclusion (Fig. 4). These evidences are insufficient to confirm that gene therapy has a definite effect on BCVA ETDRS improvement at 6 months. In the clinical practice of treating RPGR-associated XLRP, the promotion of gene therapy needs to be considered cautiously. Currently, more studies with rigorous design, sufficient sample size, and high homogeneity are required to verify the true impact of gene therapy on BCVA ETDRS in the treatment of RPGR-associated XLRP, providing a more reliable basis for clinical decision making.

Figure 4.

Changes of BCVA ETDRS at the sixth month of follow-up.

Changes of Retinal Sensitivity

A total of 96 participants were enrolled in this section of the study at the sixth month and 58 participants at the 12th month. At the sixth month (MD = 1.06, 95% CI = 0.41–1.72, P = 0.001), there was a statistically significant difference in micro retinal sensitivity changes between the experimental group and the control group, with the experimental group receiving gene therapy demonstrating a better effect. This indicates that gene therapy for RPGR-associated XLRP has a definite improvement effect on micro retinal sensitivity in the short term, serving as a positive signal for early-efficacy evaluation and providing a basis for the preliminary judgment of gene therapy effectiveness in clinical practice (Fig. 5).

Figure 5.

Changes in retinal sensitivity at the sixth and 12th months of follow-up compared to baseline. (a) Changes in retinal sensitivity at the sixth month of follow-up compared to baseline. (b) Changes in retinal sensitivity at the 12th month of follow-up compared to baseline.

At the 12th month (MD = 2.47, 95% CI = 1.55–3.40, P < 0.00001), it indicates that there is a significant difference in the changes of retinal sensitivity between the experimental group and the control group. The gene therapy still demonstrates a remarkable effect in enhancing retinal sensitivity, which, to a certain extent, suggests that the gene therapy may have relatively durable efficacy. This is conducive to promoting the long-term application of gene therapy in clinical practice. However, more long-term studies are still required for further confirmation (see Fig. 5).

Retinal Function Assessment-Responders

A total of 90 participants were enrolled in this section of the study at the sixth month and 70 participants at the 12th month. At the sixth month (OR = 2.12, 95% CI = 0.50–9.04, P = 0.31) and the 12th month (RR = 2.33, 95% CI = 0.91–5.92, P = 0.08), there was no statistically significant difference between the experimental group (gene therapy) and the control group (without gene therapy) in terms of retinal function assessment-responders (Fig. 6), however, the retinal function assessment indicates a tendency toward elevation. These results suggest that gene therapy for RPGR-associated XLRP had limited effects on this indicator at these two time points but it may still be effective in improving retinal function. Future research should conduct studies with larger samples and longer follow-up periods to deeply explore the potential effects of gene therapy on retinal function assessment responders and to clarify its clinical value.

Figure 6.

Changes in responders to retinal function assessment at the sixth and 12th months of follow-up compared to baseline. (a) Changes in responders to retinal function assessment at the sixth month of follow-up compared to baseline. (b) Changes in responders to retinal function assessment at the 12th month of follow-up compared to baseline.

Safety of Gene Therapy

Overview of Adverse Events Occurrence

A total of 96 participants were enrolled in this section of the study.

Figure 7a presents the statistics for patients with at least one TEAE (RR = 1.56, 95% CI = 0.98–2.50, P = 0.06), indicating that there was no significant statistical difference in the probability of TEAE occurrence when gene therapy was used or not for treating RPGR-associated XLRP. However, the result had a high heterogeneity (I² = 76%), indicating significant differences among different studies and low reliability of the result (see Fig. 7a).

Figure 7.

Summary of the occurrence of treatment-emergent adverse events (TEAEs). (a) The occurrence of at least once TEAE. (b) The occurrence of ocular TEAEs. (c) The occurrence of at least once SAE.

Figure 7b further summarizes the situation of patients with ocular TEAE (RR = 5.52, 95% CI = 2.86–10.63, P < 0.00001). The results show that the experimental group was significantly higher than the control group, indicating that gene therapy is associated with an increase in ocular AEs. This result has direct guiding significance for clinical practice, suggesting that when applying gene therapy to treat RPGR-associated XLRP, close monitoring of ocular conditions (such as regular fundus examinations, assessment of ocular inflammatory responses, etc.) is required, and timely intervention in ocular AEs should be carried out to ensure the ocular safety of patients. At the same time, it also prompts researchers to optimize the administration method to reduce potential damage to ocular tissues and enhance the safety of the therapy (see Fig. 7b).

Figure 7c presents the statistics for patients with at least one SAE (including ocular and non-ocular ones; OR = 3.36, 95% CI = 1.01–11.15, P = 0.05). Although the risk of SAE after using gene therapy to treat RPGR-associated XLRP is not statistically significant, it still indicates a trend of increased SAE risk in the experimental group, and SAE should still be highly vigilant in clinical applications (see Fig. 7c).

The ultimate goal of gene therapy for RPGR-associated XLRP is to maximize patient safety while pursuing efficacy and promoting its rational and cautious application.

Summary of the Occurrence of Adverse Events in Each Study

The Table 1 summarizes the occurrence of AEs after gene therapy for RPGR-associated XLRP in the five studies: M. Michaelides et al. 2024, P. Yang et al. 2025, J. Cehajic-Kapetanovic et al. 2020, L. von Krusenstiern et al. 2023, and Byron L. Lam et al. 2024 (Supplementary Material S6).

The Occurrence of Common AEs

Given that the study by P. Yang et al. 2025 solely compared AEs between central and peripheral retinal injections, the research by L. von Krusenstiern et al. 2023 only provided AE data for the experimental group, and the investigation by Byron L. Lam et al. 2024 did not report specific AE outcomes, the statistical analysis of common AE incidence was restricted to the studies by M. Michaelides et al. 2024 and J. Cehajic-Kapetanovic et al. 2020. A total of 67 participants were enrolled in this section of the study.

Fifty-two percent of patients in the experimental group experienced treatment-related TEAEs, compared with 0% in the control group. Sixty-four percent of patients in the experimental group developed surgery-related TEAEs, whereas no occurrences were observed in the control group. Eye disorder events were reported in 94% of the experimental group, which was significantly higher than the 16.1% incidence in the control group. Inflammation-related TEAEs (e.g., anterior chamber cells and intraocular inflammation) were documented in 50% of the experimental group, with no occurrences in the control group (Table 2).

Table 2.

The Occurrence of Common TEAEs

	Experimental	Control
Related to treatment, n (%)	26/50 (52.0%)	0/31 (0.0%)
Related to surgery, n (%)	32/50 (64.0%)	0/31 (0.0%)
Eye disordersb, n (%)	47/50 (94.0%)	5/31 (16.1%)
Conjunctival hemorrhage	26/50 (52.0%)	0/31 (0.0%)
VA reduced	18/50 (36.0%)	0/31 (0.0%)
Anterior chamber cell	16/32 (50.0%)	0/13 (0.0%)
Eye inflammation	8/32 (25.0%)	0/13 (0.0%)
Inflammation–related TEAEs, n (%)	16/32 (50.0%)	0/13 (0.0%)
Anterior chamber cell	16/32 (50.0%)	0/13 (0.0%)
Eye inflammation	8/32 (25.0%)	0/13 (0.0%)
Cystoid macular edema	6/50 (12.0%)	1/31 (3.2%)
Uveitis	9/50 (18.0%)	0/31 (0.0%)
Iridocyclitis	2/32 (6.2%)	0/13 (0.0%)
Chorioretinitis	8/50 (16.0%)	0/31 (0.0%)
Vitritis	1/32 (3.1%)	0/13 (0.0%)

TEAE, Treatment-emergent adverse event; VA, Visual Acuity

Discussion

Summary of Main Findings

This systematic review and meta-analysis included five articles of three clinical studies, comprehensively evaluating the efficacy and safety of AAV-mediated RPGR gene therapy in patients with RPGR-associated XLRP (Supplementary Material S3). In terms of efficacy, gene therapy demonstrated significant improvements in LLVA ≥10 ETDRS letters at the sixth month (RR = 3.79, P = 0.03) and retinal sensitivity at the sixth month (MD = 1.06, P = 0.001) and retinal sensitivity at the 12th month (MD = 2.47, P < 0.00001), suggesting the presence of a short-term therapeutic window. However, no statistically significant differences were observed in LLVA ≥15 letters or BCVA metrics, and the advantage in LLVA ≥10 ETDRS letters disappeared at 12th month, indicating instability in long-term efficacy. Regarding safety, gene therapy significantly increased the risk of ocular TEAE (RR = 5.52, P < 0.00001). The primary AEs included anterior chamber inflammatory cells (50%) and intraocular inflammation (25%). A borderline significant trend toward increased SAE risk was observed (OR = 3.36, P = 0.05). Additionally, risks of surgical and drug-related complications such as retinal detachment were substantially elevated.

In Jasmina Cehajic-Kapetanovic et al.’s trial, patient C4.1 was highlighted due to their typical clinical course of “improvement - inflammation – recovery,” distinct from other patients with modest efficacy or compromised benefits from inflammation. This patient achieved the most significant improvement in retinal sensitivity from 0.5 dB to 6.6 dB among all participants. Moreover, manual optical coherence tomography (OCT) segmentation confirmed anatomic changes including outer nuclear layer thickening and the emergence of new linear structures, suggesting potential regeneration of photoreceptor outer segments. These results not only intuitively validated the biological effects of the treatment but also provided critical clinical evidence for optimizing corticosteroid anti-inflammatory regimens. In Paul Yang et al.’s trial, 3 patients from group 6 were separately presented via figures and tables because they belonged to the highest-dose group and exhibited typical retinal pigment epithelium (RPE) changes. After receiving gene therapy at a dose of 1.99 × 10¹² vg/eye, all 3 patients showed improved retinal sensitivity but developed progressive RPE depigmentation. Notably, one patient had more severe lesions due to multiple retinotomies during subretinal injection. These findings underscored the unfavorable risk-benefit ratio of high-dose gene therapy and offered key evidence for determining the maximum tolerated dose. However, neither trial detailed differences in age, genotype, or other characteristics between the highlighted patients and other participants.

The M. Michaelides et al. trial reports improvement in visual acuity, whereas the J. Cehajic-Kapetanovic et al. trail does not. The core reasons for the differences in visual improvement between the two studies are possibly that the study by Michaelides et al. included participants aged ≥5 years including children with relatively well-preserved retinal structures. It used the AAV5 vector which contained a truncated RPGR gene and an optimized corticosteroid regimen to reduce inflammation. Additionally, through multi-dimensional sensitive endpoints (e.g., vision-guided mobility) and long-term 52-week follow-up, significant improvements were observed. In contrast, the study by Cehajic et al. only enrolled adults aged ≥18 years, among whom some had more severe disease. It adopted the AAV8 vector which expressed the full-length RPGR gene with only oral corticosteroids for prophylaxis, leading to obvious inflammation in the high-dose group. Furthermore, the follow-up duration was only 6 months with a single endpoint, thus no clear visual improvement was demonstrated.

Challenges and Opportunities of RPGR-Associated XLRP Gene Therapy

Current gene therapy approaches using AAV vectors for RPGR-associated XLRP still face substantial challenges. First, immunogenicity remains a critical issue. Studies have found that 30% to 60% of the population harbors pre-existing anti-AAV antibodies, which may neutralize vectors and reduce therapeutic efficacy.^13,14 Potential solutions include developing immune-evasive AAV variants or combining therapy with immunosuppressive agents.^15,16 Second, gene delivery efficiency and persistence present hurdles. Gene delivery efficiency and persistence are mutually interdependent. For instance, although the XIRIUS study missed its primary endpoint, post hoc analysis revealed early improvements in some patients. However, the compromised persistence caused by inflammation and improper vector distribution prevented these improvements from yielding statistically significant benefits. In the AGTC-501 trial, the high-dose cohort exhibited sustained improvements, yet the dose had to be discarded due to delivery-related risks of RPE damage. Precision and stability of gene delivery efficiency, as well as long-term maintenance of therapeutic effects, remain the core hurdles to be addressed urgently in gene therapy for XLRP. Given the large size of the RPGR gene fragment, dual-vector systems or minigene designs are required.¹⁷ However, the episomal nature of AAV genomes may lead to long-term expression attenuation. Integration-capable vectors (e.g., lentiviruses) could address this, although safety concerns persist.^18,19 Third, manufacturing processes and costs pose significant barriers. AAV vectors constructed for RPGR-associated XLRP therapy exhibit empty capsid rates as high as 90%, necessitating complex and costly purification protocols.^20,21 Emerging production systems, such as stable cell lines or serum-free cultures, have shown promise in improving yields.^22,23 Fourth, clinical translation challenges arise from RPGR mutation heterogeneity in patients with RPGR-associated XLRP, necessitating personalized treatment strategies.²⁴ Additionally, standardized retinal functional metrics (e.g., electroretinogram [ERG] and OCT) are critical to support regulatory approval.^25,26

Simultaneously, these challenges afford novel opportunities for advancing RPGR-associated XLRP gene therapy. First, vector engineering optimization could be realized through viral capsid engineering—via directed evolution or rational design—to generate next-generation AAV serotypes with augmented retinal tropism and penetration capabilities.^27,28 Additionally, promoter optimization using tissue-specific elements (e.g., photoreceptor-specific promoters) may enhance spatiotemporal precision of transgene expression while minimizing off-target effects.^24,29 Second, combinatorial therapeutic strategies and technological innovation represent another frontier. Exosome- or lipid nanoparticle-assisted delivery platforms may enhance vector trafficking across the blood-retinal barrier, improving therapeutic efficacy.^26,30 Third, expanding clinical indications and industrial scalability are further priorities. Building upon existing regulatory approvals for AAV-based retinal therapies (e.g., Luxturna for Leber congenital amaurosis), RPGR-associated XLRP represents a compelling candidate for next-generation gene therapy.^29,31 Process optimization—such as tetracycline-regulated production systems—holds promise for reducing manufacturing costs and accelerating translational adoption.^21,32

Limitations

All three trials used a different capsid with AAV8 in the Biogen trial, AAV5 in the MeiraGTx trial, and a triple mutant AAV2 capsid in the AGTC trial. This difference between the vectors is the key difference that needs to be highlighted. The AGTC and Biogen trials used the correct sequence of RPGR which was achieved by codon optimization of the transgene. Potential difficulties arise in vector production due to the presence of repetitive sequences in the RPGR transgene. For instance, these repetitive sequences are prone to replication slippage during bacterial plasmid amplification and AAV viral packaging, which leads to sequence deletions, insertions or rearrangements, generates truncated or dysfunctional transgene fragments, and thus drastically compromises the purity and efficacy of the resulting vectors.³³ Meanwhile, the repetitive sequences elevate the risk of plasmid recombination, which not only increases the technical difficulty of vector construction, but also necessitates additional stability screening steps, thereby prolonging the production cycle and raising the research costs.³⁴ Studies have demonstrated that synonymous codon optimization can attenuate the instability of the ORF15 repeat sequence. This had the additional benefit of removing the splice site within the RPGR-ORF15 splice isoform, thereby reducing the mutation rate during vector production.³⁵ In the MeiraGTx trial reported by Michaelides et al., a fundamentally different vector was used because it had a truncated RPGR with a large deletion in the ORF-15 region. This is an important functional region of the protein.³⁶ The MeiraGTx construct also maintained the wild-type exon 15 / intron 15 splice donor site. This made it highly likely to undergo aberrant splicing, as indeed was seen by the additional RPGR protein band shown in the Western blot in figure 4B of Sladen et al. (2024).³⁷ So reduced efficiency of the deleted RPGR is likely the cause of the failure of the MeiraGTx phase 3 clinical trial which showed no significant difference in navigational tests over controls.

All studies used a rhodopsin kinase promoter in the transgene. Excluding the significant difference in the RPGR coding sequence above, the vectors and doses were otherwise similar. It should, however, be noted that the results in the Biogen trial were better with the lower dose, whereas in the AGTC trial the higher dose fared better. The Biogen trial failed to meet the primary endpoint but the recruitment into this trial was severely curtailed by the coronavirus disease 2019 (COVID-2019) and the numbers recruited into each of the study arms were below the number required to show statistical significance using the microperimetry analysis approved by the US Food and Drug Administration (FDA).¹² However, the Biogen trial did show statistically significant improvements in mean microperimetry and LLVA and almost certainly would have met approval had it been allowed to recruit to the full potential of 15 to 20 patients per study arm.

It can be seen from this that the COVID-19 pandemic exerted notable adverse impacts on gene therapy clinical trials in multiple aspects. A sharp drop in initiation numbers and a rise in suspension cases were observed in non-COVID-19 trials, with at least 322 cell and gene therapy trials disrupted by July 2020 and phase 2 trials accounting for 44.8%.^38,39 At the same time, raw material and capacity diversion to COVID-19 products hindered gene therapy research and development, cold chain and apheresis center operations were restricted. Therefore, the clinical trials had to shift to remote follow-up and decentralized modes, causing high protocol deviation rates and demanding adjustments to ethical and approval processes.⁴⁰

With regard to LLVA analysis, follow-ups up to a year and beyond may be adversely affected by the development of cataract. This is a known side effect of subretinal gene therapy and will impact on LLVA. In addition, the BCVA is not likely to be improved by 15 letters in most of these patients because RPGR is a centripetal retinal degeneration in which their central visual acuity still remains good until the very late stages of the disease.⁴¹ There is, therefore, a natural ceiling effect which will prevent the BCVA from making significant gains, as, for instance, would be the case after treating macular edema with an anti-VEGF. At this point, microperimetry, which explores changes over the entire macular area before the central vision is affected, is probably a more accurate measure of improved retinal function.

With regard to the microperimetry responder analysis (which is more than 7 dB in 5 points that are not preselected), the FDA requires the points to be preselected and so technically speaking this is not a relevant endpoint. In addition, it is worth mentioning that the distribution of the vector after subretinal injection will be variable and therefore the responder analysis makes the assumption that a point chosen is exposed to vector, which may not always be the case, so the mean microperimetry is a far more reliable outcome measure as the includes gains across both treated and untreated areas, regardless of the area of detachment.

In addition, the relatively small total sample size (n = 245) and limited follow-up duration (maximum 24 months in P. Yang et al. 2025) restricted the evaluation of long-term safety and therapeutic durability of gene therapy, and incomplete safety data necessitate further validation through larger-scale investigations.

Conclusions

This study confirmed the short-term efficacy of AAV-mediated RPGR gene therapy for RPGR-associated XLRP through systematic analysis, while emphasizing the need to balance risks of ocular inflammation and other adverse events. The novelty of this research lies in demonstrating that the sixth month is a critical time point for efficacy evaluation and revealing a strong correlation between ocular TEAEs and gene therapy. Rigorously designed multicenter trials are warranted to further optimize treatment protocols and advance the safe clinical application of gene therapy for RPGR-associated XLRP.