In August, Regenxbio announced interim data for its 18-patient phase 1 trial of RGX-314, a gene-therapy-derived anti-VEGF, for wet AMD. The highest-dosed cohort (6 patients) had an encouraging response that included a mean 8-letter gain in BCVA and a mean reduction in central retinal thickness. The data from the highest dose cohort showed great improvement over the 2 lower-dosed cohorts. The company is planning to treat a fourth cohort with an even higher dose of RGX-314. In this column, we asked several world-renowned experts for their initial reaction to the news release and asked a principal investigator for the trial to respond.

Timothy G. Murray, MD, MBA, Murray Ocular Oncology and Retina, Coral Gables, Florida

Regenxbio’s RGX-314 phase 1 clinical trial has presented preliminary results for the first targeted neovascular AMD gene therapy clinical trial. As in all phase I clinical trials, treatment is targeted to evaluate safety, establish dosing parameters, and develop a foundation for appropriate phase 2 and 3 clinical trials. Uniquely, protein expression from the viral vector appears to be both dose dependent and able to achieve functional intraocular levels which appear to be sustained throughout the study.

Of interest is that cohort 3 showed the greatest signal response as well as a positive treatment effect. This unique delivery of a one-time therapy for sustainable anti-VEGF has the potential to greatly impact both the patient and physician treatment burden. The caveat to all phase 1 clinical trials is the potential to overstate effect (truly not meant to be measured). Nonetheless, the presence of protein, effect, and sustainability in the setting of apparently minimal safety concerns highlight the potential for this therapeutic approach. Ultimately, phase 2 and 3 clinical trials will be required to validate the utility of RGX-314 in the treatment of neovascular AMD.

Jay Duker, MD, New England Eye Center and Tufts University School of Medicine, Boston, Massachusetts

Phase 1 studies are primarily designed to test safety. The Regenxbio phase 1 data on its subretinal gene therapy treatment for wet AMD delivered via pars plana vitrectomy (PPV) showed no safety concerns in the 18 enrolled eyes at the 6 month time point. Given the small number of eyes and the high quality of the participating surgeons, the lack of serious adverse events is not surprising.

For most phase 1 studies, an efficacy signal is hard to determine. This is attributable to the fact that phase 1 trials typically have a small “N,” they usually lack a control group, there is no randomization between dose groups, and they feature short follow-up. This study is no exception.

In the highest dose (which the company expects to use in phase 2), a change from averaging 2 injections over a 6-month period prior to enrollment in the study, to 1 injection over a 6-month period after enrollment, does not represent much of a change. This is especially so because the parameters for reinjection were not specified and were no doubt not applied during the 6-month time period prior to enrollment.

Given that the parameters for reinjection were not specified, the fact that 3 of 6 high-dose eyes (50%) did not need a reinjection after the therapy over 6 months may or may not be significant. It will be interesting to see if this result is durable. If it is, this would be a positive efficacy signal.

The OCT changes in retinal thickness are too small to be clinically meaningful. Given that this is a cohort of eyes that had been treated prior with intravitreal anti-VEGFs, and therefore probably had little residual fluid to at the time of enrollment, these data are not particularly revealing either way. While some eyes showed visual improvement, without knowing how this coincided with their last anti-VEGF injection, both prior to enrollment and at the 6-month visit, the data are hard to interpret.

In summary, this phase 1 trial shows no safety concerns, and a measurable level of transgene in the aqueous humor in a dose-dependent fashion. While efficacy signal is suggested in some eyes at the high dose, the efficacy results are not particularly strong.

My overall take is that it is reasonable to use these results to advance to a phase 2 trial with a larger “N” and a control group. However, given the 1% to 3% rate of surgical complications for PPV in general, coupled with the extremely low rate of serious complications from intravitreal injections, I am skeptical that any subretinal delivery of gene therapy for wet AMD will prove safe enough in a large, multi-center trial for FDA approval, even if efficacy is demonstrated.

Emmett T. Cunningham, Jr., MD, PhD, MPH, Clarus Funds, Cambridge, Massachusetts

The Regenxbio therapy is intended to be a one-time subretinal injection of a proprietary adeno-associated virus 8 (AAV8) vector engineered to express a monoclonal antibody fragment, similar to a Fab, directed against VEGF-A. The study included 3 arms, each with 6 patients selected for a prior history of requiring frequent anti-VEGF injections. The doses studied spanned more than an order of magnitude increase in genome copy number (3x109, 1x1010, and 6x1010) and all patients were followed for at least 6 months. While a detailed report of the findings have yet to be shared, I find the approach and topline data important for several reasons.

First, and perhaps most important, this represents a meaningful step toward use of gene therapy to deliver antibody-like therapeutics. No one would deny that antibody therapeutics have transformed the treatment of human disease over the past few decades, and the promise of achieving sustained benefit with a one-time administration is truly exciting. Second, while the number of subjects in each cohort was too small to provide adequate power to achieve statistical significance, there did appear to be a trend toward a lower post-treatment anti-VEGF injection requirement in the cohort of patients who received the highest genome copy number - at least over the first 6 months following treatment. A beneficial effect on injection burden was further supported by maintenance or improvement in both best-corrected visual acuity and central retinal thickness in most patients in the highest dosed cohort over this time period. The company plans to enroll 6 additional subjects who will be administered more than twice the genome copy number, 1.6x1011, and all patients will be followed for 24 months. These expanded and longer-term results will be tremendously important. I was also encouraged to see that subretinal delivery of RGX-314 appeared to be well tolerated in all 18 patients. Regenxbio’s website characterizes the study sites in this trial as “6 leading retinal surgery centers across the United States,” so the extent to which subretinal delivery can be safely extended to a broader retina surgery community remains to be seen.

Jeffrey S. Heier, MD, Principal Investigator for Regenxbio’s RGX-314 phase 1 clinical trial and co-president of Ophthalmic Consultants of Boston, Boston, Massachusetts

The Regenxbio phase 1 trial of RGX-314 is the first study of gene therapy in wet AMD to demonstrate intraocular protein being expressed in a dose-dependent fashion. RGX-314 utilizes an AAV8 vector, which was shown in preclinical studies to produce high levels of sustained protein in the eye. The recent data released suggest RGX-314 gene therapy has potential as a one-time treatment for sustained delivery of anti-VEGF protein to maintain outcomes of frequent anti-VEGF injections to treat retinal diseases.

In order to standardize the subretinal delivery, great care was taken to select experienced surgeons, as well as work collaboratively to standardize and perfect the surgical procedure. Data from the first 3 cohorts (18 patients) has demonstrated that this therapy can be delivered safely.

RGX-314 was well tolerated by all subjects with no drug-related adverse events. Based on the findings to date, the trial has been expanded to include a new cohort 4 at a higher dose and dosing is under way.

Although this is an early trial in a small number of wet AMD patients, there does appear to be biologic activity and clear evidence of expression in cohort 3 as demonstrated by ocular protein levels along with maintenance of central retinal thickness with no or few anti-VEGF injections (in 3 of the 6 patients, no rescue injections were needed over 6 months). In early-phase studies, I put very little emphasis on visual acuity; that being said, it is encouraging that despite long-standing disease, several patients in the trial who required few or no rescue injections saw meaningful improvements after treatment with RGX-314. It is also encouraging in this phase 1 trial that despite a surgical intervention, no patient lost vision. Given these findings, progression to a phase 2 trial is clearly warranted. RP