A viral vector for gene therapy to treat neovascular age-related macular degeneration has demonstrated long-term safety of gene transfer and therapeutic protein production for up to four years in the eyes of affected patients. Recent Phase I trial results of the vector may lead to packaging of a protein that binds vascular endothelial growth factor.

The trial evaluated RetinoStat (Oxford BioMedica, Oxford, UK), a lentiviral equine infectious anemia virus (EIAV) vector that expressed endostatin and angiostatin in 21 patients with neovascular AMD. Results were published in the January issue of Human Gene Therapy.1

Patients were enrolled in three different dose-ranging cohorts—2.4 × 104 transduction units (TU); 2.4 × 105 TU; and 8 × 105  TU, each with three patients—and one cohort of 12 receiving the highest dose of 8 × 105 TU. The study duration was 48 weeks and patients were encouraged to enroll in a 15-year long-term follow-up study.

All of the doses were well tolerated, although one patient developed a macular hole that was treated and resolved. The study documented a dose-related increase in levels of endostatin and angiostatin in the aqueous humor, which peaked between weeks 12 and 24 in the highest-dose group.

Fluorescein angiography showed a reduction in leakage in most patients, but only one subject showed convincing evidence of anti-permeability. Lesion size, however, showed no significant change in the maximum therapy group.

Here, principal investigator Peter A. Campochiaro, MD, Eccles Professor of Ophthalmology and Neuroscience at Johns Hopkins Wilmer Eye Institute, Baltimore, discusses the Lentiviral Vector Gene Transfer of Endostatin/Angiostatin for Macular Degeneration (GEM) study.

Summary of GEM findings in his own words:
RetinoStat is an equine infectious anemia lentiviral vector that enters cells and produces endostatin and angiostatin, both of which are proteins that have strongly suppressed abnormal blood vessel growth and leakage in mouse models of wet AMD. Several studies have demonstrated that subretinal injection of adeno-associated viral (AAV) vectors is safe and well-tolerated.

The purpose of the GEM trial was to determine if subretinal injection of a lentiviral vector is safe and well-tolerated in patients with advanced wet AMD.

Lentiviral vectors differ from AAV vectors in that they incorporate into the genome, meaning that they insert into a patient’s DNA, which is the most stable situation for long-term expression. The study demonstrated that subretinal injection of an EIAV vector caused no identifiable toxicity and thus was safe.

Regarding secondary trial endpoints:
The major secondary endpoint was to determine if there was good production of endostatin and angiostatin. Levels of endostatin and angiostatin in the aqueous humor were measured at several time points after subretinal injection. We did this by inserting a needle into the front of the eye and removing a small amount of fluid used to measure levels of the proteins.

High levels of endostatin and angiostatin were present in the aqueous humor obtained from almost all patients, and the levels did not decrease over time. Many patients had measurements years after the subretinal injection of RetinoStat and the levels of endostatin and angiostatin were still high with no decrease from earlier levels.

Other secondary endpoints were the measurement of the amount of fluid in the macula by optical coherence tomography and change in visual acuity from baseline. There was no significant reduction in fluid or improvement of visual acuity after injection of RetinoStat—but most patients had very advanced wet AMD with scarring in the macula resulting in permanent loss of vision that could not be improved.

The mechanism of action of RetinoStat in his own words:
When RetinoStat is injected under the retina, it enters retinal pigment epithelium cells and photoreceptors, inserts into the DNA of those cells and begins producing endostatin and angiostatin. We know that this occurred because after the injection of RetinoStat, patients had high levels of endostatin and angiostatin in their eyes and before the injection they did not.

In mice, endostatin and angiostatin cause reduction in abnormal blood vessels and leakage of fluid from abnormal blood vessels. Our study suggests that endostatin and angiostatin may have less effect on abnormal blood vessels in humans than they do in mice.

What investigators learned about the potential for changes in visual acuity:
There were a few patients who had improvement in vision and reduced fluid after the injection of RetinoStat, but the majority did not.

How findings may inform future trials:
This study indicates that EIAV is an extremely good platform for gene therapy because it provides long-lasting production of proteins.

Next steps in RetinoStat development:
A potential next step is to package a protein that binds vascular endothelial growth factor (VEGF) in EIAV. We already know that injections of proteins that bind VEGF provide benefit in patients with wet AMD. Long-term production of such proteins in the eye after injection of a lentiviral vector could provide a long-term benefit.

REFERENCE
1. Campochiaro PA, Lauer AK, Sohn EH, et al. Lentiviral Vector Gene Transfer of Endostatin/Angiostatin for Macular Degeneration (GEM) Study. Human Gene Ther. 2017;28:99-111.