Lessons Learned from the Development of the First FDA-Approved Gene Therapy Drug, Voretigene Neparvovec-rzyl

Abstract

In the 5 years following U.S. Food and Drug Administration (FDA) approval of the first gene therapy reagent approved to treat a genetic disease, voretigene neparvovec-rzyl (Luxturna), retinal disease clinics, hospital pharmacies, operating rooms, and even health insurance entities around the world have incorporated gene therapy as a standard procedure. The success of Luxturna has helped pave the way to establish a template for developing other gene therapy reagents that promise to restore sight or halt the progression of photoreceptor cell loss in both inherited and acquired retinal diseases. Here we review lessons learned from development of a gene therapy drug for RPE65 disease and how these lessons may expedite the development of additional treatments for previously untreatable blinding conditions.

Figure 1.

Comparative anatomy. Approximate differences in sizes of the eye and lens in species commonly used for preclinical testing compared to the human eye. The size of the eye and percentage occupied by the lens are factors that dictate the route of delivery and the volume (and thus virus dose) that can be administered. Figure created from data derived from Faqui (2016) and Dejneka et al. (2004), and measurements made in the Bennett laboratory.

Figure 2.

Dose extrapolation exercise. Hypothetical data using efficacy in an affected mouse model and toxicity in an unaffected large animal (pig) model to select doses to be used in a phase 1 human clinical trial. Efficacy in the mouse (blue squares) requires subretinal administration of Dose #3 (arrow). (Approximate dose [vg] per cell is calculated based the known number of cells in retinas of different species and the area of retina that is targeted. The area targeted is a function of volume injected.) Meanwhile, in the pig (green squares), there is safety (no observed adverse effect level [NOAEL]) at Doses #1 and #2 (arrowheads). Signs of toxicity appear at Dose #3. Proposed human doses (red polygons) would span a region where there is efficacy in mice and safety in pigs. Typically, a phase 1 gene therapy safety study uses three doses. One would want to avoid doses in which there is a slim chance of efficacy or a high chance of toxicity.

Figure 3.

Accurate dosing of adeno-associated virus (AAV), devices, and surfactant. When AAV is loaded into an injection device, up to 80% of the particles bind to the “inert” surfaces of the device, thus making accurate dosing difficult. Bennicelli et al. (2008) showed that addition of a small amount of surfactant (Pluronic F68), prevents this binding and allows accurate dosing. (X) Syringe lacking surfactant.

Figure 4.

A variable that can affect the success (and potentially the safety) of gene therapy is the ratio of “full” (transgene-containing adeno-associated virus [AAV] capsids) versus “empty” (capsids alone) in the test reagent. Shown is the appearance of bands representing empty and full capsids after centrifugation of an AAV preparation on a cesium chloride (CsCl) gradient. The band containing “full” capsids can be isolated from that containing “empties.” Image provided by Dr. Shangzhen Zhou.