Future Studies

A. Challenges in Planning a Gene Therapy Clinical Trial for LCA

In the studies described above, AAV-2-mediated delivery of a wild-type version of the defective RPE65 gene was used to provide retinal function in a canine model of severe early

Figure 1 Frames in sequence from a video clip demonstrating right preferential gaze. Movie was filmed 9 months after subreti-nal delivery of AAV.RPE65 in the right eye in a mutant RPE65 dog. The dog does not see the person when the person is behind him on his left (untreated) side (A). the dog follows the person walking around him with his right (treated) eye (B-H). When the person steps outside the visual field of the dog's right eye (H), the dog turns his head all the way to the right to visualize the person with his (seeing) eye.

Figure 1 Frames in sequence from a video clip demonstrating right preferential gaze. Movie was filmed 9 months after subreti-nal delivery of AAV.RPE65 in the right eye in a mutant RPE65 dog. The dog does not see the person when the person is behind him on his left (untreated) side (A). the dog follows the person walking around him with his right (treated) eye (B-H). When the person steps outside the visual field of the dog's right eye (H), the dog turns his head all the way to the right to visualize the person with his (seeing) eye.

onset retinal degeneration. There are many steps that must be taken before this treatment can be tested in humans with LCA. With the availability of both the canine and murine (rpe65 knockout) (29,28) models for this disease, these steps can be taken using both animal models. The first and most important step is to establish safety of the treatment. Selection of regulatory elements that limit transgene expression to the target cells will minimize toxicity to other cells that might be exposed to the virus. Special attention must be paid to the possibility that the transgene could potentially be introduced to the brain via unwanted ganglion cell/optic nerve targeting (13).

The duration of the therapeutic effect should be characterized because LCA is a lifelong disease. The therapeutic window should be defined with respect to age of the animal/level of disease progression so that predictions can be made as to what the effects of treatment will be in younger and older humans with LCA.

The vector that provides the maximal therapeutic effect at minimum dose must then be selected. The recent demonstration that one can control cellular specificity, onset, and levels of transgene expression by packaging the AAV in capsids from AAVs of different serotypes may play a role in this process (3). Vector selection will be difficult. There is likely to be continued progress in developing viral vectors with slightly improved transduction qualities. At a certain point, a choice will have to be made: whether to select a particular vector as the candidate for formal toxicity studies (in anticipation of initiating a human clinical trial using this vector) or whether to perform additional efficacy tests on newly formulated vectors. The toxicity studies are rigorous, time-consuming, and expensive, and use a large number of animal subjects. It will be preferable not to have to repeat them using a second vector (unless the additional benefits of that vector warrant the effort). Once the ideal vector is selected, toxicity testing must be conducted to verify that the treatment does not cause any impairment to the eye or to the rest of the body. This must be performed using the same virus that is to be used in the proposed human clinical trial. Studies evaluating the possibility of readministration should also be performed with this vector in order to be able to predict whether it will be possible to treat the contralateral eyes of human patients at future timepoints.

B. Issues Regarding Future Patient Selection

While animal studies are in process, humans with LCA should be screened for their disease-causing mutations and individuals carrying homozygous/compound heterozygous RPE65 disease-causing mutations should be identified. The typical progression pattern of the human disease (with respect to visual function measures and death of retinal cells) should be characterized. This will allow selection of the optimal noninvasive outcome measures for evaluation of effects of treatment in a human clinical trial. These outcome measures (likely measures of ERG/pupillometry response) will be quantitative and reproducible. One question that will arise is whether the first treatments, which will be phase I safety and toxicity tests, should be performed on adult vs. pediatric subjects. Also, an exit strategy should be planned if the results are unpleasant to the patient. Oliver Sacks describes, for example, the shock that a 55-year-old man had upon seeing for the first time since he was 6 years old, when he first opened his eyes after having had cataracts, present from childhood, removed (33). It should be emphasized, however, that the treatment will be deemed successful even if it is effective in a seemingly small way. If a patient can benefit from improved navigational skills that will expand his or her independence, this in itself will be a successful outcome. The ultimate goal will be, of course, to achieve vision that will allow the patient to lead as normal a life as possible.

C. Extrapolation of Success in Canine LCA Gene Therapy Studies to Other Forms of Early Onset Retinal Degeneration

A major challenge is how to extrapolate the successful treatment of the RPE65 disease to treatment of other severe early onset retinal degenerative diseases. One approach to meet this challenge is to identify the conditions that are responsible for the success of the RPE65 dog study. Besides the virus delivery characteristics, there may be elements of the RPE65-caused disease that make it particularly amenable to treatment. Is the success due to the fact that the RPE65 mutant retinal photore-ceptors degenerate very slowly and thus there is a large therapeutic window during which the retina can be rescued? In some animal models (e.g., those with homozygous mutations in the beta subunit of rod cGMP phosphodiesterase) the degeneration is so rapid that rod photoreceptors do not even have a chance to fully differentiate before they are lost (10). In such diseases, it is likely that the treatment will have to be administered very early in life while the target cells are still viable (e.g., before postnatal day 10 in the rd1 mouse model of this disease) (8,17).

The AAV-2 vector may not have a rapid enough onset to make any impact in the disease progression of rapidly progressive degenerative diseases. One might want to employ a viral vector that results in a rapid onset of transgene expression. AAV-2 results in a slow (6-8-week) onset of transgene expression in the dog (6). In contrast, AAV-2 packaged with an AAV serotype 1 capsid (AAV-2/1) delivers transgenes that are expressed within a few days after injection in the mouse (3). If early administration with the optimal viral vector is not enough to achieve rescue, it may be possible to expand the therapeutic window by delivering the transgene at gestational timepoints. Recent progress in developing methods for in utero retinal gene transfer will expand these options (36). Finally, it will be desirable to test the rescue ability of the human RPE65 cDNA (vs. the canine cDNA) in animal models before testing this cDNA in humans. The human RPE65 cDNA was recently cloned and inserted into AAV vectors by N. Dejneka (personal communication; 2001) and promises to be useful in studies in both the RPE65 mutant dog and the rpe65-/- mouse.

In particular diseases, there may be requirements of the disease model that will have to be met by developing alternative vectors/vector strategies. For example, one challenge is how to proceed with AAV if the cDNA and/or transgene cassette is too large to be packaged. The AAV cargo capacity is a maximum of 4.8 kb. In such a situation, one would have to consider other vectors that could carry the intact transgene cassette. Options include lentivirus, which can carry a cassette of ~8 kb, and gutted adenovirus, which can carry a cassette of >30 kb. If one wanted to use AAV even though the transgene cassette was too large, one could harness the AAV dual vector approach and split the transgene construct into 2 complementary vectors, which would be trans-spliced in the target cell (30).

Another area that is likely to benefit from new technological developments is the ability to treat gain-of-function disease. Recent work involving gene correction strategies for both loss-of-function and gain-of-function disease and delivery of RNAi will likely be applied to animal models of early onset retinal degeneration using virus-mediated gene transfer.

Was this article helpful?

0 0

Post a comment