Abstract
The Foundation Fighting Blindness, RDH12 family organizations, and the RDH12 Fund for Sight convened a jointly organized workshop in Columbia, Maryland, on November 19, 2019. The purpose of the workshop was to share perspectives on what is known about the RDH12-associated retinal dystrophies (RDs) and discuss the advancement of therapies, primarily gene therapy, for people with mutations in the RDH12 gene which cause Leber congenital amaurosis 13 (LCA13). The workshop began with presentations on the RDH12 landscape, patient perspectives, the use of statistical modeling for clinical trial design, and the Foundation's My Retina Tracker Registry. An afternoon roundtable discussion focused on four key areas essential to advance research toward gene therapy clinical trials: trial design, dose projection from nonclinical to clinical studies, natural history, and regulatory considerations. In their comments, the 27 participants from academic centers, affected families, biotechnology and pharmaceutical companies, and the regulatory community highlighted a number of research priorities, including the following: systematic inventory of retrospective natural history studies and planning for a multicenter prospective study, development of large animal models, and evaluation of novel tests of functional vision in young children. Despite these research opportunities, the workshop participants agreed that the field could be ready now for a clinical trial aimed initially at testing the safety and, possibly, efficacy of RDH12 gene therapy. Advancements in this field are being fostered by the emergence of an innovative multi-stakeholder research endeavor that relies on the effective engagement of the patients.
Translational Relevance:
This initiative serves as a model for how affected individuals and their families can be research partners on the path to treatments and cures.
Francesca Sofia briefly presented the highlights of her full landscape report of
RDH12 RDs (see
Supplementary Material), from the current understanding of genetics, pathophysiology, clinical manifestations, therapeutic approaches, and research precedents from other clinical trials, including the development of Luxturna for
RPE65-associated RD.
As she summarized in her presentation, Leber congenital amaurosis (LCA) is a molecularly and clinically heterogeneous group of inherited retinal disorders characterized by vision loss, involuntary eye movements (nystagmus), and severe retinal dysfunction.
1–3 LCA, which accounts for 5% of all IRDs, is the leading cause of inherited childhood blindness in the first decade of life. It has high genetic heterogeneity, with 25 causative genes identified to date. Among those,
RDH12, which encodes a photoreceptor-specific retinal dehydrogenase, causes LCA type 13 (LCA13).
The pathogenetic mechanism associated with RDH12 mutations is not completely understood; however, it correlates with the loss of RDH12 enzymatic activity and consequent accumulation of toxic compounds in photoreceptor cells. Over 100 different mutations in RDH12 have been found to cause LCA13, which manifests with visual field constriction, loss of visual acuity, and night blindness in the first years of life and inescapably progresses to legal blindness in early adulthood.
At present, there is no treatment for LCA13, but the field of
RDH12 research is quite active, as evidenced by the numerous publications, involvement of the affected families in research, and participation of leading researchers in the workshop itself. Basic and preclinical research addressing
RDH12 disease mechanisms and/or testing potential therapeutic interventions is hampered by the lack of reliable in vivo and in vitro models. Furthermore, due to the rarity of LCA13, clinical knowledge of the disease is confined to a few natural history studies conducted by individual clinical and research institutions. Nevertheless, interviews with experts in the field (see
Supplementary Material) revealed that there are several opportunities in the therapeutic pipeline ranging from gene therapy—the most promising and advanced one on the path to cure—to other gene-agnostic approaches. Among those, neuroprotective strategies involving either gene therapy or neuroprotective compounds may slow down disease progression and provide a longer time window for treatment. Efforts are also being made in the field of optogenetics that are aimed at vision restoration by precisely exciting the neural apparatus. In addition, cell replacement therapies are being explored in IRDs and could, therefore, be applied to
RDH12 RD, although several concerns have yet to be addressed. Finally, gene editing and antisense oligonucleotides may provide additional means to tackle the dominant forms of the disease.
In order to make the most of this promising pipeline of therapies, some major gaps will have to be filled; for example, mouse models developed so far do not recapitulate the human phenotype and better animal models would be needed. In addition to this nonclinical testing ground for new therapies, in vitro preclinical testing using induced pluripotent stem cells (IPSCs) and organoids holds great promise; however, these models are relatively early in their development and require further refinement. Similarly, clinical knowledge of the disease is limited to a few natural history studies conducted by individual centers. For this reason, two steps would be beneficial: an inventory of completed clinical studies and the available data and, if warranted, a multi-center longitudinal observational study.
Any clinical study, whether observational or an experimental clinical trial, may benefit from the precedents of earlier LCA clinical trials. These studies shared several efficacy outcome measures including the following: multi-luminance mobility test (MLMT), full-field light sensitivity threshold (FST) testing, best-corrected visual acuity, pupillometry (pupillary light reflex), optical coherence tomography (OCT), and fundus autofluorescence. Safety assessments included adverse events, ophthalmic examination, physical examination, and laboratory testing. A few patient-reported outcomes have been used in trials thus far, although none has been developed specifically for the largely pediatric LCA population.
Bart LeRoy stated that many people with RDH12 mutations retain a significant amount of vision well into adulthood; therefore, he cautioned against labeling the associated disease as LCA. He and other meeting participants agreed that many adults with RDH12 mutations are genetically undiagnosed and identifying them would help populate natural history studies and clinical trials. He also emphasized that patients with RDH12 mutations exhibit areas of the retina that are much better preserved and mentioned the so-called patchy preservation of the peripapillary and peripheral retina as a predictive feature of RDH12 retinal dystrophy, a sign that should immediately call for targeted gene testing.
Wiley Chambers pointed out that natural history is normally very valuable but seems to be less valuable in RDH12, as there are many ophthalmologists—such as those who were sitting around the table—who are monitoring RDH12 natural history. Therefore, he suggested that it is not worth investing in and waiting for natural history before moving on with an interventional trial. Abigail Fahim added that a multicenter prospective observational study for RDH12 would be feasible and that the more centers involved, the better the study.
Rick Ferris highlighted the value of prospective natural history studies by describing the example of the Foundation's Usher syndrome study, RUSH2A. In this study, data are collected prospectively and specific outcome variables are being measured longitudinally with the idea of assessing not only the variability of the measure but also its rate of progression. This study is showing that longitudinal observations can be very informative but difficult due to high variability across a small number of patients who are genetically different and in different stages of the disease. He added that it is likely that RDH12 would benefit from the same systematic collection of data in which imaging can be used. However, he pointed out that investigators can identify outcome measures from retrospective studies and, at the same time, use data from a control group to learn about natural history. In this context, even if the treatment does not work, a trial would reveal insights for a future trial because it is a source of data on disease progression.
The group discussed the merits of prospective versus retrospective natural history studies. Barbara Wendelberger noted that not having a prospective natural history study is not an impediment to moving forward. Particularly in diseases whose mechanisms and natural history are poorly understood, any available information to either rule out or use what is assumed to be a good outcome measure can serve to improve their statistical model of disease progression. Everyone agreed that prospective studies are richer and better in terms of quality of data and standardization, but these studies are very expensive and demanding. By contrast, retrospective studies have lower quality because of the heterogeneity of collection methods. This might be overcome if imaging methods were standardized and a model for progression rate could be developed using retrospective data, although it would be equally challenging to collect these data. The remaining questions were as follows: What data are currently available on these measures? Is there a way to understand disease progression from the current data? How do we consolidate those data? And, do we have natural history study data that allow us to move ahead with clinical trial design?
Abigail Fahim noted that data are available from retrospective cohorts; however, they are scattered. Every patient has visual acuity, a lot have OCT, not everyone does the visual field, and they undergo imaging but not every year. This is because when doctors see the patients they do not think in terms of natural history; however, data are available, and there are data centers that might be able to handle data coming from different sources and in different formats. So, knowing what would be acceptable by the regulators is paramount to deciding what to collect and consolidate as a first step toward understanding the natural history of the disease.
In the end, it was agreed that there are two paths to follow: One is to move on with a clinical trial making use of data from existing natural history studies. The other one would be to enable a longitudinal prospective study seeking the involvement of industry or another cost-sharing arrangement.
Wiley Chambers explained that both functional and anatomical endpoints are important. Referring to previous experience in LCA, he stated that neither FST nor pupillometry would be acceptable as primary efficacy outcomes. However, a measure that shows how and if a treatment can prevent the loss of photoreceptors would provide an ideal anatomic endpoint that would logically translate into a functional outcome. On the other hand, considerable variability affects functional measures where a function is measured against a threshold that cannot always be correlated with the structural finding. Therefore, to the extent that there are clear anatomical measures, those should be employed in any RDH12 gene therapy trial.
Bart LeRoy stated that the objectively assessed MLMT developed by Spark Therapeutics (Philadelphia, PA) for the Luxturna program, full-field stimulus testing, and detailed imaging of typically preserved areas would provide the outcome measures that are needed for an RDH12 trial. FST and MLMT may be able to show results much faster than the anatomical preservation seen with OCT and other imaging modalities.
Katherine Uyhazi agreed that FST is a useful outcome measure. She added that her research institution also uses pupillometry because it is an objective measure of visual function that correlates very well with FST. Pupillometry is more useful than FST in young children because it does not require the cooperation of the patient.
Adam Dubis pointed out there were many virtues of pooling data. For one, a larger dataset would lead to better conclusions because poor-quality observations could be de-emphasized or removed. Additionally, larger data would open up the possibility of machine learning-based prediction modeling either alone or as a transfer learning exercise from other IRD datasets. The biggest challenge is accumulating enough data in one place. Other requirements include having graders trained in identifying where the boundary edges are, studying the data, and standardizing the methods. He also pointed out that data from at least 100 patients would be needed as long as multiple, evenly spaced datapoints were available. The sensitivity and specificity of this approach depend on how precisely the data going in are curated for type of data and duration between visits. Ultimately, the more data available, the better the models can be made.
Based on her experience with the study of Luxturna for RPE65-associated RD, Jean Bennett reported that Spark Therapeutics had several exploratory endpoints, which may be a useful approach. Spark thought that the electroretinograms were going to improve, but in fact they did not. However, a few exploratory endpoints, including pupillometry and light sensitivity, did show an improvement. So, the data from the phase I study were informative in terms of going forward with primary endpoints to demonstrate efficacy in phase III.
In addition to his earlier comments about endpoints and clinical trial design, Wiley Chambers raised a challenge of developing a gene therapy whose greatest benefit may be in children. He pointed out that rules within the United States would not allow children to be tested in a trial until there is some prospect for a direct benefit. Also, due to the lack of a good animal model showing success of the approach, adults ages 18 and older would be tested first in phase I. This initial study would not require many adults nor a definitive endpoint, but surely the trial would have to be done with people who could give informed consent. The obvious downside of this initial study is that adults are too far along in the disease; therefore, the results would not necessarily provide information about the product's true potential in children.
He further explained that the phase I study would require a cohort of three patients treated with the lowest dose to identify any safety problems before escalating to higher doses. If the data support overall safety at the lowest dose, the study could proceed to test a higher dose in three more patients and check again for safety problems. The study may include a third highest dose. From the first patients, one can usually assess whether or not adverse effects arise. From there, additional study patients can be randomized to the selected dose groups. Within the phase I studies, it would be necessary to demonstrate there was a prospect of benefit for children. The benefit could be restoration of some measure of vision or, usually, not getting worse.
Debra Thompson asked the group whether evidence of safety in adults, coupled with the nonclinical data showing a decrease in light damage, would be predictive that children might benefit from therapy. There was not a clear answer to this question, other than agreement that results from humans would be preferable to nonclinical data.
Kate Arkell, Retina International, Buckingham, UK
Jean Bennett, University of Pennsylvania, Philadelphia, PA, USA
Silvia Cerolini, Family Patient Representative, London, UK
Wiley Chambers, Center for Drug Evaluation, U.S. Food and Drug Administration, Silver Spring, MD, USA
Janet Cheetham, Foundation Fighting Blindness, Columbia, MD, USA
Jogin Desai, Family Patient Representative, Bangalore, India
Adam Dubis, Moorfields Eye Hospital and University College of London, London, UK
Todd Durham, Foundation Fighting Blindness, Columbia, MD, USA
Blair Ettles, MeiraGTX, New York, NY, USA
Abigail Fahim, University of Michigan, Ann Arbor, MI, USA
Rick Ferris, Foundation Fighting Blindness, Columbia, MD, USA
Mike Fiore, Family Patient Representative, Syossett, NY, USA
Allison Galloway, Family Patient Representative, Westminster, CO, USA
Jessica Imrie, Limelight Bio, Philadelphia, PA, USA
Rusty Kelley, RD Fund of Foundation Fighting Blindness, Columbia, MD, USA
Bart Leroy, Ghent University, Ghent, Belgium
Brian Mansfield, Foundation Fighting Blindness, Columbia, MD, USA
Jason Menzo, Foundation Fighting Blindness, Columbia, MD, USA
Mariya Moosajee, Moorfields Eye Hospital, London, UK
Mathew Pletcher, Family Patient Representative, Basel, Switzerland
Sean Ring, Editas Medicine, Cambridge, MA, USA
Ben Shaberman, Foundation Fighting Blindness, Columbia, MD, USA
Francesca Sofia, Science Compass, Milan, Italy
Debra Thompson, University of Michigan, Ann Arbor, MI, USA
James Tobin, Janssen Pharmaceutical Companies of Johnson & Johnson, Titusville, NJ, USA
Katherine Uyhazi, University of Pennsylvania, Philadelphia, PA, USA
Barbara Wendelberger, Berry Consultants, Austin, TX, USA
Ben Yerxa, Foundation Fighting Blindness, Columbia, MD, USA