Opus Genetics Shows Proper Planning Is Key To Adaptive Trial Design

A conversation between Opus Genetics CEO George Magrath, MD, and Clinical Leader Executive Editor Abby Proch

Finding initial safety and efficacy success in their sentinel patient, Opus Genetics is embracing its choice of an adaptive trial design for OPGx-BEST1, an ocular gene therapy.

In this Q&A, CEO George Magrath, MD, shares the team’s experience with the Phase 1/2 trial and discusses the value of clearly predefined decision pathways and structured flexibility in guiding real‑time study decisions.

The OPGx-BEST1 program uses an adaptive Phase 1/2 design. What aspects were most effective in guiding real-time trial decisions, and what lessons might be useful for others designing early-stage ocular gene therapy studies?

George Magrath, MD: One of the most valuable aspects was having clearly prespecified decision pathways for dose escalation, cohort progression, and safety review, particularly following treatment of the sentinel participant. In early ocular gene therapy studies, where patient populations are small and disease presentation can vary significantly, this type of structured flexibility is critical.

Our initial clinical data presented at the Macula Society demonstrated that OPGx-BEST1 was well tolerated in the sentinel participant, with no ocular inflammation, no treatment-related adverse events, and no dose-limiting toxicities at three months. At the same time, we observed early signals of biological activity, including a 12-letter gain in best corrected visual acuity (BCVA) and a 23% reduction in central subfield thickness (CST) in the treated eye. These early observations illustrate how an adaptive design can help guide program progression while carefully monitoring safety and emerging efficacy signals.

A key lesson for others is that adaptive features should be carefully preplanned and transparently documented so that operational flexibility supports learning without compromising trial integrity. A strong adaptive design requires cross-functional planning from the outset, bringing together clinical development and medical leads to define endpoints and decision points; biostatisticians to design the adaptive framework and interim analyses; regulatory experts to ensure alignment with agency expectations; and clinical operations to execute enrollment and protocol adaptations seamlessly. This should be supported by data management and analytics teams to enable timely, high-quality readouts, as well as independent safety oversight (e.g., a DSMB) to review emerging data. Where relevant, translational and biomarker experts can help inform mechanistic decision-making. Critically, all adaptation rules, thresholds, and governance processes should be pre-specified and aligned across teams to ensure flexibility is disciplined, data-driven, and preserves trial integrity.

How did your adaptive framework help manage patient heterogeneity and still produce meaningful efficacy and safety insights?

Inherited retinal diseases often present with significant variability in baseline visual function, retinal structure, and disease progression. The adaptive framework allows us to evaluate emerging safety and functional signals sequentially across participants rather than relying on a single static data set. This approach is particularly important in early-stage gene therapy trials where each participant can contribute meaningful insights.

The team can then act on emerging data in real time rather than waiting for a fixed dataset, enabling actions such as adjusting dose levels or schedules, refining eligibility criteria, prioritizing or pausing enrollment, expanding cohorts once safety is established, and incorporating additional endpoints or assessments as signals emerge. It can also support earlier go/no-go decisions, more efficient resource allocation, and proactive safety management, all while maintaining pre-specified guardrails — capabilities that are not possible with a traditional static dataset approach where decisions are deferred until the end of the study.

For example, in the first treated participant, we observed both functional improvement and resolution of intraretinal fluid in some retinal regions as early as one month. While these results are preliminary, the adaptive structure allows us to place these findings in context as additional participants are treated and followed over time. Ultimately, this approach helps ensure that safety and biological activity signals are interpreted appropriately despite inherent patient heterogeneity.

Opus is using a sentinel, sequential cohort–based adaptive design in which patients are treated and evaluated one at a time (or in very small cohorts) with predefined safety and activity checkpoints before proceeding. The study begins with a sentinel participant to establish initial safety and tolerability, after which subsequent participants are enrolled in a staggered fashion, allowing investigators to review emerging safety (e.g., inflammation, AEs, DLTs) and early functional/structural signals (e.g., BCVA, CST) before advancing dosing, expanding enrollment, or modifying aspects of the protocol.

In this trial, and perhaps others for rare ophthalmic diseases, how do you balance flexibility and regulatory compliance when implementing adaptive elements?

For OPGx-BEST1, the adaptive elements were defined prospectively within the protocol, including how safety reviews would occur, how dose cohorts would progress, and how emerging data would inform next steps in the study. Maintaining close communication with regulatory authorities and ensuring that all adaptive components are clearly documented helps ensure the study remains scientifically sound and compliant.

In rare ophthalmic diseases such as BEST1-associated retinal disorders, where patient populations are limited and there are no approved treatments, regulators recognize the value of thoughtful trial designs that maximize the information gained from each participant. The goal is to enable responsible learning while maintaining the transparency and rigor required for clinical development.

Many adaptive trials rely heavily on interim analyses. What best practices have you developed for interpreting early signals while maintaining statistical integrity and avoiding bias?

Interim analyses are an important part of adaptive development, but they must be interpreted carefully, particularly in early-stage studies with small sample sizes. A key best practice is ensuring that interim analyses are designed around predefined questions — such as safety confirmation, feasibility of administration, or early biological signals — rather than exploratory observations that could introduce bias.

In the OPGx-BEST1 study, the sentinel participant provided an initial opportunity to evaluate safety and early activity before advancing enrollment. At the same time, it is important to remain disciplined in how early signals are interpreted, recognizing that the full cohort data — expected in midyear 2026 — will provide a more comprehensive understanding of the therapy’s clinical profile.

Given the clean safety profile so far, how were your safety monitoring and stopping criteria structured adaptively? How might others think about confronting uncertainty in their own first-in-human trials?

The OPGx-BEST1 study uses a staged cohort structure with predefined safety observation periods and formal safety reviews before advancing enrollment. This approach allows investigators to closely monitor early participants, including the sentinel participant, before proceeding with additional dosing.

The favorable safety profile observed to date is encouraging and supports continued enrollment in the study. For others designing first-in-human trials, it is important to combine rigorous safety monitoring with clear decision frameworks that acknowledge uncertainty while allowing responsible progression of the program.

Beyond traditional endpoints like BCVA, how can adaptive designs help incorporate emerging measures of functional vision or patient experience in future ophthalmic studies?

Best corrected visual acuity remains an important endpoint in retinal disease studies, but it does not always capture the full impact of a therapy on visual function. Adaptive designs can help incorporate additional functional and anatomical measures over time, particularly as the field continues to evolve.

In the OPGx-BEST1 study, we are evaluating both functional and structural endpoints, including visual acuity and changes in retinal anatomy, such as central subfield thickness and intraretinal fluid resolution. As the program progresses and additional participants are treated, these measures will help provide a more comprehensive picture of biological activity. More broadly, adaptive frameworks can support the integration of emerging functional vision metrics and patient-centered outcomes that better reflect how therapies impact real-world visual function and quality of life for patients with inherited retinal diseases.

About The Expert:

George Magrath, MD, currently serves as CEO and board member of Opus Genetics, which was recently acquired by Ocuphire Pharma. Dr. Magrath is a board-certified ophthalmologist with extensive clinical, business, and financial experience. Most recently, Dr. Magrath was CEO of Lexitas Pharma Services, a leading CRO in ophthalmology. Lexitas experienced a substantial growth in headcount, expanded into multiple additional service offerings, including an image reading center and a significant shift into retinal diseases, and was acquired by a top-tier private equity group during Dr. Magrath’s tenure. Prior to Lexitas, Dr. Magrath served as the medical director at Hovione, where he helped develop proprietary assets in ophthalmology, dermatology, and respiratory. Dr. Magrath began his career as an equity analyst at Edison Investment Research, where he covered small and midsize pharma and biotechnology companies. Dr. Magrath holds an MAE from Johns Hopkins University, an MBA from The Citadel, and an MD from the Medical University of South Carolina.