Together with robust data-driven modeling, rethinking regulation and data use could push forward a notoriously challenging field.
The development of treatments for rare diseases faces a multitude of challenges, with an estimated 95% still lacking an FDA approved treatment. Despite some inspiring success stories and incredible advances, development is still very slow, chiefly because regulators consider each treatment development on a case-by-case basis. Through our work at Certara, which partners with clients to optimize their drug development processes, we’ve become convinced that the time has come to replace this approach with a more flexible “sandbox” strategy based on the disease etiology and type of treatment that is being developed.
The sandbox approach, borrowed from digital innovation, refers to a controlled, collaborative space where regulators, industry sponsors, patients and academic experts can test and refine new methodologies in real time. In rare disease treatment development, this method offers an environment to trial novel endpoints, adaptive designs and statistical approaches without the rigidity of traditional regulatory frameworks.
For example, instead of the traditional paradigm of having a pre-investigational new drug (pre-IND) meeting with the FDA, conducting a Phase 1 study, then having one or more additional meetings with the FDA before moving to a Phase 2/3 study, the whole program could be developed from the nonclinical to pivotal Phase 3 study in collaboration with stakeholders and discussed with the FDA. This would essentially open one IND for a continuous Phase 1/2/3 clinical study, where clinical data could be reviewed by the FDA as it is generated rather than at separate milestone meetings.
Creating a Sandbox Environment
The main idea of this approach is for treatments to be placed in a sandbox according to how they are manufactured. For example, a “small sandbox” would focus on development of small synthetic molecules, a “bio sandbox” would support well characterized biologics manufactured by a recombinant technology (e.g., monoclonal antibodies) and a “complex sandbox” would facilitate development of complex cell and gene therapy products.
The reason for separation into three “sandboxes” is that the regulatory and product development requirements are different between small molecule, biologics and cell and gene therapies. As such, it will help developers to have pre-set expectations based on the type of product being developed.
Within sandboxes, the requirements for product development should be aligned as much as possible. The developmental and regulatory framework will also be aligned—to the extent possible based on the type of treatment—with disease etiology.
We can hear the skeptical voices: It is not that simple. I absolutely agree, but we have to start somewhere, and given the latest FDA plan to phase out animal testing requirements for monoclonal antibodies, now is the perfect time to explore it.
Monoclonal antibodies to treat rare diseases can be developed in the sandbox approach that establishes the new framework for nonclinical studies, eliminates use of nonhuman primates in research and implements robust nonclinical models, speeding up product development. By using the sandbox approach to rare disease drug development as a pilot, we will gain valuable experience that can later be applied to the development of monoclonal antibodies for more common diseases.
Adding Model-Informed Drug Development Sandbox environments allows for iterative learning and validation. Regulatory agencies including the FDA and European Medicines Agency (EMA) have expressed increasing openness to flexible engagement models, often launching pilot programs or initiatives under the umbrella of model-informed drug development (MIDD).
MIDD utilizes data-driven models to integrate knowledge across biology, pharmacology and clinical sciences. These range from pharmacokinetic/pharmacodynamic (PK/PD) models to disease progression and exposure-response models, and enable predictions that guide dose selection, trial design and regulatory submissions. In the context of rare diseases, MIDD can be particularly powerful. With limited clinical trial data, models allow developers to maximize the utility of every data point, reducing reliance on large patient cohorts.
By fostering early and continuous dialogue, stakeholders in a sandbox can co-develop and stress-test modeling strategies, ensuring that novel approaches are scientifically valid and meet regulatory expectations.
When MIDD is applied within a sandbox approach, the benefits are amplified. For example, developers of a gene therapy for a pediatric neuromuscular disorder might lack sufficient long-term outcome data. Through MIDD, developers can create disease progression models using natural history data, which are then validated and iteratively refined in a sandbox setting with regulator input. These models can help justify surrogate endpoints, inform dosing strategies and provide a mechanistic rationale for accelerated approval pathways.
This synergy also supports platform development, where the knowledge gained from one rare disease program is transferable to another with shared pathophysiological features. The sandbox enables structured data sharing and modeling standardization, while MIDD provides the tools to generalize and apply the insights.
The integration of MIDD and sandbox approaches into rare disease drug development is still evolving. Key to success is transparent communication, robust model validation and continued investment in data infrastructure. Regulatory buy-in remains critical, as does patient engagement—especially in co-designing endpoints and understanding risk-benefit trade-offs.
Looking ahead, the use of real-world evidence and AI-driven modeling within sandbox-MIDD frameworks could further expand the possibilities for rare disease drug approvals. These tools can simulate long-term outcomes, assess comparative effectiveness and even identify new patient subgroups that may benefit from targeted therapies.
Together, MIDD and the sandbox approach represent a paradigm shift in rare disease drug development. By merging rigorous quantitative science with regulatory flexibility and stakeholder collaboration, they offer a viable path forward for overcoming the inherent limitations of traditional drug development in small, heterogeneous populations. Their combined use promises to accelerate innovation, reduce uncertainty and ultimately bring much-needed therapies to patients who have long been underserved.
