The Distinct Characteristics of Product Development in Myasthenia Gravis

Product development in Myasthenia Gravis (MG) has its set of challenges that stem from the nature of the disease itself: MG is rare, clinically heterogeneous, and can vary widely in antibody status, symptom pattern, severity, and treatment response. That makes it more difficult to define study populations, select endpoints, interpret results, and generate sufficiently robust evidence. The limited size of the patient population also means that data can be sparse, and real-world evidence may be less complete than in more common disease areas.

This article explores four areas that make product development in MG distinct: patient-reported outcomes, subtype-driven trial design and access, the broader relevance of MG therapies across autoimmune neurology, and the emerging role of artificial intelligence in research.

Patient-Reported Outcomes in MG Development

A defining feature of MG clinical trials is the central role of patient-reported outcomes. Clinical trials in MG are unusual in one important respect: even in interventional studies, the primary endpoint is often a patient-reported outcome measure. In many disease areas, evidence generation relies primarily on biomarkers, imaging, clinician-rated assessments, or event-based outcomes. These measures are important, but they can sit at a distance from what patients experience in daily life.

In MG, the patient experience has moved closer to the center of how the benefit of treatment is measured. This is particularly visible in the use of the Myasthenia Gravis Activities of Daily Living scale, or MG-ADL, which is used increasingly as a primary endpoint in clinical development.

MG-ADL is rooted in the realities of living with myasthenia gravis. It captures functional challenges that can shape a patient’s daily experience, including:

• speaking.
• chewing.
• swallowing.
• brathing,
• seeing clearly.
• lifting the arms.
• rising from a chair.

This matters for product development because clinical benefit is assessed through changes that are directly connected to how patients function day to day.

The use of MG-ADL as a key endpoint at the pre-approval stage creates an important opportunity: trial results can be connected more directly to real-world patient experience. This is especially relevant for real-world data sources, including patient-reported registries that already collect MG-ADL data from people living with MG. These registries can include MG patients who may not otherwise have access to clinical research or who may be underrepresented in clinical trials. When these datasets include MG-ADL, they can help contextualize clinical trial results and provide insight into the types of patients who may benefit from emerging treatments. In MG, product development is not only about measuring disease activity. It is also about understanding whether treatment meaningfully changes the daily experience of living with the disease.

Subtype Stratification in MG Trial Design

Clinical trials in MG often begin by studying a specific patient population. This may be defined by antibody status, disease presentation, or symptom pattern. For example, most new products start by restricting trial enrollment to generalized MG patients but may later expand their research to ocular MG patients. This approach can help researchers evaluate whether a therapy works in a clearly defined population before studying broader groups of patients.

Acetylcholine receptor antibody-positive, or AChR+, MG is the most common antibody-positive subtype of MG. Because it represents a large and clinically well-characterized group, AChR+ MG has often been a starting point for targeted therapy development. Beginning with this defined population for product development can support clearer study design, endpoint interpretation, and regulatory evidence generation.

Subtype-specific development does not always mean subtype-specific access over the long term. As additional evidence is generated, approved labels may sometimes expand to include broader groups of patients. VYVGART®, for example, was originally approved in the US for adults with generalized MG who were AChR+. Its US label has since expanded to include all adults with generalized MG, regardless of antibody status.

Narrower trial populations can make it easier to evaluate treatment effects and generate robust evidence. Real-world data sources, such as registries, can be a great way to evaluate patient outcomes in off-label use of a drug in a non-approved MG subtype.

At the same time, broader patient inclusion is essential to learning how innovation can benefit the full diversity of people living with MG. Product development in MG is shaped by multiple aspects, like:

• disease biology.
• who is included in the research.
• whose experience is reflected in the evidence.
• how innovation can ultimately reach the patients who need it most.

Cross-Disease Relevance of MG Therapies

Many newer therapies developed for MG are not designed to target MG in a narrow sense. Instead, they target specific immune pathways that contribute to disease processes. These pathways may also be involved in other autoimmune and inflammatory conditions, so interest in MG therapies can extend beyond the MG patient population. Examples include therapies that reduce circulating pathogenic antibodies or interrupt immune pathways involved in tissue damage.

When a therapy demonstrates benefit in one disease, researchers may investigate whether the same biological pathway is relevant elsewhere. As a result, MG can sometimes become part of a broader immunology or neurology development strategy rather than a standalone indication. However, shared biology does not guarantee shared clinical benefits. Each disease still requires its own clinical studies, endpoints, safety evaluation, and understanding of which patients are most likely to respond. Progress in MG research may therefore have implications far beyond myasthenia gravis itself. For patients, this can be encouraging because it reflects growing investment in targeted immune therapies. For researchers and drug developers, it also highlights the importance of understanding not only the diagnosis, but the underlying biology that connects diseases across traditional categories.

AI-Enabled Evidence Generation in MG

MG is a highly heterogeneous disease. Patients can differ in antibody status, symptom presentation, disease severity, treatment history, comorbidities, and risk of exacerbation. This complexity creates challenges for traditional analytical approaches. Machine learning may help researchers evaluate multiple variables simultaneously, such as antibody status, symptom patterns, MG-ADL scores, treatment history, and healthcare utilization, and identify patterns that can be difficult to detect using conventional methods alone. As real-world datasets grow, they can create the foundation for researchers to ask more sophisticated questions:

• Who is at a higher risk of worsening?
• How do symptoms, treatments, and outcomes cluster over time?

One area where AI, particularly machine learning, is gaining attention is the analysis of patient registries. In rare diseases such as MG, challenges including limited sample sizes, incomplete data, and imbalanced outcomes can affect model performance and interpretation. Registries can provide insight into how patients experience disease over time, including symptoms, treatment use, healthcare utilization, and outcomes. As registries grow in size and longitudinal depth, they create new opportunities to explore questions that may not be easily addressed through clinical trials alone.

AI in MG will not replace clinical judgment or predict individual outcomes. Instead, it has the potential to:

• help researchers better understand patterns across patient populations.
• identify factors associated with disease worsening.
• generate new questions for future study.

In this way, machine learning can support more patient-centered evidence generation and help make real-world data more actionable. For patients, this matters because better use of real-world data may help research reflect the variability of lived experience in MG and support more targeted questions about symptoms, treatment needs, and disease burden. As MG research continues to evolve, new analytical approaches may help researchers learn more from the growing volume of patient-generated and real-world evidence. For patients, AI has the potential to make their reported symptoms, treatment experiences, and day-to-day disease burden more visible within the evidence used to guide future research.

Conclusion

Product development in MG requires a careful balance between scientific rigor and patient relevance. The disease’s heterogeneity, limited data, and evolving treatment landscape all shape how trials are designed, how evidence is generated, and how value is demonstrated. As MG continues to advance, successful development will depend on approaches that are sufficiently precise to capture the complexity of the disease and sufficiently grounded to reflect the needs of the patients living with it.