The best and worst of times in therapy development for myasthenia gravis

Abstract

Within the last 5 years the U.S. Federal Drug Administration (FDA) has approved complement and neonatal Fc receptor (FcRN)-inhibitors for treatment of generalized myasthenia gravis; and several other therapies are in late stage clinical trials or under regulatory review. Questions, however, about which patients are most likely to benefit from which therapies, and the relative effectiveness of these very expensive drugs, results in uncertainty around the place that they should occupy in the existing therapeutic armamentarium. MGNet (a Rare Diseases Clinical Research Consortium funded by the National Institute of Neurological Diseases and Stroke) held two meetings during the Myasthenia Gravis Foundation of America (MGFA) International Conference to discuss the most critical needs for clinical trial readiness and biomarker development in the context of therapy development for MG. The authors provide a summary of these discussions, but not a consensus opinion, and offer a series of recommendations to guide focused research in the most critical areas. We welcome ongoing discussion through comments on this publication.

Introduction

The current state of therapeutic development for myasthenia gravis might arguably be described as both “the best of times” and the “worst of times”. Within the last 5 years, we have seen U.S. Food and Drug Administration (FDA) approval of drugs that target complement activation¹ and neonatal Fc receptor (FcRN)-FcRN-mediated antibody recycling² with phase 3 clinical trials demonstrating the efficacy of these compounds for patients with generalized disease, including those often regarded as refractory to more traditional therapeutics. Other novel therapies are in the advanced stages of clinical trials or are under review for approval by regulatory authorities.^3–5 The emergence of this array of novel therapeutics, however, also raises a series of questions. Since we currently lack information about which patients are most likely to benefit from which therapies and about the relative effectiveness of these very expensive drugs, there is significant uncertainty around the place that these novel drugs should occupy within the existing therapeutic armamentarium.⁶

These were topics of discussion at the annual meeting of MGNet, held in May 2022 as part of the Myasthenia Gravis Foundation of America’s (MGFA) 14^th International Conference on Myasthenia Gravis and Related Disorders. MGNet is a Rare Diseases Clinical Research Consortium funded by the National Institute of Neurological Diseases and Stroke (NINDS) that is dedicated to the study of myasthenia gravis. All MGNet investigators, NIH program officers, and clinician scientists, who are widely accepted as leaders in clinical research, were invited by the authors to participate in two 75-minute sessions to discuss the most critical needs for clinical trial readiness and biomarker development in the context of therapy development for MG. Here we summarize these discussions with the goal of stimulating discussion and helping to chart a course forward for the field. We welcome any additional comments to be posted at the MGNet website Http://rarediseasesnetwork.org/mgnet.

Therapeutic Questions and Needs

Corticosteroids are broadly recognized as a mainstay of therapy for patients with MG.⁶ They have a relatively rapid onset of action, usually result in a favorable response, and are inexpensive. However, their use is also associated with significant side effects, the burden of which is poorly documented.⁷ Despite their usefulness, as many as 30% of patients may either have an inadequate response to steroids or develop intolerable side effects.⁸ There was uniform agreement among attendees that there is an urgent and pressing need to develop an alternative to corticosteroids for treating MG. An orally administered agent that is fast acting, associated with few side effects, better tolerated, and relatively inexpensive is considered a critically important goal for the field. As a complement to short-acting immunotherapeutics, it was noted that little progress has been made in developing symptomatic therapies. Pyridostigmine is effective in some patients but may be limited by gastric side effects (even with concurrent selective muscarinic antagonist use), highlighting the need for additional research in this area.

The currently FDA-approved complement inhibitors (eculizumab and ravulizumab) and FcRN antagonists (efgartigimod) are welcome additions to the array of therapeutic options for patients with acetylcholine receptor antibody positive generalized myasthenia, but there is considerable uncertainty around how to optimally integrate these treatments into clinical practice. Like intravenous immunoglobulin, the route of administration (subcutaneous or intravenous), and prohibitive cost make these less appealing as first-line agents. Predictors of individual treatment response to a particular therapeutic would greatly facilitate therapeutic decisions, and the development of such predictive biomarkers was widely felt to be an important research priority (see below). Moreover, the potential role (if any) of these agents in seronegative patients or those with purely ocular disease is unknown. Consensus guidelines could certainly help to fill the evidentiary gap and might also address open questions such as whether thymectomy should be offered to patients with purely ocular disease; whether there is a role for combination therapy that incorporates both traditional and novel therapeutic agents; and whether particular therapeutic approaches might be more or less effective based on disease duration at the time of administration. For example, might B-cell depletion be more effective early in disease? Prospective studies to fill these knowledge gaps would also be most welcome.

The yearly cost of eculizumab is $653,100⁹ and ravulizumab is $458,000,¹⁰ while efgartigimod is $225,000¹¹. The very high price of recently approved treatments for MG has brought financial considerations to the forefront of many clinicians’ minds. The Institute for Clinical and Economic Review concluded that both efgartigimod and eculizumab should be priced below $20,000 per year, well below the US prices.^9,11 Given the paucity of data on the costs, benefits and risks of various treatment options for MG, there is an urgent need for studies to address these important health economic considerations.

Clinical Trial Design

There was near uniform dissatisfaction, with no contradicting opinions, with existing clinical trial endpoints – a frustration that is also shared by patients (personal communications and Reference 11).¹² Although important to measure, the MG-Activities of Daily Living (MG-ADL) was thought not to assess the breadth of relevant daily activities. For example, inability to reflect emotions because of facial weakness, which often is a source of frustration, is not assessed. Further, some patients minimize their limitations. One participant described a patient with significant proximal arm weakness, who responded that they had no difficulty brushing their teeth explaining that they rested their forehead on the bathroom counter to avoid raising their arms. As such, the MG-ADL score was considered “not appropriate” as a primary endpoint for phase 3 trials and was considered “a bar too low” for approval of a therapy, especially given the threshold of a 2-point improvement for clinically meaningful improvement.¹³ This view is all the more important given its instrumental role as the primary outcome measure in pivotal phase 3 clinical trials that have led to the approval of new, very costly treatments.

In addition to developing more meaningful patient reported outcomes and standardizing other traditional quantitative measures of disease severity, it was recognized that minimizing side effects and avoiding hospitalization are important concerns for patients.¹⁴ And yet, we have a relatively poor understanding of factors that drive hospitalization and our tools for characterizing the frequency of adverse events are limited. Clinical trials collect extensive data on adverse events, but meaningful signals are often lost in the noise of (overly inclusive) reporting.¹⁵ There is limited data in clinical practice on these issues and the information that does exist is restricted to academic medical centers, which generally adhere to consensus guidelines. (We discuss novel approaches to the collection of data acquired during clinical practice below).

There was also general agreement that MG clinical trials have largely focused eligibility criteria on the population that is feasible to enroll and in whom a therapeutic effect might be demonstrated. While understandable, this leaves unanswered important questions about the therapeutic utility of new agents in populations in need of therapeutic options, but who were not included in pivotal clinical trials. Phase 3 trials, for example, almost uniformly exclude the seronegative population, patients with purely ocular MG, those with unstable disease, and those in whom changes in dose of background immunosuppressive therapy are still being made. Under-representation based on race, ethnicity and socioeconomic status further compound the problem. The generalizability of study results to populations not adequately represented in clinical trials is likely to become even more acute as global development shifts to geographies without access to the most recently approved (and expensive) therapeutic agents.

Innovative trial design has been lacking in MG with all Phase 3 studies using traditional single-agent, parallel-group designs. There was significant interest among attendees in considering a platform trial approach in which multiple therapeutics could be evaluated using a common infrastructure.¹⁶ In addition to the sample size savings of a shared placebo group across treatment arms in a platform trial, randomization could be weighted in favor of active treatment arms making trials more attractive to patients and facilitating faster recruitment. The significant challenge of industry acceptance of such trials, as well as the need for an initial investment in capital to design and launch a platform trial, were acknowledged. The availability of meaningfully effective treatments for MG patients, even if associated with a relatively high risk of side effects, obviously complicates clinical trial design insofar as there is a need to ensure that all patients have access to standard of care.

Registries and Opportunities to Leverage the Electronic Health Record

Several registries have emerged with potential to fill some knowledge gaps. Developed with the support of the MG Foundation of America, the first patient-based but not physician verified registry, the MG Patient Registry, has provided unique insights into gender differences in adverse effects and severity of MG.¹⁷ More recently, pharmaceutical companies have established patient registries to collect data - also leading to a concern for “registry fatigue”, confusion among patients and limited sample sizes. In addition, differential access to these registries based on internet accessibility and outreach to some communities may further compromise broad characterizations of the patient experience. Provider-based registries are challenging because of effort involved in submission of extensive, reliable data, while purely patient registries call into question diagnosis confirmation, unfamiliarity with diagnostic methods and treatments as well as likely recall bias. The MGBase ¹⁸ may provide a physician-friendly platform for global collaboration and has the highly successful MSBase as a model.¹⁹

While the electronic health record (EHR) was built to support clinical documentation and billing, it also holds enormous potential to facilitate research, and to overcome some of the limitations that are inherent to provider-based registries.²⁰ A key challenge in the United States is for us, as neurologists providing care to MG patients, to successfully transition from clinical documentation via free-form text to the systematic collection of structured data at the point of clinical care. This would permit collection of data from a much broader swath of the MG patient population that is more diverse, inclusive, and representative. Indeed, a nascent effort is underway to develop an MG-Toolkit, initially within the Epic EHR system (Verona, WI), but ideally across multiple medical record platforms. The MG-Toolkit aims to standardize data elements that are core to the clinical phenotyping and classification of MG: documentation of medications used to treat MG including dose, duration, tolerability, side effects and therapeutic efficacy, as well as systematic use of widely accepted clinical outcome measures. Such an approach to data collection would greatly facilitate collection of Phase 4 data to help close some of the knowledge gaps that remain at the end of Phase 3 clinical development programs. Other countries with integrated health care systems may be more capable of utilizing their systems to identify adverse effects and therapeutic responses based on reduced hospital admissions or use of other medications for MG.

A role for biomarkers in therapy development and clinical practice

Historically, the biomarker landscape in MG has been dominated by diagnostic markers, including serological (e.g acetylcholine receptor and muscle specific kinase antibodies) and electrophysiological (e.g. single fiber electromyography). Largely lacking, however, are predictive biomarkers, which might identify those patients most likely to benefit from a particular therapy.²¹ Notable exceptions are acetylcholine receptor antibody positivity identifying those who might stand to benefit from complement inhibition or thymectomy; and muscle specific kinase antibody status perhaps predicting those more likely to respond better to B cell ablation. However, even for these treatments, there are those who respond poorly or not at all, and the reasons for this are unknown or unexplored. A surprisingly simple, yet not assessed, biomarker is the impact of body mass and percent body fat, which likely contribute to drug levels, complications, and the underlying autoimmune pathology, which also may impact therapeutic response from concomitant disease or medications. A recommendation was made to collect the body mass index for all research subjects. This was based on its common acquisition in clinical practice and the increasing appreciation of greater body mass index placing individuals at risk for autoimmune diseases.²²

Notably lacking are prognostic biomarkers, which inform the future course of disease (e.g. future mild course of disease or likelihood of future hospitalization). Availability of such biomarkers might help to select those patients in whom a more aggressive treatment strategy from the outset might be more appropriate.

Conclusion

Since the last MG International conference in 2017, there has been a dramatic acceleration of MG research spanning basic understanding of the pathophysiology, greater characterization of clinical and biological phenotypes, increased rigor in clinical trial methodology and ultimately FDA approval of novel medications. The field is privileged to be in this situation because of the intellectual property benefits conferred on rare diseases, a good understanding of basic pathophysiology, and a growing collaborative network of scientists and clinicians across the globe. But we also face a series of challenges that highlight research priorities (Table) and opportunities to enhance not only the development of additional novel therapeutics, but also to empower optimal use of these agents in clinical practice. Success will bring us ever closer to the aspirational goal of precision medicine in which we have the tools to deliver the right drug to the right person with minimal side effects, and at the lowest costs to patients and society.

Table.

Research Priorities

1.	Development of new outcome measures that better capture the patient perspective on clinical status on all aspects of disease and its treatment, including impact of non-oral routes of administration and side effects
2.	Development of biomarkers that fit for purpose with a well-defined context of use
3.	Updating international consensus guidelines for the treatment of myasthenia to help guide clinicians’ use of new (and typically very expensive) therapeutic agents
4.	Development and widespread dissemination of electronic health record tools to permit collection of structured data at the point of clinical care
5.	Generation of health economic data that captures both the direct costs of therapies as well as the potential fiscal savings accrued through a reduction in the frequency and severity of side effects

ABBREVIATIONS

FDA

Federal Drug Administration

FcRN

neonatal Fc receptor

NIH

National Institutes of Health

NINDS

National Institute of Neurological Disorders and Stroke

MG

Myasthenia Gravis

MG-ADL

MG-Activities of Daily Livin