Two questions were at the heart of a recent scientific conference held in Toronto, Canada: How can MS clinical trials be better designed to detect early signals of potential benefit and speed testing, especially for progressive MS? And how can trials be made more inclusive so that there will be information about the safety and benefits for a wider and more diverse range of people when the therapy is approved?
The answers could speed the development of better treatments and pave the way for more personalized medicine to stop all forms of MS. Progress is being made on many different fronts, including among drug regulatory agencies that are expanding possibilities for novel clinical trial designs and are establishing requirements that trials have recruitment plans to include more diverse participants.
The conference was convened as part of the ongoing work of the International Advisory Committee on Clinical Trials
, which is jointly sponsored by ECTRIMS
and the National MS Society. Conference co-chairs Ruth Ann Marrie, MD, PhD (University of Manitoba), Jeremy Chataway, MD, PhD (University College London) and Maria Pia Sormani, PhD (University of Genoa) and others led discussions and are working toward publishing recommendations for pushing this work forward in the global MS community.
NOVEL CLINICAL TRIAL DESIGNS
Most often in MS, therapies are tested one at a time. Typical testing involves stages of phase 1 (testing for safety), phase 2 (testing for safety and hints of effectiveness), and finally phase 3 (testing for safety and more robust signs that the therapy has benefits). Only after phase 3 trials can a therapy be evaluated for marketing approval by drug regulators. These steps can take decades, with no guarantee that the experimental therapy will work or will be approved.
One issue particularly in progressive MS is that it usually takes a long time to detect changes that may indicate benefit of a therapy. A measure that has been used is brain shrinkage, or atrophy. According to conference participants, detecting slowing of brain atrophy takes a minimum of 18 months. Research is under way to develop blood tests or other biomarkers that can detect benefit more quickly so that trials needed for approval can begin earlier.
Marcus Koch, MD, PhD (University of Calgary) explained that “futility” trials are a way to screen out unpromising candidate therapies quickly in phase 2 trials. These trials can use small numbers of participants and no control group. Outcomes of the treated participants are compared to what is known about the “natural history” of how groups of people are likely to progress if they did not receive an effective treatment. Dr. Koch’s team has applied this method to test generic therapies that might be repurposed for progressive MS. Promising candidates would then merit more rigorous testing.
Testing many therapies at once:
Dr. Chataway discussed novel clinical trial designs that combine stages and test multiple therapies at once – called multi-arm, multi-stage trials (MAMS) that can adapt as they go along. This design can speed the identification and testing of promising therapies. All of the treatment arms are compared to one control group. The control group is also treated according to standards of care, so no one is on placebo.
At specified times during the trial, researchers may drop a therapy that doesn’t seem promising. People in that arm may be re-assigned to a new treatment arm. Promising-looking drugs stay in the trial, with more people joining the existing participants. In this way, what would normally be two or more consecutive trials are delivered in one, saving time and money.
These types of trials are new to the MS field but have been adopted for other conditions such as cancers. A large trial like this is getting underway in the United Kingdom, called the OCTOPUS trial
, which focuses on progressive MS. Presenters Susan Scott and Judy Beveridge (individuals with MS who are engaged with designing OCTOPUS), and Emma Gray, PhD (UK MS Society) described how this trial was designed with ongoing input from people with MS. Incorporating viewpoints of people with MS in trial design was an important theme emphasized throughout the conference.
Dr. Peter Feys (University of Hasselt, Belgium) noted the complexity of designing rehabilitation trials. In the past these trials have generally been small and their results were sometimes hard to interpret. One challenge is that being involved in a trial itself is motivating and can alter behavior, and so can coaching and the bond that develops with the rehab therapist. Care must be taken in choosing the control intervention, and there are ethical considerations in having people in the control group wait for something that may help them. Some other variables among participants, such as dosing, intensity, and practice may also change outcomes of a trial. He noted that there are lessons to be learned from the stroke rehabilitation community.
Non-traditional trial statistics:
In most MS clinical trials, statistics help determine if the differences between those arms are unlikely to be explained by chance alone. The statistical framework most often used is called a frequentist approach, which at the outset assumes that the differences are random unless they reach a specific threshold of significance.
Using alternative statistical frameworks could save time and reduce the number of participants needed. One is called Bayesian statistics. Dr. Sormani explained that this approach incorporates prior information about the question, such as previous trial outcomes or rates of progression, and the result gives you the probability that your idea (that the therapy works) is correct.
She noted that this approach may be especially useful when it would be hard to recruit the numbers needed for a more conventional trial, such as the relatively rare pediatric MS. Conference participants noted that drug regulators are beginning to warm up to the use of the Bayesian framework for clinical trials, but more work is needed to normalize its adoption.
Registries and “Pragmatic” trials:
There is a growing list of studies that enroll people who are seen at MS centers and consistently track their health status, visits, treatments, treatment switches, and other factors over time. These are called registries. Pragmatic trials are designed with specific questions in mind and people are randomly assigned to receive a particular type of therapy. They can answer more general questions about care, such as is there a time when it is safe to discontinue therapy?
Several presenters described pragmatic trials in MS (funded through PCORI), stroke and COVID. One advantage of these trials is that they incorporate normal health visits and standard monitoring so some of their costs are covered by healthcare systems or insurers. These studies more closely reflect real-world conditions and the effectiveness of therapies in a broader range of people than standard drug trials.
Fredrik Piehl, MD, PhD (Karolinska Institutet) described a trial called COMBAT-MS
, which uses the Swedish MS registry and Kaiser Permanente Southern California registry to compare the safety and effectiveness of rituximab to all other available MS therapies. Ellen Mowry, MD (Johns Hopkins University) described the TREAT-MS
pragmatic trials that are comparing outcomes in people who are started on latest generation, more powerful therapies versus people who are started out with less risky, more moderately effective therapies and escalating from there if needed.
Conference attendees discussed what treatment questions would be best answered through pragmatic trials These include symptom management studies, combining standard treatment with a rehabilitation strategy, comparing models of care, and looking at the impact of comorbidities (health conditions in addition to MS) on progression. Gaining longer-term consent from participants and regulatory boards and establishing minimum standards of data gathered by all MS registries would increase the ability to use and combine patient registries to answer questions.
Dr. Sormani and David Kent, MD, MSc (Tufts Medical Center) explained that most trials measure average responses to treatments, combining results of people who may respond very well with people who do not. It is difficult to know who would best respond to the treatment but this is a key clinical question. Dr. Kent noted that looking at outcomes from subgroups of individuals would require many more trial participants.
Dr. Sormani discussed new statistical methods that look at characteristics of participants at their entry into the trial. This could help to define in advance individuals who are expected to respond to a therapy in different aspects of function (such as walking or cognition), rather than obtaining averages.
To determine if a therapy works, measures of effectiveness are established before the trial begins. Outcome measures are often stand-ins (surrogates) for the true goal of the therapy. For example, step counts are a stand-in for the goal of better mobility. MRI scans can detect whether there are fewer active MS lesions in the brain, as a stand-in for slowing or stopping the disease process. Surrogate measures can give hints of effectiveness in both phase 2 and phase 3 trials, and also can be planned to give hints before a trial ends. In phase 3 trials, regulators usually expect to see clinical outcomes as well – such as fewer relapses or better functioning. There is a need for better outcome measures that can more quickly and reliably predict benefits.
Panelists and participants discussed emerging outcome measures such as more refined and focused imaging techniques to detect nerve damage, atrophy, and repair. Gary Cutter, PhD (University of Alabama, Birmingham) discussed some statistical issues that may arise when combining measures or using too many, which may dilute the outcomes. There was also discussion about how to establish the minimum number of participants needed when using specific outcome measures, and how to establish a minimum threshold for a particular outcome that would predict benefit.
Another panel focused on outcomes in rehabilitation trials. Optimally these should be established based on the deficit being targeted, noted panelist Marcia Finlayson, PhD, OT (Queen’s University), but this can be complex. Even measuring the engagement and participation of people enrolled in trials is difficult and often overlooked, noted panelist Ulrik Dalgas, PhD (Aarhus University). Using wearable activity trackers could help with that. These devices could also detect adherence to a program, as well as the common phenomenon of people in a control group increasing their activity levels just because they were enrolled in a trial.
Several participants mentioned a useful measure being incorporated into rehabilitation trials, the Canadian Outcome Performance Measure
. This surveys an individual’s opinion about their own functioning in many areas of daily living.
HOW TO INCREASE BROAD PARTICIPATION IN CLINICAL TRIALS
Typical clinical trials are very specific about who may participate, such as specifying a type of MS, types or frequency of disease activity, and age ranges. They often exclude people who have other health conditions. Trials indirectly exclude people if the trial is only conducted in large cities, or if in-person visits must be done during typical work hours and it would be difficult for some to take off from work. In addition, there is typically low participation among people of color due to institutional distrust, or due to geographic regions with low population diversity.
For these and other reasons, when a therapy is eventually approved, we are often left not knowing how a wider range of people with MS may benefit from it, or if there might be additional safety issues.
MS is not a white person’s disease:
Mitzi Williams, MD (Joi Life Wellness Group Multiple Sclerosis Center, Newnan, Georgia) presented on growing evidence showing that MS is much more prevalent among Black people than previously believed. Despite this, there is low enrollment of Black people in MS trials, and the ancestry, race and ethnicity of trial participants are rarely reported for phase 3 clinical trials. Dr. Williams described recent advances including the establishment of the National African Americans with MS Registry (NAAMSR
), the Alliance for Research in Hispanic MS (ARHMS
), and the successful enrollment of the CHIMES trial
, the first MS trial to focus on Black and Hispanic people.
Lilyana Amezcua, MD, MS (University of Southern California) presented on how social determinants of health – such as economic stability, employment, income, expenses, neighborhood, education, food and other factors, combined with access to medical care – impact people’s overall wellness and their MS. This is especially true for people who are Black or Hispanic/Latinx, who generally experience worse outcomes from MS than people who are white. Yet most studies do not collect or report data on participants’ social determinants of health.
Barbara Bierer, MD (Harvard Medical School’s Multi-Regional Clinical Trial Center) focused on improving the rigor of global trials. She noted that trials should reflect the population affected by the condition under study. Her team has developed toolkits to enhance recruitment strategies. They suggest recruiting from within communities, and not in one central location, while considering feasibility, health literacy, site selection and available support. Presenters Elena Hernandez Martinez de Lapiscina, MD, PhD (European Medicines Agency) and Jennifer Panagoulias, RAC (Angelman Syndrome Biomarker and Outcome Measure consortium) noted that drug regulatory agencies are evolving. They are now requiring diversity plans for trials and rationales for why a trial has restrictive eligibility criteria (such as excluding comorbidities or having upper age limits).
Lack of diversity in rehab trials:
Dr. Finlayson discussed literature reviews of rehabilitation trials, revealing a systemic lack of effort to recruit more broadly, and few reports of characteristics of participants. Many trials recruited from only one source, usually an MS clinic. Depending on the intervention being tested, many trials exclude people with greater disability, older age, and cognitive impairment. Other impediments include living in a rural area, being unable to pay for study-related costs, and being unable to get to the study site. Dr. Finlayson said it would be helpful if researchers would report on recruitment methods that worked and didn’t work, and if funders and publishers would set expectations for reporting participant characteristics.
Conference participants discussed ideas for increasing diversity in trials. One idea is to run a smaller, parallel trial that includes participants excluded for age or comorbidities, or a more diverse population not available in the location of the main trial. Other ideas involved reimbursing for expenses such as childcare, tapping peer group influencers, translating materials, using video instead of just print, using wearable sensors and televisits, having a trial “buddy” to help with communication and compliance, and taking blood samples and doing safety evaluations locally.
The conference organizers are evaluating discussions and the thoughtful feedback from participants. They plan to publish recommendations for strategies to improve clinical trial designs and to enhance trial diversity. These will help guide researchers, regulators, and funders seeking to find more effective therapies and rehabilitation approaches for all types of people who are living with MS.