This week I did a Digesting Science with children/young adults who had MS/Demyelinating disease about the impact of COVID-19 on DMT that was organised by Alison. I learned that I am useless at Origami and my Immune Cell was rubbish….but I’ll let Alison explain all that. Children with MS need treatment, so what should we use and how should we test agents?.
I have lifted this debate from the Multiple Sclerosis Journal. I thought it may be a bit of light relief from COVID-F-ing-19. If you are a parent or someone with MS, especially if you are a younger reader, what are your thoughts this issue?
The ethics of conducting randomized controlled trials of disease-modifying therapies (DMTs) in paediatric MS is debated in light of currently available data. Approximately 5000 paediatric patients have MS worldwide and their access to DMT approved for adult MS varies by country. Recognizing the gap in children care, regulatory agencies have mandated the conduct of trials to approve drugs under the age of 18 years including in paediatric MS; however, agencies have modelled their trial design requirements to those of the adult MS field (i.e. phase 3 randomized clinical trials (RCTs) with clinical endpoints).
The three RCTs in paediatric MS that have been completed or are about to be completed encountered similar challenges, the top one being the very prolonged time to complete enrolment (about 5 years) despite modest sample sizes (150–200 subjects) and high numbers of participating sites. This highlights how challenging it is to conduct RCTs in paediatric MS. While in the first RCT, prolonged enrolment was rewarded by drug approval in children 10 years and above, approval was in fact delayed by 8 years in children compared to adults, and children under 10 years of age were not studied. Knowing this, can we continue to argue that RCTs of DMT approved in adults are ethical to conduct in children with MS?
Most physicians who care for paediatric MS patients agree that determining the appropriate dose and acquiring safety information in young patients are more critical than determining drug efficacy. As such, an international consensus on clinical trials in paediatric MS recommended performing pharmacokinetic/pharmacodynamic and extension studies to inform on key data needed for clinical management of young patients. In fact, the international group recognized that safety was not appropriately evaluated in trials of 100–150 paediatric patients with MS receiving the drug under study for 2 years; hence, recommendations were made for long open-label extension studies (at least 5 years) following RCT or the creation of registries to capture long-term side effects/toxicities that are less common and would be missed in a 2-year trial.
In the past 2 years, new information coming from RCT (PARADIGMS) and large observational studies in children with MS has further questioned the ethics of performing RCT in children for drugs approved in adult MS. As these studies have confirmed comparable DMT effectiveness in children compared to adult studies, the rational of conducting RCTs with clinical endpoints in children with MS has become weaker, especially because typical designs use placebo or low-efficacy DMT for the control arm in order to lower the sample size. The international study group has clarified that the use of a placebo for trials of agents approved for relapse prevention in adult MS is not ethical. The use of low efficacy control treatment also likely would not reach equipoise in most cases.
Efforts are underway to discuss international recommendations for paediatric MS trials with regulatory agencies so future studies of MS drugs more effectively identify short- and long-term safety and appropriate dosing in children, especially under 10 years of age.
Two-armed Active Comparator Trials Are Unethical in Paediatric Multiple Sclerosis
Most medicines used by children internationally are unlicensed or off-label, with no randomized control trials (RCTs) data in more than 50% of interventions used in children compared with adults. Despite the growing therapeutic options for adults with relapsing remitting multiple sclerosis, few of these treatments have been specifically studied in paediatric multiple sclerosis. Until recently, children with multiple sclerosis have been treated based on extrapolation from adult clinical trials on the assumption that multiple sclerosis mechanisms and pathobiology are shared across the age span. However, children absorption, distribution, metabolism, and excretion of drugs differ from adults. Off-label medicine use in children and adolescents may lead to under- or over-dosing and poor adherence. Extrapolation from adult trials is not always appropriate since other variables may need to be considered, such as growth, cognitive development, and immune system maturity. Even the assumption that the pathobiology of the disease is the same can be questioned, some studies suggesting a distinct cellular profile in younger children with prominent activation of the innate immune response, as opposed to the typical activation of the adaptive response in older patients.
It is only just that treatment decisions in paediatric multiple sclerosis are guided by good clinical evidence. RCTs yield an estimate of the effect of an intervention that is unbiased and consistent and thus provide more reliable evidence than observational studies which are susceptible to selection biases and other confounding factors. Ethical decisions need to take into consideration the specific research questions, type of intervention, and population. A specific study design cannot be considered unethical in general but each of the study designs can be unethical in specific circumstances (related to the intervention, the study population, etc.). For instance, placebo-control RCTs are acceptable when there is no standard treatment, the net therapeutic advantage of the standard treatment has been questioned, or effective treatment exists, but is not available due to high cost or short supply. If none of these can be assumed, the use of placebo would be considered unethical and standard of care should be used as a comparator. Besides, two-armed comparator RCTs may be a useful tool to test many interventions, including non-pharmacological interventions (i.e. physiotherapy, cognitive rehabilitation, among others) and thus this study design cannot be considered unethical per se.
Setting up an RCT is challenging in pediatric multiple sclerosis mainly due to the small patient’s population and the under-developed research infrastructures, but these challenges do not make RCTs unethical. On the contrary, there are several arguments to support that we should be gathering evidence from such trials specifically in the paediatric population instead of making treatment decisions based on adult trials. The biological arguments include not only the fact that children are not “small adults” as suggested above but also that different clinical outcomes in children may be relevant compared to adults. For example, outcomes heavily relying on motor function, commonly used in adult multiple sclerosis trials, may not be useful in pediatric onset multiple sclerosis where patients take longer to reach irreversible motor disability milestones whereas cognitive outcomes may be more relevant. Moreover, specific outcomes such as days off school, missing out on social activities, or response to immunization are not considered in adult multiple sclerosis trials and should be considered in children. Second, the inherent nature of an RCT, where the study population is not randomly assigned, may preclude generalization/extrapolation.
Children experience two- to threefold more frequent relapses than adults early in the disease course and a higher rate of lesion accumulation on brain MRI than adults. Overall, this makes placebo-controlled RCTs of immunomodulatory agents proven to be beneficial in adults (or to have shown some effect in small groups of children) inappropriate. Two-armed comparator trials are then natural alternative to study a disease modifying treatment effect. It reduces risk to individual patients and thus enhances study enrollment.
Despite all the difficulties and fears regarding the logistics of setting up a two-armed active comparator trial (which requires larger sample sizes than placebo-controlled trials) in the relatively small cohort of children with multiple sclerosis, a recent trial, comparing intramuscular interferon beta-1a with oral fingolimod, has proven that this type of trial design is feasible in this population. In this trial, it was possible to document efficacy with a modest sample size. Furthermore, sample size issues can potentially be tackled with different strategies: by using different primary outcomes such as MRI endpoints or different clinical outcomes (i.e. time to event instead of the annualized relapse rate),by reinforcing international collaborations, and, for drugs yet not approved in adults, by including children above the age of 12 years old in adult RCTs.
To summarize, ethical considerations should apply to specific research questions, type of intervention, population, and not a specific study design. Two-armed comparator trials should be an available option for obtaining good quality unbiased evidence to help guide treatment decisions in pediatric multiple sclerosis, especially when biological factors or specific clinical questions cannot be extrapolated from adult trials.
Two-armed Active Comparator Trials Are Unethical in Paediatric Multiple Sclerosis
There is a clear need for guidance that is specific to the treatment decisions for patients with pediatric multiple sclerosis (MS). Controlled clinical trials are clearly the gold standard for determining medication efficacy, including in special populations. However, with the current status of MS clinical trial design, we argue that conducting two-armed comparator trials in paediatric MS is unethical.
Paediatric MS patients represent a vulnerable population. Medical ethics dictate that research must have beneficence, requiring investigators to have the welfare of the research participant as the goal of the research. Clinical trial designs in MS therapeutics traditionally compare a new treatment to a standard treatment. With the design of new treatments to improve care, there is a high degree of certainty that the standard treatment may be inferior. Any inferior treatment may be associated with increased relapse, an adverse event that is an unacceptable risk.
MS relapse has direct physical and emotional affects and may alter the disease course, particularly in the context of ongoing neurodevelopment. Relapses in pediatric MS lead to missed school days, motor and gait impairments requiring physical or occupational therapy, and have been associated with slowed cognitive processing. The adverse psychological impact of MS relapses has been well-documented in adults and which may be amplified when experienced in children and adolescents, as well as impacting their families. Those with paediatric MS randomized to MS therapies less effective in controlling relapses experience (e.g. as in the PARADIGMS STUDY) have lower levels of self- and parent-reported quality of life.
We remain in critical need of therapeutic advances, and therefore must consider alternatives to clinical trials in paediatric MS. In this special case, we suggest that evidence supports the reliance instead on the clinical trials from adult samples. Treatments established in adult relapsing remitting MS samples can be confidently extended to those with paediatric onset, given the similarities between disease features and course. Furthermore, the effect of age on the course and the treatment response in paediatric MS can be effectively modeled from data available in adult relapsing–remitting multiple sclerosis (RRMS) clinical trials. Indeed, referencing data from adult clinical trials are consistent with the recommendations by the International Paediatric Multiple Sclerosis Study Group (IPMSSG), which specifically advises trials to take advantage of prior information about these medications in the adult population and to extrapolate from the adult data to the pediatric age group.5
For example, given that adult MS clinical trials in RRMS typically enroll patients from age 18 to 55 years and pediatric MS trials enroll patients aged 10–17 years, it is reasonable to extrapolate paediatric MS outcomes from young adult MS data. A Baysesian meta-analytic predictive approach for annualized relapse rate (ARR) for paediatric MS based on three historical MS trials (FTY720 Research Evaluating Effects of Daily Oral therapy in Multiple Sclerosis (FREEDOMS I), FREEDOMS II, and Trial Assessing Injectable Interferon vs FTY720 Oral in Relapsing–Remitting Multiple Sclerosis (TRANSFORMS)) led to results that were very close to the ARR observed in the paediatric MS clinical trial, PARADIGMS.
Another alternative to conducting controlled two-armed comparator clinical trials in paediatric MS is to utilize real-world observational data. Using observational data from a number of large cohorts of adult MS patients, several studies have demonstrated that high efficacy disease-modifying treatment (DMT) for adults with MS when initiated at the start of treatment as opposed to an escalation approach lead to better long term outcomes. Is there similar data for paediatric MS? While the number of large databases including long-term follow-up for pediatric MS patients are limited, the US Network of Paediatric MS Centres is one example for which treatment outcomes with different DMTs were effectively compared. For example, in an observational registry study of n = 744 pediatric MS patients started on DMT, investigators established that newer (orals and infusions) DMTs were superior to older (injectable) medications, a finding consistent with the PARADIGMS study.
Does this mean that no active comparator clinical trial should be conducted in a paediatric MS sample? No. In fact, there are some trial designs that are reasonable to consider and ethically sound. For example, studies comparing two highly efficacious agents such as in a non-inferiority study are reasonable. In this design, a medication shown to be highly effective in controlling MS relapses such as a B cell depleting therapy is compared to fingolimod, already shown to be highly effective against paediatric RRMS.
Another acceptable design approach would be to enroll adolescent MS patients along with adults in clinical trials that appear to be safe. For example, in the trial of satralizumab in neuromyelitis optica spectrum disorder, adolescents aged 12 years and older were enrolled as were adults. Studies in epilepsy also have successfully enrolled adolescents along with adults in Phase III clinical trials. However, as recommended by the IPMSSG, we must perform pharmacokinetic studies that take into account the lower weights of children and adolescents. Further we should not neglect the safety of DMTs in the youngest children with MS, those less than 10 years of age.
In summary, controlled trial designs to evaluate MS treatments in which one effective agent is compared to one highly likely to be inferior currently pose an unacceptable risk for our pediatric patients. Alternatives include extrapolating data from young adults and utilizing prospective observational designs to inform treatment choices. We can perform open label safety studies and carefully document the clinical outcomes in those with paediatric MS of all ages. Future designs comparing a new treatment to a higher efficacy comparator may be more acceptable, and also designs may extend to younger ages to include adolescents as an approach successfully employed in other central nervous system (CNS) disorders. While imperative to move forward, we need to continue to carefully balance the risks to our patients with vigilance to ensure their wellbeing across our research endeavors.
Now will say the views of the debators may not represent their personal views and they may be taking a certain stance for the purposes for debate.
What do you think would be done with dimethyl fumarate and ocrelizumab as an examples of ongoing trials in paedeatric MS