AJMC®: Given the number of drug classes and mechanisms of action available in the multiple sclerosis [MS] treatment spectrum, can you talk about the mechanisms of MS treatment and your process when it comes to matching an agent or drug class to the individual patient?
Leist: When we look at the patient population with MS, in general, there are probably somewhere between 10,000 and 15,000 individuals that are diagnosed with MS every year; a very large majority of patients have been on therapy for some time. Regarding newly diagnosed patients, it is worth noting that they are more likely to be young women. Keep in mind, somewhere between two-thirds and three-quarters of all MS patients are women. At the time of diagnosis, the patient is probably in her early 30s, so there are considerations surrounding pregnancy, family planning that may also come into play.
That said, the average age of patients with MS in the United States with MS is over 50 years and has been on 2 or more MS medications, bringing the issue of treatment sequencing into play. These consideration include the kind of MS medications a given patient has been on, what the individual’s clinical course has been, and also comorbidities and aging immune system. When I consider MS medications, I think of the impact of the medication on the immune system, and I include in the consideration the degree of effectiveness of a given medication. This latter point includes data from clinical trials, registry and other post-approval sources of such information, and clinical experience.
It has become increasingly clear that early in MS, that is, in newly diagnosed patients, there is a time limited window of opportunity, which, if used to institute effective treatment, can alter the disease trajectory and potentially significantly improve the long-term outcome. From this point of view, if we are thinking of the life journey of a patient with MS, these very early months/years after diagnosis are crucial to long-term control of the disease and prevention of avoidable disability. This does not mean that we don’t want to treat the patients with more established MS with effective therapy, but it puts to question an approach based on step wise treatment escalation, as this approach requires to fail or acquire additional, potentially irreversible neurologic injury before going on to a more effective medication.As always, though, when we are older, we have more comorbid conditions that also come into play.
AJMC®: What factors do you weigh when evaluating the potential of a particular agent or drug class when making treatment decisions?
Leist: I often look for studies that include an active comparator that allows me to gauge the relative efficacy, albeit under the constraints of the populations studied in a given trial. I also consider the mode of action, that is, how a medication work is thought to work. Further, I consider what mechanistic class a medication falls into like interferons, S1Ps [Sphingosine-1-phosphate receptor modulator] as within a class. Obviously, the individual products on class may have non-overlapping attributes. Along this line, while they are both monoclonal antibodies, I consider anti-CD20 agents and the anti-VLA-4 [very late antigen 4] agents as different medications. Currently in development are new classes of medications: the BTK [Bruton tyrosine kinase] inhibitors that have current are use in oncological condition. Consideration of the mode of action supersedes consideration of a specific product within that class.
When considering modes of action, I keep close watch on what try to achieve. Normally, I try to achieve disease control in a given patient—no disability progression, no relapses, no MRIs. I want to achieve this with other considerations in mind, such as: If somebody wants to get pregnant, is it appropriate for this person to be immunosuppressed with a medication that is dosed on an ongoing basis?
The prime directive is control the disease at the lowest potential risk costs. Within the interferons, we have high-dose and low-dose interferons, something played out with vigorous discourse during the first 9 to 12 years of the treatment era in MS. There was also a back-and-forth with respect to whether glatiramer acetate is less effective than the interferons. We know that these head-to-head trials ended in an essential efficacy stalemate.
As a clinician, I look at head-to-head trials and my personal experience to see whether my efficacy stratification of agents is appropriate or not. When we are looking at the most recent discussions around anti-CD20 and teriflunomide, or S1P and teriflunomide, there seems to be a titration of outcomes that are either met or not met in the head-to-head comparison. That also gives us some path to consider relative efficacy of agents, as there will never be trials to compare every agent against all the others. The first level for such a comparison for me is comparing a relapse-based outcome. This is followed by disability and MRI-based outcomes. Naturally, I keep in mind that current disability measures may only assess part of the true disability impact. We lack a tool to reliably assess cognitive outcomes in trials.
Looking at these 3 cardinal outcomes, they afford us a base for relative comparisons, at least from a clinical point of view, and establish a stratification. That doesn’t mean the patients who are currently stable on a given medication need to be changed, but it means that there needs to be careful monitoring for disease activity over time. Again, the rationale for these considerations is that time to effective control matters, as a delay in reaching this can mean avoidable disability and its downstream consequences for the patient.
AJMC®: How would you assess selection of agents within the S1P inhibitor class, particularly in light of the fact that the clinical trials differed so greatly in design and populations?
Leist: Focusing at S1P inhibitors as a mode, there is a growing number of choices. As a class, S1P therapies have proven to be an effective treatment approach for relapsing forms of multiple sclerosis. With this background of efficacy, one can focus on the attributes that differentiate the members of this class and which may make one agent preferred over another in a specific patients. The need for a first dose observation versus the availability of a dose titration for most patients is a quite apparent differentiator, but, that said, the first does observation may provide the opportunity for teaching and may invest the patient in the treatment.
On one hand, a shorted half-life permits more ready elimination of the agent when needed, eg, in the case of complications or an unplanned pregnancy, the shorter half-life also requires better treatment adherence by the patient in order to avoid the need for new dose titration.
Availability of data from active comparator phase 3 trials and patient cohorts with select disease phenotypes are of value when considering patients with specific disease attributes. Specificity for select receptor subtypes with newer agents, data regarding white matter and gray matter atrophy, information on effects on neurofilament light levels, novel disability data and fatigue outcomes, are also available for select agents belonging to this mode of action.
AJMC®: What are your thoughts on indirect treatment comparisons when we don’t have head-to-head trials but are looking within classes and are assessing trial data against each other?
Leist: In clinical trials, patients generally get selected on the basis of having some level of disease activity or progression of disability. If you look at a practice, such as the one that I lead, many of the patients are stable on their current treatment and may not have the level of disease activity needed to be eligible for a specific trial. We need to keep in mind that, as much information as clinical trials give us, they may not be completely reflective of the practice experience with that particular agent. In trials, there are patients that are more clinically active than the average patient, but, on the other hand, the very active patients may also not be represented in a clinical trial. Keep in mind that monitoring is important. Because there are never going to be comparisons of everything against everything, we are dependent on other comparisons. Some of these comparisons come out of the experience of large databases, where groups of patients on different agents are compared to each other on the basis of, for example, matched propensity scores. Naturally, there is never a total control of all possible confounders.
There are other approaches that have been taken: looking at system utilization on the basis of payer databases, we can see whether there are differences in the utilization of care and support services—getting canes and wheelchairs— between groups of patients on different agents. If we look at the hierarchy of comparisons, clinical trials provide information on a relatively homogeneous and healthy population, this may not be the case for cohorts entered into community based studies. Nevertheless, both approaches provide important information. Where questions arise is if an agent in the clinician’s hands doesn’t perform at the level reported in community-based studies. Many of these observational studies are driven significantly by the questions that are asked. Nevertheless, there will never be exhaustive comparisons of all agents against each other. Registry-based comparisons and evaluations of other important sources of additional information have a clear role to play.
AJMC®: What are the managed care implications of multiple mechanisms of action, particularly with regard to the nuances that you pointed out when attempting to indirectly compare agents/classes?
Leist: Once a patient has stepped through a given mode of action and had either significant, immitigable side effects or breakthrough disease activity, I view this mode of action as excluded from further consideration as treatment in this particular patient. That’s where a formulary approach of clinical consideration may not necessarily aligned. Frequently, agents within a mode of action may change their position in formularies in subsequent plan years. This may lead to a mandate for patients to step through a medication of a mode of action that they previously failed. It is important that patients are not mandated to step again through a mode of action that they previously failed.
Additionally, certain agent classes, based on their clinical data and mode of action, are viewed as higher efficacy. That doesn’t necessarily mean that every patient needs to start on the highest efficacy medication. There are considerations applicable to an individual patient that may make one medication preferred over another, but in order to achieve higher efficacy these agents need to be accessible to patients on a first line basis. There is a possibility that less effective agents may be viewed as more financially attractive to plans within a given plan year, but patients inadequately treated for years will eventually cost more, both with respect to lost revenue because of transition of the patients to plan products with a lower contribution margin or in costs because of increased system use.
AJMC®: Is there a way that individualized care and population health approaches can work together?
Leist: It is interesting when we look across the spectrum of autoimmune diseases, the American the College of Rheumatology has been relatively clear with respect what efficacy milestone should be achieved in treatment of RA. It is their position that radiographic silence should be achieved of RA in a patient within 6 months of the diagnosis of RA, if I properly understand their guidelines. In RA, as in many rheumatological conditions, disease modifying therapies are added to platform agents. At the start, a more generic agent is initiated, and specialty medications are added subsequently. The goal is clear from the very beginning: radiographic silence of the disease.
In MS, we have additional challenges. The disease is highly variable. Some patients have a milder course of the disease, but at the current time, we do not have tools to identify these patients with certainty at the beginning. At the other end, there are patients with catastrophic disease activity from the beginning. Also, patients transition from the relapsing-remitting clinical phenotype to secondary progressive MS. This transition is often apparent only in retrospect. Similarly, primary progressive MS, where a person never has an attack, may differentiate itself only with hindsight. At the same time, we have observations of agents, like the anti-CD20 medications, which can be used in relapsing and primary progressive patients. Why am I mentioning this? I do think there’s a very significant overlap in the biology in these clinical phenotypes, but we don’t have a fuller understanding for what drives the progressive component of the disease as of yet. Until we have this information, it is very difficult to come to a standard pathway for the treatment of MS.
That said, we have clear guidance; if you look at the last 20 to 25 years since the introduction of treatments in MS, 2 things have emerged: Better control of the disease at an individual patient level leads to a better outcome, and better adherence leads to a better outcome. These are the areas where plans and providers can work together. Early control of the disease may afford a better outcome, preserve the patient in an employed environment, reduce system utilization, and permit individuals with MS to live with as few disease-related impediments as possible.
On the other hand, it also means this patient doesn’t transition to government plans. The patient doesn’t transition into nursing homes. I’m practicing in Pennsylvania where, at this point in time, many of the nursing home products and home health products, such as wheelchairs, have become insured products. If I think of the provider and plan, it is important to keep it in mind the newly evolving population health and how early decisions will potentially affect downstream cost, respectively downstream outcomes, in these patients.