Over the past year, NeuroSens has explored some of the clinical considerations when sequencing therapies in patients with relapsing-remitting MS. The series reviewed the mechanisms of action of disease-modifying therapies in targeting T and B lymphocytes (Part 1), initiating treatment (Part 2), evaluating treatment response (Part 3), treatment planning throughout the disease course (Part 4), lateral switching versus escalation (Part 5), safety considerations (Part 6), and general principles when sequencing therapies (Part 7). The series attracted over 1,000 page-views (as such things are measured online),
With each installment, we surveyed readers about their approach to sequencing, and their perceptions of the injectable, oral and infusion agents currently available to treat RRMS. The following is a summary of our readers’ responses.
While much has been made of the modes of action (MOA) of DMTs, only 17% of survey respondents said that MOA was an important consideration when choosing an initial therapy; 50% said it was somewhat important, and 33% said it was not very important. In part, this appears to be due to uncertainty about the target for targeted therapies: 17% of respondents thought that MS was primarily a T cell-mediated disease, and 0% thought it was primarily a B cell-mediated disease. The largest proportion (42%) thought that the underlying pathophysiology differed in individual patients, while 33% thought that the immune target changed during the course of MS.
As a result, 67% of respondents said their initial goal of therapy was to reduce inflammation, rather than target a specific immune mechanism, reduce relapses, or prevent long-term physical/cognitive worsening. However, 67% believed that a drug’s MOA became more important once it was time to sequence therapies – but primarily so as to avoid limiting subsequent treatment options (58%). As noted previously in this series, there are theoretical concerns about cumulative toxicities with sequential therapies.
Starting treatment
The most common approach to treatment among survey respondents was to start any DMT, reassess in 1-2 years, and plan to switch patients with an inadequate response (50%). The preferred starting therapy at the time of the survey was teriflunomide (33%), with fewer respondents opting for glatiramer acetate (28%), dimethyl fumarate (22%), or interferon-beta (17%). The main reasons: patient acceptance of the regimen (43%), and the availability of long-term safety data (41%), rather than efficacy (6%). This reinforces an observation made previously in the series – that drug efficacy is not clinicians’ prime consideration when starting treatment.
Once a treatment has been started, few clinicians (13%) said they would switch if a patient had relapses alone. Rather, a combination of clinical and MRI findings (73%) would be needed to motivate a switch, which is more in keeping with the Canadian treatment optimization recommendations (Freedman et al. Can J Neurol Sci 2013;40:307-323) rather than the Rio or other criteria. Most clinicians (73%) evaluate treatment response with an MRI obtained annually. However, if a patient has been deemed stable for 5 years, 47% said they would only order another MRI if there were new or worsening symptoms; 40% would continue with annual MRIs.
In practice, “stable disease” in a patient on an initial therapy was variously defined: 31% said it was no clinical or MRI evidence of disease activity (NEDA), 19% said it was no relapses or MRI changes, 6% said it was no new/enlarging MRI lesions, and 6% said it was no EDSS change. However, the highest proportion (38%) said that stability was largely patient-determined – there could be some clinical or radiological worsening, but the patient would be considered stable if he/she remained satisfied with the regimen. Interestingly, these proportions scarcely changed when respondents were asked to define stable disease in a patient on a second-line therapy: 38% would find some amount of disease activity acceptable if the patient remained satisfied.
Moreover, 93% of respondents thought that a majority of their current RRMS patients were stable at two years. For patients on treatment for 2-5 years, 67% of respondents felt that most patients remained stable. For patients on treatment for >5 years, 56% of respondents said they thought a majority of their patients continued to have stable disease.
Switching treatments might be expected in patients with breakthrough disease, however, respondents were fairly evenly split: 56% said they would switch a majority of their patients with breakthrough disease, and 44% said that most breakthrough patients would not be switched. This suggests that breakthrough may be necessary, but is often not sufficient, to motivate a change in therapy.
Thus, patient management appears to be largely reactive, with most patients expected to remain stable. A change in therapy would appear to be more likely if the patient complains of worsening symptoms or poor tolerability. This would help to explain the low rate of switching (11% in one survey [Salter et al. Patient Prefer Adherence 2014;8:971-979]) in clinical practice.
Switching therapies
When a change in therapy is deemed necessary, 50% of respondents said they would switch to a higher-efficacy DMT. However, 40% would raise the issue with the patient, and plan to switch at the next visit if the response continued to be suboptimal, while 10% would switch to another front-line therapy.
When switching patients with breakthrough disease, most clinicians would opt for fingolimod (50%) for a majority of their patients, followed by natalizumab (26%) and alemtuzumab (13%). (Ocrelizumab was not available at the time of the survey.) The choice of agent was governed by clinical need; in effect the goal was to match the incremental worsening of disease with an incremental gain in efficacy. The alternative interpretation is that an incremental worsening of disease offsets the incremental concerns about safety. The most important safety concerns among survey respondents were progressive multifocal leukoencephalopathy (PML) (50%), persistent lymphopenia (33%) and opportunistic infections (17%).
Sequencing
As noted above, a common concern is that the choice of a given DMT will limit future treatment choices. Respondents said the easiest front-line treatment to switch from was glatiramer acetate (67%). When switching from an injectable DMT, the safest next treatment in sequence was perceived to be teriflunomide (56%), followed by fingolimod (33%). When escalating from a front-line oral, the safest switch was perceived to be fingolimod (80%).
For patients sequenced to fingolimod, the primary safety concerns were opportunistic infections (40%), infections (20%) and lymphopenia (40%); since lymphopenia is the mode of action of fingolimod, it may be assumed that the concern was the need for a washout period to allow for lymphocyte recovery before initiating a lymphocyte-depleting therapy, such as alemtuzumab or ocrelizumab. That said, 64% of respondents said fingolimod was the easiest agent to switch from (if additional switches were necessary), compared to 22% for alemtuzumab and 14% for natalizumab.
As for the other high-efficacy agents, the primary safety concern with natalizumab was PML (100%); for alemtuzumab it was secondary autoimmunity (100%); and for ocrelizumab it was long-term B cell depletion (55%) and malignancies (45%).
The first DMTs became available a quarter-century ago, but for the first decade, clinicians were generally skeptical about the merits of therapy. Treatment is now well-accepted, but clinicians appear still to be hesitant about employing the increasingly potent agents that are now on offer. A large proportion of MS patients continue to start on an injectable, and many begin and end their treatment course on one or another front-line therapy. This has been criticized by some as ‘therapeutic inertia’ (Saposnik et al. BMC Neurol 2016;16:58). But it may more closely reflect clinicians’ uncertainties about the therapeutic target, and whether the long-term benefits of treatment will justify the potential for harm. Long-term outcomes and safety data will help to address some of these concerns. A biomarker – of disease activity, of treatment response – would be helpful; and too, a better understanding of the cause(s) of MS.
Comment
Dr. Daniel Selchen: The survey results from the NeuroSens sequencing series offer some interesting insights into the current state of MS treatment in Canada.
It is not surprising that mode of action (MOA) was not an important consideration in treatment decisions, given that 100% of patients were started on the four classes of drugs currently available in Canada for first-line use. There is considerable expert opinion emerging that these agents are not ideal for the 10-15% of MS patients with aggressive presentations. The availability of more potent possibilities for first-line treatment (ocrelizumab, cladribine, ozanimod?) might make MOA and sequencing issues a higher priority.
The observation that efficacy is not the major concern of physicians for the initial treatment is again not surprising, as up to now there has been no opportunity to use high efficacy drugs at onset. More disheartening (to me) is the notion that after initiation of treatment, stability is frequently defined by patient satisfaction. This may account for the very high number of patients felt to be stable both at two years and 2-5 years by survey respondents – which almost certainly does not reflect the proportion who would be found to be stable by looking at clinical and MRI outcomes.
Similarly, the fact that 44% of respondents felt that most breakthrough patients would not be switched seems to reflect a bygone era, when switching generally did not make any difference. This probably does reflect the phenomenon of “therapeutic inertia” which may be a characteristic shared by neurologists and patients in Canada. In patients who are switched, the major consideration in the survey appears to be safety and convenience rather than efficacy.
The survey results suggest that we have made considerable gains in accepting the need for treatment in multiple sclerosis. I suspect that 15 years ago, the proportion of respondents who felt treatment was necessary/valuable would have been much lower. The results also reflect that we still have a long way to go.
Neurologists seem still to be much more conservative than other subspecialties treating autoimmune disorders in tolerating breakthrough disease. In my opinion, this reflects an excessive focus on the risk of treatment, as opposed to the risk of disease. There is increasing evidence that disease risks are very large (e.g. 50% of patients with EDSS 3 are unemployed). There were several “big data” papers at the recent ECTRIMS that suggest that early, effective treatment delays the time to progression/disability significantly (Kalincik et a. ECTRIMS 2017; abstract P732; Casanova Estruch et al. ECTRIMS 2017; abstract P331). There is also emerging evidence that biological markers – particularly serum neurofilament – may offer us the opportunity both to assess disease severity and to monitor treatment effectiveness.
These developments will hopefully lead to earlier, more aggressive first- and second-line treatment of MS patients, either with induction or escalation therapies. But, as suggested by this series, we will need to learn a lot more about the risks (and benefits?) of sequencing and of different treatment strategies.
Click here to read Part 1 in this series
Click here to read Part 2 in this series
Click here to read Part 3 in this series
Click here to read Part 4 in this series
Click here to read Part 5 in this series
Click here to read Part 6 in this series
Click here to read Part 7 in this series