Comment by Dr. Jiwon Oh
NeuroSens survey on sequencing – Part 5
Most RRMS patients begin their treatment journey with a platform therapy, either an injectable or a first-line oral (teriflunomide or dimethyl fumarate in Canada) (see Part 2 of this series). Many will remain on the platform long after the train has pulled out. Only a minority will have their DMT regimen optimized, even when it is recognized that the current treatment has failed (see Part 4 of this series).
When considering treatment optimization, the emphasis is typically on identifying which patients are at risk of early progression of disability. The strategy is then to escalate therapy as needed, in effect titrating the efficacy to minimize the perceived higher risks associated with high-efficacy agents. This is an approach favoured by regulators, who often limit the use of higher-efficacy agents until the patient has failed one or more DMTs. Needless to say, this is not evidence-based: the majority of subjects in phase III trials of high-efficacy DMTs were treatment-naïve, with only the CARE-MS II of alemtuzumab enrolling patients with prior treatment failure.
The strategy of “titrating efficacy” rests on two assumptions. The first is that a majority of individuals with RRMS can make do with a less effective therapy – they can be managed as if they had essentially benign disease, unless they subsequently declare otherwise. This approach is optimistic at best, given the known long-term physical and cognitive sequelae of MS.
Secondly, it is assumed that the clinical/radiological factors at presentation are sufficiently prognostic to stratify mild, “grey-zone” and severe MS. Only 5-10% of RRMS patients are typically considered at high risk of early progression and may be escalated early; the remainder are less of a concern, and are often maintained on the first-choice therapy for a decade or more, until the onset of SPMS.
However, a substantial proportion of patients with milder presentations will, in retrospect, be found to have had aggressive disease. “Aggressive MS” has been variously defined. Using the criterion of time to EDSS 6 within five years of MS onset, an analysis of the B.C. database estimated that 4% of RRMS-onset patients have aggressive MS (numbers corrected to exclude PPMS) (Menon et al. J Neurol Neurosurg Psychiatry 2013;84:1192-1198). A seemingly similar definition – EDSS 6 by age 40 – actually tripled the estimate to about 13%. Menon and colleagues applied the same definitions to the London database and arrived at substantially higher estimates: 9% of RRMS patients reached EDSS 6 within five years of MS onset, and 22% reached EDSS 6 by age 40 (numbers corrected to exclude PPMS) (Menon et al. ECTRIMS 2014; abstract P227). The proportion of RRMS patients meeting any criteria for aggressive MS (including early onset of SPMS) was 27%.
Thus, a high number of patients – perhaps 1 in 4 – will require a higher-efficacy therapy from the outset or will be candidates for an early switch. And among the less worrisome patients, most will become worrisome in time. In an analysis of a contemporary DMT-treated cohort, 55.3% of relapsing-onset patients developed clinically significant disability worsening within 10 years; and 24.2% transitioned to SPMS within 20 years (UCSF-Epic Team. Ann Neurol 2016;80:499-510).
There are three general strategies for managing patients with an inadequate response. Combination therapy has the potential of hitting multiple targets, but has been shown to be ineffective. In the CombiRx trial, an interferon + glatiramer acetate provided no additional clinical benefit compared to either agent alone (Lublin et al. Ann Neurol 2013;73:327-340). Adding teriflunomide to an interferon did not improve ARR (Freedman et al. Neurology 2012;78:1877-1885). The triple combination of interferon-β+ azathioprine + oral predisone in the ASA study, and interferon + methotrexate + methylprednisolone in the ACT study, were no more effective than interferon alone (Havrdova et al. Mult Scler 2009;15:965-976; Cohen et al. Neurology 2009;72:535-541). Natalizumab and rituximab have been used as add-on therapies with an interferon (Radue et al. J Neurol Sci 2010;292:28-35. Naismith et al. Neurology 2010;74:1860-1867). The combination was more effective than interferon alone; the more relevant question, however, was whether the interferon provided any added benefit, which neither study examined.
A lateral switch to another front-line therapy (e.g. an injectable to an oral) is often employed, in keeping with an incrementalist approach to efficacy. This may be reasonable in patients with poor tolerability (Dorr & Paul. Curr Treat Options Neurol 2015;17:354), although it may not be optimal. An MSBase analysis found that subjects who were stable on an injectable showed no disease reactivation after switching to a front-line oral (Spelman et al. Eur J Neurol 2016;23:729-736). The alternative view, however, is that switching produced no additional benefit with respect to relapse reduction or disability progression during the period of a narrowing therapeutic window.
Patients with an inadequate response to a front-line therapy may be considered as the most obvious candidates for escalation to an agent such as fingolimod, ocrelizumab or alemtuzumab. Natalizumab may be less favoured since increasingly its use is limited to rescue therapy for patients with rapidly evolving MS; its utility in sequencing is problematic due to the risk of PML, both during treatment and after discontinuation, and the risk of disease reactivation following natalizumab withdrawal (Giovannoni et al. Pract Neurol 2016;16:389-393).
The benefits of escalation have been demonstrated in Phase III trials using active controls. The annualized relapse rate (ARR) was significantly lower with fingolimod vs. interferon-β-1a IM in TRANSFORMS (0.16 vs. 0.33); with ocrelizumab vs. interferon-β-1a SC in OPERA I/II ((0.16 vs. 0.29); and with alemtuzumab vs. interferon-β-1a SC in CARE-MS I (0.18 vs. 0.39) (Cohen et al. N Engl J Med 2010;362:402-415. Hauser et al. N Engl J Med 2017;376:221-234. Cohen et al. Lancet 2012;380:1819-1828).
Improvements were also seen when interferon-treated subjects were switched in extension studies. For TRANSFORMS, the time to first relapse was doubled when patients switched to fingolimod (Meng et al. Contemp Clin Trials 2015;41:69-74); and the proportion of patients with ongoing disease activity (≥1 relapses and ≥1 Gd+ lesions) was reduced from 13.7% to 3.0% in the year after switching (Cohen et al. J Neurol 2013;260:2023-2032). ARR was reduced from about 0.25 to about 0.10 after switching to ocrelizumab in the interim analysis of the OPERA I/II extension (Naismith et al. AAN 2017; abstract S31.004). Similarly, ARR declined from 0.39 to 0.11 after switching from interferon to alemtuzumab in the CARE-MS I extension (Oreja-Guevara et al. AAN 2017; abstract P5.360).
A matched retrospective analysis by the MSBase group compared subjects with ongoing disease activity during treatment who had switched either to an injectable (interferon or glatiramer acetate) or to fingolimod (He et al. JAMA Neurol 2015;72:405-413). Patients switched to fingolimod had a lower risk of relapse (hazard ratio 0.74) and disability progression (HR 0.53), a higher rate of disability improvement (HR 2.0), and a lower discontinuation rate than those switched to another injectable. A separate study reported that the discontinuation rate was also lower when switching from an injectable to fingolimod rather than to another oral DMT (Korn et al. AAN 2017; abstract P2.397).
A less evident advantage of escalation is the impact of switching on brain volume loss (BVL), which has emerged as one of the most important determinants of disability progression. All three escalation agents have demonstrated significant effectiveness on this endpoint. In TRANSFORMS, the BVL rate was significantly lower with fingolimod versus interferon-β (-0.31% vs. -0.45%) at 1 year; the BVL rate was reduced 50% (to -0.22%) after interferon-treated patients were switched to fingolimod (De Stefano et al. CNS Drugs 2017;31:289-305). The rate of BVL was normalized (< -0.27%) in year 3 of treatment with fingolimod. In the observational MS-MRIUS study, the BVL rate was < -0.4%/year (below the range seen in untreated MS patients) in 58.8% of fingolimod-treated patients in the first 16 months of therapy (Weinstock-Guttman et al. AAN 2017; abstract P4.388).
In CARE-MS I, the median BVL rate was -0.59% and -0.25% in years 1 and 2 of alemtuzumab, and remained low over the next four years of follow-up (-0.17% in year 6) (Traboulsee et al. AAN 2017; abstract P2.104). In the active control arm, the mean BVL rate was reduced from -0.94% and -0.50% during the two years of interferon-β, to -0.01% in year 4 of alemtuzumab (Pelletier et al. AAN 2017; abstract S12.003). In both CARE-MS trials, the median BVL rate was normalized in the year following the second course of alemtuzumab (Traboulsee 2017).
A less robust effect on BVL was seen with ocrelizumab in the OPERA studies (Hauser 2017). There was a -22.8% difference in brain volume change at week 96 favouring ocrelizumab versus interferon-β in OPERA 1; but a non-significant -14.9% difference in OPERA II. When re-baselined from weeks 24-96, a subgroup analysis of early MS patients reported a -25.8% reduction in BVL with ocrelizumab vs. interferon-β, although the BVL rate remained high in the ocrelizumab group (-0.69%) (Traboulsee et al. AAN 2017; abstract P6.338).
The greater efficacy that can be achieved with higher-potency DMTs is generally not disputed. What is more contentious is whether the clinical gains are worth the risks of escalation. This issue will be examined in part 6 of this series.
Comment
Dr. Jiwon Oh: Treatment approaches for relapsing-remitting MS have changed substantially in the last decade due to a combination of factors, including: an increase in available treatment options; accumulating evidence that early treatment optimization results in better clinical outcomes; and the regular use of MRI to monitor subclinical disease activity.
In general, when optimizing treatment in patients with more active, aggressive RRMS, clinicians tend to take one of two approaches: treatment escalation/titration, or initiating a higher-efficacy therapy from onset. Both approaches have different merits and drawbacks. The treatment escalation approach is often favoured by both regulatory bodies and risk-averse clinicians as this approach may be the safer option when considering the potential side effects that can be associated with some of the higher-efficacy disease-modifying treatments (DMTs). With this trial-and-error approach, DMTs that are safer than the most highly-efficacious agents are used initially, as they may be able to control a particular patient’s disease. If the patient fails a specific DMT, then treatment is escalated to a more efficacious DMT.
Another approach that is being utilized more often is that of initiating a higher-efficacy DMT from the onset. This approach has the benefit of rapid control of MS-related disease activity, and minimizing the risk of disability accumulation in both the short- and long-term. However, this approach also carries with it the risk of exposing a patient to the rare, but potentially serious adverse events that are known to be associated with some of the higher-efficacy DMTs, such as progressive multifocal leukoencephalopathy (PML) and autoimmune events.
An important caveat to this discussion is that, in the real-world, treatment approaches can be determined by restrictions related to insurance coverage, rather than a particular neurologist’s recommendation.
Although there is no right or wrong approach to treatment optimization, evidence continues to accumulate suggesting the benefit of early control of MS-related disease activity, particularly in patients with more active, aggressive disease. Long-term epidemiological studies support the existence of a “window of treatment opportunity”, which implies that if MS patients are not treated early, before the accumulation of modest disability, DMTs may not be efficacious to prevent the relentless accumulation of disability that is seen later in the course of the disease. Furthermore, real-world evidence from MSBase and other registries also supports the importance of early treatment optimization.
Taken together, regardless of the particular approach, the importance of rapid treatment optimization in the clinical management of people living with RRMS is apparent. One of the difficulties with implementing this treatment philosophy in clinical practice is the lack of an accurate biomarker of disease activity in patients with early RRMS. At the current time, a combination of clinical and MRI measures is utilized – however, these measures are only modestly predictive of disease activity, at best. The development of better clinical tools that accurately predict disease activity, and enable monitoring of treatment response to specific DMTs, remains one of the greatest unmet needs in the clinical care of MS patients. Such tools would enable true treatment personalization in patients with RRMS, which would ultimately result in preventing disability accumulation, and improving clinical outcomes in patients with this potentially debilitating disease.
NeuroSens survey on sequencing – Part 5