Comment by Dr. Paul Steven Giacomini
NeuroSens survey on sequencing
The year 2013 was the twentieth anniversary of the first published trial of a disease-modifying therapy (DMT) in multiple sclerosis, an occasion marked by the arrival of two new oral agents in an expanding armamentarium of MS therapies. Teriflunomide got the green light in Canada and Europe (previously approved in the U.S.); and dimethyl fumarate (DMF) was approved in Canada and the U.S. (but is still pending in Europe).
Laquinimod, fast-tracked by the FDA four years ago, appears to be idling at the curb.
Later in December, the infusion drug alemtuzumab will learn its fate following a three-thumbs-down review by FDA staff (www.fda.gov/downloads/Advisory), and a mixed blessing from the Peripheral and Central Nervous System Drugs Advisory Committee. The camel-designing committee voted that the trials were too flawed to be able to properly assess alemtuzumab, but decided that there was substantial evidence that the drug was effective, and unanimously agreed that safety concerns should not block approval in the U.S.
More treatment options raise the question of whether traditional approaches need to be updated. There is an urgency to the clinical decision-making since there is evidence to suggest that a DMT will be most effective in the first 2-5 years of treatment (Scalfari et al. Brain 2010;133:1914-1429). So there is a limited time window in which to find the most effective therapy for an individual RRMS patient.
The injectables may be adequate for responders at lower risk of progression. However, long-term use may not be the optimal strategy since opinion is divided about whether the interferons slow the accumulation of disability (Shirani et al. JAMA 2012;308:247-256; Bermel et al. Mult Scler 2010;16:588-596). As for glatiramer acetate, the product label makes no claims about an effect on disability, and a meta-analysis concluded that the drug does not have a significant effect on disability progression (La Mantia et al. Cochrane Database Syst Rev 2010 May 12;(5):CD004678).
In any event, the advent of first-line oral therapies is likely to make injectables a tough sell to newly-diagnosed patients. Teriflunomide and DMF would appear to be reasonable starting therapies, with efficacy demonstrated in the TEMSO, TOWER and DEFINE trials (O’Connor et al. N Engl J Med 2011;365:1293-1303; Miller et al. AAN 2013, abstract S01.004; Gold et al. N Engl J Med 2012;367:1098-1107). However, it is unclear if they will be more effective than the injectables. In the TENERE trial, teriflunomide was non-inferior to subcutaneous interferon beta-1a using the curious endpoint of treatment failure (Vermersch et al. Mult Scler 2013, epublished October 14, 2013). In the post-hoc analysis of the CONFIRM trial, there was no statistically significant difference between twice-daily DMF and glatiramer acetate with respect to annualized relapse rate, new T1 lesions or time to disability progression (Fox et al. N Engl J Med 2012;367:1087-1097).
Plan B
At least 30% of patients started on an injectable DMT will have an inadequate treatment response (Portaccio et al. Eur Neurol 2008;59:131-135; Tremlett & Oger. Neurology 2003;61: 551-554). Many if not most patients will switch treatments or discontinue therapy altogether within the first two years (Wong et al. Can J Neurol Sci 2011;38:429-433). With first-line oral therapies, it is unknown if persistence rates will be higher. The two-year discontinuation rates were 27% in TEMSO, and 31% in DEFINE for teriflunomide and DMF, respectively, and may be higher in clinical practice. Indeed, in the TENERE study, 25% failed treatment, defined as relapse or discontinuation, within six months of starting teriflunomide (Vermersch 2013).
Thus, a substantial proportion of RRMS patients will be candidates for a second therapy. Two switching strategies have been described (Freedman et al. Can J Neurol Sci 2013;40: 307-323). A lateral switch from one injectable to another may produce an objective response; the subjective response, however, may be to refuse another needle. A lateral switch to an oral therapy may be more acceptable to patients, an absence of data notwithstanding. This approach may be more justified for patients who stop treatment due to poor tolerability rather than a lack of efficacy.
For patients with an inadequate response to a first-line injectable or oral agent, the more reasonable strategy may be to escalate therapy. Patients who fail to respond to teriflunomide or DMF would be unlikely to embrace an injectable, so the options would be to continue oral therapy with fingolimod or to initiate natalizumab. While no clinical trials have directly compared fingolimod and natalizumab, an observational cohort study reported that the two treatments had similar efficacy with respect to the proportion of patients who are relapse-free and/or progression-free (Braune et al. J Neurol 2013; epublished September 6, 2013).
A number of open-label studies have indicated that there are clinical benefits when patients are switched from an injectable to fingolimod. In the TRANSFORMS extension, there were significant improvements in the annualized relapse rate (ARR) and MRI lesion number in the 12 months after switching from intramuscular interferon-beta-1a to fingolimod (Khatri et al. Lancet Neurol 2011;10:520-529). In the phase IIIb FIRST study, ARR was reduced 55% within four months of switching to fingolimod (Comi et al. ECTRIMS 2013; abstract P513). The MSBase study reported a 45% reduction in time to first relapse in the year after initiating fingolimod compared to a lateral switch to another injectable; in addition, the discontinuation rate was reduced 78% (Spelman et al. ECTRIMS 2013; abstract P1096). A switch to an oral therapy has also been shown to improve patient-reported satisfaction ratings (Cree et al. ECTRIMS 2013; abstract P1010; Bergvall et al. ECTRIMS 2013; abstract P637; Edwards et al. ECTRIMS 2013; abstract P555). The benefits of escalating from a front-line oral to fingolimod have not been determined.
Treatment escalation to second-line natalizumab has also been shown to be effective in patients with breakthrough disease. A study involving five MS centres found that 86% of patients with highly active disease were relapse-free one year after switching to natalizumab (Putzki et al. Eur J Neurol 2010;17:31-37). A single-centre study reported that relapse rates decreased 77% after switching to natalizumab (Castillo-Trivino et al. PLoS One 2011;6:e16664a). A smaller retrospective study also found that natalizumab was more effective than interferon-beta-1a with respect to relapses, EDSS change and number of Gd-enhancing lesions at 12- and 24-month follow-up (Lanzillo et al. Acta Neurol Scand 2012;126:306-314). An Italian observational study reported that escalation to natalizumab was more effective than switching from one injectable to another; a benefit was not seen at 12 months but was significant at 24 months (Prosperini et al. Mult Scler 2012;18:64-71). Treatment persistence has also been shown to improve after switching from an injectable to natalizumab (Jokubaitis et al. PLoS One 2013;8:e59694; free full text at www.ncbi.nlm.nih.gov/pmc/articles/PMC3602083/pdf/pone.0059694.pdf). This may be due to patient satisfaction with treatment, or the fact that dose-taking is supervised. Treatment cessation in the study was attributed to poor drug tolerability of the injectables. While natalizumab may be better tolerated, there are also powerful incentives to staying on the drug: the difficulties associated with stopping natalizumab, and the limited options thereafter.
As a result of these factors, when considering the long-term treatment plan, natalizumab may be less desirable than fingolimod as a second-line therapy. Fingolimod has the requirement of a six-hour observation period at first dose; thereafter, patients can remain on therapy indefinitely. In contrast, the natalizumab treatment course will be limited in duration due to the risk of progressive multifocal leukoencephalopathy (PML). While PML has been much discussed, it could be argued that the risk remains underappreciated in practice.
Over 400 cases of PML have now been reported, with about 10 new PML cases identified per month. Moreover, in the past few years there has been an “incidence creep”, suggesting that risk-stratification efforts are not adequately reducing the risk. The three known risk factors are duration of natalizumab exposure, prior immunosuppressant use and anti-JC virus antibody status. In the last published analysis, the overall PML risk was 2.1 cases per 1000 (Bloomgren et al. N Engl J Med 2012;366:1870-1880). That risk is now estimated to be 3.3 cases/1000, a 57% increase (Biogen Idec, data on file, September 2013). PML risk in anti-JCV Ab-positive patients, no prior immunosuppressant exposure and 1-24 months of natalizumab exposure has increased from 0.56 to 0.7 cases/1000, a 25% increase. With 25-48 months of drug exposure, the estimated risk in anti-JCV Ab-positive patients with prior immunosuppressant use is now 1.12%. It is unknown if the PML risk will differ in patients previously treated with a first-line oral therapy rather than an injectable.
As a result of safety concerns associated with cumulative drug exposure, the most prudent course is to limit the duration of natalizumab. While patient re-consent is recommended in some countries if continuing natalizumab beyond 24 months, the sufficiency of this strategy enters medicolegal territory that is only now being explored in the litigious U.S. (Weisman R. Boston Globe, September 10, 2013; www.bostonglobe.com).
If natalizumab must be discontinued, a practical consideration is the difficulty in doing so. Any clinical gains may be lost soon after natalizumab withdrawal, according to a meta-analysis of three trials (O’Connor et al. Neurology 2011;76:1858-1865). In the prospective TY-STOP study, the proportion of patients experiencing a relapse in the 12 months following natalizumab discontinuation was 35.6% (Clerico et al. AAN 2013; abstract P01.197). This increase in disease activity may be due in part to the pro-inflammatory effects of natalizumab in the periphery (Frisullo et al. Mult Scler 2011;17:556-566; Kivisakk et al. Neurology 2009;72:1922-1930, free full text at www.ncbi.nlm.nih.gov/pmc/articles/PMC2690969/pdf/6636.pdf). In effect, natalizumab may keep the barbarians at the gate, but meanwhile the barbarians are proliferating.
Maintaining control may be a challenge when natalizumab is withdrawn and the barbarians enter. De-escalation may not be the preferred strategy since many patients will have failed to respond to a prior injectable. Indeed, in the TY-STOP study, 17.8% of patients switched to another DMT or immunosuppressant experienced early relapse (Clerico 2013). Similarly, relapses occurred in 17% of patients receiving monthly pulsed IV methylprednisolone as a short-term bridging therapy (Borriello et al. Eur J Neurol 2012;19:783-787). A three-month course of IVMP followed by glatiramer acetate was shown to be ineffective in reducing disease activity, with 55.5% of patients developing Gd-enhancing lesions within six months (Magraner et al. J Neurol 2011;258:1805-1811).
Fingolimod may be used as rescue therapy but it is not uniformly effective. The French ENIGM survey reported that 22% of patients switched from natalizumab to fingolimod experienced a relapse (Cohen et al. ECTRIMS 2013; abstract P623). However, the mean washout period was 17 weeks, and studies using a shorter washout period have reported better disease control. The relapse risk appears to be highest in the first 8-12 weeks after discontinuing natalizumab (Havla et al. J Neurol 2013;260:1382-1387; Hernandez Clares et al. ECTRIMS 2013, abstract P553). In the TOFINGO study, the proportion of patients free of active T2 lesions after two months of fingolimod was 61.2% with an eight-week washout period, compared to 32.0% with a 16-week washout period (Kappos et al. ECTRIMS 2013, abstract 167). Both the MSBase group and the German Competence Network Multiple Sclerosis group now recommend a maximum two-month washout period when switching patients from natalizumab to fingolimod to reduce the relapse risk (Jokubaitis et al. ECTRIMS 2013; abstract 179; (Havla et al. Ther Clin Risk Manage 2013;9:361-369; free full text at www.ncbi.nlm.nih.gov/pmc/articles/PMC3794889/pdf/tcrm-9-361.pdf).
In the coming year, it is likely that the treatment paradigm will shift in favour of initiating treatment with an oral therapy. For patients who do not achieve an adequate response to a first-line agent, there is a narrow window of opportunity for initiating Plan B. Fingolimod may be the most suitable option for appropriate candidates and has no inherent limitations with respect to duration of therapy once it has been initiated. As with other therapies, duration will be determined by treatment response, adverse effects and patient tolerability/preferences. Natalizumab or an experimental agent would then be kept in reserve for patients who are not candidates for fingolimod, and those who fail to respond or develop clinically significant adverse effects while on fingolimod.
Additional studies are needed to more fully characterize the benefits and risks of different sequencing strategies. Finding the right sequence may take some trial and error. But identifying the most effective regimen is crucial in the first five years post-diagnosis, when treatments may have their greatest impact on long-term outcomes in RRMS.
Comment:
Dr. Paul Giacomini, MD, FRCPC, Associate Director, MS Clinic, Montreal Neurological Hospital and Institute, Assistant Professor, Department of Neurology and Neurosurgery, McGill University, Montréal, Québec: The advent of new oral therapies, coupled with a greater understanding of MS pathophysiology, is certainly reshaping the way neurologists and their patients start and escalate treatment. Defining suboptimal treatment response, as suggested by Freedman et al., provides physicians and patients with criteria they can use to help decide when to consider escalating therapy. When tolerability is the principal reason for switching therapies, certainly ‘lateral’ switching to an alternative first-line agent is reasonable. But persistent disease activity despite first-line therapies should trigger a discussion about escalation. That being said, none of the available therapies are capable of completely suppressing disease activity. Even the superior ability of the second-line therapies, such as fingolimod and natalizumab, to suppress focal inflammation needs to be counterbalanced by concerns regarding long-term safety. Risk stratification strategies, such as JCV index, have been helpful in identifying natalizumab patients at increased risk of developing PML. Furthermore, recent work showing a potential relationship between L-selectin and PML risk (Schwab et al. Neurology 2013;81:865-871) may help further refine risk stratification strategies for patients taking natalizumab. However, the risk of PML will never be able to be fully mitigated. Similarly, despite fingolimod’s acceptable long-term safety to date, there have been a few isolated cases of disseminated herpetic infections, and patients still need to be monitored for other potential safety concerns.
Patient risk tolerance is of paramount importance whenever discussing treatment escalation, as patients may have a different perspective than their physician when it comes to balancing the risks of their illness versus the risks of the medications. Identifying a suboptimal treatment response should always trigger a discussion about escalation with the patient. But MS is a lifelong, chronic illness, and any therapeutic decisions need to respect the individual patient’s risk tolerance over both the short and long term.
NeuroSens survey on sequencing
This poll is currently disabled.