When switching from one disease-modifying therapy (DMT) to another in patients with RRMS, consideration must be paid to the potential short- and long-term consequences of administering drugs in sequence.
As a general rule, drug-drug interactions are of lesser concern than the pharmacodynamic effects of sequential agents on the immune system, and the potential for cumulative toxicities (see Part 3, Comment by Dr. Paul Giacomini).
Among the injectable agents, the longest half-life is seen with subcutaneous peginterferon beta-1a (range 2-6 days), suggesting that a 2-4 week washout period would be sufficient when switching from any injectable to another agent. Dimethyl fumarate is metabolized before it reaches the systemic circulation; the half-life of its active metabolite, monomethyl fumarate, is about one hour.
In contradistinction is teriflunomide, which requires 8-24 months to clear the drug. Accordingly, an accelerated elimination protocol is recommended prior to initiating another DMT. The regimen is oral cholestyramine 4-8 g tid, or activated charcoal 50 g bid, for 11 days. The disadvantage of such a protocol is that it can be perceived by clinicians and patients as onerous. The advantage is the assurance that the drug has been eliminated within two weeks, allowing another treatment to be initiated within a reasonable time frame.
The main toxicities of interest with front-line therapies when considering sequencing are lymphopenia and hepatotoxicity. A CBC with lymphocytes, and liver function testing, may be advised prior to initiating another agent. The general recommendation is to allow lymphocytes and liver enzymes to normalize before starting the next drug in sequence.
A concern is persistent lymphopenia, defined as ALC < 0.5×109/L for >6 months, which has been reported to occur in 2.5% of patients on DMF for >6 months (Fox et al. ECTRIMS 2015; abstract P606). The incidence outside of the clinical trial setting has been reported to be higher. A single-centre analysis reported that 17% of DMF-treated patients developed Grade >2 lymphopenia that did not resolve unless treatment was discontinued (Longbrake et al. Mult Scler J Exp Transl Clin 2015;1:pii: 2055217315596994). A separate study found that 13% of patients discontinuing DMF due to low lymphocyte counts had persistent lymphopenia >6 months despite no subsequent exposure to another DMT (Winslow et al. ECTRIMS 2016; abstract EP1520). Similar results have reported at other centres (Baharnoori et al. ECTRIMS 2016; abstract P724; MacDougall et al. ECTRIMS 2016; abstract P624).
Prolonged lymphopenia is believed to be associated with an increased risk of progressive multifocal leukoencephalopathy (PML) (Khatri et al. Mult Scler Relat Disord 2015;4:377-379), which has been reported in a small number of patients treated with DMF or fingolimod. However, ALC appears to offer only a partial explanation. Two PML cases have been reported in psoriasis patients without severe lymphopenia who had received other formulations of DMF (Nieuwkamp et al. N Engl J Med 2015;372:1474-1476; Hoepner et al. Neurol Neuroimmunol Neuroinflamm 2015;2:e85-e85). One suggestion is that ALC may not adequately reflect the effect of DMF on immune subsets (Balak & Hajdarbegovic. N Engl J Med 2015;373:582-583). Indeed, a retrospective study reported that 9% of DMF-treated MS patients had severe CD8+ lymphopenia despite an ALC >0.8×109/L; the authors suggested that CD8+ count may be a better predictor of PML risk in patients taking fumarates (Spinelli et al. AAN 2017; abstract P5.372). Whether the same applies to other DMTs is unknown. The implication is that caution is needed when switching from DMF to fingolimod, which reduces lymphocyte count as its mode of action, or to natalizumab, since DMF may constitute prior immunosuppressant use, a known PML risk factor.
Current European guidelines recommend that DMF discontinuation be considered in patients with persistent lymphopenia (www.ema.europa.eu/docs/en_GB/document_library/Press_release/2015/10/WC500196017.pdf). However, the time to lymphocyte recovery following DMF discontinuation has not been established (Tecfidera product monograph. Biogen Canada, February 17, 2017). Since another DMT should not be initiated until lymphocyte counts normalize, there may be a considerable delay before therapy can be reinstituted, during which the patient will be at risk of renewed disease activity.
Switching to a second-line agent
With natalizumab, there is a risk of PML during treatment and after switching to another agent. For example, most cases of PML with fingolimod have been attributed to prior natalizumab exposure (Putzki N. ECTRIMS 2014; abstract FC3.1). The one PML case reported with ocrelizumab was also associated with prior natalizumab use. Whether prior treatment with lymphocyte-depleting agents, such as teriflunomide or DMF, contributes to PML risk during natalizumab exposure has not been determined.
When switching from a front-line therapy to fingolimod, the two adverse events of interest are cardiovascular effects at first dose, and infection risk. In the phase IIIb FIRST study of fingolimod in real-world use, 85% of patients had received a prior DMT, such as an interferon (70.3%), glatiramer acetate (30.9%) and natalizumab (10.5%) (Gold et al. J Neurol 2014;261:267–276). The incidence of second-degree atrioventricular block was 0.9% in patients without a pre-existing cardiac condition; all conduction abnormalities were asymptomatic. This incidence falls within the range of second-degree AV block reported in the pooled analysis of phase III trials, and the observational START study of cardiac safety (range 0.2-1.5%) (DiMarco et al. Mult Scler Relat Disord 2014;3:629-638; Limmroth et al. ECTRIMS 2016; abstract P726).
These findings suggest no increased risk of cardiac abnormalities when switching from a front-line therapy to fingolimod. It should be noted, however, that there is a low risk of hypertension (4.2%) with teriflunomide (Aubagio product monograph, Sanofi Genzyme, January 25, 2017), so blood pressure monitoring may be advised after accelerated elimination of teriflunomide and prior to the first dose of fingolimod.
The overall safety of switching from a DMT (an injectable, natalizumab or mitoxantrone) to fingolimod was examined in an analysis of the Spanish registry (Meca-Lallana et al. ECTRIMS 2016; abstract P1170). A total of 13.5% of patients experienced an adverse effect; the incidence of severe adverse events was 1.6%. In the MSFIRST substudy of the MSBase registry, there was no significant difference between fingolimod and interferons/glatiramer acetate with respect to the incidence of serious adverse events, including infections (relative risk 2.06), herpes zoster (RR 1.0), non-melanoma skin cancers (RR 0.79), or malignancies (RR 1.47) (Haartsen et al. ECTRIMS 2016; abstract P704).
LONGTERMS data indicate that the long-term risk of infection (incidence rate 65.8 vs. 92.0 in the core studies) or serious infections (IR 1.0 vs. 1.1 in core studies) declines slightly with continued fingolimod exposure (Cohen et al. AAN 2015; abstract S4.0006). There is a higher rate of herpes zoster (2.0% vs. 1.0% with placebo in controlled studies) and varicella zoster virus (VZV) infections (11 vs. 6 per 1000 patient-years with fingolimod), but these rates have not increased with cumulative drug exposure (Arvin et al. JAMA Neurol 2015;72:31-39). Vigilance has been advised for serious opportunistic infections, including PML and cryptococcal infections, most notably in patients with prior exposure to natalizumab or immunosuppressants (Khatri BO. Ther Adv Neurol Disord 2016;9:130-147).
There are limited data on the safety of switching from a front-line oral agent to fingolimod. In an analysis of Canadian patients (n=149) switching from teriflunomide or DMF to fingolimod, the rate of discontinuation due to adverse events was 4.9% (Grand’Maison et al. ECTRIMS 2015; abstract P532). There were few discontinuations due to cardiac symptoms or infections (0.9% and 1.9%, respectively).
The CARE-MS II trial of alemtuzumab enrolled patients previously treated with a DMT: 70% had received one, 22% had received two and 9% had received three or more prior therapies (Coles et al. Lancet 2012;380:1829-1839). Prior therapies included an interferon or glatiramer acetate; recent natalizumab exposure was an exclusion criterion. The incidence of serious adverse events with alemtuzumab was low (0.16). The rate of herpes zoster infection was 6%, compared to 3% in the previously-untreated cohort in CARE-MS I (Cohen et al. Lancet 2012;380:1819-1828). This difference may be due, in part, to prior DMT exposure, or may relate to the timing of antiviral drug use in the two trials. The rate of herpes simplex infection was lower in CARE-MS II compared to CARE-MS I (10% vs. 13%). The rates of immune thrombocytopenia purpura (1%) and hepatotoxicity (4%) were the same in the two CARE-MS study populations.
Data are generally lacking on the safety of switching from a front-line oral to alemtuzumab, but there are a few theoretical concerns. Liver function testing after teriflunomide or DMF may be advised to reduce the potential risk of cumulative hepatotoxicity. Changes in renal function have been observed with DMF and alemtuzumab (including rare cases of anti-glomercular basement membrane disease) (Meyer et al. Mult Scler Relat Disord 2013;2:60-63), so urinalysis may be advised following DMF discontinuation.
About 26% of patients in the OPERA trials of ocrelizumab were previously treated (Hauser et al. N Engl J Med 2017;376:221-234), but at this juncture it has not been determined if prior DMT exposure is associated with any increased risks. There are theoretical concerns with long-term suppression of B cell function, but it is unknown if there are increased risks associated with prior exposure to B cell-depleting agents, such as teriflunomide. Long-term observational data will help to clarify the risks.
As noted previously (see Part 3), there are potential concerns with repeated manipulation of immune function, which may produce as-yet undetermined cumulative or synergistic effects on immune subsets, most notably in aging patients with a longer duration of MS. Since data are lacking, it may be prudent to limit the number of drug exposures during the treatment course. This may be accomplished by switching early to an effective therapy that can be maintained and well-tolerated over the longer term.
What are the optimal sequencing options? This issue will be explored in part 7 of this series.
Comment (En français)
Dr. Marc Girard: Clinical practice for neurologists specializing in multiple sclerosis has changed drastically over the last 20 years. Neurologists have gone from a complete lack of treatment, to drugs that are increasingly effective but which pose significant risks that can affect the survival of their patients. Neurologists no longer have the option of expanding their medical skills to become familiar with the hematological, hepatic, cardiac, renal and endocrine risks associated with the use of the numerous treatments now available. They must also discuss the risks of pregnancy, since treatment is often directed at a female population of reproductive age; and they must plan for vaccination, as some vaccines may become contraindicated with certain treatments.
Neurologists must also be on the lookout for opportunistic infections. Most know how to recognize PML, and are aware of strategies to reduce PML risk during natalizumab use. But with the proliferation of new treatments with new modes of action, neurologists will be exposed to new complications, such as listeriosis (Rau et al. Int J Mol Sci 2015;16:14669-14676), or nocardiosis (Sheikh-Taha & Corman. Mult Scler 2017;23:872-874) in patients treated with alemtuzumab, or Cryptococcus infections in patients treated with fingolimod (Carpenter et al. Mult Scler 2017;23:297-299).
Management is becoming more complicated as neurologists are increasingly faced with sequential treatment choices, with the risk of additive effects on immune function from a previously-used treatment. More than ever, neurologists treating MS should expect to see the unexpected.
The solutions will come from real-world data of large groups of patients. The hope is that approaches based on risk stratification will be developed, as they now exist to manage the risk of PML with natalizumab. Throughout the treatment course, it is essential to discuss the different options available with the patient – the final decision about risk rests with the patient, not the doctor. A shared decision approach is essential.
Dr Marc Girard: Le type de pratique pour un neurologue spécialisé en sclérose en plaques a changé drastiquement depuis 20 ans. D’une absence complète de traitement, les neurologues ont maintenant accès à des médicaments de plus en plus efficaces mais qui présentent des risques non négligeables pouvant même toucher la survie de leurs patients. Les neurologues n’ont donc plus le choix d’élargir leurs compétences médicales et de devenir familiers avec les risques hématologiques, hépatiques, cardiaques, rénaux et endocriniens qui sont associés à l’utilisation des nombreux traitements maintenant disponibles. Ils doivent aussi discuter des risques de grossesse puisque ces traitements s’adressent fréquemment à une population féminine en âge de procréer. Ils doivent prévoir la vaccination sachant que certains vaccins pourront devenir contre-indiqués avec certains traitements.
Les neurologues devront dorénavant être toujours aux agents d’une infection opportuniste. La vaste majorité sait reconnaître la PML et les stratégies pour en diminuer le risque dans l’utilisation du natalizumab. Mais avec la multiplication de nouveaux traitements utilisant de nouveaux modes d’action, les neurologues seront exposés à de nouvelles complications comme la listériose (Rau et al. Int J Mol Sci 2015 ;16 : 14669-14676) ou la nocardiose (Sheikh-Taha & Corman. Mult Scler 2017 ;23 : 872-874) chez les patients traités avec l’alemtuzumab ou encore des infections à Cryptococcus chez les patients traités avec le fingolimod (Carpenter et al. Mult Scler 2017;23 :297-299).
La gestion est d’autant plus compliquée que les neurologues font de plus en plus face à des choix de traitements séquentiels avec le risque que les effets immunitaires d’un traitement précédemment utilisé se surajoutent au nouveau traitement choisi. Plus que jamais, les neurologues traitant des cas de sclérose en plaques doivent s’attendre à voir à tout moment de l’inattendu.
Les solutions passeront par les données real world obtenues grâce au suivi de grand groupe de patients. On peut souhaiter que des approches basées sur la stratification du risque seront développées comme il existe maintenant pour gérer le risque de PML avec le natalizumab.
Il faut aussi ne jamais oublier de discuter des différentes options avec notre patient. La décision finale quant aux risques qu’il désire courir pour sa santé est la sienne et non celle du médecin. Plus que jamais, il est essentiel qu’une approche basée sur le shared decision soit en tout temps utilisée.
NeuroSens survey on sequencing – Part 6