From injection therapies to natalizumab: views on the treatment of multiple sclerosis

Abstract

Discoveries of the mechanisms that underlie the pathogenesis of multiple sclerosis have been acquired at an impressive rate over the last few decades and, as a consequence, a growing number of treatments are becoming available for this disease. This review first analyzes the experience from the early stages of the disease-modifying therapies, then, expanding on the concept of early treatment for improved outcomes, it focuses on natalizumab and its major complication, progressive multifocal leukoencephalopathy. We offer views on the risks associated with the use of natalizumab by underscoring the importance of the JC virus serology and by providing preliminary data on our experience with the extended interval dosing of natalizumab. This approach, which advocates individualized treatment plans, raises the question of the minimum effective natalizumab dose. Extended interval dosing suggests efficacy can be maintained while providing advantages of costs and convenience over regular monthly dosing. More data examining this strategy are necessary.

The early stages of multiple sclerosis therapy

Until the early 1990s, with the approval of interferon beta as a disease-modifying therapy for multiple sclerosis (MS), neurologists, limited in the availability of well-studied therapies, relied on the use of glucocorticosteroids, adrenocorticotropic hormone (ACTH), and immunosuppressive agents to mitigate the effects of disease-related exacerbations. Conversely, over the last few years, the formidable challenge of reducing disease activity has been significantly impacted by the emergence of a series of novel agents working through unique mechanisms of action [Birnbaum, 2010]. Notwithstanding this important progress, the efficacy of each of the agents is now recognized to be associated with corresponding adverse events of various magnitudes and charact6er [Birnbaum, 2010].

Interferon-based therapies followed by the introduction of glatiramer acetate have ushered in a new era of MS therapeutics characterized by modest efficacy, but with imminent safety and acceptable tolerability in most patients [Giovannoni, 2011]. The early placebo-controlled clinical investigations yielded objective, evidence-based efficacy data, yet a lasting clinically meaningful response is lacking for a substantial percentage of patients.

Several studies attempted to demonstrate superiority of either drug class over the other or greater efficacy of higher dose and frequency therapy over weekly interferon [Mikol et al. 2008; O’Connor et al. 2009; Durelli et al. 2002; Panitch et al. 2002], and collectively they have supported the relative equivalency of their disease-modifying effects, inducing disease remission in only a proportion of patients. It is possible that randomizing patients to a treatment leads to leveling of the treatment effects of the medications, as the group analysis yields an average efficacy that is the result of responders and nonresponders mixed together. Once the nonresponders are reassigned to the other platform therapy, better control of the disease has been observed [Coyle, 2008b].

A change in treatment strategy

A major paradigm shift in the treatment of MS derived from the concept of applying the same injection therapies to patients at the time of their first clinical attack of inflammatory demyelination, the clinically isolated syndrome (CIS) [Jacobs et al. 2000; Tintoré, 2007; Coyle, 2008a; Bates, 2011; Stüve et al. 2008]. Ultimately, data from the study of all of the injection agents appear to corroborate the hypothesis that earlier treatment intervention provides greater benefits on disease activity than waiting until the diagnosis by McDonald criteria is confirmed, and offering treatment at that later time [Jacobs et al. 2000; Tintoré, 2007; Coyle, 2008; Bates, 2011; Stüve et al. 2008]. Furthermore, increasing evidence was mounting that early inflammation and MRI metrics of lesional burden were inextricably linked with the subsequent transition from a relapsing to a progressive course of disease [Fisniku et al. 2008]. However, while the disease activity seems to be impacted by early initiation of therapy, progression of disability does not appear to be altered with interferon-based therapies [Kinkel et al. 2006; Kappos et al. 2009].

In any case, the concept has evolved from ‘multiple events in space and time’ to the declaration of the MS diagnosis in those presenting with a single attack, when mimicking conditions have been excluded and the presence of MRI lesions are compelling (shape and location) in their being reminiscent of the disease process [Polman et al. 2005; Swanton et al. 2006, 2007]. With the goal to further simplify the diagnostic criteria yet maintaining specificity, in 2010 a revision of the radiological evidences for dissemination in space and in time was proposed. A single MRI scan that demonstrates the concomitant presence of T2 or fluid attenuated inversion recovery (FLAIR) and contrast-enhanced lesions, would suffice to support the diagnosis of MS [Montalban et al. 2010].

There is now an emerging literature to support yet another diagnostic designation of MS; this is suggested in patients who undergo imaging studies for reasons other than searching for MS, but are nonetheless found to have a radiologic signature of inflammatory demyelination, astrogliosis, white and gray matter lesions, and tissue destruction (gray or black holes). This new designation, the radiologically isolated syndrome (RIS) [Okuda et al. 2009], perhaps represents an opportunity not heretofore available, to recommend treatment intervention on the basis of a ‘working diagnosis’ of a multiphasic and highly disabling disorder, prior to the onset of a first clinical attack. Even though recommendations for therapy in this condition are thus far not established, recent data highlighting the prognostic value of asymptomatic spinal cord lesions provide further evidence that such a concept may not be so provocative [Okuda et al. 2011].

Natalizumab and its potential complication, progressive multifocal leukoencephalopathy

A profound advance in MS therapeutics was achieved with the introduction of the monoclonal antibody, natalizumab, for the management of relapsing forms of the disease [Polman et al. 2006]. Here for the first time we have a drug with a highly characterized mechanism of action that influenced a principal step in the underpinnings of the pathobiology of MS: the exuberant trafficking of mononuclear cells from the blood into the central nervous system (CNS). The magnitude of combined clinical and radiologic efficacy of this agent had not previously been documented with any prior disease-modifying strategy [Miller et al. 2007]. Despite early enthusiasm, the linkage of this agent to the development of progressive multifocal leukoencephalopathy (PML) [Langer-Gould et al. 2005] culminated in natalizumab being removed from the market in February 2005. Following the development of a risk management strategy, the drug was reintroduced into the marketplace in June 2006. Since that time, the postmarketing experience has substantiated the impressive efficacy of natalizumab, but also confirmed the ability of this agent to increase the development of PML in a small percentage of patients [Sangalli et al. 2011; Kappos et al. 2011; Biogen Idec, 2011]. The dynamics of the JC virus infection, the host immune response (the presence of JC virus IgG appears to portend a higher risk of PML than patients who are seronegative), and the influence of antecedent treatments (e.g. the use of immunosuppressant or chemotherapeutic agents) have yielded important knowledge that will help refine our ability to more intelligently understand and counterbalance the benefits and risks of intensive immunotherapy for individual patients.

As of 4 October 2011, based on the 170 cases reported worldwide, the overall risk of PML in patients exposed to natalizumab was estimated to be 1.82 per 1000 patients (95% confidence interval 1.56–2.11) [Biogen Idec, 2011]. This risk appears to be augmented when subgroup analysis considers the JC virus IgG status and the use of immunosuppressive therapy prior to the onset of natalizumab therapy [Gorelik et al. 2010; Sandrock et al. 2011] (Table 1). The concept of PML being an opportunistic infection arising during immune suppression dates back to its earlier description as a disease and the impaired immune surveillance provoked by natalizumab therapy is at least a partial explanation for causality [Warnke et al. 2010]. Details on how the JC virus is transformed to an agent with the capability of invading and targeting the CNS are not entirely clear [Warnke et al. 2010], but the direct relationship between the duration of therapy with natalizumab and incidence of PML has now been established [Sandrock et al. 2011] (Table 1). The currently available data do not permit an accurate calculation of PML risk beyond the initial 3-year period on natalizumab therapy, because of the limited numbers in this group. However, it is worth noting that the incidence of PML appears to be trending downward compared with the occurrence of PML in patients in the 2–3-year treatment epoch [Biogen Idec, 2011] (Table 1). One could speculate that some patients are resistant to PML regardless of the duration of exposure to natalizumab, and it is probable that those are patients who have never encountered and do not harbor the virus in their body. Alternately, it is conceivable that some patients are destined to never develop PML, irrespective of carrier status.

Table 1.

Incidence of progressive multifocal leukoencephalopathy (PML) in patients treated with natalizumab (as of 4 October 2011).

	Number of Patients	95% confidence interval per 1000 patients
Total number of patients exposed to natalizumab	88.100	N/A
Total cases of confirmed PML worldwide	170	N/A
Total number of death in patients with PML	33 (19%)	N/A
Estimated overall risk of PML	1.82 per 1000	1.56–2.11
	Estimates by treatment epoch per 1000 patients	95% confidence interval per 1000 patients
Patients having received
Infusions 1–12	0.04	0.01–0.11
Infusions 13–24	0.55	0.38–0.77
Infusions 25–36	2.01	1.60–2.49
Infusions 37–48	1.50	1.05–2.06
Patients in the clinical trials	1.00	0.20–2.80
Patients without prior exposure to immunosuppressive medications*
Infusions ≤ 24	0.19	0.10–0.33
Infusions ≥ 25	1.37	0.97–1.90
Patients with prior exposure to immunosuppressive medications*
Infusions ≤24	0.66	0.32–1.20
Infusions ≥ 25	4.30	2.90–6.20

^*Immunosuppressive medications include mitoxantrone, azathioprine, methotraxate, cyclophosphamide, mycophenolate, cladribine, rituximab and do not include corticosteroids. Data are from Kappos et al. [2011] and Biogen Idec [2011].

About 1 year ago, a novel enzyme-linked immunosorbent assay (ELISA)-based serology test for JC virus was introduced by Biogen Idec [Gorelik et al. 2010] and a negative test, namely the absence of anti-JC virus antibodies, would indicate noncarrier status. The false-negative rate was found to be 2.5% (five out of 204) in one dataset [Gorelik et al. 2010]. Acquisition of data on the prevalence of the positive serology for JC virus is in progress and the preliminary reports in the MS population demonstrate an approximate 54% [Sandrock et al. 2011] with a seroconversion rate of 2–3% per year [Biogen Idec, 2011]. The JC virus serology test has now become commercially available, and when combined with the risks posed by duration of natalizumab therapy and prior exposures to immunosuppressive medications, the positivity of this test could function to stratify the patients at risks of developing PML [Sandrock et al. 2011].

Possible strategies to minimize the risks of PML

The recruitment for participation in clinical trials over the last 5 or 6 years have targeted patients in earlier stages of the disease compared with studies performed 10 or 15 years ago and this may account for better treatment responses [Goodin, 2008]. However, the treatment effects of natalizumab were unprecedented at that time and the strong interest to continue utilizing this medication, while attempting to reduce its risks, led some authors to propose a suspension of therapy of natalizumab that could theoretically permit the temporary reconstitution of immune surveillance for long enough to halt PML before it developed. When a 3–6-month break was applied to patients reaching 1 year on monthly dosing of natalizumab [West and Cree, 2010], it was observed that suspension of therapy with natalizumab posed risks for reactivation of the disease and this strategy did not seem to be supported by significant benefits [West and Cree, 2010; Vellinga et al. 2008]. Similar conclusions were drawn by another study during which MS patients, relapse free and with negative contrast-enhancing lesions on MRI for 12 months while treated with natalizumab, underwent a 24-week treatment interruption. Reactivation of disease was detected clinically and radiologically starting approximately 12 weeks postsuspension, and resumption of natalizumab provided better control of the disease over injection therapies or intravenous (IV) methylprednisolone [Fox et al. 2011]. However, the potential for recurrence of clinical and MRI activity after cessation of natalizumab therapy may not apply to all patients, as there is growing recognition that clinical disease activity prior to initiation of natalizumab therapy is the best predictor of clinical and paraclinical disease activity following cessation of treatment. Specifically, patients with a high relapse rate before initiation of natalizumab appear to have the highest risk of disease re-activation, whereas patients with a low relapse rate have relatively a low risk [O’Connor et al. 2011]. Therefore, continuous treatment with natalizumab may not be necessary for all patients with MS.

The rationale for a monthly dosing derives from pharmacokinetic data [Khatri et al. 2009] but the minimal effective dose is not known and it may well be that different patients have different needs. It has been calculated that with a plasma concentration of natalizumab of less than 1 µg/ml, the saturation of α4-integrin falls below 50% and this is reached 82 days after the last dose of natalizumab [Khatri et al. 2009; Tan et al. 2011]. Variations in the rate of clearance have been observed among patients [Khatri et al. 2009], thus supporting the concept that less frequent dosing of natalizumab would be appropriate for some patients. Preliminary evidence indicates that over time there is an accumulation of natalizumab in the serum of patients treated with monthly regimen [Foley, 2011]. This would suggest that perhaps treating patients with an extended interval dosing of 8 weeks may be feasible to prevent a persistent saturation, thus restoring some degree of immune surveillance. The concepts behind immune reconstitution are discussed in a recently published editorial [Berger, 2011].

At our institutions, MS patients who have been on therapy with natalizumab for 1 year at the regular monthly dosing, are counseled on the opportunity to change the administration of the drug to an extended interval dosing (EID) that could be every 6 or 8 weeks. Upon approval of the Institutional Review Board, a retrospective review of all patients with MS treated with natalizumab between October 2006 and December 2009, we identified 203 consecutive patients: 112 remained on the monthly dosing, while 91 patients were switched to EID. At the time of the data collection, 84 patients had been on therapy for longer than 12 months and 28 patients had been on therapy for more than 24 months. Out of 111 surveillance MRI scans, 99 remained unchanged. Of these stable scans, 81 were from patients on EID and 18 on the monthly dosing (Figure 1). Five clinical relapses occurred, one of which was on EID [Alderazi et al. 2011]. The collection of data for the analysis between October 2006 and September 2011 is in progress and we plan to submit our findings to the peer-reviewed literature shortly.

Figure 1.

Bar-graph representation of the number of MRI scans in patients on natalizumab on extended interval dosing (EID; left-graph) and standard dosing (right graph). Stable was defined as having no T2/ fluid attenuated inversion recovery (FLAIR) lesions compared with prior scan and no contrast-enhanced lesions. Worse was defined as either having new T2/FLAIR lesions as compared with the last scan, or having gadolinium enhancement(s).

This alternative strategy aimed at reducing the exposure to the medication, hence theoretically cutting down the risks of PML, stems from the idea that MS is, by its very nature, extremely heterogeneous. Disease courses vary from mild to very aggressive and clinicians of all specialties of medicine are often confronted with the need to tailor the treatment to the specific cases. As mentioned above, the flexibility of different dosing of interferon therapy for MS patients has been investigated [Mikol et al. 2008; O’Connor et al. 2009; Durelli et al. 2002; Panitch et al. 2002] and data from the trials as well as experience in clinical practice support the view that some patients achieve disease stability, hence seem to display a treatment effect with the lowest doses of interferon [Bates, 2011; Stüve et al. 2008]. Different formulations of interferon have been available for clinical use (30 µg weekly versus 22 µg three times weekly versus 44 µg three times weekly for interferon beta-1a, and 250 µg for interferon beta-1b), and alternative dosing regimens have been under investigation for glatiramer acetate too [Cohen et al. 2007; US National Institutes of Health, 2011]. The same principle of a diversified treatment that is customized to patients could be applied to the use of natalizumab and in our opinion an EID is a reasonable option to be offered to patients. Selection of the most appropriate treatment regimen for individual patients should be part of the practice of medicine and retrospective reviews can shed light on the appropriateness of those treatment initiatives [Khan, 2009].

When considering the stratification of patients provided by the status of their JC virus serology, it is intuitive that those who are positive run the highest risks for PML. In this latter group switching to the EID could be considered in view of the fact that a regimen with every other month dosing results in half of the drug exposure per equivalent period of treatment. At this point there are no data to support the claim that increasing the dosing interval would impact the risks of PML, but an advantage on costs, logistics and convenience could be envisioned.

If one looks at the incidence of PML in the pivotal clinical trial a value of 1 per 1000 patients was estimated (Table 1). With the wide use of the medication in the postmarketing phase, the incidence of PML has steadily continued to rise and has now reached the historical maximum with a much tighter confidence interval, indicating more accurate data. We believe that an initiative ought to be taken in the attempt to decrease this upward trend. A clinical trial that could compare the regular monthly dose to the EID and assess PML as an outcome would require thousands of patients to demonstrate an effect and it is not feasible. On the other hand, observational studies have the great potential to provide robust data [Khan, 2009] and should this practice become widespread [Pawate et al. 2011] the data could be analyzed retrospectively and the PML incidence evaluated in large cohorts.

Learning from the experience acquired in the injection therapies [Panitch et al. 2002; Coyle, 2008a] and continuing to collect data on real-life treatment strategies [Castillo-Trivino et al. 2011] will advance our abilities to treat MS.

Our observation of the maintained treatment effects with the EID of natalizumab is going to be evaluated in the setting of a randomized clinical trial. A study is under way in Europe where not only EID but also reduced dosing (150 mg) is going to be tested in subcutaneous versus intravenous administrations [EU Clinical Trials Register, 2011], and better understanding of the minimal required dose is expected.

Future perspectives

The change of treatment strategy introduced with the injection medications in CIS patients, according to which early treatment is likely to result in more profound responses to therapy and longer lasting beneficial effects, have paved the way for an additional concept to be further developed. Early control of the disease would lead to better outcomes and it would make sense to use the therapeutic agents with the highest chances of stopping the disease progression early on; once a good control of the pathological process is obtained and the potential risks of complications increase, then the switch to an alternative regimen or safer medication could be made. This would be a principle of induction followed by maintenance therapy as it was applied in the recent past in a study with mitoxantrone followed by glatiramer acetate in patients with MS [Vollmer et al. 2008]. The complication of PML has led to the recommendation (see prescribing information) of natalizumab being proposed to MS patients as a second-line agent and patients would become appropriate for therapy with natalizumab after having failed at least one of the first-line agents [Castillo-Trivino et al. 2011]. If the validity of the JC virus serology and other risk factor are confirmed, it would be possible to stratify the patients’ risks for PML and natalizumab could be considered an equally appropriate first-line choice.

Natalizumab has been the first monoclonal antibody applied to the treatment of MS and the field has now opened to a whole host of monoclonal antibodies [Bielekova and Becker, 2010]. These molecules have the advantage of being extremely selective with their targets and create models that provide insights into the mechanisms of the disease pathogeneses that will in turn result in the design of ever more specific medications.

If one would view MS as a wild horse that is hard to tame, they would now appreciate the increased availability of tools and opportunities to use different approaches that are now at their disposal.

Major challenges still lie ahead. For instance, the mechanisms underlying the degenerative component of the disease process remain difficult to define and as a consequence remain difficult to treat.