Treating Multiple Sclerosis with Monoclonal Antibodies: A 2010 Update

Mathias Buttmann


Expert Rev Neurother. 2010;10(5):791-809. 

In This Article


The most significant change in the past 2 years regarding natalizumab (NATA; Tysabri®, Biogen Idec/ELAN, MS, USA) relates to the occurrence of progressive multifocal leukoencephalopathy (PML) in patients having received NATA monotherapy. As of 9 February 2010, 35 PML cases in patients receiving NATA monotherapy among more than 64,000 patients treated in the post-marketing setting were reported by the manufacturer, corresponding to an overall incidence of 0.52 (95% CI: 0.36–0.73) in 1000 patients,.[201] Other severe infectious complications were only very rarely observed.[2–4] A number of key questions arose with these monotherapy-associated PML cases.

Does the Risk of PML Increase with Treatment Duration?

Figure 1 shows the risk of PML as of 9 February 2010, stratified according to treatment period.[201] The 95% confidence intervals increase with later treatment periods, as absolute patient numbers are currently too low to allow a stringent risk assessment in patients having received NATA for more than 2 years. These data provide evidence for an increase of risk with ongoing treatment, possibly peaking during the third year of treatment. However, currently the relative risk in all cohorts stays within the risk frame of the 2-year Phase III trials, which had been the basis for NATA approval, and it seems to be lower during the first 2 years of treatment than assumed by the clinical studies (compare left column in Figure 1). This calculation does not take into account that the incidence of PML was observed to be higher in European than in US patients, as discussed later. For a more stringent risk assessment after more than 2 years of treatment, further post-marketing data are awaited. Biogen Idec and ELAN started publishing monthly updates on PML incidence in December 2009.[201]

Figure 1.

Progressive multifocal leukoencephalopathy incidence in natalizumab-treated patients stratified according to natalizumab treatment period. Calculated from worldwide clinical trial and post-marketing data. Until 9 February 2010, 35 progressive multifocal leukoencephalopathy cases were confirmed worldwide. Dots represent the calculated risk per 1000 patients, bars represent 95% confidence intervals.[201]

Is the Risk of PML Predictable in Individual MS Patients?

There is currently no validated method available in clinical practice that would allow assessment of the individual risk of PML before or during NATA treatment. However, Biogen Idec and ELAN recently announced a serological assay detecting serum antibodies against JC virus (JCV) to be launched in 2010, potentially capable of stratifying patients according to PML risk.[202] The assay was said to be highly sensitive and specific for the presence of JCV in the body. With this assay, approximately 50% of tested individuals were positive, and approximately 50% negative, which is a lower rate of positive-testing individuals than in some, but not all, published cohorts tested with other assays, where the rate of subjects testing positive was up to 90%.[5–7] Higher JCV specificity of this assay and/or testing of a younger cohort might possibly explain the rather low rate of positive-testing individuals, as the likelihood of JCV positivity increases with age.[7] The assay was validated with blood samples from PML patients.[202] The companies announced in January 2010 that all 11 NATA-treated MS patients who developed PML, and for whom blood samples from 1–3 years before diagnosis of PML were available, retrospectively tested positive before clinical development of PML. Provided that all later PML patients develop the disease based on an endogenous reinfection,[7] positive-testing patients would be at an approximately twofold increased PML risk in comparison with the unstratified overall MS population currently receiving NATA, where approximately 50% are JCV-seropositive. By contrast, patients testing negative would be at a dramatically lower PML risk. A twofold increased risk of PML in positive-testing patients would not generally exclude the use of NATA and – at least over 2 years – still be within the approval-relevant PML risk frame of the Phase III program (see previous), but it would influence the individual benefit–risk assessment. The JCV seroconversion rate of negative-testing individuals was reported to be in the low single-digit percent range per year.[202] The companies announced that the assay will be accompanied by treatment protocols, additionally taking into account further risk factors, such as immunosuppressive pretreatment or duration of NATA treatment, which were extracted from the huge pool of post-marketing data.[202] A prospective validation of the assay in a planned cohort of overall 9000 patients was announced to start in March 2010 (NCT01070823, NCT01070836).

Detection of JCV DNA in body fluids or immune cells might theoretically be considered as another approach allowing individual risk assessment. However, according to recent results, the presence of JCV DNA detected by quantitative PCR in plasma, serum, peripheral blood mononuclear cell and urine samples did not change in 1094 MS patients included in the Safety of Tysabri® Redosing and Treatment (STRATA) trial within 48 weeks after the initiation of NATA treatment,[8] in an independent cohort of 597 patients (no plasma samples included)[9] and in finally published studies on further large cohorts (summarized in[10]). JCV DNA frequencies in these cohorts were comparable to previously published healthy study populations. In contrast to this large body of evidence, a small prospective study in 19 patients with relapsing–remitting MS (RRMS) recently reported strongly increasing frequencies of JCV DNA, but not of BK virus DNA as a control, in urine and peripheral blood mononuclear cell samples and, to a weaker extent, in plasma samples during NATA treatment, suggesting that subclinical JCV reactivation frequently occurs in NATA-treated patients.[11] A recent prospective study in 200 MS patients where JCV DNA was not detected in plasma and cerebrospinal fluid (CSF) samples at a single time point before NATA initiation reported that JCV DNA became detectable after 4–18 months of NATA treatment in three patients (two only in CSF, one only in plasma) who did not develop PML. Furthermore, JCV DNA was no longer detectable in these patients after NATA cessation.[12] Although similar observations were made for BK virus DNA in five further patients of this cohort, the lack of a control group left open the question of whether these observations reflected a NATA-specific effect or NATA-unrelated fluctuations of viral DNA levels, as might be suggested by the large trials cited earlier. In a retrospective analysis, only one of three PML patients in the NATA Phase III program had constantly elevated JCV DNA plasma levels before PML symptoms began, and another patient had high JCV DNA plasma levels but no PML.[13] In contrast to urine samples, where JCV DNA can be detected in approximately a third of the population, JCV DNA in plasma samples is usually detectable in less than 1% of a healthy population or of untreated and NATA-treated MS patients. Of interest, the presence of JCV viremia or viruria in HIV-infected patients with PML did not correlate with PML survival.[14,15] Furthermore, JCV excretion into urine was not associated with the immunological status in a small cohort of patients, although both JCV excretion into urine and presence in blood samples were more frequently observed in HIV patients without PML than in healthy controls.[15] Together, it is not yet clear whether the detection of JCV DNA in body fluids and immune cells is suitable for determination of the risk of PML in individual patients, although the vast majority of studies provided no such evidence to date.

T-cell-mediated immune reactions determine the outcome of PML in HIV patients,[15–17] and a modulation of T-cell-mediated antiviral immune responses, including but not restricted to JCV, during NATA treatment of MS patients without PML has been described.[11,18] Bioenergetic monitoring of T-cell-mediated immunity in general by the commercially available ImmuKnow® assay (Cylex Inc., MD, USA) during NATA treatment has been suggested as a possible biomarker for the determination of PML risk.[19,20] To perform this assay, a heparin anticoagulated whole-blood sample is stimulated with the T-cell-activating mitogen phytohemagglutinin for 15–18 h. Subsequently, cells expressing CD4, which are mainly T-helper cells and Tregs, are isolated by positive magnetic bead selection. Thereafter, isolated CD4+ cells are lysed to release intracellular ATP, an energy carrier considered to reflect the energy state of the cells. ATP is then finally quantified, with a low ATP level indicating immunosuppression. This method has been shown to provide information on cellular immunity in patients receiving immunosuppressive treatment after renal transplantation.[21,22] Whether the ImmuKnow assay allows the determination of the risk of PML during NATA therapy in MS patients remains to be further evaluated.

How do Relative Risk & Prognosis of PML in NATA-treated MS Patients Compare with non-NATA PML Cases?

In the vast majority of cases, PML occurs in individuals who are immunocompromised, due to either disease or immunoactive medication. Infection with HIV is still the most frequent condition underlying PML.[23] According to some studies the incidence of HIV-associated PML decreased from 3.3–10 cases per 1000 person-years at risk in the era before highly active antiretroviral therapy (HAART), to 1–1.3 cases during the HAART era,[24,25] while others found an unaltered incidence upon the arrival of HAART.[26] However, there is a consensus that the survival rate of HIV-associated PML improved with the advent of HAART. While PML took a rapid fatal course within a few months in more than 90% of patients during the pre-HAART era,[25,27] the 1-year survival rate improved to 33–56% during the HAART era.[25,28,29] This seems to imply that the current immune status determines the individual outcome of PML in this patient cohort. In line with this interpretation, it was demonstrated that the presence of CD8+ cytotoxic T cells specific to the JCV capsid protein indicates a better prognosis.[16,17] Immune reconstitution inflammatory syndrome (IRIS) with neurological symptoms has been described as a rare complication of HAART in HIV-infected patients,[30] and 0.9% of patients initiating HAART were found to develop neurological IRIS.[31] Occurrence of IRIS with neurological symptoms in this patient cohort does not require opportunistic infections such as PML as a prerequisite; however, unmasking or deterioration of HIV-associated PML upon initiation of HAART has been described.[31,32] Potential CNS IRIS after NATA withdrawal in MS patients in the absence of PML, as characterized by multiple gadolinium (Gd)-enhancing MRI lesions, most often in previously normal-appearing brain areas, has recently been reported in a series of seven MS patients where it responded well to corticosteroid treatment.[33] The overall clinical outcome of HAART-associated IRIS was also found to be favorable in the majority of HIV patients,[31] while NATA-associated IRIS in MS patients with PML may cause severe secondary damage to the CNS.[34,35]

As of 29 July 2008, 76 rituximab (RTX)-associated PML cases were recorded in the manufacturer's global safety database (69 oncological patients, six with autoimmune disorders, one unknown).[36] In a published series of 57 cases, 90% took a fatal course within a median time of 2 months after diagnosis.[37] This indicates a very aggressive PML course in patients receiving RTX, where important confounding factors are active and where rapid immune reconstitution cannot be achieved owing to the irreversibly long-lasting, profound effect of the drug on the immune system. The series of 57 cases included 52 patients with lymphoproliferative disorders, a figure that has to be seen in relation to more than 500,000 patients having received RTX for the treatment of hematological disorders. Furthermore, lymphoproliferative diseases bear an inherent risk of PML,[36] impeding an estimation of the purely RTX-related PML risk. In addition, most of the RTX-associated PML cases within the series of 57 patients had immunosuppressive pre- or co-treatment with alkylating agents (n = 39) or purine analogues (n = 26),[37] making the risk assessment for RTX even more difficult. Two of the 57 patients were treated for systemic lupus erythematosus, which also carries a RTX-unrelated risk of PML.[38] Single PML cases in this series were observed in patients with rheumatoid arthritis (RA; plus chemo- and radio-therapy for oropharyngeal cancer),[36] autoimmune pancytopenia and immune thrombocytopenic purpura. The reported confounding factors in RTX-treated patients prohibit an inherently biased direct comparison of PML incidences and fatality rates under NATA and RTX, although confounding factors in the form of immunoactive pre- or even co-treatment were also present in many MS patients experiencing NATA-associated PML, particularly in those from Europe where the incidence of NATA-associated PML was found to be higher than in the USA (USA: 11 out of 37,319; EU: 21 out of 23,142; rest of the world: three out of 4157, as of 9 February 2010).[201] According to current knowledge, MS itself does not carry an increased risk of PML; however, it is interesting to note that the US cohort mentioned in the last sentence where a lower PML incidence was observed than in patients from the EU comprised both patients with MS or Crohn's disease, while in the EU cohort only MS patients were included. In contrast to the USA, where NATA was approved for the treatment of Crohn's disease in January 2008, NATA was not approved for the treatment of Crohn's disease in the EU as of February 2010. One NATA-treated patient with Crohn's disease who developed PML was reported in the Phase III program, and no further cases have been published to date.

There is only limited public knowledge on the course of NATA-associated PML in MS patients. Four single cases have been described in case reports so far. Only two of them had received NATA monotherapy, the other two patients had received parallel IFN-β treatment within the Safety and Efficacy of Natalizumab in Combination with Interferon β-1a in Patients with Relapsing–Remitting Multiple Sclerosis (SENTINEL) trial.[34,35,39,40] According to Biogen Idec, eight out of 35 (23%) patients died from PML associated with NATA monotherapy as of mid-February 2010.[203,204] The final rate of these 35 patients not surviving NATA-associated PML may be higher than 23%, as this cohort also included recently diagnosed cases where the final outcome is not yet known. However, it is already safe to assume that the PML course in NATA-treated MS patients is significantly more favorable than in RTX-treated patients with lymphoproliferative disorders. Furthermore, available data might indicate that an early-as-possible diagnosis and treatment of PML, including early NATA withdrawal, improves the outcome of PML, which would highlight the importance of clinical vigilance.

Are 'Drug Holidays' Suited to Reduce the Risk of NATA-associated PML?

Although the mechanism by which NATA increases the risk of PML is unknown, it may be speculated that an impairment of immune surveillance of the CNS due to NATA treatment could be a crucial causative factor for PML. As outlined above, restoration of immune functions shows an overall beneficial effect on the clinical course of PML in HIV patients, and possibly also on the course of PML in NATA-treated MS patients, which might indirectly support this hypothesis. It may be deduced that pausing NATA after an initial treatment period could allow restoration of the immune system, and thereby possibly decrease the risk of PML during further NATA treatment after a gap. This therapeutic strategy is called 'drug holidays' and is a matter of discussion, as it is currently unclear whether NATA drug holidays decrease the risk of PML when redosing the drug.

The STRATA trial was an observational follow-up study in 1094 participants of the Natalizumab Safety and Efficacy in Relapsing Remitting Multiple Sclerosis (AFFIRM), SENTINEL and Natalizumab in Combination with Glatiramer Acetate in Patients With Relapsing–Remitting Multiple Sclerosis (GLANCE)[41] trials, investigating the safety and efficacy of NATA redosing after a treatment gap of at least 1 year (median: 85; range: 57–157 weeks).[42] The median number of infusions before NATA withdrawal was 25 (range: 1–37). Two PML cases (after 34 and 35 doses) among 1094 patients have been reported in this study so far.[42] Cotreatment of patients with IFN-β or glatiramer acetate within the trial program before the NATA treatment gap might have influenced later PML incidence, and hence further information on both PML cases is awaited. Although the number of PML cases in this study was too low to allow a firm estimation of PML frequency upon NATA redosing after a treatment gap of at least 1 year to be made, this study might challenge the view that drug holidays are suited to decrease the risk of PML. Alternatively, it might pose the question as to whether 1 year is a long enough a gap to allow full restoration of immune functions, along with a risk reduction after NATA redosing. At least a normalization of immune cell counts and ratios of subpopulations in peripheral blood and CSF 14 months after cessation of NATA was recently described in a cross-sectional analysis.[43] While NATA withdrawal bears the risk of a return of disease activity with resulting permanent disability, particularly in patients with highly active disease before NATA initiation,[42] the STRATA study has not so far obtained evidence for a loss of NATA efficacy after redosing. While short-term usage of NATA may lead to a rebound of disease activity, there is now increasing evidence that a sudden rebound after prolonged treatment is unlikely; however, the issue of rebound of disease activity after NATA withdrawal needs further investigation.[42–46]

A recent consensus paper by a group of German MS experts did not find enough evidence to currently recommend drug holidays as a general treatment strategy.[47] However, given the increasing risk of PML with increasing treatment duration, termination of NATA treatment should always be considered as an option, based on an individual benefit–risk assessment and available therapeutic alternatives.

What is the Optimum Treatment Strategy for NATA-associated PML?

Plasma exchange and, probably more effectively, immune adsorption were used to rapidly remove NATA from the circulation after a diagnosis of PML.[34,35] Rapid recovery of immune functions, allowing immune cells to enter the CNS and fight JCV, was the main rationale behind this strategy. However, as most MS patients with PML develop IRIS approximately 4 weeks after NATA removal, which seems to frequently cause significant secondary damage to the CNS, it remains open as to whether accelerated NATA removal is an ideal strategy, how it should be applied and which flanking therapeutic measures are best suited to limit the secondary damage due to IRIS. On the other hand, delayed development of IRIS 3 months after NATA withdrawal was also observed in a PML patient who was not treated with plasma exchange.[39] Apparent IRIS has been treated with high-dose corticosteroid pulse therapy.[34,35] It is currently a matter of discussion as to whether prophylactic treatment with intravenous immunoglobulins or other measures might dampen the overshooting immune response after NATA withdrawal in MS patients with PML. Prophylactic steroid treatment is currently not recommended by the manufacturer as it may possibly impede a desired antiviral immune response.

Antiviral treatment of HIV-associated PML with cidofovir has not proven to be effective in a meta-analysis of published studies.[48] Due to its anti-JCV activity in vitro,[49] cytarabine, which, however, poorly crosses the BBB, was tested in a randomized, controlled trial in patients with HIV-associated PML where it failed to show a beneficial effect.[50] Based on its in vitroantiviral activity and its excellent permeability of the BBB, the antimalarial agent mefloquine may represent another, more promising PML treatment strategy.[51] A small trial in 40 subjects with HIV-associated PML, investigating mefloquine in comparison to standard therapeutic care, was started in September 2008. The estimated primary completion date as of August 2009 was December 2009, according to the NIH website[205] in February 2010 (NCT00746941; Biogen Idec, ELAN). Furthermore, Biogen Idec recently announced efforts to identify additional agents for the treatment of PML by scanning a large pharmacological library for agents with in vitroanti-JCV activity and known BBB permeability. As JCV uses 5-HT2A serotonin receptors for glial cell entry in vitro,[52] and 5-HT2A blockers may inhibit this process in vitro,[53] the serotonin-blocking antidepressant mirtazapine was also used for experimental treatment of PML.[34,35,54,55]

Which Measures were taken by the Regulatory Authorities?

After initial US FDA approval in November 2004 and voluntary market withdrawal by the manufacturer in February 2005 owing to three PML cases in the Phase III clinical trial program, NATA was approved by the FDA in June 2006 as a second-line monotherapy for RRMS – with a black box warning highlighting the risk of PML, and under the restriction of a post-marketing safety program including the TOUCH prescribing program.[1] A 'Dear Healthcare Professional' letter and a change of the prescribing information, relating to potential liver toxicity, were released in February 2008. Serious liver injury was reported to the FDA in six patients from 2004 through to June 2008, four of them having occurred after the first dose.[56] In November 2009, information on the increasing risk of PML with increasing treatment duration and therapy of NATA-associated PML, including the occurrence of IRIS, was added to the prescribing information.[206] As of February 2010, the FDA did not put additional restrictions on the use of Tysabri.

Natalizumab was approved as a first- or second-line monotherapy for patients with highly active RRMS in the EU in June 2006.[207] In March 2008, the EMEA recommended an update of the product information concerning potential liver toxicity.[208] In September 2008, an update was decided upon to enhance the vigilance about PML.[209] In view of 23 PML post-marketing cases, a re-evaluation of the benefits and risks of NATA was started by the EMEA in October 2009, and completed in January 2010.[210] The EMEA recommended an update of the product information, including information on the increase of PML risk after 2 years of treatment and advice on PML management. In addition to MRI scans "at the first sign of any symptoms indicative of the possibility of PML", yearly routine MRI scans of all NATA-treated patients were suggested. Furthermore, written informed consent before treatment initiation and written informed reconsent after 2 years based on an in-depth discussion between neurologist and patient were requested.[211]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.