REVIEWER: Paul Giacomini, MD, FRCPC, Associate Director, Multiple Sclerosis Clinic, Montreal Neurological Hospital and Institute, Assistant Professor, Department of Neurology and Neurosurgery, McGill University, Canada
Switching studies
Effect on brain volume change
Updated safety data
The high-efficacy disease-modifying therapies (DMT; natalizumab, fingolimod, alemtuzumab) used to treat relapsing-remitting multiple sclerosis are often reserved for patients with severe disease at presentation or those with an inadequate response to other agents. However, these therapies are increasingly being employed earlier in the disease course. In part this is due to the growing recognition that first-generation injectable DMTs appear to have less impact on disability progression. In addition, there is increasing evidence that more potent suppression of inflammatory disease activity has the potential to reduce CNS tissue damage, slow the rate of brain volume loss and improve long-term physical and cognitive outcomes. However, the benefits of this more aggressive approach need to be weighed against the risk of adverse effects with higher efficacy agents.
The following summarizes some of the key issues addressed in studies presented at AAN 2015 on the use of high-efficacy agents.
Direct-comparison studies of fingolimod versus interferon-beta-1a IM (TRANSFORMS) and alemtuzumab versus interferon-beta-1a (CARE-MS I, CARE-MS II) have shown that high-efficacy agents are significantly more effective than interferons.
A post-hoc analysis of fingolimod phase III data reported that annualized relapse rates (ARR) were significantly reduced in the subgroups with a previous inadequate response to interferon-beta or glatiramer acetate (Nicholas et al. AAN 2015; abstract P3.248). In addition, a matched-cohort study using data from two large observational studies (PANGAEA, PEARL) found that ARR was significantly lower in fingolimod-treated patients compared to those treated with interferons or glatiramer acetate (Duerr et al. AAN 2015; abstract P3.253).
In the two-year extension of the CARE-MS I trial, which compared alemtuzumab and subcutaneous interferon-beta-1a in treatment-naïve patients, ARR was reduced 69% when the high-dose interferon-beta cohort was switched to alemtuzumab (Hartung et al. AAN 2015; abstract P7.270); ARR at two years was 0.12 compared to 0.39 during interferon treatment. Moreover, EDSS scores were stable or improved in 73%, and 17% achieved a sustained reduction in disability, as measured by the EDSS. A separate analysis reported that the proportion of patients free of new Gd+ lesions increased from 80.8% in year 2 of interferon therapy to 96.2% in the second year after switching to alemtuzumab; the proportion of patients free of new/enlarging T2 lesions increased from 59.8% in year 2 of interferon therapy to 81.7% in year 2 of alemtuzumab (Barkhof et al. AAN 2015; abstract P7.261). The rate of new lesion formation declined from 0.31/year to 0.04/year with alemtuzumab.
CARE-MS II is the only phase III study to date to examine treatment efficacy in patients with breakthrough disease (Coles et al. Lancet 2012;380:1829-1839). At four years, ARR with alemtuzumab remained low (0.23) and 76% did not experience six-month confirmed disability progression (Havrdova et al. AAN 2015; abstract P7.276). Two-thirds of patients had stable or improved EDSS scores, including 41% with a sustained reduction in disability as measured by the EDSS. Overall, 68% only required two courses of alemtuzumab, and 24% received a third course. For the cohort that switched from interferon-beta-1a to alemtuzumab, ARR declined from 0.52 to 0.15, a 71% reduction (Fox et al. AAN 2015; abstract P7.278). The proportion of patients with stable/improved EDSS scores increased from 59% to 69%, and an additional 15% showed improvement in six-month confirmed EDSS scores.
A database analysis examined outcomes following various treatment switches (McQueen et al. AAN 2015; abstract P3.272). ARR was reduced by 0.44 after switching to natalizumab, but increased 0.34 for those switching to another DMT. However, long-term treatment with natalizumab is generally not advised due to the increasing risk of progressive multifocal leukoencephalopathy (PML) with chronic drug exposure. An important issue in clinical practice is the risk of increased clinical and radiological disease activity following natalizumab withdrawal, which has been well documented (O’Connor et al. Neurology 2011;76:1858-1865; Clerico et al. JAMA Neurol 2014;71:954-960).
Clinical worsening often occurs even after switching from natalizumab to fingolimod (Sempere et al. Acta Neurol Scand 2013;128:e6-e10; Rinaldi et al. Mult Scler 2012;18:1640-1643). This may be due in part to the delayed onset of action of fingolimod, and a shorter washout period (1-2 months) has been proposed (Havla et al. Ther Clin Risk Manag 2013;9:361-369) to potentially reduce the risk of clinical/radiological worsening. A small study examined the effect of the washout duration (60 vs. 90 days) in patients switching from natalizumab to fingolimod (Clares et al. AAN 2015; abstract P3.291). Patients received methylprednisolone 1 g/month during the washout. At one year, 3 of 10 patients in the long washout group experienced a relapse compared to 1 of 10 in the short washout group. EDSS scores were stable in both groups.
Another strategy to minimize the risks associated with natalizumab withdrawal would be to taper the medication. This approach was examined in a one-year phase IV study comparing tapered versus immediate withdrawal (Weinstock-Guttman et al. AAN 2015; abstract P3.287). The taper regimen was two additional courses of natalizumab at 6 weeks and 8 weeks (i.e. over 14 weeks). IV methylprednisolone was permitted. At one year, the number of relapses was three-fold higher (28 vs. 8) in the immediate-discontinuation group. Most relapses occurred in the first three months of discontinuation. Immediate discontinuation was also associated with greater MRI activity.
A number of studies have reported an accelerated rate of brain volume loss in MS patients, and this brain atrophy is a significant predictor of physical and cognitive disability and poorer quality of life (Filippi et al. AJNR Am J Neuroradiol 2010;31:1171-1177; Giorgio et al. Neurol Sci 2010;31:S245-248). As a result, preservation of brain volume is emerging as an important goal of treatment.
The significance of brain volume changes was shown in a five-year longitudinal study (Preziosa et al. AAN 2015; abstract P6.142). MS patients demonstrated a significant accumulation of T2 lesions, brain and spinal cord atrophy, and diffusivity abnormalities in white-matter tracts during the study period. Loss of white matter integrity was predictive of physical disability, whereas grey and white matter damage and brain volume loss were predictive of cognitive worsening. In the ASCEND trial, grey matter volume was moderately correlated with cognitive dysfunction as assessed with the Symbol Digit Modalities Test (SDMT) (Kapoor et al. AAN 2015; abstract P6.143). SDMT findings have also been reported to be predictive of work disability and sick leave (Karrenbauer et al. AAN 2015; abstract P3.222).
A recent meta-analysis of four trials of interferons, glatiramer acetate and fingolimod reported a lower rate of brain volume loss with treatment compared to placebo, with results generally more robust with fingolimod (Tsivgoulis et al. PLoS One 2015;10:e0116511). Additional data on the impact of high-efficacy agents on brain volume are now available. In the STRATA study, a long-term extension of three phase III trials of natalizumab, a subset (268/716, 37%) of patients had evaluable MRI data from two scans at least one year apart (Goodman et al. AAN 2015; abstract P7.260). MRIs were obtained after 4.7-6.7 years, and 5.6-7.6 years of natalizumab exposure. The mean percent brain volume change (PBVC) was -0.253 in the interval between MRIs (mean 319 days). Overall, 71% of patients had brain volume decreases less than 0.4%, which is similar to the normal range of annual brain volume loss in healthy adults.
An analysis of patients from the two CARE-MS extensions examined the impact of alemtuzumab on brain volume loss at four-year follow-up (Coles et al. AAN 2015; abstract P7.263). Brain volume loss was measured by the change in brain parenchymal fraction (BPF). For the CARE-MS I cohort, the mean rate of BPF loss was -0.59% in year 1, and normalized to -0.25% in year 2. Although 75% of patients received no additional courses of alemtuzumab, the rate of BPF loss was further reduced during the two years of the extension (-0.19 in year 3, and -0.15% in year 4). Among previously-treated patients in CARE-MS II, the median BPF loss declined to -0.48% at year 1, -0.22% in year 2, and -0.10% in year 3, and stabilized at -0.19% in year 4.
A separate study looked at brain volume loss in the cohort of CARE-MS patients who switched from interferon-beta SC to alemtuzumab (Cohen et al. AAN 2015; abstract P7.264). In CARE-MS I, the median yearly rate of BPF loss was -0.5% after two years of interferon treatment. Following a switch to alemtuzumab, the median yearly rate of BPF loss was reduced to -0.13% in the second year of the extension. For the interferon group in CARE-MS II, the median yearly rate of BPF loss was reduced from -0.35% at the end of the core phase, to -0.06% after two years of alemtuzumab.
Natalizumab is generally well-tolerated and patients report a high degree of satisfaction during the treatment course (Watson et al. AAN 2015; abstract P3.238). The principal concern is the risk of PML, and the development of the JCV index assay has been proposed as a risk stratification tool (Plavina et al. Ann Neurol 2014;76:802-812). A single-centre study looked at the stability of the JCV index over time (Hoyt et al. AAN 2015; abstract P4.036). At baseline, 40% of patients exceeded a JCV antibody index of 0.9, and 28% exceeded the threshold of 1.5. During the 21-month observation period, 9% of patients crossed over the >1.5 index threshold, suggesting that one-third of anti-JCV Ab+ patients may be at higher risk of developing PML.
PML cases have also been reported in MS patients treated with fingolimod and dimethyl fumarate. It is not known if prior treatment with one of these agents influences the PML risk after switching to natalizumab. Most of the PML cases seen with fingolimod have occurred in patients discontinuing natalizumab, which has generally been attributed to subclinical PML at the time of switch. However, one case of PML has been reported in a fingolimod-treated patient with no prior natalizumab exposure.
To date, no cases of PML have been reported to date with alemtuzumab. The most common adverse effects are infusion-associated reactions (IAR), which were somewhat less frequent during the second treatment course in the CARE-MS II extension (71% vs. 84%); serious IARs occurred in 1.4% of patients during the first and second courses (Havrdova et al. AAN 2015; abstract P7.276). The annual incidence of thyroid adverse effects was 5% in year 1, 10% in year 2, 22% in year 3, and 12% in year 4. The cumulative incidence of immune thrombocytopenic purpura (ITP) after four years was 2.5%. For both CARE-MS studies, the overall incidence of infections declined from 59.9% in year 1 to 46.0% in year 4 (Henson et al. AAN 2015; abstract P7.265). The number of infections by treatment course was 1.34 events/patient-year after the first course, declining 30% to 0.939 with the second course. For the subset of patients receiving three or four courses, the infection incidence was 0.772 and 0.622 events/patient-year, respectively. Most infections were mild to moderate and none led to treatment discontinuation. Common infections were nasopharyngitis, urinary tract infections, and upper respiratory tract infections.