Background
Subcutaneous (SC) peginterferon beta-1a, a pegylated form of interferon (IFN) beta-1a, is a new treatment for relapsing multiple sclerosis (MS) [
1]-[
3]. Pegylation (modification via attachment of polyethylene glycol [PEG] molecules), has been used to increase half-life and reduce the immunogenicity of certain drugs with either increased or sustained efficacy [
4]-[
9]. Results from the first year (placebo-controlled period prior to cross-over onto active drug) of the 2-year pivotal Phase 3, multicenter, randomized ADVANCE study in patients with relapsing-remitting MS (RRMS) showed that peginterferon beta-1a (125 μg SC) administered every 2 or 4 weeks significantly improved clinical endpoints, with a reduction in annualized relapse rate (primary endpoint), and risk of relapse and 12-week confirmed disability progression (both secondary endpoints), and several MRI-related endpoints versus placebo, with a safety profile reflecting established IFN beta-1a therapies [
10]. Furthermore, every 2 week dosing provided numerically greater reductions relative to placebo versus every 4 week dosing across relapse and all MRI endpoints [
10].
Magnetic resonance imaging (MRI) of lesions in the brain is an important tool for monitoring MS disease activity and progression [
11],[
12], providing a sensitive measure of the effects of treatment on focal inflammatory activity in the central nervous system [
13], which initiates the demyelination and axonal damage associated with MS disease progression [
14],[
15]. Lesions detected on MRI scans have been shown to correlate with MS prognosis [
16]-[
18]; thus early reduction in MRI lesions is an important treatment goal [
12].
With the development of effective biological therapies, no evidence of disease activity (NEDA; absence of clinical [defined as no relapses and no onset of 12-week confirmed disability progression] and MRI [no gadolinium-enhancing lesions and no new or newly-enlarging T2 hyperintense lesions] disease activity) is now achievable in many patients with RRMS [
19]. As treatments have improved, annualized relapse rates recorded in clinical studies have decreased to the point where there is a need to consider alternative endpoints to evaluate new treatments for RRMS [
20]. NEDA has been proposed for use as an endpoint in MS studies [
20]. It has been evaluated in several MS studies to date [
13],[
19],[
21],[
22], with some variation in its definition to allow for minimal MRI activity [
23]. Although complete NEDA would be the ideal treatment aspiration, for many physicians, local guidelines and/or switch recommendations apply (e.g. two or more new or newly-enlarging T2-lesions or even a clinical relapse with incomplete remission before escalation is recommended) [
24].
The objectives of the present analyses were to explore the effect of peginterferon beta-1a every 2 and 4 weeks on MS disease activity as measured by brain MRI lesions during Year 1 of the ADVANCE study versus placebo (at Week 48 ([Year 1], as previously reported [
10], and at the earlier time point of Week 24), with presentation of new post-hoc statistical analyses for every 2 week versus every 4 week dose regimen comparisons at these time points, and to determine the effect of peginterferon beta-1a on the combined endpoint of NEDA via post-hoc investigations of Year 1 data (looking at a range of time-scales to evaluate timing of NEDA impact: baseline to Week 48, baseline to Week 24, and Week 24 to Week 48). Sensitivity analyses of disease activity were also carried out to include an allowance for patients who did not undergo all MRI measures during the study, but had no evidence of disease activity on any recorded assessment (last observation carried forward [LOCF] approach), and using alternative definitions that mimic thresholds used in clinical practice to guide decision-making on treatment continuation due to efficacy, or treatment switch due to inadequate efficacy (no clinical activity with minimal MRI activity).
Methods
Study design and participants
We conducted a randomized, multicenter (at 183 sites in 26 countries), double-blind, placebo-controlled, parallel group Phase 3 study of peginterferon beta-1a in patients with RRMS [
10]. Key eligibility criteria were diagnosis of relapsing MS as defined by the McDonald criteria [
25], an age of 18–65 years, a score of 0–5 on the Expanded Disability Status Scale (EDSS [
26]), and at least two clinically documented relapses in the previous three years, with at least one of these relapses having occurred within 12 months prior to randomization. Key exclusion criteria were progressive forms of MS, pre-specified laboratory abnormalities, and prior interferon beta treatment for MS exceeding four weeks or discontinuation less than six months prior to baseline. The protocol was approved by each site’s institutional review board or ethics committee (a list of ethics committees for participating sites is provided in Additional file
1). The ethics committee at Johns Hopkins University (the site of the Principal Investigator) was Johns Hopkins Medicine Institutional Review Board, Baltimore, MD, USA. The study was conducted according to the International Conference on Harmonization Guidelines for Good Clinical Practice and the Declaration of Helsinki. Every patient provided written informed consent prior to study entry.
Study procedures and endpoints
The methods of the ADVANCE study have been published previously [
10]. Here, we focus on information specific to the analyses presented.
Patients were randomly assigned in a 1:1:1 ratio to receive SC injections of placebo (Year 1 only), peginterferon beta-1a at a dose of 125 μg every 2 weeks, or peginterferon beta-1a 125 μg every 4 weeks. Patients received either peginterferon beta-1a or placebo every 2 weeks to maintain the blind; those randomized to the peginterferon beta-1a every 4 week group received alternate injections of placebo and peginterferon beta-1a every 4 weeks. Participants randomized to receive peginterferon beta-1a at beginning of study, and those re-randomized to active drug after first year, underwent dose escalation over the first four weeks of the treatment to mitigate influenza-like symptoms (starting dose 63 μg; 94 μg at Week 2; target dose 125 μg at Week 4 and thereafter) [
10]. The current analysis evaluates the effects of receiving peginterferon beta-1a vs. placebo in Year 1 (baseline to 48 weeks) of the ADVANCE study. All study management and site personnel, investigators, and patients were blinded to treatment assignment including dose escalation. Each site used separate examining and treating neurologists, thereby maintaining rater blinding for all treatment groups.
MRI scans were obtained at screening, and at Weeks 24, and 48, and were evaluated in a blinded manner at a central MRI reading center (NeuroRx Research, Montreal, Canada).
The number of new or newly-enlarging T2 hyperintense lesions at Week 48 compared with baseline was a secondary endpoint of the study. Tertiary MRI endpoints included the number of new or newly-enlarging T2 hyperintense lesions at Week 24, the number of gadolinium-enhancing (Gd+), new T1 hypointense, and new active (sum of Gd+ plus non-enhancing new or newly-enlarging T2 hyperintense) lesions, at Weeks 24 and 48, and the volume of T2 and T1 hypointense lesions, at Weeks 24 and 48.
Post-hoc NEDA analyses during Year 1 of the ADVANCE study were carried out and summarized as data from baseline to Week 48, baseline to Week 24 and Week 24 to Week 48 using the following definitions: NEDA was defined as absence of both clinical [no relapses and no onset of 12-week confirmed disability progression over the interval] and MRI [no Gd+ lesions and no new or newly-enlarging T2 hyperintense lesions] disease activity during the respective time periods; MRI-NEDA only was defined as no Gd+ lesions at Weeks 24 and 48 and no new or newly-enlarging T2 lesions at Week 48 compared with baseline (for data summarized from baseline to Week 48), no Gd+ lesions at Week 24 and no new or newly-enlarging T2 lesions at Week 24 compared with baseline (for data summarized from baseline to Week 24), and no Gd+ at Week 48 and no new or newly-enlarging T2 lesions at Week 48 compared with Week 24 or the closest previous visit before Week 48 (for data summarized from Week 24 to Week 48); clinical-NEDA only was defined as no relapses and no onset of 12-week confirmed disability progression over the intervals specified (including 12 weeks confirmation at Week 60). Sensitivity analyses were also carried out on data from baseline to Week 48 to include an allowance for patients with missing measures of MRI disease activity (analysis 1 below) and using definitions with an allowance for minimal MRI activity (analyses 2 and 3 below):
-
Sensitivity analysis 1: definitions described above were applied using LOCF (i.e. patients were not excluded from overall NEDA analyses if they did not have complete MRI scan data during Year 1, and were not excluded from MRI-NEDA only analyses if they only had partial MRI scan data during Year 1; patients were excluded from MRI-NEDA only analyses if they were missing all MRI scan data during Year 1).
-
Sensitivity analysis 2: sum of number of new or newly-enlarging T2 lesions at Week 48 compared with baseline and Gd+ lesions at both Week 24 and Week 48 ≤2.
-
Sensitivity analysis 3: ≤1 new or newly-enlarging T2 lesion at Week 48 compared with baseline and no Gd+ lesions at Weeks 24 and 48.
Statistical analysis
Statistical tests were two-sided. Multiple comparisons adjustments were performed on the primary and secondary endpoints for the study, with an overall type I error rate of 0.05. All primary and secondary endpoints for the study were statistically significant for both dose frequencies [
10]. However no multiple comparisons adjustments were performed on tertiary endpoints or post-hoc analyses, so nominal p-values are reported here for each endpoint.
MRI results were derived from patients who had any MRI data (a majority of the overall study intention-to-treat [ITT] population [randomized patients who received at least one dose of study treatment]) [
10]. Negative binomial regression was used for analysis of new or newly-enlarging hyperintense lesions on T2-weighted images (adjusted for baseline number of T2 lesions) and new active lesions (adjusted for baseline number of Gd+ lesions); multiple logit regression was used for the analysis of Gd+ and new T1-hypointense lesions (adjusted for baseline number of respective lesions) [
10]. Analysis of covariance (ANCOVA) was used for analysis of lesion volumes, adjusted for respective baseline lesion volumes [
10]. Comparisons of active treatment dosing regimens versus placebo were pre-specified; every 2 week versus every 4 week comparisons were post-hoc. Post-hoc NEDA and disease activity sensitivity analyses (no clinical activity with minimal MRI activity) proportions were calculated directly based on the definitions using data from all eligible patients. A logistic regression model was used to calculate odds ratios (ORs) and corresponding p-values.
Discussion
The MRI outcomes of peginterferon beta-1a observed here are consistent with the robust benefits on clinical outcomes after one year of treatment (reductions in risk of 12-week confirmed disability progression [every 2 and 4 weeks 38%], risk of relapse [every 2 weeks 39%; every 4 weeks 26%], and annualized relapse rate [every 2 weeks 36%; every 4 weeks 28%]), as reported previously for the Phase 3 ADVANCE study [
10]. Peginterferon beta-1a every 2 weeks provided superior outcomes for MRI lesion-based endpoints (including numbers of new or newly-enlarging T2 lesions, new T1 lesions, Gd+ lesions, and new active lesions, and volumes of new or newly-enlarging T2 lesions, and new T1 lesions) versus placebo and versus peginterferon beta-1a every 4 weeks, with statistical significance observed at the first brain MRI scheduled at Week 24 of treatment, and greater reductions through to Week 48 of the ADVANCE study. Peginterferon beta-1a every 4 weeks also provided superior results versus placebo at Weeks 24 and 48 for numbers of new or newly-enlarging T2 lesions and new active lesions, and the volume of T2 lesions, but not for numbers of Gd+ or T1 lesions, nor volume of T1 lesions.
An early decrease in brain inflammation, as shown by reductions in MRI disease activity, is particularly desirable in MS, as it may impact on associated injury to myelin and axons, with consequences for the accumulation of disability [
12]. Indeed, MRI lesion measures have been shown to be strong predictors of clinical prognosis for MS patients [
15]-[
18],[
27],[
28].
This study also carried out post-hoc analyses of ADVANCE Year 1 data using composites of clinical and radiological indicators of disease activity, which are commonly used as endpoints in MS clinical trials [
29]-[
33]. Efficacy data show that proportions of patients with NEDA and MRI-NEDA only) were significantly higher with peginterferon beta-1a every 2 weeks from baseline to Week 48, from baseline to Week 24 and from Weeks 24 to 48, versus placebo and peginterferon beta-1a every 4 weeks. Furthermore, the sensitivity analyses of disease activity conducted support the robustness of the results of the primary NEDA analyses, indicating that the effects of peginterferon beta-1a are sustained when including an allowance for patients who did not undergo all MRI measures during the study, but had no evidence of disease activity on any recorded assessment, and when applying definitions with allowances for minimal MRI activity. These observations are important because brain MRIs within 12 months of starting a disease-modifying therapy are a sensitive tool in predicting response to treatment, along with the early identification of patients who will develop long-term disability progression [
15],[
27],[
28],[
34]. MRI allowance is pertinent to current practice, as scans obtained 6–12 months after starting a new therapy may include MRI activity occurring before efficacy of a new therapy is apparent. Of note, fewer lesions were seen in the present study from Weeks 24–48, versus baseline to Week 24, and comparative ORs for NEDA and MRI-NEDA were higher from Week 24 to 48, compared with baseline to Week 24 and baseline to Week 48. Therefore, an MRI baseline established 6 months after treatment initiation may more effectively demonstrate differences between treatment arms. Further, activity on MRI may also indicate a need to switch to a treatment with an alternative mode of action, with the aim of avoiding irreversible clinical neurological disability [
34],[
35]; however this need may be more appropriately assessed by establishing a baseline for new disease activity (e.g. at 24 weeks) thus avoiding switch to potentially less safe treatments.
NEDA analyses in the ADVANCE study reflect efficacy levels previously reported in other Phase 3 trials of agents approved for the treatment of relapsing forms of MS, however, it is important to consider that direct comparisons across studies may not be advisable due to differences between study design, patient populations, MRI techniques and MRI reading techniques used across different studies. In the FREEDOMS II study in RRMS, at Month 24 significantly more patients had freedom from indicators of new inflammatory activity (i.e. no gadolinium-enhanced T1 lesions and no new or newly-enlarged T2 lesions) on fingolimod (0.5 mg 50%; 1.25 mg 63%) compared with those on placebo (26%); no composites of clinical-NEDA or complete NEDA were reported in the FREEDOMS II study [
36]; although no OR were reported, the relative differences between placebo and actives arms are similar to the findings in the present study.
It is important to compare methods of analysis (e.g. LOCF or non-LOCF), sensitivity of MRI analyses (e.g. computerized or visual read of scans) and numbers of scans when trying to estimate efficacy by endpoints like NEDA, or FMDA (Freedom from Measured Disease Activity), as termed by others, among different treatment options; it would therefore be optimal to make comparisons versus placebo-treated patients and to also compare OR results, rather than just absolute numbers.
Further analyses will be necessary to determine whether these early patients who had NEDA in the ADVANCE study remain NEDA over the longer term (i.e., after 96 weeks). A multicenter, dose-frequency blinded minimum 2-year extension of the ADVANCE study is also underway (ATTAIN; NCT01332019), which may allow for the provision of 4-year data.
Competing interests
This study was sponsored by Biogen Idec Inc. (Cambridge, MA, USA). DLA has received consultancy fees from Biogen Idec Inc. and holds stock in NeuroRx Research, which performed the central MRI analysis. PAC has provided paid consultation services to Vertex, Vaccinex, Prothea, Abbott and MedImmune, and has provided consulting services to Biogen Idec Inc. and other MS companies as an unpaid consultant. PAC has also received grant support from Biogen Idec Inc., and Novartis. BCK has received honoraria for lecturing, travel expenses for attending meetings, and financial support for research from Bayer Health Care, Biogen Idec Inc., Genzyme/Sanofi Aventis, Merck Serono, Mitsubishi Europe, Novartis, and TEVA. SIS, AD, YZ, SL, XY, BS and SH are employees of Biogen Idec Inc.
Authors’ contributions
All authors were involved in revising the manuscript critically for important intellectual content. All authors read and approved the final manuscript. DLA: contributed to the conception of this work, and was involved in the interpretation of data. PAC and BCK were involved in interpretation of data. SIS, AD, YZ, SL, XY, BS, and SH contributed to the design of this work, and provided analysis and interpretation of data.