Abstract
Objective: To evaluate the efficacy and safety of interferon beta-1b (IFNβ-1b) in subjects with secondary progressive multiple sclerosis (SPMS).
Methods: This 3-year, multicenter, double-blind, placebo-controlled, randomized trial of IFNβ-1b included 939 subjects from the United States and Canada with SPMS and Expanded Disability Status Scale (EDSS) scores ranging from 3.0 to 6.5. Subjects were randomly assigned to receive either placebo or IFNβ-1b (250 μg or 160 μg/m2 body surface area), administered subcutaneously every other day. The primary outcome was time to progression by ≥1.0 EDSS point (0.5 point if EDSS score was 6.0 to 6.5 at entry) confirmed at 6 months. Secondary outcomes included mean change in EDSS score from baseline, relapse-related measures, MRI activity, and a standardized neuropsychological function test.
Results: There was no significant difference in time to confirmed progression of EDSS scores between placebo-treated patients and either of the IFNβ-1b treatment groups. However, IFNβ-1b treatment resulted in improvement on secondary outcome measures involving clinical relapses, newly active MRI lesions, and accumulated burden of disease on T2-weighted MRI. Effects were similar for both IFNβ-1b treatment groups. Neutralizing antibodies to IFNβ-1b were detected in 23% of 250-μg and 32% of 160-μg/m2 recipients, but their presence did not consistently affect clinical or MRI outcomes. IFNβ-1b was also well tolerated at both doses.
Conclusions: Although no treatment benefit was seen on the time to confirmed progression of disability, relapse- and MRI-related outcomes showed significant benefit with both dosing regimens tested, a result consistent with the outcomes of earlier clinical trials.
Most subjects with multiple sclerosis (MS) initially experience recurrent symptoms, accompanied by transient disability, that reflect partially reversible focal inflammation and demyelination.1 Subjects who remain clinically stable between these recurrent attacks (relapses) are classified as having relapsing-remitting (RR) MS.2 Within 10 years, approximately 50% of RR subjects will transition to the secondary progressive (SP) phase of MS. This phase is distinguished from RRMS by gradual progression of disability, either between acute relapses or in the absence of relapses.2 In general, focal inflammation is considered to be less pronounced during the SP phase of MS, and increasing disability presumably reflects cumulative and irreversible axonal loss.3,4⇓
Interferon β (IFNβ) has been shown to benefit both patients with an initial demyelinating episode who are at high risk to develop MS and patients with relapsing forms of MS (either RRMS or SPMS with ongoing relapses). For example, IFNβ-1a administered once weekly delays the onset of a second demyelinating event.5,6⇓ In addition, IFNβ has been consistently shown to reduce clinical relapses and MRI activity in patients with relapsing forms of MS,7–12⇓⇓⇓⇓⇓ the results of these trials leading to the approval of IFNβ-1b and IFNβ-1a for use as disease-modifying therapies for RRMS in North America, Europe, and other parts of the world. Subsequent trials13,14⇓ and reviews of these and other studies15 suggest that higher-dose or more frequently dosed IFNβ or both offers greater and more robust clinical benefits than once-weekly IFNβ.
Four clinical trials of IFNβ in subjects with SPMS have been conducted to date.16–18⇓⇓ The European trial of IFNβ-1b in SPMS16 included 360 patients allocated to receive 250 μg of IFNβ-1b and 358 allocated to receive placebo. A highly significant treatment effect was noted on the measure of confirmed progression, representing a delay in progression of 9 to 12 months over 2 to 3 years of study. Other disability-related outcomes showed a positive benefit from therapy. Benefits on secondary clinical and MRI outcome measures were also observed, including a 30% reduction in mean overall relapse rate and a 65 to 78% reduction in new enhancing MRI lesions. The benefits of IFNβ-1b were independent of baseline Expanded Disability Status Scale (EDSS) score or the presence of relapses either before or during the trial. These results led to the approval of IFNβ-1b in Europe and Canada as a disease-modifying therapy for subjects with SPMS.
Two clinical trials of IFNβ-1a in patients with SPMS have also been reported.17,18⇓ Neither trial showed a beneficial effect on the outcome of time to confirmed EDSS progression. However, in the Secondary Progressive Efficacy Clinical Trial of Recombinant Interferon β-1a in MS (SPECTRIMS) Study, a benefit of therapy on disease progression was seen when the subgroup of patients who had experienced prestudy exacerbations was analyzed separately.19 The International MS Secondary Progressive Avonex Controlled Trial (IMPACT) Study included the Multiple Sclerosis Functional Composite (MSFC) as the primary endpoint based on concerns over the sensitivity and reliability of the EDSS to detect a treatment outcome.18 IMPACT was able to show a modest effect on the MSFC, although this was largely attributable to benefit on the 9-Hole Peg Test. Both the SPECTRIMS and the IMPACT trials showed a treatment benefit on multiple secondary clinical and MRI outcomes. In this article, we report the design and the clinical and MRI results of a second trial to assess the efficacy and safety of IFNβ-1b in SPMS, conducted in North America and completed in 2000, which complements and extends the results of the earlier European trial of IFNβ-1b.16
Methods.
Objective.
We sought to evaluate the efficacy and safety of IFNβ-1b at two doses in subjects with SPMS.
Thirty-five centers (see the Appendix) participated in this double-blind, placebo-controlled trial of IFNβ-1b for the treatment of SPMS. The trial consisted of four parallel-treatment groups with a planned 3-year study period. Clinical evaluations were scheduled every 12 weeks, and brain MRI was performed yearly. A frequent-scan cohort of 163 patients at five sites was studied with gadolinium (Gd)-enhanced brain MRI every 4 weeks for 3 years. All relapses were assessed at either scheduled or unscheduled visits. In both cases, details of relapses and other non-MS-related medical events were documented. Subjects were monitored for adherence to all study procedures throughout the trial unless they withdrew their consent or were lost to follow-up.
At each scheduled visit, patients underwent physical and neurologic examinations, assessment for adverse events, concomitant medications, and basic laboratory testing for safety assessment. Unscheduled visits typically occurred for suspected relapse or adverse event. In cases of suspected relapse, the patient was seen within 14 days by both the treating and the evaluating physicians and underwent physical and neurologic examination with recording of EDSS, Functional System (FS), and Scripps20 scores. A relapse was defined as the appearance of a new, or reappearance of a previous, neurologic symptom, present for ≥48 hours and not attributable to fever, infection, or withdrawal of corticosteroid therapy. Severity of relapses was determined by change in the Scripps score, with mild defined as a decrease of 0 to 7 points, moderate a decrease of 8 to 15 points, and severe a decrease of >15 points. A specific numerical change in the score was not required for confirmation of a relapse by the examining physician.
The study was supervised by a Steering Committee of investigators and sponsor staff. Members of the Steering Committee, with the exception of the study statistician, were blind to treatment assignment. As required by protocol, an Independent Data and Safety Monitoring Board (IDSMB) reviewed interim safety and efficacy data every 6 months.
Subjects and treatment.
This study was approved by the institutional review boards of each participating center, and all subjects provided informed consent. Patients eligible for randomization included those who met the following entry criteria: 1) age 18 to 65 years, 2) clinically definite or laboratory-supported definite MS21 of at least 2 years’ duration, 3) a history of at least one relapse followed by progressive deterioration sustained for at least 6 months, 4) an EDSS score at screening of 3.0 to 6.5 inclusive, and 5) an increase in EDSS score of at least 1.0 point in the 2 years prior to screening (at least a 0.5-point increase for subjects with a screening EDSS score of 6.5). Subjects were excluded if they 1) received treatment with systemic corticosteroids or adrenocorticotropic hormone within 60 days before the screening visit, 2) were previously treated with any IFNβ, monoclonal antibody, cladribine, or total lymphoid irradiation, or 3) received cytotoxic or immunosuppressive therapy, glatiramer acetate, or other investigational drug within 6 months before the screening visit.
Subjects were randomly assigned to receive one of the following treatments injected subcutaneously every other day: IFNβ-1b 250 μg (8.0 million international units [MIU]), IFNβ-1b 160 μg (5 MIU)/m2 body surface area, or one of two placebo treatments that lacked active drug but were otherwise identical in composition, appearance, and volume to the corresponding IFNβ-1b dosing arm. Randomization allocation was by blocks of six, such that subjects received IFNβ-1b 250 μg (8.0 MIU), IFNβ-1b 160 μg (5 MIU)/m2 body surface area, or placebo in a ratio of 1:1:1. At the start of the study, each site received an adequate number of blocks, based on assumed patient recruitment, to ensure sequential patient numbering within the site. The randomization schedule was generated by the Biostatistics and Data Management Group of Berlex Laboratories (Richmond, CA) using an SAS program (Cary, NC).
The 160-μg/m2 treatment arm (and its placebo) were included in the study to assess the possibility that body mass might influence the response to IFNβ. Recipients of the 250-μg dose initiated treatment at 0.25 mL (62.5 μg of IFNβ-1b) and increased dosing by 0.25 mL each week, until the maximum dose was achieved. The concentration of IFNβ administered to recipients of the 160-μg/m2 dose was 375 μg/mL. Depending on body surface area, the initial dose for these subjects ranged from 0.15 to 0.25 mL and was escalated by 0.10 to 0.25 mL each week to the maximum dose of 0.50 to 1.0 mL (187.5 to 375 μg).
In addition to dose escalation, ibuprofen 400 mg was administered three times daily every day that study drug was administered for the first 7 weeks to reduce drug-related adverse events and to minimize any consequent unblinding. Thereafter, ibuprofen was administered at the discretion of the treating neurologist and patient. Standardized treatment with systemic glucocorticosteroids (1 g of methylprednisolone administered IV for 5 days) was permitted for clinical relapses at the discretion of the treating physician.
Efficacy and blinding.
To avoid unblinding of treatment assignment, separate treating and examining physicians were employed. Treating physicians were responsible for the general medical care of each subject, safety assessments, and treatment of relapses. Examining physicians were responsible for completing standardized neurologic evaluations and were not permitted access to previous examination results or any other information that could potentially unblind them to treatment assignment. For this reason, injection sites were concealed when subjects were in the presence of the examining physician.
Prior to the study, all examining physicians completed training to review standardized procedures for assessment of ambulation and for determination of individual FS scores and EDSS scores. This training included a review of videotaped neurologic assessments, manuals, and written guidelines. Whenever possible, the same examining physician performed all scheduled neurologic assessments for a given subject throughout the study.
A questionnaire to evaluate the success of blinding was completed by the subjects, treating physicians, and examining physicians at week 12 and on completion of the study. Each was asked to indicate whether the subject was receiving placebo, IFNβ-1b, or “don’t know.”
The primary outcome measure was the number of days from the start of treatment to the first recorded increase of ≥1.0 point from the baseline EDSS score (≥0.5 point if the baseline EDSS score was 6.0 to 6.5) confirmed at two consecutive scheduled examinations spanning ≥6 months from the onset of progression. Confirmed progression could also be established at not less than 70 days if the first increase in EDSS score was detected at the penultimate study visit.
Secondary and tertiary clinical and MRI outcome measures of efficacy are outlined in table 1. These included a variety of relapse-related and MRI-related measures, interventions, social handicap, quality of life, and depression.
Table 1 Secondary and tertiary outcome measures
At each scheduled visit, serum samples were obtained from each subject for the determination of neutralizing antibody (NAb). NAb activity was determined using the MxA assay.22 This assay is based upon the inhibition (by the patient’s serum) of the capacity for IFNβ-1b to induce the synthesis of the MxA protein in cell culture. The titer (expressed in neutralizing units [NU]) obtained in this assay is defined as the serum dilution that reduces the activity of 10 IU/mL of IFNβ-1b to an activity of 1 IU/mL, corresponding to 90% inhibition. Subjects were considered to be “eventually” NAb positive if they had NAb titers of ≥20 NU in two consecutive serum samples. Similarly, reversion to stable NAb-negative status was defined by two consecutive titers below 20 NU and no more appearance of two consecutive positive titers for the remaining observation period. Cross-sectional and longitudinal analyses of NAb influence were performed as described elsewhere.23 For the longitudinal analyses, two definitions for eventually Nab-positive patients were used: “once positive, always positive,” where subsequent changes in NAb status are ignored; and “all switches considered,” where changes in NAb status are included in the analysis.23
Statistical analyses.
Determination of sample size for this study included assumptions that the proportions of placebo and IFNβ subjects who experienced confirmed progression at 3 years would be 50% in the placebo group and 35% in the treated groups. Three hundred subjects were required in each treatment arm, including an adjustment for 10% of subjects who were lost to follow-up, to meet these assumptions in a two-sided log-rank test at an α level of 0.05 and 95% statistical power. The α adjustments for seven interim analyses of efficacy required by study protocol (p = 0.0007 to 0.0016) and a final analysis of efficacy (p = 0.048) were based on a previously reported method.24 All statistical analyses were based on the intention-to-treat population, including all data from all subjects as randomized to 3 years or loss to follow-up.
For continuous data, comparability of treatment groups was determined using a two-way analysis of variance (ANOVA) with treatment, study site, and treatment-by-study-site interaction included in the model. The Cochran-Mantel-Haenszel (CMH) test stratified for study site was used to compare treatment groups for categorical data. The CMH ANOVA test, with equally spaced weights, was used for ordinally scaled variables.
Clinical outcomes of efficacy.
Efficacy analyses were based on the intention-to-treat population, comprising all data of all patients as randomized without any restrictive criteria. All patients are included in all summary tables to the extent of available data. Missing data were not replaced. The two placebo groups were pooled for statistical comparisons with the active treatment arms.
The distributions of time to confirmed progression by treatment group were described by the Kaplan-Meier product-limit approach, and differences between curves were assessed using the logrank test. Change in mean EDSS score was assessed using an ANOVA for change in mean EDSS score, and a logistic regression was used to assess the proportion of subjects progressing. Subject-specific annual relapse rates were computed by dividing the number of confirmed relapses by the time on study for each subject. Treatment group differences in annual relapse rate, annual rate of moderate and severe relapses, mean relapse duration, and days spent in moderate and severe relapse were assessed using an ANOVA model. Nominal categorical variables, such as the presence or absence of a response, were tested for general association with the CMH test stratified for study site. Changes in categorical variables with an ordinal scale by treatment group were tested for differences using the CMH ANOVA test with equally spaced weights. Significance of change in continuous variables by treatment group was assessed with a two-way ANOVA with study site, treatment group, and their interaction as the fixed effects in the model. Statistical plans for other outcome measures were defined prospectively and are available in the supplementary information (tables E1–E8 on the Neurology Web site at www.neurology.org).
Results.
Study population.
Figure 1 shows the disposition of 939 subjects who were randomly assigned to receive IFNβ-1b 250 μg (n = 317), IFNβ-1b 160 μg/m2 (n = 314), or one of the two placebos (n = 308). These three groups were well balanced for baseline demographics, disease characteristics, and MRI variables (table 2). The mean duration of follow-up was 998 days for the IFNβ-1b 250-μg group, 1,013 days for the IFNβ-1b 160-μg/m2 group, and 1,003 days for the placebo group. The average body surface area was 1.83 m2 for the 250-μg group and 1.84 m2 for the 160-μg/m2 group. The mean assigned doses were 250 and 300 μg, whereas the mean administered doses were 206 and 220 μg, respectively. Although the mean assigned and administered doses were slightly higher in the variable-dose group, the differences were small and not statistically meaningful.
Figure 1. Disposition of subjects. AE = adverse event; PD = protocol deviation; PoD = progression of disease; W/C = withdrew consent (includes lost to follow-up and noncompliance).
Table 2 Summary of principal baseline characteristics*
At a planned interim analysis, the IDSMB recommended early termination of the trial based on the results of a stochastic curtailment analysis that indicated that continuing the trial was unlikely to change the results of the primary efficacy outcome. Over 75% of subjects completed the study on their assigned treatment, and approximately 85% of the anticipated data set was available for analysis.
Outcome measures of efficacy.
Primary outcome.
There was no difference in the comparison of the time to confirmed EDSS progression in the pooled IFNβ-1b vs placebo recipients (logrank test: p = 0.71) (figure 2) or for the individual comparisons of 250 μg vs placebo (p = 0.61) and 160 μg/m2 vs placebo (p = 0.26). This result was unchanged following statistical adjustments for center, baseline EDSS, and duration of MS. There was also no difference in the primary outcome in the subgroup of subjects who completed the study on assigned treatment.
Figure 2. Time to 6-month confirmed progression. The time to 6-month confirmed progression was not delayed in interferon β-1b (IFNβ-1b)-treated patients (p = 0.712 for all IFNβ-1b vs placebo). Time to the 30% quantile was 750 days for the placebo group, 981 days for the 250-μg group, and 668 days for the 160-μg/m2 group. The differences between the 250-μg and placebo groups (p = 0.606) and between the 160-μg/m2 and placebo groups (p = 0.261) were not significant.
Secondary outcomes.
Table E-3 (available on the Neurology Web site at www.neurology.org) outlines the results from the secondary measures of efficacy. There was no apparent effect of treatment on the change in mean EDSS score from baseline to endpoint.
A significant treatment benefit was observed on the annual relapse rate for both the pooled IFNβ-1b group and the 250-μg arm, but not for the 160-μg/m2 arm considered alone. Overall, IFNβ-1b subjects showed a 36% reduction in annual relapse rate (43% for the 250-μg group).
A strong treatment effect was observed for absolute change in T2-weighted lesion area between baseline and endpoint. The median change from baseline to endpoint was 10.9% for the placebo group, 0.4% for the 250-μg group (p < 0.0001 vs placebo), and 0.8% for the 160-μg/m2 group (p < 0.0001 vs placebo). A similarly strong treatment effect was observed on the annual newly active lesion rate in the frequently scanned MRI cohort, with an overall 71% reduction in newly active lesions in the combined treated group compared with placebo (p < 0.001).
The composite neuropsychological score, corrected for practice effects by standardization against a matched healthy control cohort of 141 individuals, showed no differences between treatment groups at any time point. The patterns of gradual decrease over time appeared to be similar between the treatment groups.
Tertiary outcomes.
The tertiary outcome measures further supported the efficacy of 250 μg of IFNβ-1b treatment (table E-4). However, the results for the 160-μg/m2 arm were generally less impressive. Time to first relapse was increased in the 250-μg treatment group compared with placebo (p = 0.010) (figure 3), and, correspondingly, the proportion of subjects who were relapse-free during the trial was increased (71 vs 62%; p = 0.018). Intervention with systemic steroids was also less frequent in the 250-μg treatment arm.
Figure 3. Time to first relapse. The time to experiencing the first on-study relapse was prolonged in interferon β-1b (IFNβ-1b)-treated patients (p = 0.023 for all IFNβ-1b vs placebo). Time to the 30% quantile was 487 days for the placebo group, 1,051 days for the 250-μg group, and 810 days for the 160-μg/m2 group. The difference between the 250-μg and placebo groups was also significant (p = 0.010); the difference between the 160-μg/m2 and placebo groups was not (p = 0.195).
NAbs.
NAbs to IFNβ-1b, confirmed with at least two consecutive positive titers, were found for one placebo subject, 22.7% of subjects in the 250-μg group, and 32.5% of subjects in the 160-μg/m2 arm. Almost all subjects who developed NAbs did so within the first 24 months of the trial. Of these, 54.2% in the 250-μg and 54.9% in the 160-μg/m2 group later reverted to stable NAb-negative status during the 3-year observation period.
There was no adverse impact of NAb on relapse rate when evaluated using cross-sectional analyses. Indeed, the initial relapse rate in patients destined to become NAb positive appeared to be lower than in those patients who never developed NAb (figure 4). With use of longitudinal analyses, an effect of switching to NAb-positive status was noted (figure 5), but this was not consistent between the two dosing groups or across different cut-off NAb titers or definitions of NAb positivity. When analyzed according to the definition of “once positive, always positive,” there was some evidence for a treatment-attenuating effect of NAb (p = 0.01); however, the effect was not robust, as reflected by the large 95% CI (28%, 897%). There was no attenuating effect of NAb on relapse rate in the analyses where the “all switches considered” definition was applied.
Figure 4. (A) Cross-sectional analysis of annualized relapse rates (cut-off ≥ 20; 250 μg of interferon β-1b [IFNβ-1b]). *p = 0.003 vs neutralizing antibody (NAb)-negative subgroup; **p = 0.005 vs NAb-negative subgroup. (B) Cross-sectional analysis of annualized relapse rates (cut-off ≥ 20; 160 μg/m2 of IFNβ-1b). *p = 0.01 vs NAb-negative subgroup.
Figure 5. Longitudinal analysis of effect on neutralizing antibody (NAb) titer on percentage estimated increase in relapse (cut-off ≥ 20; mean ± SE). p = 0.01 for the 250-μg once positive, always positive group.
NAb had no consistent influence on the change in EDSS from baseline or on MRI outcomes using either cross-sectional or longitudinal analyses. Longitudinal analyses revealed no discernable pattern of effects related to a switch from NAb-negative to NAb-positive status for MRI lesion area and annual newly active lesion rate.
Treatment was safe and generally well tolerated (table E-3). Adverse events led to discontinuation of therapy in 4% of placebo, 9% of 250-μg, and 10% of 160-μg/m2 recipients. Serious adverse events were no more frequent in the treated or placebo groups. Death occurred in seven subjects during the study, but none was considered to be related to treatment. Three suicide attempts were seen, all in the IFN-treated groups. There was no association between new or worsened depression and treatment with IFNβ-1b as reflected by analyses of quarterly monitoring using the Beck Depression Inventory, spontaneous adverse event reporting, or use of antidepressant medications.
Flu-like symptoms and injection-site reactions were common during the early weeks of treatment but became less frequent through the course of the study. Injection site necrosis was reported in 5% of pooled IFNβ-1b recipients.
Leukopenia, in particular lymphopenia, was more common in treated subjects. Small increases in liver transaminases were seen after initiation of treatment, but these were not clinically meaningful, and there was a trend toward normalization over approximately 12 months. Incidence of grade 3 or 4 toxicity was ≤2% in all groups for serum glutamic-oxaloacetic and serum glutamic-pyruvic transaminases (table E-4).
Results of the blinding survey indicated that evaluating physicians were successfully blinded to treatment allocation at both 12 and 156 weeks. At 156 weeks, evaluating physicians were rarely able to guess the correct treatment allocation for any individual patient. Patients and treating physicians were more likely to guess treatment allocation correctly (table E-5).
Discussion.
The North American trial of IFNβ-1b in SPMS failed to show a treatment benefit on the primary outcome of time to sustained progression defined as a confirmed 1.0-point increase in EDSS score (or 0.5-point increase for EDSS score of 6.0 to 6.5). Nevertheless, IFNβ-1b treatment consistently showed benefits on both relapse- and MRI-related secondary and tertiary outcomes assessed in this study (see tables E-1 and E-2). These benefits are similar to those seen in other trials of IFNβ in RRMS and SPMS.7–12,16–18⇓⇓⇓⇓⇓⇓⇓⇓ Treatment with IFNβ-1b was well tolerated over the 3-year study duration, and no new or unexpected adverse events were observed.
Since the introduction of IFNβ-1b in 1993 as the first therapy to modify the course of RRMS, numerous studies in RRMS have shown a consistent profile of benefits. Reductions in relapse rate ranging from 18%10 to 34%, 9,11⇓ together with substantial beneficial effects on MRI activity and the accumulation of disease burden, clearly demonstrate the utility of IFNβ in the treatment of RRMS. In SPMS, the trials undertaken are also consistent in reporting relapse-related and MRI treatment effects comparable with those seen in the RRMS trials. Less consistent findings between the trials, however, have been observed in the effect of IFNβ on EDSS progression.
In the current trial, there is an apparent dissociation between the nonsignificant findings on the primary outcome and the highly significant findings on the majority of other outcomes. Several potential explanations can be postulated for the discrepancy between the primary outcome of this study and that of the European trial of IFNβ-1b in SPMS.16 Among these are differences in the inclusion criteria, the reliability of EDSS assessment in the trial, and differences in the definition of the primary endpoint in terms of the period required for confirmation of progression.
It is generally accepted that the inflammatory features of MS—clinical relapses and Gd-enhancing lesions—decline spontaneously during the SP phase, despite the gradual progression of disability that is associated with axonal loss and declining brain parenchymal volume. This suggests that the underlying pathology responsible for progression may be temporally dissociated from relapses and Gd-enhancing lesions. Thus, it is conceivable that the subject population enrolled in this study tended more toward the less inflammatory, more progressive stage of the disease and therefore was less responsive to IFNβ-1b treatment. In addition, most of the secondary and tertiary outcomes that were favorably affected in this trial were relapse related (e.g., attack rate, attack-free status, time to first relapse, and new MRI activity), and it is to be expected that these results would be consistent with the findings on relapse rate. The numerous limitations of the EDSS are well known, 25 and consequently, the trial was designed with standardized training to minimize variability in EDSS measurement. However, unlike the European trial, EDSS training was carried out only at the start of the trial. In addition, confirmed progression in this trial was required after 6 months compared with 3 months in the European trial.
These results suggest that there is a point in the evolution of SPMS beyond which treatment with IFNβ has only limited benefit on the pathology underlying progression. The dichotomy between the positive benefits of treatment on the inflammatory components of the disease and the lack of benefit on the progressive elements—in contrast with the earlier European trial—suggests that more subjects in this study had passed that point. Kappos et al.26 explore this question in greater detail. This hypothesis would support the concept that earlier treatment of MS would be most appropriate regardless of the phase of MS.
Unlike earlier studies that have suggested a relationship between IFNβ and depression, this study provided no support for the notion that treatment with IFNβ-1b is associated with depression, either according to the Beck Depression Inventory or as a reported adverse event. Neither was there benefit on cognitive measures, perhaps as a consequence of the disease duration or the cognitive battery used to assess change over time in this study. Our results do not support the view that cognitive testing is a responsive measure in subjects who have entered more advanced stages of SPMS. Also apparent was the lack of difference in perceived quality of life between the treatment groups. These results suggest that even when there is a robust treatment effect on relapse-related outcomes, these benefits are insufficient to substantially improve quality of life in the absence of a benefit on disability progression. This is perhaps not surprising, given that over the long term, the principal driver of decline in quality of life is disease progression.27
Analyses to examine the effects of NAbs to IFNβ-1b did not show a consistent effect on change in EDSS scores, relapse rates, or MRI endpoints in this trial. The lack of a consistent effect on change in EDSS score is not surprising, given the primary outcome of this study. Although an effect of NAb on relapse rate was suggested by longitudinal analyses, this was apparent only for some cut-off titers using the “once positive, always positive” definition. No effect was noted for analyses using an “all switches considered” approach. Longitudinal analyses allow a more direct evaluation of whether NAb-positive status is associated with diminished treatment efficacy than cross-sectional analyses as in the former each patient serves as his or her own control.
In both the European trial of IFNβ-1b in SPMS23 and the trial of IFNβ-1b in RRMS,28 the effects of NAb were likewise inconsistent when studied using the same methodology. Outcomes that appeared to be influenced by NAb status in that study did not seem to be influenced by NAb status in a similar manner in the current study. However, one common finding was that NAbs were transient. Approximately half of all subjects in both studies who developed NAbs subsequently reverted to stable NAb-negative status for the remainder of the trial, indicating that NAbs are a transient phenomenon in a substantial proportion of IFNβ-1b-treated patients. Interestingly, over 95% of reversions occurred in the group continuing to receive treatment. Full details of these analyses will be reported separately.29–32⇓⇓⇓
This study did not show a benefit of IFNβ-1b on the primary endpoint of time to confirmed progression, despite treatment benefits on the secondary and tertiary clinical and MRI outcome measures of efficacy. The contrast between the lack of an effect on disability progression in this study and the positive outcome of the European clinical trial of IFNβ-1b16 may indicate a heterogeneity in response to IFNβ-1b that depends on baseline disease characteristics. This question has been explored elsewhere,26 leading to the conclusion that IFNβ-1b provides therapeutic benefit on disability progression only in those patients who remain in the inflammatory phase of their disease.
Appendix
Members of the North American Study Group.
Writing Committee: Hillel Panitch (Burlington, VT), Aaron Miller (Brooklyn, NY), Donald Paty (Vancouver, Canada), Brian Weinshenker (Rochester, MN). Centers and principal investigators: Bernard Gimbel MS Center (Mary Ann Picone); Blake Neurology Professional LLC (Cynthea Blake); Bowman Gray School of Medicine (Douglas Jeffery); Carolinas Medical Center (Michael Kaufman); Dallas Neurologic Clinic (J.T. Phillips); Duke University Medical Center (Barrie Hurwitz); Henry Ford Hospital (Stanton Elias); Hopital Notre Dame (Pierre Duquette); Kaiser Permanente, San Diego (Jay Rosenberg); Lahey Hitchcock Clinic (H. Stephen Kott); MS Center at Shepherd (William Stuart); Maimonides Medical Center (Aaron Miller); Mayo Clinic (Brian Weinshenker); Medical College Wisconsin (Lorri Lobeck); Montreal Neurologic Institute (Yves Lapierre); Ohio State University (Kottil Rammohan); Oregon Health Sciences University (Ruth Whitham); SUNY at Stony Brook (Patricia Coyle); Neurology Center, Houston (Martin Rusinowitz); Thomas Jefferson University (Robert Knobler); Tulane University Medical Center (Antonio Stazio); University of British Columbia, Vancouver (Joel Oger); University of Alabama (John Whitaker [deceased]); University of Arizona (William Sibley); University of Chicago (Barry Arnason); University of Maryland (Hillel Panitch); University of New Mexico (Corey Ford); University of South Florida (Peter Dunne); University of California at Irvine (Stanley van den Noort); UCLA (Lawrence Myers); UCSF-Mt. Zion (Donald Goodkin); University Hospital, London, Ontario (George Ebers, George Rice); Vanderbilt University (Subramaniam Sriram); Washington University School of Medicine (Anne Cross); Yale University School of Medicine (Timothy Vollmer). Independent Data and Safety Monitoring Board: Henry F. McFarland (chair; Gaithersburg, MD), Stuart D. Cook (Newark, NJ), Thomas R. Fleming (statistician; Seattle, WA), Stephen C. Reingold (New York, NY), Jerry Wolinsky (Houston, TX).
Study Steering Committee:
Donald Goodkin (study principal investigator and Steering Committee chair; San Francisco, CA), Aaron Miller (Brooklyn, NY), Lawrence Myers (Los Angeles, CA), Hillel Panitch (Baltimore, MD, and Burlington, VT), Donald Paty (Vancouver, BC, Canada) William Sibley (Tucson, AZ), Brian Weinshenker (Rochester, MN), Lorianne Masuoka (Richmond, CA), Lisa Bedell (statistician; Richmond, CA).
Acknowledgments
Funded by Berlex Laboratories, Montville, NJ.
The authors thank the patients for their participation. They also thank Lorianne Masuoka for supervision of the study, Douglas Goodin for critical review of the manuscript, and Christopher James for editorial support.
Footnotes
-
Additional material related to this article can be found on the Neurology Web site. Go to www.neurology.org and scroll down the Table of Contents for the November 23 issue to find the link for this article.
-
↵*Members of the North American Study Group on Interferon beta-1b in Secondary Progressive MS are listed in the Appendix on page 1794.
-
Members of the writing committee and advisory board have received honoraria to lecture at conferences or participate at advisory meetings, but none of them holds any financial interest in the pharmaceutical company.
- Received December 11, 2003.
- Accepted September 27, 2004.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hastening the Diagnosis of Amyotrophic Lateral Sclerosis
Dr. Brian Callaghan and Dr. Kellen Quigg