FormalPara Key Summary Points

Diroximel fumarate (DRF) and dimethyl fumarate (DMF) are orally administered fumarate disease-modifying therapies (DMTs) for relapsing forms of multiple sclerosis (MS).

No Evidence of Disease Activity (NEDA-3) is a composite efficacy endpoint used in clinical trials for MS; however, efficacy assessments may be confounded by the time taken for the effect of a DMT to become established.

This analysis used data from the 96-week, open-label, phase 3 EVOLVE-MS-1 study of DRF to re-baseline NEDA-3 outcomes for patients who had previously received DRF as part of the preceding EVOLVE-MS-2 study (“prior DRF”); outcomes were compared with those for patients who had received DMF in EVOLVE-MS-2 (“prior DMF”) and patients who were newly enrolled (“de novo”).

At week 48, NEDA-3 estimates were 72.3% (prior DRF), 72.1% (prior DMF), and 62.1% (de novo); at week 96, estimates were 50.2% (prior DRF), 48.2% (prior DMF), and 36.5% (de novo).

After re-baselining at approximately 7 weeks, approximately half of DRF-treated patients achieved NEDA-3 at week 96, compared with 36.5% of patients who had not been re-baselined.

These outcomes support that re-baselining may be valuable for mitigating the influence of disease activity prior to the onset of efficacy when assessing the effect of DMTs.

Introduction

Diroximel fumarate (DRF) is an orally administered fumarate disease-modifying therapy (DMT) approved in the USA for adult patients with relapsing forms of multiple sclerosis (MS) and in Europe for adult patients with relapsing–remitting MS (RRMS) [1, 2]. The safety, tolerability, and exploratory efficacy of DRF were evaluated in the 96-week, open-label, phase 3 EVOLVE-MS-1 study (NCT02634307) [3, 4]. As of December 31, 2023, 44,297 patients had been treated with DRF, corresponding to 61,780 patient-years of exposure [5].

DRF and dimethyl fumarate (DMF) have a common active metabolite, monomethyl fumarate (MMF). Oral administration of DRF 462 mg and DMF 240 mg produces bioequivalent exposure to MMF, and therefore DRF and DMF are expected to have similar efficacy and safety profiles [2, 6, 7]. However, DRF demonstrated improved gastrointestinal (GI) tolerability compared with DMF in the 5-week, randomized, double-blind, phase 3 EVOLVE-MS-2 study (NCT03093324), with lower rates of GI adverse events (AEs) and more favorable quality of life outcomes [8, 9].

No Evidence of Disease Activity (NEDA-3), a composite endpoint of clinical and magnetic resonance imaging (MRI) outcomes, is increasingly evaluated in MS studies, and comprises assessments of relapses, disability progression, and MRI activity (gadolinium-enhancing and new or newly enlarging T2 lesions). Efficacy and NEDA outcomes may be confounded by initial disease activity, e.g., on MRI, that occurs before the full effect of a DMT has been achieved. One possible approach to account for such activity in clinical trials assessments is by re-baselining for a duration after the initiation of treatment, thereby accommodating for the time taken until the onset of efficacy, as utilized in other studies [10,11,12,13]. The therapeutic effect of fumarates on disease activity may take several weeks to become established; reductions in serum neurofilament light chain were recorded approximately 10 weeks after starting DMF in the open-label, phase 4 TREMEND study, and reductions in MRI activity occur as early as 7 weeks after initiation of DRF treatment [14, 15].

The EVOLVE-MS-1 study population comprised patients who had previously received DRF or DMF in the EVOLVE-MS-2 study and then rolled over into EVOLVE-MS-1, and patients who were newly enrolled in the DRF program and had no recent fumarate treatment. Based on this, it is possible to re-baseline efficacy data from EVOLVE-MS-1 for patients who had prior fumarate treatment. Therefore, in order to account for the duration between the initiation of treatment and the onset of efficacy, the objective of this analysis was to assess NEDA-3 outcomes in patients newly enrolled in EVOLVE-MS-1 and those who had previously received DMF or DRF in EVOLVE-MS-2 and were re-baselined after approximately 7 weeks.

Methods

Study Design and Patients

The study design for EVOLVE-MS-1 was reported previously [3, 4]. Briefly, EVOLVE-MS-1 was an open-label, single-arm, 96-week, phase 3 study to evaluate DRF safety, tolerability, and exploratory efficacy endpoints in adults with RRMS (Fig. 1) [3, 4]. Patients entered the study either after completing EVOLVE-MS-2, a randomized, 5-week, double-blind, head-to-head, phase 3 study of DRF and DMF, or as newly enrolled in the DRF clinical development program [3, 4, 8].

Fig. 1
figure 1

EVOLVE-MS-1 study design. Adapted from “Diroximel fumarate in patients with relapsing–remitting multiple sclerosis: final safety and efficacy results from the phase 3 EVOLVE-MS-1 study” [4], licensed under CC BY-NC 4.0 (http://creativecommons.org/licenses/by-nc/4.0/). BID twice daily, DMF dimethyl fumarate, DRF diroximel fumarate, MRI magnetic resonance imaging, W week. aAdapted from [8]; http://creativecommons.org/licenses/by-nc/4.0/. bAdapted from [3]; http://creativecommons.org/licenses/by-nc/4.0/. cMRI performed within 1 week after visit 2 (week 1) for rollover patients

Full size image

EVOLVE-MS-1 eligibility criteria required patients to be 18–65 years old with a confirmed diagnosis of RRMS, and no history of clinically significant recurring or active GI symptoms within 3 months of screening [4]. Patients were neurologically stable with no evidence of relapse in the 30 days before screening [4]. Prior treatment with DMTs was permitted; exclusion criteria for newly enrolled patients included the use of teriflunomide within 2 years of visit 2 (week 1); natalizumab within 2 months of visit 2; alemtuzumab; fingolimod within 90 days of visit 2; daclizumab within 6 months of visit 2; or B cell-depleting therapies within 12 months of screening [4].

EVOLVE-MS-1 was approved by local and central ethics committees and conducted in accordance with International Council on Harmonisation Guidelines for Good Clinical Practice and the Declaration of Helsinki [4]. All patients provided written, informed consent [4]. Ethical approval was not applicable for this analysis as data were based on existing outputs from the EVOLVE-MS-1 study.

Re-Baselining and Assessments

This analysis assessed outcomes in EVOLVE-MS-1 based on patients who had rolled over from receiving DRF or DMF in EVOLVE-MS-2 (“prior DRF” and “prior DMF” groups, respectively), or who had newly enrolled in EVOLVE-MS-1 (“de novo” group). Patients who rolled over from EVOLVE-MS-2 had a treatment duration of ≥ 5 weeks on DRF or DMF in EVOLVE-MS-2 and rolled over into EVOLVE-MS-1 within 7 days of treatment completion. MRI was performed within 7 days of the start of EVOLVE-MS-1. Therefore, patients who received DRF in both studies were on study medication for up to 7 weeks before efficacy outcomes were assessed and were thus re-baselined after approximately 7 weeks.

Baseline characteristics and demographics for the prior DRF and prior DMF groups are reported and based on assessments at baseline of EVOLVE-MS-2; the corresponding characteristics for the de novo group are given on the basis of assessments at baseline of EVOLVE-MS-1.

MRI scans were performed at baseline before each study start (approx. 7 weeks apart) and at weeks 48 and 96 in EVOLVE-MS-1. Relapses and disability were assessed in EVOLVE-MS-1. NEDA-3 was defined as having no relapse, no 24-week confirmed disability progression (CDP), and no MRI activity (i.e., no new or newly enlarging T2 lesions and no new gadolinium-enhancing lesions). Proportions of patients who were free from CDP, were relapse free, and had NEDA-3 were estimated using the Kaplan–Meier product limit method. Analyses were performed using SAS 9.4.

Safety outcomes reported for the prior DRF, prior DMF, and de novo patient groups included AEs, AEs leading to discontinuation, and serious AEs.

Results

Patients

Overall, 1057 patients were included in EVOLVE-MS-1, of whom 239 (22.6%) had received DRF in EVOLVE-MS-2 and then rolled over into EVOLVE-MS-1 (“prior DRF” group); 225 (21.3%) had received DMF in EVOLVE-MS-2 and rolled over into EVOLVE-MS-1 (“prior DMF” group); and 593 (56.1%) were newly enrolled (“de novo” group). Mean (standard deviation [SD]) age was 43.8 (11.0) years in the prior DRF group at EVOLVE-MS-2 baseline, 43.5 (9.8) years in the prior DMF group at EVOLVE-MS-2 baseline, and 41.5 (11.0) years in the de novo group at EVOLVE-MS-1 baseline (Table 1). Baseline characteristics were generally consistent between the patient groups. There were 71 (29.7%) discontinuations in the prior DRF group, 61 (27.1%) in the prior DMF group, and 125 (21.1%) in the de novo group. Median (min, max) duration of exposure to fumarate (DRF or DMF) across EVOLVE-MS-2 and EVOLVE-MS-1 was 708 (51, 751) days in the prior DRF group and 708 (57, 732) days in the prior DMF group; for the de novo group, median (min, max) duration of exposure to DRF in EVOLVE-MS-1 was 673 (1, 702) days.

Table 1 Baseline demographics and disease characteristics in DRF rollover patients, DMF rollover patients, and de novo patients
Full size table

NEDA-3 Outcomes

Week 48 NEDA-3 Component Outcomes

At week 48, estimated proportions of patients with no relapse were 89.0% in the prior DRF group, 88.0% in the prior DMF group, and 87.1% in the de novo group.

Estimated proportions of patients free of 24-week CDP were 95.0% in the prior DRF group, 93.3% in the prior DMF group, and 96.1% in the de novo group.

From baseline to week 48, the proportions of patients with no new or newly enlarging T2 lesions were 70.5% (n = 146/207; prior DRF), 68.0% (n = 136/200; prior DMF), and 50.6% (n = 264/522; de novo). At week 48, proportions of patients free of gadolinium-enhancing lesions were 94.7% (n = 196/207) in the prior DRF group, 95.5% (n = 191/200) in the prior DMF group, and 90.6% (n = 473/522) in the de novo group.

Week 96 NEDA-3 Component Outcomes

At week 96, estimated proportions of patients with no relapse were 83.3% in the prior DRF group, 81.7% in the prior DMF group, and 82.3% in the de novo group.

Estimated proportions of patients free of 24-week CDP were 92.4% in the prior DRF group, 92.7% in the prior DMF group, and 93.9% in the de novo group.

From baseline to week 96, the proportions of patients with no new or newly enlarging T2 lesions were 64.2% (n = 111/173; prior DRF), 62.1% (n = 105/169; prior DMF), and 46.6% (n = 220/472; de novo). Mean (SD) number of new or newly enlarging T2 lesions from baseline to week 96 was 1.8 (6.5) for prior DRF, 1.6 (3.8) for prior DMF, and 4.4 (13.9) for de novo. Most patients were free of gadolinium-enhancing lesions at week 96, with rates of 90.2% (n = 156/173) in the prior DRF group, 92.9% (n = 157/169) in the prior DMF group, and 90.7% (n = 431/475) in the de novo group. Mean (SD) number of gadolinium-enhancing lesions at week 96 was 0.2 (0.7) in prior DRF patients, 0.2 (0.8) in prior DMF patients, and 0.5 (3.8) in de novo patients.

NEDA-3

From baseline to week 48, Kaplan–Meier estimates of proportions achieving NEDA-3 were 72.3% in the prior DRF group, 72.1% in the prior DMF group, and 62.1% in the de novo group (Fig. 2a). Corresponding estimates at week 96 were 50.2% in the prior DRF group, 48.2% in the prior DMF group, and 36.5% in the de novo group (Fig. 2b). Compared with the de novo group, the proportion of patients achieving NEDA over the study period was significantly higher in the prior DRF group (p < 0.001) and the prior DMF group (p = 0.004).

Fig. 2
figure 2

NEDA-3 outcomes after re-baselining in prior DRF and prior DMF patients, and without re-baselining in de novo patients, from baseline to week 48 (a) and to week 96 (b). Proportions free of Gd+ lesions and free of N/NE T2 lesions were observed. Proportions relapse free, free of 24-week CDP, and achieving NEDA-3 were estimated using Kaplan–Meier product limit method. CDP confirmed disability progression, DMF dimethyl fumarate, DRF diroximel fumarate, Gd+ gadolinium-enhancing, NEDA No Evidence of Disease Activity, N/NE new/newly enlarging

Full size image

Safety Outcomes

AEs occurred in 212 (88.7%) patients in the prior DRF group, 207 (92.0%) patients in the prior DMF group, and 519 (87.5%) patients in the de novo group; most were mild to moderate in severity. The most common AEs in the prior DRF and prior DMF groups were MS relapse (20.1% and 20.0%, respectively), upper respiratory tract infection (17.6% and 15.6%), lymphopenia (14.6% and 16.9%), and flushing (13.8% and 12.9%); in the de novo group, the most common AEs were flushing (38.1%), MS relapse (19.1%), and nasopharyngitis (14.5%). Lymphopenia was reported in 51 (8.6%) patients in the de novo group. AEs led to discontinuation of DRF in 23 (9.6%) patients in the prior DRF group, 13 (5.8%) patients in the prior DMF group, and 49 (8.3%) patients in the de novo group. Serious AEs occurred in 12.1% (n = 29) of patients in the prior DRF group, 11.1% (n = 25) of patients in the prior DMF group, and 11.6% (n = 69) of patients in the de novo group.

Discussion

No Evidence of Disease Activity has become established as a useful outcome measure in clinical studies of DMTs [16]. While separate clinical and radiological parameters are valuable for evaluating disease activity, combining them allows for a broader, and at the same time more stringent, assessment of the impact of disease activity on patients. In an observational study of a longitudinal cohort of patients with clinically isolated syndrome or RRMS, NEDA at 2 years was predictive of disability progression at 7 years [17]. The efficacy of several DMTs has been demonstrated by increased proportions of patients achieving NEDA-3 status at 1–2 years compared with placebo or active comparator in phase 3 studies [16, 18,19,20]. NEDA status has also been assessed as part of real-world comparisons of DMF with other DMTs, with favorable outcomes for DMF versus interferon beta-1a [21], and no significant differences observed between DMF and teriflunomide or fingolimod, in separate studies [22, 23].

This analysis of the EVOLVE-MS-1 study of DRF assessed NEDA-3 outcomes over 96 weeks in newly enrolled patients and in patients after re-baselining to account for clinical and radiological disease activity that occurred during the preceding EVOLVE-MS-2 study. At week 96, 50.2% of DRF-treated patients achieved NEDA-3 after re-baselining, compared with 36.5% of patients who were newly enrolled to EVOLVE-MS-1 and who had not been re-baselined. Of patients who had previously received DMF in EVOLVE-MS-2, 48.2% achieved NEDA-3 at week 96 on DRF in EVOLVE-MS-1, a similar rate to that estimated for the prior DRF group.

Outcomes for separate components of NEDA-3 were generally consistent between the prior DRF, prior DMF, and de novo patient groups. However, one notable difference was observed in the rates of patients who were free of new or newly enlarging T2 lesions at week 96, which were 64.2% and 62.1% in prior DRF and prior DMF groups, respectively, and 46.6% in the de novo group. Given the consistency observed between groups across other assessments, it is likely that the T2 lesion outcomes contributed significantly toward the lower NEDA-3 rates for de novo patients compared with the prior DRF and prior DMF groups. Previous studies showed that MRI activity is the most common reason for failure to achieve NEDA-3 [13]. We hypothesize that in the relatively short period of approximately 7 weeks, new or newly enlarging T2 lesions were the most sensitive measurement of carryover disease activity that occurred prior to the onset of efficacy of DRF.

Re-baselining was proposed as a monitoring strategy for NEDA outcomes, based on the fact that the onset of DMT biological activity is often delayed in comparison to DMT half-life [24]. Consequently, the initial disease activity that occurs before achieving the full effect of a DMT may result in the underestimation of its efficacy and needs to be accounted for to avoid misrepresenting early efficacy outcomes. For example, for DMF, a re-baselining period of 3–6 months was previously suggested as being appropriate [24].

The re-baselining period of approximately 7 weeks used in the current analysis was notably shorter when compared with the previous recommendations and the periods of 6 and 12 months used in other studies [11,12,13, 24]. MS DMTs included in prior analyses that re-baselined at week 24 or week 48 have an onset of efficacy within 2–3 months; therefore, re-baselining at week 24 disregards 3 months of disease activity that occurred while the drug had already taken effect, and re-baselining at week 48 disregards 9 months of disease activity that occurred after the drug had taken effect. A key strength of the study design of EVOLVE-MS-1 is that it allowed for re-baselining after approximately 7 weeks in the patients who rolled over from EVOLVE-MS-2, whereas the earliest time point at which most other phase 3 studies can re-baseline patients is typically not until week 24, as that is often the time point of the earliest post-baseline MRI scan. One other published phase 3 study that utilized a similar NEDA re-baselining approach as in our study was the open-label, phase 3b CASTING study of ocrelizumab in patients with RRMS, which used MRI re-baselining 8 weeks after ocrelizumab initiation [10]. Our study and the CASTING study would likely have shown higher rates of NEDA with the use of a longer re-baselining period (i.e., 24-week or 48-week) owing to the high reliance of NEDA on MRI parameters; however, these longer re-baselining periods may overestimate NEDA because of the aforementioned reasons.

Taken together, based on evidence from analyses of study populations as well as the protocols employed in clinical practice by the authors of this paper, re-baselining after about 3 months in the real-world setting may be a good compromise to avoid over- or underestimation of treatment effects. This strategy may be useful to judge an individual patient’s response to treatment, and the evidence shown here and discussed above supports the authors’ recommendation for this interval.

Safety outcomes from the current analysis were consistent with those previously reported for the overall EVOLVE-MS-1 population. However, some differences were identified in the rates of specific AEs between patient groups, and these variations might be reflective of the longer duration on fumarate treatment for the prior DRF and prior DMF groups, as well as the dynamics by which certain AEs typically occur. Flushing occurred more frequently in the de novo group (38.1%) than in either of the prior DRF (13.8%) or prior DMF (12.9%) groups. Flushing symptoms are often reported during the first month on DMF and improve thereafter, consistent with the current observation of higher rates of flushing in the de novo patients who had more recently initiated fumarate treatment [25]. Lymphocyte counts have been reported to decrease during the first year on fumarate treatment and subsequently plateau [4, 26]. This aligns with the higher rates of lymphopenia observed in the prior DRF (14.6%) and prior DMF (16.9%) groups in this analysis, compared with the de novo group (8.6%), given their longer cumulative duration of exposure to fumarate treatment. Across the patient groups, there were low rates of serious AEs and discontinuation of DRF treatment due to AEs.

One limitation of this analysis is that the study design did not assess disease activity within the first 7 weeks before re-baselining in the prior DRF or prior patients with DMF, and therefore it was not possible to determine NEDA-3 without re-baselining in those two groups. In addition, since this was a post hoc analysis and NEDA-3 was an exploratory efficacy endpoint of EVOLVE-MS-1, the study had not been specifically developed to assess the effect of re-baselining on NEDA-3 outcomes.

Conclusion

In EVOLVE-MS-1, after re-baselining after about 7 weeks, approximately half of the DRF-treated patients achieved NEDA-3 at week 96, compared with 36.5% of newly enrolled patients who achieved NEDA-3 at the same time point having not been re-baselined. For clinicians who consider re-baselining as a standard practice for monitoring the efficacy of MS DMTs, a re-baselining period of 2 or 3 months from DMT initiation may be a reasonable interval to use for gauging the effectiveness of the DMT.