You are here
Impact of a switch to fingolimod on depressive symptoms in patients with relapsing multiple sclerosis: An analysis from the EPOC (Evaluate Patient OutComes) trial
Journal of the Neurological Sciences, June 2016, Pages 190 - 198
Depression is common in patients with multiple sclerosis (MS), may confound evaluation of therapeutic effectiveness and may be impacted by MS-specific treatments.
First, to assess the impact on depressive symptoms of a switch to fingolimod versus remaining on an injectable disease-modifying therapy (iDMT) in a post-hoc analysis of prospectively collected data from the EPOC study. Secondly, to investigate the underlying Beck Depression Inventory-II (BDI-II) factor structure in patients with MS, and estimate treatment differences using the resulting subscales.
EPOC was a 6-month, open-label study assessing patient-reported outcomes after switch from iDMT to oral fingolimod 0.5 mg versus remaining on iDMT in 1053 patients with relapsing–remitting MS.
At end of study (EOS), a greater proportion of patients on fingolimod versus iDMT no longer had BDI-II scores indicating depression (p < 0.001). Fewer mildly and moderately symptomatic patients developed severe depressive symptoms, and fewer severely symptomatic patients continued to have scores indicating severe depression at EOS on fingolimod versus iDMT (p = 0.027, p = 0.038, p = 0.030, respectively). Two BDI-II subscales were identified and labelled Somatic and Affective; fingolimod demonstrated more reduction on both subscales at EOS versus iDMTs (p < 0.0001 and p = 0.0001, respectively).
A switch to fingolimod versus remaining on/switching to another iDMT was associated with an improvement in depressive symptoms in patients with relapsing–remitting MS.
- Multiple sclerosis (MS)-specific treatments may cause or exacerbate depression.
- Switching therapies may therefore impact on depression in relapsing-remitting MS.
- Injectable disease-modifying treatments (iDMT) and oral fingolimod were compared.
- Depressive symptoms improved after switch to fingolimod vs switch to/stay on iDMT.
Keywords: BDI-II, Depression, Fingolimod, Multiple sclerosis, Patient-reported outcomes, Satisfaction.
Patients with multiple sclerosis (MS) are at high risk for depression; lifetime prevalence of depression in this patient group has been estimated to be as great as 50% . Patients with MS may experience concomitant impairments in cognitive function and health-related quality of life (HRQoL) that can be debilitating . Depression has been shown to be a strong determinant of HRQoL in this patient group , , and . Furthermore, initiation of treatment with interferon-beta (INFβ) may precipitate or exacerbate depression in patients with MS  and . Together, these findings highlight the importance of assessing changes in depression over time in studies investigating the initiation or switching of MS therapies.
The Beck Depression Inventory-II (BDI-II) provides a patient-reported measure of the severity of depressive symptoms during the preceding 2 weeks . This instrument was developed for assessing symptoms corresponding to diagnostic criteria for depressive disorders in the Diagnostic and Statistical Manual of Mental Disorders (DSMIV)  and . It consists of 21 questions, scores for which are summed to generate a total score that can be categorized according to depression severity. Subscales reflecting the somatic, affective and/or cognitive components of depression have been identified by factor analysis in general and non-MS populations  and . Although not specifically validated for use in MS, the BDI-II is one of the most widely used patient-reported outcome (PRO) tools for measuring the severity of depression, and has been validated for use in outpatients and the general population  and .
Two prospective, randomized clinical trials have assessed the impact of fingolimod MS therapy on BDI-II scores, including a phase II study and the phase IV study to Evaluate Patient OutComes (EPOC; NCT01216072)  and . The 6-month, phase II study evaluated the impact of fingolimod 1.25 and 5.0 mg versus placebo on patient-reported HRQoL and depression outcomes in relapsing-remitting MS (RRMS) . Changes in depression (BDI-II score) favored fingolimod 1.25 mg over placebo (p < 0.05), with categorical analysis showing that both fingolimod doses reduced the proportion of patients with BDI-II scores indicative of clinical depression at 6 months versus placebo (both p < 0.05). The 6-month EPOC trial assessed the impact on treatment satisfaction and other PROs after a switch from an approved, first-line, injectable disease-modifying therapy (iDMT; either glatiramer acetate, subcutaneous [SC] INFβ-1b, intramuscular [IM] INFβ-1a or SC INFβ-1a) to oral fingolimod 0.5 mg compared with remaining on or switching to another iDMT in patients with RRMS  and . Improvements in treatment satisfaction, convenience, fatigue and HRQoL favored switching to oral fingolimod compared with staying on or switching to another iDMT. There was also a pronounced improvement in BDI-II scores in patients who received fingolimod at Month 6 compared with those who received iDMT (p < 0.001).
The present post-hoc analysis of prospectively collected data from the EPOC study aimed  to evaluate the impact of a switch to fingolimod versus remaining on an iDMT on changes in categorical BDI-II scores after 6 months of treatment, and  to investigate the underlying factor structure of the BDI-II in patients with MS, and estimate treatment differences using the resulting subscales.
2.1. Study design and participants
EPOC was a 6-month, randomized, open-label, phase IV study conducted at 152 centers in North America and six centers in Canada. Details of study design, rationale and baseline characteristics of the patient population have been reported previously  and .
Patients were adults (aged 18–65 years), had RRMS in accordance with the 2005 McDonald criteria  and an Expanded Disability Status Scale (EDSS) score of 0–5.5 . All were treated continuously with an approved, first-line iDMT for at least 6 months prior to screening and were naïve to fingolimod (FTY720; Gilenya®, Novartis Pharma AG, Basel, Switzerland). Key exclusion criteria were consistent with those in phase III clinical trials of fingolimod.
Eligible patients were randomized in a 3:1 ratio to receive once-daily oral fingolimod 0.5 mg or an iDMT for 6 months. Therapy was switched directly without a washout period. Following consultation with a physician, patients randomized to receive an iDMT either remained on the same iDMT or switched immediately to another iDMT. The iDMTs included INFβ-1b (Extavia® or Betaseron®) 0.25 mg injected SC every other day; INFβ-1a (Avonex®) 30 μg injected IM once weekly; INFβ-1a (Rebif®) 22 or 44 μg injected SC three times weekly; or glatiramer acetate (Copaxone®) 20 mg injected SC once daily.
The study was conducted in accordance with ethical principles laid down in the 2008 Declaration of Helsinki and written informed consent was obtained from each patient before enrolment.
2.2. Study assessments
The EPOC trial assessed several clinically relevant PROs that have been previously reported. The primary outcome measure was the Treatment Satisfaction Questionnaire for Medication V 1.4 (TSQM). This post hoc analysis further explores the BDI-II study results .
The BDI-II contains 21 multiple-choice questions, each scored from 0 to 3 and summed to generate a total score ranging from 0 to 63. Total BDI-II scores were classified categorically according to criteria proposed by Beck et al. . Total scores > 13 were considered indicative of clinical depression; scores of 14–19 indicated mild depression, 20–28 indicated moderate depression and 29–63 indicated severe depression . Patients scoring ≤ 13 were considered to have no/minimal depression. The BDI-II was completed at the screening visit (described as baseline hereafter) and at 3 and 6 months after treatment randomization. If a patient scored > 0 on the question assessing suicidal thoughts or wishes, the investigator was required to evaluate their need for clinical intervention and suitability to continue in the study. Data on concomitant antidepressant usage were available at baseline and post-baseline.
2.4. Statistical analyses
All data included were collected prospectively. These post-hoc analyses were conducted using the full analysis set, which comprised all randomized patients to whom study medication was assigned1 (Fig. 1). Baseline data for this population have already been published .
Changes in the proportions of patients with depressive symptoms in each treatment arm were assessed from baseline to end of study/early termination (EOS/ET; Month 6) in patients with data at both time points (Table 1; Fig. 1). Missing data were imputed using the last observation carried forward (LOCF) method. Between-group differences for changes in depression severity category over time for patients with data available both at baseline and EOS/ET were assessed using the Chi-square test. Treatment group comparisons were not adjusted for multiplicity of testing. A sensitivity analysis was conducted using the multiple imputation method (based on 500 imputations) to explore the impact of using a different method to handle missing data.
|Fingolimod 0.5 mg (n = 768)||iDMT (n = 240)|
|Age, years (SD)||45.9 (9.77)||45.0 (9.95)|
|Men/women, n (%)||186/582 (24.2/75.8)||47/193 (19.6/80.4)|
|Race, n (%)|
|Caucasian||626 (81.5)||193 (80.4)|
|Black||110 (14.3)||41 (17.1)|
|Native American||3 (0.4)||1 (0.4)|
|Other||27 (3.5)||5 (2.1)|
|Mean duration of MS symptoms, years (SD)||12.0 (8.23)||11.8 (8.37)|
|Number of relapses within last year, mean (SD)||0.8 (1.21)||0.8 (1.34)|
|Number of relapses within last 2 years, mean (SD)||1.4 (2.06)||1.4 (1.96)|
|Baseline EDSS score||2.44 (1.33)||2.42 (1.33)|
|Previous MS treatment|
|iDMT at screening, n (%)|
|Glatiramer acetate||254 (33.1)||83 (34.6)|
|IFNβ-1a IM||201 (26.2)||54 (22.5)|
|IFNβ-1a SC||193 (25.1)||61 (25.4)|
|IFNβ-1b||120 (15.6)||42 (17.5)|
|Corticosteroid used in the 6 months before screening, n (%)||150 (19.5)||51 (21.3)|
|Number of corticosteroid courses used in the 6 months before screeninga, mean (SD)||1.7 (2.34)||1.9 (3.35)|
|Primary reason for switching iDMT, n (%)|
|Dissatisfaction with mode of administration||479 (62.4)||150 (62.5)|
|Tolerability issues||107 (13.9)||31 (12.9)|
|Lack of efficacy||90 (11.7)||36 (15.0)|
|Patient request||69 (9.0)||15 (6.3)|
|Other||19 (2.5)||6 (2.5)|
|Compliance||3 (0.4)||2 (0.8)|
|Concerns regarding safety with current product||1 (0.1)|
|Patient disposition, n (%)|
|Completed study||708 (92.2)||225 (93.8)|
|Discontinued study||60 (7.8)||15 (6.3)|
|Subject withdrew consent||8 (1.0)||6 (2.5)|
|Discontinued owing to AE(s)||37 (4.8)||3 (1.3)|
|Unsatisfactory therapeutic effect||3 (0.4)||3 (1.3)|
|Lost to follow-up||4 (0.5)|
|Abnormal test procedure result||2 (0.3)|
|Abnormal laboratory value||1 (0.1)|
a The mean number of corticosteroid courses was calculated only for patients who took corticosteroids.
AE = adverse event, EDSS = Expanded Disability Status Scale, iDMT = injectable disease-modifying therapy, IFNβ = interferon-beta, IM = intramuscular, SC = subcutaneous.
The factor structure of the BDI-II was evaluated by factor analysis using baseline data from the EPOC study full analysis set (following the PROC FACTOR procedure in SAS) . Initial common factor extraction was performed using the principal component method, and estimates of loadings were obtained using varimax rotation, retaining factors with eigenvalues > 0.5 and items with loading scores > 0.4 within the domains. Items loading on multiple factors were grouped with the highest-loading factor. Based on this factor analysis, two subscales of depression were derived. Changes in scores from baseline to each post-baseline time point and EOS/ET on the newly developed subscales were compared between treatment arms using an analysis of covariance (ANCOVA) adjusted for the corresponding baseline scores using the LOCF approach. The least-squares (LS) mean difference and Cohen's effect size for the difference between treatment arms are reported.
Between-group differences in the change in BDI-II scores (total and subscale scores) from baseline to Month 6 were assessed using an observed case approach. A repeated measures ANCOVA was conducted with treatment, month, and treatment by month as factors, and corresponding baseline as a covariate. The model assumed an unstructured within-subject covariance matrix. In a sensitivity analysis, a second repeated measures ANCOVA model included concomitant antidepressant usage and its interaction with treatment as additional factors. The LS mean difference and 95% confidence interval (CI) for the difference between treatment arms are reported.
Supportive analyses were conducted to evaluate the impact of baseline suicidal ideation and disease severity on depression. The impact on BDI-II score of suicidal ideation at baseline (as determined by a score > 0 on BDI-II question 9) was assessed by ANCOVA with suicidal ideation at baseline (yes/no), treatment, and the treatment by suicidal ideation interaction as factors, and baseline score as a covariate (LOCF)·Spearman's correlation coefficient was used to assess the relationship between EDSS at baseline and BDI-II score during the course of the study.
BDI-II scores at baseline were similar for patients in fingolimod and iDMT arms (full analysis set: mean score [SD], 11.7 [9.13] versus 11.0 [8.97], respectively; subgroup with both baseline and post-baseline measurements: mean score [SD], 11.7 [9.13] versus 11.0 [9.03]; Table 2). Comparable proportions of patients receiving fingolimod and iDMT had BDI-II scores indicative of clinical depression at baseline (full analysis set, 36.2% versus 33.5%, respectively, p = 0.419; subgroup with both baseline and post-baseline measurements: 35.8% versus 32.9%, respectively, p = 0.413; Table 2).
|Baseline BDI-II data||Fingolimod 0.5 mg||iDMT|
|Full analysis set|
|Number of patients||778||257|
|BDI-II score, mean (SD)||11.7 (9.13)||11.0 (8.97)|
|Category of depression severity, n (%)|
|None/minimal||496 (63.8)||171 (66.5)|
|Mild||137 (17.6)||39 (15.2)|
|Moderate||104 (13.4)||33 (12.8)|
|Severe||41 (5.3)||14 (5.4)|
|Subgroup with both baseline and post-baseline measurements|
|Number of patients||768||240|
|BDI-II score, mean (SD)||11.7 (9.13)||11.0 (9.03)|
|Category of depression severity, n (%)|
|None/minimal||493 (64.2)||161 (67.1)|
|Mild||132 (17.2)||34 (14.2)|
|Moderate||103 (13.4)||31 (12.9)|
|Severe||40 (5.2)||14 (5.8)|
BDI-II = Beck Depression Inventory-II, iDMT = injectable disease-modifying therapy.
Baseline BDI-II scores for patients in the iDMT arm previously treated with GA and INFβ were similar (full analysis set: mean score [SD], 11.2 [8.91] versus 10.9 [9.03], respectively; subgroup with both baseline and post-baseline measurements: 11.0 [8.93] versus 11.0 [9.11], respectively).
Discontinuation rates were similar for both fingolimod and iDMT arms (9.6% and 12.9%, respectively; Fig. 1); based on the adverse event and severe adverse event reports, no patient discontinued due to suicidal thoughts/wishes. There were no differences in baseline characteristics between patients in the fingolimod and iDMT arms overall and among those who discontinued the study. However, although representing < 10% of those analyzed, significantly more patients in the iDMT versus the fingolimod arm were missing BDI-II measurements at either baseline or EOS (23/263 [8.7%] versus 21/789 [2.7%], respectively).
3.1. Changes in depression category over time
A greater proportion of patients on fingolimod versus iDMT no longer had BDI-II scores indicating depression at EOS/ET (50.5% versus 25.3%, respectively, p < 0.001; Fig. 2a). The severity of depression reduced to none/minimal at EOS/ET in a greater percentage of patients on fingolimod with mild and moderate depression at baseline (p = 0.008 and p = 0.012 versus iDMT, respectively; Fig. 3a). Although the same overall pattern was observed for patients with severe depression at baseline, the difference was not statistically significant (p = 0.195).
Among patients with BDI-II scores consistent with clinical depression at baseline, switching to fingolimod reduced progression to severe depression. Fewer patients with mild or moderate depression at baseline developed severe depression by EOS/ET in fingolimod versus iDMT treatment arms (p = 0.027 and p = 0.038, respectively; Fig. 3d). In addition, fewer patients with severe depression at baseline remained severely depressed at EOS/ET in fingolimod versus iDMT treatment arms (p = 0.030).
There was no difference between treatment arms in the proportion of patients with no/minimal depression at baseline who developed clinical depression during the study (p = 0.633; Fig. 2b).
Antidepressant usage was similar in both treatment arms is shown in Table 3.
|Fingolimod 0.5 mg (n = 768)||iDMT (n = 240)|
|Therapeutic class, n (%)|
|Selective serotonin reuptake inhibitors||219 (28.5)||79 (32.9)|
|Non-selective monoamine uptake inhibitors||43 (5.6)||25 (10.4)|
|Other antidepressants||168 (21.9)||62 (25.8)|
|Antidepressants||2 (0.3)||2 (0.8)|
3.2. Factor analysis
Factor analysis of baseline data revealed two BDI-II subscales for MS, labelled as Somatic and Affective based on items that they included (Table 4). The somatic and affective components of the BDI-II contributed equally to the variance of baseline assessments of depression in patients with MS (50.5% versus 49.5%, respectively).
|BDI-II item No.||BDI-II item description||Eigenvalue scores|
|Factor 1 (scores in bold = Somatic)||Factor 2 (scores in bold = Affective)|
|4||Loss of pleasure||0.54837||0.42141|
|12||Loss of interest||0.48441||0.48713|
|15||Loss of energy||0.70327||0.21515|
|16||Changes in sleeping pattern||0.52589||0.17002|
|18||Changes in appetite||0.45472||0.26401|
|20||Tiredness or fatigue||0.74254||0.15301|
|21||Loss of interest in sex||0.43877||0.21563|
Factors were segregated into the BDI-II subscales using eigenvalue loading scores > 0.4 (indicated in bold), obtained through varimax rotation.
BDI-II = Beck Depression Inventory-II.
The fingolimod group showed greater reductions in scores on both BDI-II subscales at EOS/ET (Fig. 4).
3.3. Descriptive statistics: impact of primary reason for treatment switch
BDI-II scores for patients who cited the primary reason for treatment switch as dissatisfaction with mode of administration were similar at baseline for fingolimod and iDMT treatment arms (full analysis set: mean score [SD], 11.2 [8.98] and 11.4 [9.37], respectively; subgroup with both baseline and post-baseline measurements: 11.2 [8.98] and 11.4 [9.48], respectively). At Month 6 (LOCF), mean BDI-II scores were lower than at baseline in both treatment arms (mean change from baseline [SD] for subgroup with both baseline and post-baseline measurements: fingolimod, − 3.3 [7.51]; iDMT, − 1.1 [6.45]). Among patients who cited other primary reasons for treatment switch, mean BDI-II scores at baseline were higher for the fingolimod than the iDMT arm (full analysis set: mean score [SD] 12.5 [9.32] and 10.4 [8.31], respectively; subgroup with both baseline and post-baseline measurements: 12.5 [9.33] and 10.2 [8.22], respectively). At Month 6, mean BDI-II scores were lower than at baseline for the fingolimod arm (mean change from baseline, − 3.2 [7.58]) and remained similar to baseline values for the iDMT arm (mean change from baseline, 0.3 [7.05]) in the subgroup with both baseline and post-baseline measurements.
3.4. Sensitivity analysis: impact of concomitant antidepressant usage
Antidepressant use was similar in fingolimod and iDMT arms (full analysis set: 21.9% versus 26.2%, respectively). Concomitant antidepressant usage was associated with the change in total BDI-II score (p = 0.0024); however, the fingolimod treatment effect remained significant after adjustment for antidepressant usage (LS mean treatment difference [CI], − 2.1 [− 3.28, − 1.00]). The same was seen for the Affective subscale (LS mean treatment difference [CI], − 0.9 [− 1.47, − 0.34]; Fig. 5a). For the Somatic subscale, adjustment for antidepressant usage revealed a significant interaction between antidepressant usage and treatment arm (p = 0.0773; threshold for significance = p < 0.1); while there remained a significant difference in favor of fingolimod over iDMT for patients not taking an antidepressant (− 1.8 [− 2.51, − 1.13]), this effect was diminished and not significant for patients taking antidepressant medication (− 0.7 [− 1.83, 0.52]; Fig. 5d).
3.5. Sensitivity analysis: impact of missing data
Findings from the sensitivity analysis of BDI-II total and subscale scores showed that results obtained by the multiple imputation and the LOCF methods were consistent (LS mean treatment difference at Month 6 [CI]: total BDI-II score, − 2.7 [− 3.7, − 1.7], p < 0.0001; Somatic, − 1.7 [− 2.3, − 1.1], p < 0.0001; Affective, − 1.0 [− 1.5, − 0.5], p < 0.0001).
3.6. Supportive analyses
3.6.1. Impact of suicidal thoughts/wishes at baseline
Patients with suicidal thoughts/wishes based on question 9 of the BDI-II were distributed relatively evenly across both treatment arms at randomization (7.9% [61/768] randomized to fingolimod, 7.5% [18/240] to iDMT). In this small patient subgroup, the direction of change was consistent with that for patients without suicidal thoughts/wishes at baseline (Cohen's effect sizes for: total BDI-II score, − 0.221; Somatic, − 0.149; Affective, − 0.239). No treatment by suicidal thoughts/wishes interaction was observed from ANCOVA analysis (total BDI-II score, p = 0.642; Somatic, p = 0.426; Affective, p = 0.769).
3.6.2. Correlation analysis
The Spearman rank correlation coefficients between measures of depression (total BDI-II and Somatic and Affective subscale scores) and EDSS score at baseline for the total population were low and ranged from 0.1704 to 0.2036.
RRMS patients who switched from an iDMT to fingolimod 0.5 mg demonstrated greater improvement in BDI-II-assessed depressive symptoms than those who remained on their prior treatment or switched to another iDMT. These treatment differences were evident when assessing changes in categorical BDI-II summary scores, as well as MS-specific subscale scores derived through factor analysis. Findings are consistent with the previously reported pre-specified analysis of BDI-II as a secondary outcome measure of the EPOC study , and add to data from large randomized, controlled clinical trials that demonstrated the efficacy of fingolimod in patients with RRMS  and . The implications for patients with RRMS extend beyond the observed improvement in depression. Indeed, depression has been reported to be a major predictor of impaired self-perceived health status in patients with MS . As such, a switch to fingolimod may have a positive impact on HRQoL by virtue of its association with an improvement in depressive symptoms.
EPOC was the first prospective clinical trial to assess the effects of switching from injectable to oral therapy on patient satisfaction and HRQoL in MS . Reported improvements in satisfaction with convenience, side effects and effectiveness, as well as in fatigue, depression and HRQoL, for fingolimod relative to iDMTs serve as evidence of the positive impact of treatment switch from the patient perspective . Furthermore, a previous EPOC trial post hoc analysis showed that BDI-II-assessed depressive symptoms improved following a switch to fingolimod versus remaining on glatiramer acetate, SC IFNβ-1a or SC IFNβ-1b, but not IM IFNβ-1a  and . Thus, data are broadly in line with findings of greater levels of patient satisfaction and improved HRQoL associated with oral versus injectable therapies in other disease states , , and . One previous prospective study of 121 consecutive RRMS patients looked at the impact of starting IM IFNβ-1a on BDI-II scores in MS and found no significant change in score after 12 months of treatment . However, willingness to enter the study and to switch therapies may influence BDI-II scores and therefore the contribution of this to the results cannot be ruled out.
Factor analysis of the baseline EPOC cohort identified two BDI-II subscales reflecting the somatic and affective components of depression. To our knowledge, this analysis represents the first investigation into BDI-II subscales as they pertain to patients with MS. Several previous factor analysis studies in other populations have identified either two or three components within the BDI-II, usually comprising a combination of Somatic, Affective and/or Cognitive domains  and . The two-factor structure in the current MS population was most similar to that found by Beck et al. in a psychiatric outpatient population; ‘crying’ and ‘loss of interest’ being the only two items classified differently in these two models . Loadings within each factor also showed a similar pattern, with, for example, ‘tiredness or fatigue’ and ‘loss of energy’ being the highest loading items on the somatic-centered factor in both models. Less pronounced similarities in factor structure and loadings were also apparent versus the general population  and . The parallels between factor structures for MS and other populations is an interesting finding, and should be replicated in an independent MS population.
Benefits of a switch to fingolimod versus remaining on or switching to another iDMT were evident for both the Somatic and Affective subscales. Of note, Somatic scores were higher at baseline and showed a numerically greater change in response to treatment than Affective scores. Antidepressant usage also had an impact on the treatment effect for the Somatic but not the Affective subscale; there was a comparable improvement in Somatic scores for both treatment arms for patients taking concomitant antidepressant medication. Although patient numbers within this subgroup were not large, the latter finding suggests that antidepressants may be contributing to improvements in the somatic symptoms of depression in the present study.
Questions have been raised regarding the potential for the original BDI to overestimate the prevalence and severity of depression in MS owing to inclusion of items capturing somatic concerns common in both depression and progressive neurological disorders ; however, a recent evidence-based management guideline publication on psychiatric disorders in MS concluded that there was insufficient evidence to determine definitively whether somatic/neurovegetative symptoms affect the accuracy of the BDI . This relationship has not been specifically assessed using the BDI-II. Unfortunately, it was not possible to evaluate whether the prevalence or severity of depression was overestimated in the current study in the absence of appropriate patient and normal control populations in which to compare BDI-II scores. Future research should assess the ability of the BDI-II to discriminate depressed and non-depressed patients with MS.
Limitations of this study and analysis include the open-label study design. As patients knew which medication they were taking, both change in active drug and administration route could have contributed to improvements in BDI-II scores. Therefore, it is not possible to determine whether this is a direct effect of fingolimod or would be found when switching to any oral therapy. Furthermore, based on data in major depressive disorder, some patients (approximately 26–28%) with depression would be expected to respond following switch, regardless of the type of treatment to which they were switched . While switching from injectable to oral therapy may have improved drug tolerability for some patients it is unlikely that this accounts for the observed magnitude of improvement in BDI-II scores; however, this could be evaluated in a switching study involving two oral agents. Similarly, it is important to consider that prior dissatisfaction with route of administration may have had an impact on the degree of depression experienced following treatment switch. Although the present trial was not designed to evaluate this variable, descriptive data suggest that the relationship between reasons for treatment switch and depression is not clear cut. Further study of the relationship between depression/satisfaction and reasons for treatment switch in patients with MS is warranted. The lack of a washout period in the present trial means that it was not possible to assess baseline depression in the absence of therapy. While inclusion of a washout would be expected to result in the generation of more robust evidence regarding the effects of a switch to fingolimod, such trial design would need to consider the impact of patient concerns regarding disease progression on measures of depression, even during short-term cessation of treatment where rescue medication was available. The within-subject, repeated-measures and categorical nature of the analysis could also be subject to regression towards the mean or within-subject measurement variability. Nonetheless, greater BDI-II improvements were seen with fingolimod in most cases across outcome definitions and sensitivity and supportive analyses.
In conclusion, a switch to fingolimod versus remaining on or switching to another iDMT was associated with an improvement in depressive symptoms in patients with RRMS in the EPOC study. Although this study does not answer the question as to whether this is a direct or indirect effect of fingolimod, these data add to the growing body of evidence regarding the efficacy of fingolimod in patients with RRMS and also highlight the importance of considering patient-reported measures of depression when treating this patient population. Further research to evaluate potential effects of fingolimod and other treatment comparators on depression in patients with MS is warranted. Sequential, Multiple Assignment, Randomized Trial (SMART) designs similar to those employed in trials of antidepressants could be employed in future MS studies in order to help elucidate the complex relationships between MS treatments and outcomes.
This study was funded by Novartis Pharmaceuticals Corporation. Oxford PharmaGenesis (Oxford, UK) provided editorial support for this manuscript. Funding for this support was provided by Novartis Pharmaceuticals Corporation.
Conflict of interest statement
SF Hunter performs research for Abbott, Acorda Therapeutics, Avanir Pharmaceuticals, Bayer, Biogen Idec, Genzyme/Sanofi, Lilly, Novartis, Roche, Synthon, Teva Pharmaceuticals and XenoPort; has received a grant from Genzyme for a research study; is a speaker for Acorda Therapeutics, Avanir Pharmaceuticals, Bayer, Biogen Idec, Novartis, Questcor Pharmaceuticals and Teva Pharmaceuticals; and consults for Bayer, Biogen Idec, Osmotica Pharmaceutical and Roche.
M Agius has received honoraria from Biogen Idec, Genzyme, Novartis, Teva Neuroscience, Bayer and Questcor, and has received research grant support from Actelion, Acorda, Biogen Idec, GSK, MedImmune, NEXT, Novartis and Roche.
DM Miller has received research support from the National Multiple Sclerosis Society and Novartis.
G Cutter has served on data and safety monitoring committees for Apotek, Biogen, Cleveland Clinic, Eli Lilly, GSK, Medivation, Merck, Modigenetech (Prolor), the National Institute of Neurological Disorders and Stroke, the National Multiple Sclerosis Society, the Eunice Kennedy Shriver National Institute of Child Health and Human Development, the National Heart, Lung, and Blood Institute (Protocol Review Committee), Ono Pharmaceuticals, Sanofi-Aventis and Teva Pharmaceuticals, and has served on consulting and advisory boards for Alexion Pharmaceuticals, Abbott, Allozyne, Bayer, Celgene, Consortium of MS Centers (grant), Coronado Biosciences, DioGenix, MedImmune, Klein Buendel Inc., Novartis, Nuron Biotech, Receptos, Somnus Therapeutics, Spinifex Pharmaceuticals and Teva Pharmaceuticals. G Cutter is employed by the University of Alabama at Birmingham and is President of Pythagoras Inc., a private consulting company located in Birmingham, AL, USA.
L Barbato was an employee of Novartis during preparation of the manuscript and holds stock in Novartis; he is currently employed by AbbVie.
K McCague, X Meng and N Agashivala are employees and stockholders of Novartis Pharmaceuticals Corporation.
P Chin was an employee of Novartis during preparation of the manuscript and holds stock in Novartis; he is currently employed by Genentech.
E Hollander has consulted for Coronado Biosciences, Roche and Transcept Pharmaceuticals.
All authors made substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data; were involved in drafting the manuscript and revising it critically for important intellectual content; and had final approval of the version to be published. The authors take full responsibility for the content of the paper and thank Yanxin Wang and Nan Sun (RPS Strategic Solutions, a division of PRA International) and Stan Li (Minimax Information Services, Belle Mead, NJ, USA) for statistical support, and Harriet Crofts, PhD (Oxford PharmaGenesis, Oxford, UK) for medical writing support, editorial assistance, and collation and incorporation of comments from all authors. This work was funded by Novartis Pharmaceuticals Corporation.
-  A.D. Sadovnick, R.A. Remick, J. Allen, et al. Depression and multiple sclerosis. Neurology. 1996;46:628-632 Crossref
-  The Goldman Consensus Group. The Goldman consensus statement on depression in multiple sclerosis. Mult. Scler.. 2005;11:328-337
-  S. Fruehwald, H. Loeffler-Stastka, R. Eher, et al. Depression and quality of life in multiple sclerosis. Acta Neurol. Scand.. 2001;104:257-261
-  S. D'Alisa, G. Miscio, S. Baudo, et al. Depression is the main determinant of quality of life in multiple sclerosis: a classification-regression (CART) study. Disabil. Rehabil.. 2006;28:307-314 Crossref
-  J.P. Reese, G. Wienemann, A. John, et al. Preference-based health status in a German outpatient cohort with multiple sclerosis. Health Qual. Life Outcomes. 2013;11:162 Crossref
-  L.K. Neilley, D.S. Goodin, D.E. Goodkin, et al. Side effect profile of interferon beta-1b in MS: results of an open label trial. Neurology. 1996;46:552-554
-  D.C. Mohr, W. Likosky, A.C. Boudewyn, et al. Side effect profile and adherence to in the treatment of multiple sclerosis with interferon beta-1a. Mult. Scler.. 1998;4:487-489
-  A.T. Beck, R.A. Steer, G.K. Brown. Manual for the Beck Depression Inventory-II. (The Psychological Corporation, San Antonio, TX, USA, 1996)
-  A.T. Beck, R.A. Steer, R. Ball, et al. Comparison of Beck Depression Inventories - IA and - II in psychiatric outpatients. J. Pers. Assess.. 1996;67:588-597 Crossref
-  S. Vanheule, M. Desmet, H. Groenvynck, et al. The factor structure of the Beck Depression Inventory-II: an evaluation. Assessment. 2008;15:177-187 Crossref
-  E.A. Storch, J.W. Roberti, D.A. Roth. Factor structure, concurrent validity, and internal consistency of the Beck Depression Inventory-Second edition in a sample of college students. Depress. Anxiety. 2004;19:187-189 Crossref
-  X. Montalban, G. Comi, P. O'Connor, et al. Oral fingolimod (FTY720) in relapsing multiple sclerosis: impact on health-related quality of life in a phase II study. Mult. Scler.. 2011;17:1341-1350 Crossref
-  M. Cascione, D. Wynn, L.M. Barbato, et al. Randomized, open-label study to evaluate patient-reported outcomes with fingolimod after changing from prior disease-modifying therapy for relapsing multiple sclerosis: EPOC study rationale and design. J. Med. Econ.. 2013;16:859-865 Crossref
-  E. Fox, K. Edwards, G. Burch, et al. Outcomes of switching directly to oral fingolimod from injectable therapies: results of the randomized, open-label, multicenter, evaluate patient OutComes (EPOC) study in relapsing multiple sclerosis. Mult. Scler. Relat. Disord.. 2014;3:607-619 Crossref
-  C.H. Polman, S.C. Reingold, G. Edan, et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald criteria”. Ann. Neurol.. 2005;58:840-846 Crossref
-  J.F. Kurtzke. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology. 1983;33:1444-1452
-  SAS Institute Inc. SAS/STAT® 9.3 User's Guide. (Cary, NC, USA, 2011)
-  J.A. Cohen, F. Barkhof, G. Comi, et al. Oral fingolimod or intramuscular interferon for relapsing multiple sclerosis. N. Engl. J. Med.. 2010;362:402-415 Crossref
-  L. Kappos, E.W. Radue, P. O'Connor, et al. A placebo-controlled trial of oral fingolimod in relapsing multiple sclerosis. N. Engl. J. Med.. 2010;362:387-401 Crossref
-  B. Cree, D. Kantor, B. Steingo, et al. Patient- and physician-reported outcomes after therapy switch from glatiramer acetate to fingolimod versus remaining on glatiramer acetate. Mult. Scler.. 2013;19(11 Suppl.):8-597
-  K. Edwards, H. Crayton, J. Calkwood, et al. Patient- and physician-reported outcomes after therapy switch from interferon beta to fingolimod versus remaining on interferon beta therapy. Mult. Scler.. 2013;19(11 Suppl):8-597
-  L.H. Jensen, K. Osterlind, C. Rytter. Randomized cross-over study of patient preference for oral or intravenous vinorelbine in combination with carboplatin in the treatment of advanced NSCLC. Lung Cancer. 2008;62:85-91 Crossref
-  R.H. Osborne, R. De Abreu Lourenco, A. Dalton, et al. Quality of life related to oral versus subcutaneous iron chelation: a time trade-off study. Value Health. 2007;10:451-456 Crossref
-  C. Twelves, S. Gollins, R. Grieve, et al. A randomised cross-over trial comparing patient preference for oral capecitabine and 5-fluorouracil/leucovorin regimens in patients with advanced colorectal cancer. Ann. Oncol.. 2006;17:239-245
-  H. Zephir, J. De Seze, T. Stojkovic, et al. Multiple sclerosis and depression: influence of interferon beta therapy. Mult. Scler.. 2003;9:284-288 Crossref
-  M.A. Whisman, J.E. Perez, W. Ramel. Factor structure of the Beck Depression Inventory-Second edition (BDI-II) in a student sample. J. Clin. Psychol.. 2000;56:545-551 Crossref
-  D.C. Mohr, D.E. Goodkin, W. Likosky, et al. Identification of Beck Depression Inventory items related to multiple sclerosis. J. Behav. Med.. 1997;20:407-414 Crossref
-  S.L. Minden, A. Feinstein, R.C. Kalb, et al. Evidence-based guideline: assessment and management of psychiatric disorders in individuals with MS: Report of the Guideline Development Subcommittee of the American Academy of Neurology. Neurology. 2014;82:174-181 Crossref
-  A.J. Rush, M.H. Trivedi, S.R. Wisniewski, et al. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am. J. Psychiatry. 2006;163:1905-1917
a Advanced Neurosciences Institute, 101 Forrest Crossing Blvd, Suite 103, Franklin, TN 37064, USA
b Neurology Department, Barrow Neurological Institute, Phoenix AZ 85013, USA
c The Mellen Center, Cleveland Clinic Foundation, 1950 East 89th Street, Cleveland, OH 44195, USA
d University of Alabama at Birmingham, 1400 University Boulevard, Birmingham, AL 35223, USA
e Novartis Pharmaceuticals Corporation, 1 Health Plz, East Hanover, NJ 07936, USA
f Albert Einstein College of Medicine and Montefiore Medical Center, 111 East 210th Street, Bronx, New York, NY 10467, USA
⁎ Corresponding author.
1 With the exception of one patient found subsequently to be ineligible for inclusion according to the study protocol.
© 2016 Elsevier B.V., All rights reserved.