Treatment satisfaction, adherence and behavioral assessment in patients de – escalating from natalizumab to interferon beta

Interferon beta (IFN B 1B) and glatiramer acetate (GA) are the most widely prescribed first line disease modifying drugs for relapsing remitting multiple sclerosis (RRMS). They have proven to be efficacious in reducing relapse frequency and disability progression in the short term. NTZ, a novel drug approved in Europe as second line therapy, exerts a robust anti-inflammatory activity in RRMS, but is associated with an increased risk of progressive multifocal leukoencephalopathy (PML) [1‐8] which limits its prolonged use in the majority of cases. However, NTZ cessation is followed by a reactivation of disease activity [9] and no established guidelines exist on patient risk stratification and management. A few recent reports have explored the possible role of first line MS disease modifying therapies as de-escalating compounds after NTZ cessation [10‐12], showing a moderate capability to limit recrudescence of disease activity. Besides efficacy, patient-reported outcomes (PRO) are crucial when addressing the selection of a specific treatment, especially in the long term. Here we present changes in treatment satisfaction and adherence, as well as in behavioral assessments including measures of cognitive functions, fatigue and quality of life after NTZ de-escalation to IFN B 1B in a small group of patients in clinical practice.

Nineteen patients were included (NTZ =10, IFN B 1B = 9).

Table 1 summarizes baseline measures related to demographics and clinical parameters that were similar between groups. At baseline, all clinical variables were similar between groups (PASAT p = 0.68, FSMC p = 0.93, FAMS p = 0.33), except for EQ-VAS measure that was lower in the IFN B 1B group (p = 0.04).

17/19 patients completed the study: one IFN B 1B-patient withdrew consent (day 34) because she could not comply with study procedures; one NTZ-patient opted for an oral treatment (day 139). Patient (#9) switched from IFN B 1B to rescue treatment with GA at day 69 due to systemic side effects.

Table 2 summarizes the results of behavioral evaluation and treatment assessment in the two study groups at the three time points.

Within-group comparisons of behavioral and treatment assessments at t1 vs t0, t2 vs t0, and t2 vs t1 showed no significant differences (p values ranging from 0.08 to 1). The interaction analysis of treatment effect over time showed no significant changes for cognitive function (PASAT), fatigue (FSMC), and quality of life (FAMS, EQ-VAS) (p = 0.8, p = 0.9, p = 0.2, respectively), as well as treatment satisfaction (TS-VAS) (p = 0.6) (Figure 1). Nonetheless, coherent trends in favor of NTZ were observed for the majority of these outcomes. The main reasons for treatment satisfaction were “convenience of use” and “no MS illness recall” for NTZ and “optimal safety” for IFN B 1B.

Persistence on treatment was 90% (9/10) for NTZ, and 78% for IFN B 1B (7/9).

Treatment adherence over the study period was high overall and showed a similar trend between groups (p = 0.9, Table 2). Main reasons for missing IFN B 1B injections were: “forget to administer” (75%) and “concomitant fever” (9%). One patient didn’t receive one NTZ infusion because of concomitant hospitalization for gastroenteritis.

Clinical-radiological outcomes have been reported in detail in Gobbi et al. [12], and are summarized in Table 3. Briefly, a trend towards a superior efficacy of NTZ compared to IFN B 1B was found in the majority of the clinical and radiological outcome parameters.

Adverse events were within the expected range in both groups. Skin reactions occurred in 44% of patients in the IFN B 1B group, and self-limiting infections occurred in both groups at comparable rates (33% for IFN B 1B vs 50% for NTZ, p = 0.65). None of the patients continuing NTZ developed PML symptoms or signs during the course of this study.

PML risk stratification in patients undergoing NTZ treatment can be performed based on treatment duration, history of previous immunosuppression and, more recently, on serology for JCV [3‐6]. The decision to discontinue NTZ may be derived from an unfavorable risk/benefit ratio calculated on the basis of probabilistic clinical reasoning [23]. Patients may also independently interrupt NTZ for fear of developing PML. Overall, in clinical practice up to 60% of patients with JCV seropositivity discontinue treatment [23]. Recently, some studies have been conducted that address the efficacy of first line compounds for MS after NTZ discontinuation, showing a partial efficacy for glatiramer acetate and IFN B 1B [10, 11]. Adherence to treatment and how treatment might change the quality of life are key issues of patient outcomes in chronic conditions where long-term treatments are required. The awareness of this concern has grown among neurologists to such an extent that patient-reported outcomes (PRO) are nowadays included even in larger MS clinical trials (ClinicalTrials.gov ID NCT00751881). To the best of our knowledge, no previous report on satisfaction of patients de-escalating from NTZ to a first line, injectable disease modifying agent for MS is available.

Behavioral measures as well as treatment satisfaction and adherence in the IFN B 1B group were comparable to those in the group with continued NTZ treatment. Overall, NTZ was coherently favored in the majority of the outcomes, even though this trend did not reach statistical significance in any of the measures taken alone. A larger study would be necessary to validate this finding.

We evaluated changes in cognitive performance in the two groups over time. At study entry both groups obtained high PASAT scores suggesting preserved orbito-frontal pathways involved in working memory maintenance, manipulation and monitoring processes. We then found that both treatments were associated with non-significant improvement of cognitive performances over time. Our results are in line with Fischer’s findings [24], showing a positive effect of IFN B 1B on cognitive domains investigated also with the PASAT test in a wide sample of RRMS patients. This result, however, differs from previous studies [25], which found NTZ to be more effective than IFN B 1B in reducing cognitive deterioration. Several factors are likely to contribute to explain discrepancies between our and previous findings, including limited cognitive impairment at baseline, short duration of follow-up, and tests used for the assessment of cognitive performance. A “practice effect” might have also contributed, however the use of parallel forms of PASAT should have mitigated it at least to some extent.

Fatigue assessment also showed a trend favoring NTZ after 6 and 12 months of NTZ de-escalation to IFN B1 B, and compared to continued NTZ treatment. The role of interferons in MS fatigue is controversial. Some authors found interferons to aggravate fatigue [26, 27], while others showed interferons to be irrelevant to fatigue [28, 29]. On the other hand, escalation to NTZ has been shown to be associated with a perceived reduction of fatigue [30, 31]. In our study, fatigue might have been influenced by disease recurrence in some patients in the IFN B 1B arm.

QoL on IFN B 1B was comparable to previous treatment with NTZ and to continued NTZ therapy at 6 and 12 months’ time points. This result may be related to the substantial satisfaction with the treatment expressed by patients on IFN B 1B, which was similar to that of the NTZ group (p = 0.6). Interestingly, the main reasons for treatment satisfaction were “convenience of use” and “no MS illness recall” for NTZ and “optimal safety” for IFN B 1B. Treatment safety remains a crucial concern for patients undergoing NTZ treatment, and possibly counterbalances the disadvantage of IFN B 1B administration route and side effects, when assessing overall quality of life. Indeed, in our clinical practice the fear of PML is sometimes followed by development of disabling mental distress ultimately associated to a significant worsening of quality of life. Similar findings were also reported by van Rossum et al., showing a correlation between unsafe feelings related to the risk of PML, and anxiety in MS patients treated with NTZ [32].

Adherence to treatment reached values above 90% in the IFN B 1B group at each time point, which is greater than the mean value described in the literature [33‐35]. Also, a relatively high proportion of patients persisted on treatment in both groups (IFN B 1B nearly 80% vs NTZ 90%), similarly to the findings from larger trials that evaluated MS therapies over similar timeframes [36, 37]. An association between adherence to disease modifying agents and MS patients’ clinical outcome has been outlined in several works, with lower risks of relapses, hospitalizations and reduced quality of life in patients with regular intake of medications [35, 38].

The main limitation of our study is represented by the small sample size. High retention and adherence to IFN B 1B treatment may reflect a bias of a patient sub-population that is willing to participate in a clinical study. Moreover, our sample of patients is fully representative of the MS patient population treated with NTZ in southern Switzerland as all patients receiving NTZ in this region are treated at our site, and clinical and radiological characteristics of the patients included in the study were similar to remaining NTZ treated patients at our site (data not shown). Moreover, we postulate that our study population is well representative of a typical NTZ treated population as included patients experienced over 90% reduction of annualized relapse rate under NTZ treatment compared to the two years before NTZ initiation, which is in line with the known efficacy of NTZ in active MS patients [39, 40]. A second limitation of this sub-analysis is represented by the lack of specific scales for assessment of mood disorders, even though the FAMS questionnaire includes items related to physiological wellness. In addition, cognitive outcome should be considered with caution since we evaluated only the most frequent cognitive dysfunctions in MS (working memory, sustained attention, and information processing speed) using a single test (PASAT) as measure of cognitive impairment. Albeit, this test is a validated instrument widespread applied as screening test in MS clinical research. Finally, the study lacks a placebo arm, which was avoided for ethical reasons. These limitations may be partially mitigated by the randomized, prospective design with quarterly MRI monitoring.

In conclusion, our study suggests that de-escalation to IFN B 1B at NTZ cessation is feasible, well accepted and tolerated by the majority of patients, and seems to be associated with substantial stability of cognitive and fatigue parameters.