Study: Does the patient know best? Quality of life assessment in multiple sclerosis trials

Dimethyl fumarate (DMF, BG-12, Tecfidera^®, Biogen Idec), recently received regulatory approval as treatment for relapsing forms of multiple sclerosis (MS) following completion of the DEFINE and CONFIRM Phase 3 trials.^1,2 Both trials were placebo-controlled: the latter also included a glatiramer acetate (GA) treatment arm as an internal reference. Advantageous aspects of DMF include potent efficacy, oral route of administration, and positive safety profile. The principal side effects of DMF are cutaneous flushing and gastrointestinal symptoms, reported by approximately one-third of DMF-treated patients in the Phase 3 trials, with the highest incidence in the first month of treatment and decreasing thereafter.

In this issue of Multiple Sclerosis Journal, Kappos et al. report the results of quality of life (QOL) assessments in the DEFINE trial.³ Short Form-36 (SF-36), and EuroQol-5D (EQ-5D) index and visual analog scale (VAS) were administered at baseline and 24, 48, and 96 weeks, and global impression of well-being was measured by a visual analog scale (VAS) which was performed at baseline and then every 12 weeks. SF-36 physical and mental component scores (PCS and MCS) were lower at baseline in the trial participants than the general US population and were inversely related to increasing disability measured by the Expanded Disability Status Scale, as demonstrated in some but not all previous studies.⁴ At two years, DMF-treated patients reported higher SF-36 PCS and, less consistently, MCS compared to placebo patients. Mean scores improved over two years with DMF treatment with a dose effect and worsened in the placebo group. Significantly greater proportions of patients had a clinically meaningful five-point improvement on SF-36 subscales in the DMF groups than in the placebo group (absolute increase of up to 8% improvements in PCS and up to 10% in MCS subscores). Significant benefit of DMF relative to placebo was also seen on the EQ-5D index and VAS scores and global impression of well-being VAS, again with improvement over two years with DMF and worsening with placebo.

Also in this issue of Multiple Sclerosis Journal, Kita et al. report the results of QOL assessment in the CONFIRM trial, which employed the same measures.⁵ Similar benefits on QOL of DMF compared to placebo were demonstrated, though the results were not as robust, especially the dose effect and impact on the SF-36 MCS. The benefits of GA on the various QOL endpoints were largely comparable to those of DMF.

Clinician-reported outcomes have been criticized, including recently by regulatory agencies,⁶ as not adequately reflecting the complex picture of MS. Patient-reported outcome measures (PROMs) might contribute substantially to the assessment of new treatments. Therefore, the authors of these publications are to be commended for reporting QOL results. Comparable data have been collected and reported in detail for relatively few MS clinical trials and, when included, different outcome measures frequently were applied, leading to the heterogeneous picture discussed by Kappos et al. in their paper.³

These two articles illustrate some of the issues encountered with PROMs. Although these results support the overall benefit on QOL of DMF compared to placebo, the magnitudes of treatment effects on mean changes in the measures and on the proportions of patients with clinically meaningful change were modest. One explanation may be the inclusion of generic QOL measures only. While generic scales remain useful to compare QOL and the factors that determine it across disorders, they can show ceiling or floor effects and neglect disease-specific impairments in MS in, for example, visual or bladder function. Disease-specific instruments, such as the Multiple Sclerosis Impact Scale-29 (MSIS-29) or MSQOL-54, may be more sensitive to factors that affect QOL in patients with MS.^4,7Recently, Neuro-QOL, was proposed as a new approach to measure QOL in different neurologic conditions with computer-adapted-test-based and short form questionnaires.⁸ Both patients and health-care providers were involved in the development process. At this time, however, the utility of Neuro-QOL in MS clinical trials remains unproven.

These studies also illustrate that the factors that determine QOL are incompletely understood. For example, somewhat surprisingly, GA and DMF produced similar benefits on QOL in CONFIRM. One might have expected an advantage for DMF based on its better efficacy on some clinician-assessed measures and oral route of administration. One potential explanation is that medication side effects and selective dropouts impacted QOL. The observations that study drug discontinuation was somewhat more frequent in the DMF groups (28–30%) vs 25% in the GA group and that only 64% of randomized patients overall completed the trial on assigned treatment support this possibility.

Finally, many studies have shown that patient-reported QOL is strongly affected by psychological factors,⁴ which also may account for one of the major measurement issues in QOL research, the so-called response shift,⁹ i.e. patients tend to adapt to a progressive condition and QOL may remain stable despite increasing disability. Studies of this phenomenon in MS do not exist. External factors, such as social support, also can affect QOL. It has been emphasized that merely being in a clinical trial may have a beneficial effect disease activity, even if patients are in a placebo group.¹⁰ Similarly, the comprehensive follow-up and support provided to the patient by the study team could artificially improve QOL. So-called pragmatic trials and real life cohort studies might provide a more realistic estimate of the effects of a medication on QOL. Nevertheless, despite these issues, because any medical intervention in the end should aim to enhance well-being, QOL measures should continue to be included in trials that test new therapies.
Click here to read more