REPORT FROM THE ACTRIMS-ECTRIMS – BOSTON, MA, SEPTEMBER 10-13, 2014 – The assessment of treatment response has evolved over the past two decades of DMT use, with early trials focusing on relapse rates and disability progression. Indeed, the pivotal trial of glatiramer acetate did not evaluate the effect of treatment on MRI (Johnson et al. Neurology 1995;45:1268-1276).

Since that time, individual clinical (relapses, EDSS) and radiological measures have become the gold standard for evaluating treatment effects, although a number of authors have reported that combining metrics may have prognostic value (Sormani et al. Mult Scler 2013;19:605-612).

These three domains of assessment – relapses, EDSS progression (disability) and MRI – were first used as a combined metric in a post-hoc analysis of the AFFIRM cohort of natalizumab-treated patients (Havrdova et al. Lancet Neurol 2009;8:254-260). “Disease activity-free” (DAF) was defined as no relapses, no sustained disability progression and no MRI activity (no Gd+ lesions, no new/enlarging T2 lesions). A newer term for DAF now in use is no evidence of disease activity, or NEDA.

In a special presentation at ACTRIMS/ECTRIMS, it was noted that NEDA rates have been calculated for most of the DMTs, although somewhat different definitions may be used (Hartung H-P. ECTRIMS 2014; abstract PS 9.1). For the AFFIRM dataset, the two-year NEDA rate was 37% (vs. 7% with placebo; Havrdova 2009). The NEDA rates were 33% with fingolimod (vs. placebo 13%) in FREEDOMS (Kappos L. AAN 2011); 28% with dimethyl fumarate 240 mg BID (vs. placebo 15%) in DEFINE (Giovannoni G. AAN 2012), and 18% for DMF (vs. glatiramer acetate 12%, placebo 7%) in CONFIRM (Havrdova E. AAN 2013); 23% with teriflunomide 14 mg (vs. placebo 14%) in TEMSO (Freedman M. AAN 2012); 39% with alemtuzumab (vs. IFNb-1a 27%) in CARE-MS I (Cohen et al. Lancet 2012), and 32% (vs. IFNb-1a 14%) in CARE-MS II (Hartung H-P. AAN 2013).

NEDA rates necessarily decline over time as patients experience a clinical or radiological event. However, one area of confusion is that a number of recent studies have calculated NEDA cross-sectionally over a limited time period (e.g. during year 4 of treatment; Giovannoni et al. ECTRIMS 2014; abstract FC3.5) rather than longitudinally. Thus, in these analyses, NEDA may represent patients with prior disease activity; an increasing NEDA rate during extended therapy may reflect a treatment effect and/or the natural history of MS.

The predictive value of NEDA was evaluated in a 7-year follow-up of treated CIS and RRMS patients (Rotstein et al. ECTRIMS 2014; abstract P763). For the entire cohort, the NEDA rate  was 46.5% at one year, 27.0% at two years, and 7.9% at seven years. However, the positive predictive value for EDSS change <0.5 at 7 years was 71.7% at year 1, and 79.3% at year 2, suggesting that it is a good measure for predicting long-term outcomes.

A further refinement of the NEDA concept is to include brain volume loss (“NEDA-4”) since brain atrophy has been shown to be highly prognostic of long-term physical and cognitive disability. This metric was used in an analysis of patients in the FREEDOMS and FREEDOMS II trials (Kappos et al. ECTRIMS 2014; abstract FC1.5). Since annual percentage brain volume change (PBVC) in healthy individuals is estimated to be -0.1 to -0.3%/year, a cut-off PBVC value of -0.4%/year was used. At two years, the proportion of patients with NEDA-4 was 19.7% with fingolimod versus 5.3% with placebo. Between-group differences were significant (odds ratio 4.41) in favour of fingolimod. The authors concluded that adding brain volume loss produces a more stringent combined metric without decreasing the sensitivity to treatment effects.

These studies represent an ongoing effort to use combined metrics in assessing treatment response and improving outcomes in individual patients.

Guest Reviewer: Dr. Daniel Selchen, St. Michael’s Hospital, Toronto, Ontario.