Comparisons of therapies in different scenarios help complete the puzzle

Escalation of therapy is the most widely used strategy for managing active relapsing-remitting multiple sclerosis. The difficulty in tailoring the choice of treatment to the needs of patients with multiple sclerosis reflects the complexity of the underlying decision matrix, which consists of factors such as treatment goals, available therapies, route of administration, family planning, safety and efficacy, in the context of individual disease severity.

Thus, the choice of the right treatment for a concrete patient at a given time depends on multiple circumstances – including prior treatment history, disease activity and personal preference. From the article by Dr Lorscheider and colleagues, published in this issue of Multiple Sclerosis Journal,¹ we learn that while both fingolimod and natalizumab are associated with substantial reduction in relapse frequency, natalizumab offers superior control of relapse activity and improved chance of recovery from previously accrued disability when compared to fingolimod, if used in patients who have experienced relapses despite being treated with platform injectable therapies (interferon β or glatiramer acetate) during the preceding 1 year. This study, which utilised the data from the Swiss Federation of Common tasks of Health Insurances, used methodology that is the gold standard for assessing treatment effectiveness in observational data sets. The authors studied 358 patients, matched closely on a propensity score that was derived from a comprehensive list of potential confounders, in order to minimise inclusion bias. They used paired analyses and pairwise censoring of follow-up to eliminate attrition bias. Lorscheider et al. have chosen robust disability endpoints, including 12 month confirmed improvement in Extended Disability Status Scale score, and have applied internal processes to ensure data quality and completeness. The source data set represents 65% of the Swiss multiple sclerosis population. Although the study did not estimate the vulnerability of its outcomes to potential unmeasured confounders (such as magnetic resonance imaging (MRI) activity), the authors assure us about the validity of their results through a number of converging sensitivity analyses, including analyses using different matching strategies and different definitions of prior disease activity.

This study replicates the results of a number of other studies, which showed superior effect of natalizumab on relapse incidence and some of them also on improvement of disability in patients with previously active disease. These include studies completed in the French Observatoire Français de la Sclérose en Plaques (OFSEP) cohort,² a two-centre study in northern Italy,³ a multicentre Italian study⁴ and a study from the global MSBase registry.⁵ In contrast, a study conducted in the population-based Danish National Multiple Sclerosis Registry did not find any differences in relapse and disability outcomes among 928 matched patients treated with fingolimod or natalizumab.⁶

The differences in the results of observational studies may in some readers raise doubts about the validity of real-world evidence. In fact, validity of such evidence strongly depends on the quality of the underlying data and on the used statistical methodology. Reassuringly, the common denominator of the studies cited above as well as the study published in the issue of Multiple Sclerosis Journal is the high degree of rigour applied to study designs and analyses. Before the reader sets out on a journey to explore the differences in their statistical methodologies, we should contemplate the different contexts in which natalizumab and fingolimod had been compared. An interesting pattern emerges. The Danish study enrolled patients with variable disease and treatment histories, including patients with relatively low prior relapse frequency and limited treatment possession. On the other hand, the studies that showed superiority of natalizumab over fingolimod were all enriched for patients with previously high disease activity, and several of these studies even required recent failure of injectable platform therapies for patient inclusion.

In summary, the currently available literature suggests that the magnitude of the difference observed between therapies of unequal anti-inflammatory potential is determined by the degree of previous inflammatory activity experienced by the studied patients. Therefore, the choice of natalizumab or fingolimod should, in addition to safety and convenience considerations, be guided by prior disease activity and treatment history. Treatment-naïve patients or those with previously stable disease may equally benefit from treatment with fingolimod or natalizumab. In contrast, patients with a relatively active disease and a history of suboptimal treatment response should opt for the more potent treatment option without hesitation. Such hypothesis is intuitive and reflects the approach already practised by many neurologists. However, conclusive evidence that will confirm that treatment escalation strategies should be driven by prior disease activity is still needed.

continue