Interferon-β treatment and the natural history of relapsing-remitting multiple sclerosis
Version of Record online: 13 AUG 2007
Copyright © 2007 American Neurological Association
Annals of Neurology
Volume 63, Issue 1, pages 125–126, January 2008
How to Cite
Koch, M., Mostert, J., De Keyser, J., Tremlett, H. and Filippini, G. (2008), Interferon-β treatment and the natural history of relapsing-remitting multiple sclerosis. Ann Neurol., 63: 125–126. doi: 10.1002/ana.21185
- Issue online: 29 JAN 2008
- Version of Record online: 13 AUG 2007
We read with interest Trojano and colleagues'1 observational study. Although longer-term data from clinical practice can be useful to help us ascertain disease progression in multiple sclerosis, we would like to express our concern that the study may contain systematic biases that could throw doubt on the validity of its findings.
The control and interferon-β (IFN-β)–treated groups were imbalanced in several major ways for which no amount of statistical adjustment can compensate. In particular, control patients were included at their initial visit, whereas treated patients were included after an observation period. Patients with secondary progression during this period were presumably included in the control group. This produced a fundamental imbalance, with the control group being biased by inclusion of progressed patients and the treated group by inclusion of patients with a sustained relapsing-remitting disease course. This alone could yield a positive association between IFN-β treatment and maintenance of a relapsing-remitting course. Furthermore, 23% of control patients had concomitant diseases such as cancer and psychiatric disease, which can worsen disability independent of the disease process of multiple sclerosis. Twenty percent of control patients discontinued IFN-β for adverse events; these patients should have been included in the treatment group for an intention-to-treat analysis. The final group imbalance is that 19% of control patients voluntarily refused IFN-β treatment; these patients likely differ from those who accepted IFN-β treatment in ways not easily measurable.
Outcome assessments were not blinded. Blinding is essential to avoid detection bias, particularly because conversion to secondary progression can be difficult to establish, especially if the clinical course is confounded by relapses.2
The numbers of treatment deviations, withdrawals, and losses to follow-up were not reported. Attrition bias, therefore, cannot be excluded as an explanation of the results. It is also unclear whether follow-up time was equal in the two groups; the numbers of patients at risk at each year in each group should have been reported with the survival plots.
The definition of secondary progression as “a steady worsening of symptoms and signs for at least 6 months” is inadequate because one cannot be confident of unremitting changes in disability scores with less than 1 year of follow-up.3 Likewise, time to Expanded Disability Status Scale scores of 4 and 6 had to be confirmed at 6 months. This is not the same as “irreversible” because one study has indicated that worsening to Expanded Disability Status Scale score of 6, confirmed at 6 months, was not sustained after 2 years by almost one-third of patients.4
Marcus Koch MD*, Jop Mostert MD*, Jacques De Keyser MD, PhD*, Helen Tremlett PhD, Graziella Filippini MD, * Department of Neurology, University Medical Centre, University of Groningen, Groningen, the Netherlands, Faculty of Medicine (Neurology), University of British Columbia, Vancouver, British Columbia, Canada, Unità di Neuroepidemiologia, Fondazione Istituto Neurologico “Carlo Besta,” Milan, Italy.