Neurology -- Correspondence for Polman et al., 60 (1) 37-43

Correspondence published:

Reply to Letter to the Editor: Chris H Polman, J Petkau, A Thompson, D H Miller, R White, F Dahlke and L Kappos (25 April 2003)

Neutralizing antibodies during treatment of secondary progressive MS with interferon ß-1b: Gavin Giovannoni (25 April 2003)

Reply to Letter to the Editor 25 April 2003

Chris H Polman
VU Medical Centre Amsterdam The Netherlands,
J Petkau, A Thompson, D H Miller, R White, F Dahlke and L Kappos
Send Correspondence to journal:
Re: Reply to Letter to the Editor

CH.Polman{at}vumc.nl Chris H Polman, et al.

We are grateful for Dr Giovannoni's comments on our manuscript, which underline the difficulties associated with interpreting data on the impact of neutralising antibodies (NABs) to interferon beta.
We would like to respond to some of the issues raised, especially those regarding the power of our study and the appropriateness of the longitudinal approach to the analyses.
Of course the study was not powered to measure the impact of NABs. We are not aware of any study reported so far that was specifically powered to measure this. But the NAB+ subgroup we describe is one of the largest studied so far and due to the detailed clinical and MRI follow-up is well suited to allow a better (though not definitive) understanding of the potential impact on NABs. We do not necessarily agree that an understanding of the impact of NABs can only be obtained from analysing the primary endpoint for the efficacy analysis of the study; exploratory analyses on secondary endpoints (like relapse rate and MRI measures) can contribute important information and did so in our study.
With respect to the appropriateness of the longitudinal approach to the analyses, which we applied to overcome the problem of potential confounding baseline variables (observed and unobserved) in cross-sectional comparisons between NAB+ and NAB- patients, Dr Giovannoni raises the issue of carry-over effects of treatment. He argues that after a patient becomes NAB+, there is a transition period when the NAB impact is not yet fully manifested, and vice versa if a patient reverts from being NAB+ to NAB-. We agree that this is valid concern which also impacts on the interpretation of cross-sectional analyses. If the length of this transition period was known, our approach to the analyses could be adjusted accordingly, but at this point in time we are not aware of data that provide reliable information on how to view such a transition period. We also agree with Dr Giovannoni that longitudinal analyses are less suitable for assessing impact on non-longitudinal responses, such as disease progression (described either as yes/no, or as time to); that is exactly the reason our paper also included analyses of outcomes such as relapses and MRI measures where repeated observations over time are available on each of the patients.
In conclusion, even though we agree that the points raised are important and might be used to develop even more sophisticated analyses, they do not invalidate our main conclusion that high titres of NABs do have clinical impact, but that this impact is rather limited, a conclusion based on a combination of endpoints analysed both longitudinally and cross-sectionally.

Neutralizing antibodies during treatment of secondary progressive MS with interferon ß-1b 25 April 2003

Gavin Giovannoni
University College London UK
Send Correspondence to journal:
Re: Neutralizing antibodies during treatment of secondary progressive MS with interferon ß-1b

g.giovannoni{at}ion.ucl.ac.uk Gavin Giovannoni

The efficacy of IFNb in secondary progressive MS is less clear than in relapsing remitting MS [1,2]. Therefore, it is not surprising that Polman et al. did not show an impact in neutralising anti-IFNb antibodies (NABs) on disease progression [3]. This trial is under powered to detect an effect of NABs on disease progression [3].
In this study 100/360 (28%) IFNb-1b treated patients developed NABs [3]. Applying the results of the study and assuming that NAB+ patients behave as if on placebo, one would expect at least 49.8% of the 100 NAB+ patients compared to a maximum of 38.9% of the 260 NAB- patients to progress over the 3 years of the study. At a level of significance of 0.029, which was used in the original study, a two-sided test would only have a 35% chance of detecting a significant difference between NAB+ and NAB- patients (Fisher Exact Test). This calculation overestimates the power of the study as it ignores the observed therapeutic effect of IFNb- 1b before the development of NABs [3].
Polman et al. argues that a cross-sectional analysis is less desirable as it fails to take into account inherent differences in the clinical response from patient to patient. They suggest a longitudinal approach in which clinical responses are compared within individual patients during NAB- and NAB+ periods [3]. As this limits the number of patients available for comparison it would be useful to know the power of this method to detect an effect of NABs on disease progression. As there is a refractory period of 6-10 months after terminating IFNb treatment before disease activity returns to baseline [4], it would be incorrect to assume that once patients become NAB- that there are no carry-over effects of treatment. Does the longitudinal analysis control for carry-over effects? As disease progression typically occurs over a prolonged period the duration of the observation periods available for comparison is important. In the longitudinal analysis what was the mean follow-up time in the NAB- and NAB+ phases? Was their sufficient time in the NAB- and NAB+ phases to confirm disease progression and to take into account the delayed effects of NABs? A cross-sectional analysis is the standard method for assessing therapeutic efficacy in phase III clinical trials, why should we use a longitudinal analysis to demonstrate the converse or loss of efficacy?
References
1. European Study Group on interferon beta-1b in secondary progressive MS. Placebo-controlled multicentre randomised trial of interferon beta-1b in treatment of secondary progressive multiple sclerosis. Lancet 1998;352:1491-1497.
2. Secondary progressive efficacy clinical trial of recombinant interferon-beta-1a in MS (SPECTRIMS) study group. Randomized controlled trial of interferon-beta-1a in secondary progressive MS: clinical results. Neurology 2001; 56:1496-1504.
3. Polman C, Kappos L, White R, et al. Neutralizing antibodies during treatment of secondary progressive MS with interferon b-1b. Neurology 2003; 60:37-43.
4. Richert ND, Zierak MC, Bash CN, et al. MRI and clinical activity in MS patients after terminating treatment with interferon beta-1b. Mult Scler. 2000 Apr;6(2):86-90.