Lack of support for association between the KIF1B rs10492972[C] variant and multiple sclerosis

To the Editor

In their recent communication, Aulchenko et al.¹ suggested that the rs10492972[C] variant of KIF1B increases susceptibility to multiple sclerosis. In an attempt to replicate this observation, we genotyped this variant in eight case-control and three trio-family collections (in total 22,854 individuals were considered, comprising 8,391 cases, 8,052 unrelated controls and 2,137 trio families). None of these studies showed evidence for a statistically significant association; more than half of the studies showed a trend in the opposite direction (Fig. 1). Based on the odds ratio (OR) reported by Aulchenko et al.¹ (OR = 1.35), each of the collections we studied had >80% power to demonstrate association at the 5% significance level, except for the two smaller Australian studies; a population where association with this KIF1B variant has already essentially been excluded². We also found no evidence for association with rs10492972[C] in analyses that considered all of our data together or those that pooled our new data with the allele counts reported by Aulchenko et al.¹ (final P = 0.1). Given the P value originally reported by Aulchenko et al.¹ (P = 2.5 × 10⁻¹⁰), it is important to consider why this association has not been replicated.

Figure 1.

Odds ratio for the rs10492972(C) allele. OR > 1.0 are consistent with the original study by Aulchenko et al.¹. The area of the symbol is proportional to the number of cases included in the respective analysis. The error bars indicate the 95% confidence interval (see Supplementary Note for more detail regarding our new collections). Analysis was performed using PLINK⁴ for the individual collections and UNPHASED⁵ for the combined analyses.

The genesis of the original claim for association is important. The genome-wide association study performed by Aulchenko et al.¹ was predicated on the notion that in a small genetically isolated population, risk alleles that are rare in the general population may have become concentrated and thereby can be more readily detected. However, there was limited power in using this approach due to the small sample size (45 cases and 195 controls), and no significant associations were identified in the initial genome-wide association study¹. In this setting, the odds that a modestly associated variant (P = 0.0004) from a candidate gene is genuinely associated with the disease are unfavorable³. The possibility that perhaps this variant is relevant in The Netherlands, Sweden and Canada (the populations studied by Aulchenko et al.¹) but not elsewhere in the world seems unlikely considering the allele frequencies we have observed. In all of the collections we tested, the observed allele frequency was comparable with that seen in the European CEU HapMap samples (frequency = 0.34). However, in the study from Aulchenko et al.¹, although all of the case groups showed a HapMap-consistent frequency for rs10492972[C], the frequency of this allele was reduced in the control groups (allele frequency in Dutch isolate controls was 0.21 and the allele frequency in pooled controls was 0.27). This is the reverse of what would be expected if the risk allele had been concentrated in the Dutch population. After testing, there is no meaningful allele frequency difference between the cases in our new data and those originally reported; however, a significant difference in the allele frequency between the two control groups was observed (P = 3.5 × 10⁻¹⁶).

The Swedish population is the only one considered in the original report1 that has been directly studied here. In the Swedish samples considered by Aulchenko et al.¹ (826 subjects and 997 controls), modest apparent association was reported, whereas in the nonoverlapping Swedish samples we typed (1,239 subjects and 736 controls), no association was found. Comparing these two Swedish data sets indicates that this divergence of results stems almost exclusively from a difference in allele frequency between the control groups.

*. IMSGC members

Australia

David R. Booth, Robert N. Heard, Graeme J. Stewart

University of Sydney, Institute for Immunology and Allergy Research, Westmead Millennium Institute, Westmead Hospital, NSW 2145, Australia

Mathew Cox, Rodney J. Scott, Jeannette Lechner-Scott

Hunter Medical Research Institute, University of Newcastle, Callaghan, NSW 2308, Australia

Belgium

An Goris, Rita Dobosi, Bénédicte Dubois

Section for Experimental Neurology, Katholieke Universiteit Leuven, 3000 Leuven, Belgium.

Finland

Janna Saarela, Virpi Leppä

Public Health Genomics Unit, National Institute for Health and Welfare and Institute for Molecular Medicine Finland (FIMM), University of Helsinki, Helsinki, Finland

Leena Peltonen

The Wellcome Trust Sanger Institute, Hinxton, Cambridge, UK and the Institute for Molecular Medicine Finland (FIMM), University of Helsinki, Helsinki, Finland

Tuula Pirttila

Department of Neurology, Kuopio University Central Hospital, Finland

France

Isabelle Cournu-Rebeix, Bertrand Fontaine

INSERM, UMR_S975, Paris, France. UPMC Univ Paris 06, UMR_S975, Centre de Recherche Institut du Cerveau et de la Moelle, CNRS 7225. Fédération des maladies du système nerveux, Pitié –Salpêtrière Hospital, AP-HP, Paris, France

Italy

Laura Bergamaschi, Sandra D’Alfonso

Department of Medical Sciences and Interdisciplinary Research Center of Autoimmune Diseases, University of Eastern Piedmont, Novara, Italy

Maurizio Leone

Clinica Neurologica, AOU Maggiore della Carità, Novara, Italy

Norway

Åslaug R Lorentzen

Department of Neurology, University of Oslo, and Institute of Immunology, Oslo University Hospital, Norway

Hanne F Harbo

Department of Neurology, Oslo University Hospital and University of Oslo, Norway

Elisabeth G Celius

Department of Neurology, Oslo University Hospital, Norway

Anne Spurkland

Institute of Basal Medical Sciences, University of Oslo, Norway

Sweden

Jenny Link, Ingrid Kockum, Tomas Olsson, Jan Hillert

Department of Clinical Neuroscience, Karolinska Institutet, CMM, L8:04, SE-171 76, Stockholm, Sweden.

UK

Maria Ban, Amie Baker, Anu Kemppinen, Stephen Sawcer, Alastair Compston

University of Cambridge, Department of Clinical Neuroscience, Addenbrooke’s Hospital, Hills Road, Cambridge, CB2 2QQ, UK

Neil P Robertson

Department of Neurology, University Hospital of Wales, Heath Park, Cardiff CF14 4XW, UK.

US

Philip L. De Jager,

Program in Translational NeuroPsychiatric Genomics, Department of Neurology, Brigham and Women’s Hospital, Boston MA, USA and Harvard Medical School, Boston MA, USA Broad Institute, Cambridge MA, USA

David A. Hafler

Department of Neurology, Yale University School of Medicine, New Haven CT, USA Broad Institute, Cambridge MA, USA

Lisa F. Barcellos

Division of Epidemiology, School of Public Health, University of California at Berkeley, Berkeley, CA, USA

Adrian J. Ivinson

Harvard NeuroDiscovery Center, Harvard Medical School, Boston, MA, USA.

Jacob L. McCauley, Margaret A. Pericak-Vance

John P. Hussman Institute for Human Genomics, The University of Miami Miller School of Medicine, Miami, FL, USA

Jorge R. Oksenberg, Stephen L. Hauser

Department of Neurology, University of California San Francisco, San Francisco, CA, USA

David Sexton, Jonathan Haines

Center for Human Genetics Research, Vanderbilt University Medical Center, Nashville, TN, USA.