Introduction Overt blood brain barrier (BBB) disruption, as de- tected by visible gadolinium (Gd) enhancement on MRI, is a consistent finding in new multiple sclerosis (MS) lesions in patients with relapsing MS. [1]. His- tological studies of active MS lesions demonstrate the extravascular deposition of fibrinous exudates, indicative of BBB disruption, in conjunction with perivenular inflammation, lymphocyte infiltration and oedema [2]. Additional evidence of BBB disrup- tion in chronic, non-active lesions has come from histological studies, including the demonstration through confocal laser microscopy of disruption of the intercellular network of tight junctions which form the physical BBB [3], and the widespread his- tological finding of extravascular deposition of fibrin in non-active MS lesions [2, 4, 5]. MRI evidence of low Derek Soon Daniel R. Altmann Kryshani T.M. Fernando Garin Giovannoni Frederick Barkhof Chris H. Polman Paul O’Connor Bruce Gray Michael Panzara David H. Miller A study of subtle blood brain barrier disruption in a placebo-controlled trial of natalizumab in relapsing remitting multiple sclerosis Received: 15 May 2006 Received in revised form: 18 June 2006 Accepted: 25 July 2006 Published online: 3 February 2007 j Abstract Natalizumab, an anti- a4 integrin antibody, significantly reduces the number of visibly enhancing multiple sclerosis (MS) lesions. In this substudy of a 2-year trial of natalizumab mono- therapy versus placebo, contrast- enhanced imaging investigated for subtle blood brain barrier (BBB) leakage in relapsing remitting (RRMS) patients, and whether such leakage is modified by na- talizumab. After 24 weeks on treatment, 40 patients from 3 centres (27 on natalizumab and 13 on placebo) were studied. T 1 weighted images were obtained before and at set timepoints up to 46 minutes after gadolinium (Gd)- DTPA (0.3mmol/kg to 18 patients, 0.15mmol/kg to 22). Paired re- gions of interest were placed around non-enhancing lesions and contralateral normal appearing white matter (NAWM). BBB leak- age was inferred through post-Gd T 1 weighted signal intensity (SI) change. SI change was greater in T 2 non-enhancing lesions than paired NAWM at all timepoints (P < 0.005), indicating BBB leak- age in lesions. No significant dif- ference in inferred BBB leakage was observed between treatment arms as measured by SI change of lesions (P > 0.05 for all time- points, joint test P = 0.24), or in SI change of NAWM (joint test P = 0.37). T 1 hypointense and isointense lesions exhibited simi- lar SI changes (joint test P = 0.12). There is evidence of a subtle BBB leakage within visibly non- enhancing lesions in RRMS that was not modified by a4 integrin blockade in this substudy cohort. j Key words natalizumab Æ Tysabri Æ BBB Æ MS ORIGINAL COMMUNICATION J Neurol (2007) 254:306–314 DOI 10.1007/s00415-006-0356-z D. Soon Æ K.T.M. Fernando G. Giovannoni Æ D.H. Miller (&) Dept. of Neuroinflammation Institute of Neurology Queen Square London WC1N 3BG United Kingdom Tel.: +44-207/829-8771 E-Mail: d.miller@ion.ucl.ac.uk D.R. Altmann London School of Hygiene and Tropical Medicine, United Kingdom F. Barkhof Æ C.H. Polman VU Medisch Centrum Amsterdam, The Netherlands P. O’Connor Æ B. Gray St Michael’s Hospital Toronto, Canada M. Panzara Biogen Idec Cambridge, MA, USA