Jérôme Hodel1, Olivier Outteryck1, Céline Dubron1, Bastien Dutouquet1, Mohamed Amine Benadjaoud1, Emeline Duhin1, Sébastien Verclytte1, Marc Zins1, Alain Luciani1, Alain Rahmouni1, Jean-Pierre Pruvo1, Patrick Vermersch1, Xavier Leclerc1. 1. From the University of Lille, CHU Lille (J.H., O.O., C.D., B.D., E.D., J.P.P., P.V., X.L.), Lille, France; Departments of Neuroradiology (J.H., C.D., B.D., J.P.P., X.L.) and Neurology (O.O., E.D., P.V.), Hôpital Roger Salengro, Rue Emile Laine 59037 Lille, France; Department of Radiology (S.V.), Hôpital Saint Philibert, Lille, France; Departments of Neuroradiology (J.H.) and Medical Imaging (A.L., A.R.), AP-HP, Hôpitaux Universitaires Henri Mondor, Créteil, France; Faculty of Medicine, Université Paris Est Créteil, Créteil, France (J.H., A.L., A.R.); Inserm, CESP Centre for Research in Epidemiology and Population Health, U1018, Radiation Epidemiology Team, Villejuif, France (M.A.B.); Department of Radiology, Hôpital Saint Joseph, Paris, France (M.Z.).
Abstract
PURPOSE: To determine diagnostic precision with magnetic resonance (MR) imaging of the brain, the most predictive MR imaging features, and the added value of comparison with previous data for the diagnosis of asymptomatic progressive multifocal leukoencephalopathy (PML) associated with natalizumab (NTZ). MATERIALS AND METHODS: This retrospective study was approved by the institutional review board, and written informed consent was obtained. Eleven consecutive patients with multiple sclerosis (MS) who had received a definitive diagnosis of asymptomatic NTZ-associated PML (NTZ PML, 18 brain lesions) underwent 3-T MR imaging. The control group included 40 patients with MS but without PML who were treated with NTZ. Three readers independently performed blinded analysis of MR images. First, the readers were asked to detect NTZ PML lesions without comparing current images with previously obtained MR imaging data by evaluating MR images for the following features: U fiber and/or cortex involvement, lesion signal intensity and borders, and occurrence of punctate lesions. Second, they reassessed NTZ PML lesions with all the previous MR imaging data available. Diagnostic precision with MR imaging was assessed with and without comparison with previously obtained data. Logistic regression analyses were performed to identify the association of MR imaging features with NTZ PML. RESULTS: Overall interobserver agreement was good (κ = 0.76; 95% confidence interval [CI]: 0.71, 0.81). Hyperintensity on diffusion-weighted images and involvement of U fibers were the most predictive features (odds ratio, 33.7; 95% CI: 4.9, 229.7 [P < .0001] and odds ratio, 8.7; 95% CI: 1.2, 61.4 [P = .03], respectively), while punctate lesions were exclusively observed in patients with NTZ PML. Comparison with previous MR imaging data improved specificity of MR imaging for the detection of NTZ PML lesions (from 88% to 100%, P = .05). CONCLUSION: Recognition of the most predictive imaging features and comparison with previous MR imaging data may facilitate the detection of asymptomatic NTZ PML.
PURPOSE: To determine diagnostic precision with magnetic resonance (MR) imaging of the brain, the most predictive MR imaging features, and the added value of comparison with previous data for the diagnosis of asymptomatic progressive multifocal leukoencephalopathy (PML) associated with natalizumab (NTZ). MATERIALS AND METHODS: This retrospective study was approved by the institutional review board, and written informed consent was obtained. Eleven consecutive patients with multiple sclerosis (MS) who had received a definitive diagnosis of asymptomatic NTZ-associated PML (NTZ PML, 18 brain lesions) underwent 3-T MR imaging. The control group included 40 patients with MS but without PML who were treated with NTZ. Three readers independently performed blinded analysis of MR images. First, the readers were asked to detect NTZ PML lesions without comparing current images with previously obtained MR imaging data by evaluating MR images for the following features: U fiber and/or cortex involvement, lesion signal intensity and borders, and occurrence of punctate lesions. Second, they reassessed NTZ PML lesions with all the previous MR imaging data available. Diagnostic precision with MR imaging was assessed with and without comparison with previously obtained data. Logistic regression analyses were performed to identify the association of MR imaging features with NTZ PML. RESULTS: Overall interobserver agreement was good (κ = 0.76; 95% confidence interval [CI]: 0.71, 0.81). Hyperintensity on diffusion-weighted images and involvement of U fibers were the most predictive features (odds ratio, 33.7; 95% CI: 4.9, 229.7 [P < .0001] and odds ratio, 8.7; 95% CI: 1.2, 61.4 [P = .03], respectively), while punctate lesions were exclusively observed in patients with NTZ PML. Comparison with previous MR imaging data improved specificity of MR imaging for the detection of NTZ PML lesions (from 88% to 100%, P = .05). CONCLUSION: Recognition of the most predictive imaging features and comparison with previous MR imaging data may facilitate the detection of asymptomatic NTZ PML.
Authors: J Hodel; O Outteryck; S Verclytte; V Deramecourt; A Lacour; J-P Pruvo; P Vermersch; X Leclerc Journal: AJNR Am J Neuroradiol Date: 2015-12-03 Impact factor: 3.825
Authors: M A Schmidt; R A Linker; S Lang; H Lücking; T Engelhorn; S Kloska; M Uder; A Cavallaro; A Dörfler; P Dankerl Journal: Clin Neuroradiol Date: 2017-03-06 Impact factor: 3.649
Authors: Martijn T Wijburg; Iris Kleerekooper; Birgit I Lissenberg-Witte; Marlieke de Vos; Clemens Warnke; Bernard M J Uitdehaag; Frederik Barkhof; Joep Killestein; Mike P Wattjes Journal: JAMA Neurol Date: 2018-07-01 Impact factor: 18.302