1118677 research-article20222022 MSJ0010.1177/13524585221118677Multiple Sclerosis JournalI Meaton, A Altokhis MULTIPLE SCLEROSIS MSJ JOURNAL Original Research Paper Paramagnetic rims are a promising diagnostic Multiple Sclerosis Journal 2022, Vol. 28(14) 2212­–2220 imaging biomarker in multiple sclerosis DOI: 10.1177/ https://doi.org/10.1177/13524585221118677 13524585221118677 https://doi.org/10.1177/13524585221118677 © The Author(s), 2022. Isobel Meaton*, Amjad Altokhis*, Christopher Martin Allen, Article reuse guidelines: Margareta A Clarke , Tim Sinnecker, Dominik Meier, Christian Enzinger, sagepub.com/journals- Massimiliano Calabrese, Nicola De Stefano , Alain Pitiot , Antonio Giorgio, permissions Menno M Schoonheim, Friedemann Paul, Mikolaj A Pawlak , Reinhold Schmidt, Cristina Granziera , Ludwig Kappos , Xavier Montalban, Àlex Rovira , Jens Wuerfel* and Nikos Evangelou*; on behalf of the MAGNIMS study group Abstract Correspondence to: Background: White matter lesions (WMLs) on brain magnetic resonance imaging (MRI) in multiple N Evangelou Mental Health and Clinical sclerosis (MS) may contribute to misdiagnosis. In chronic active lesions, peripheral iron-laden macro- Neurosciences Academic phages appear as paramagnetic rim lesions (PRLs). Unit, School of Medicine, University of Nottingham, Objective: To evaluate the sensitivity and specificity of PRLs in differentiating MS from mimics using Queen’s Medical Centre, Nottingham NG7 2UH, UK. clinical 3T MRI scanners. nikos.evangelou@ Method: This retrospective international study reviewed MRI scans of patients with MS (n = 254), MS nottingham.ac.uk Isobel Meaton mimics (n = 91) and older healthy controls (n = 217). WMLs, detected using fluid-attenuated inversion Amjad Altokhis recovery MRI, were analysed with phase-sensitive imaging. Sensitivity and specificity were assessed for Christopher Martin Allen Nikos Evangelou PRLs. Mental Health and Clinical Results: At least one PRL was found in 22.9% of MS and 26.1% of clinically isolated syndrome (CIS) Neurosciences Academic Unit, School of Medicine, patients. Only one PRL was found elsewhere. The identification of ⩾1 PRL was the optimal cut-off and University of Nottingham, Nottingham, UK had high specificity (99.7%, confidence interval (CI) = 98.20%–99.99%) when distinguishing MS and Margareta A Clarke CIS from mimics and healthy controls, but lower sensitivity (24.0%, CI = 18.9%–36.6%). All patients Institute of Imaging Science, Vanderbilt University with a PRL showing a central vein sign (CVS) in the same lesion (n = 54) had MS or CIS, giving a speci- Medical Center, Vanderbilt ficity of 100% (CI = 98.8%–100.0%) but equally low sensitivity (21.3%, CI = 16.4%–26.81%) University, Nashville, TN, USA Conclusion: PRLs may reduce diagnostic uncertainty in MS by being a highly specific imaging diagnos- Tim Sinnecker tic biomarker, especially when used in conjunction with the CVS. Dominik Meier Medical Image Analysis Center AG and Department of Biomedical Engineering, University Basel, Basel, Keywords: Multiple sclerosis, MRI, CIS, biomarkers Switzerland Christian Enzinger Department of Neurology, Date received: 11 January 2022; revised: 3 July 2022; accepted: 12 July 2022 Medical University of Graz, Graz, Austria Massimiliano Calabrese Neurology Unit, Department Introduction recent changes to the MRI criteria decreased the diag- of Neuroscience, Biomedicine and Movement The need for accurate, early diagnosis and considera- nostic specificity,7–9 as WMLs can be present in other Sciences, University of tion of early treatment of multiple sclerosis (MS) conditions such as migraine,10 neuromyelitis optica Verona, Verona, Italy introduces challenges for clinicians.1 The 2017 modi- spectrum disorder (NMOSD)11 and central nervous Nicola De Stefano Antonio Giorgio fied McDonald diagnostic criteria2 necessitate typical system vasculitis.12 Department of Medicine, clinical symptoms and the presence of white matter Surgery and Neuroscience, University of Siena, Siena, lesions (WMLs) on magnetic resonance imaging There is growing acceptance of the role of the central Italy (MRI). These criteria shorten time to diagnosis3 and vein sign (CVS) in diagnosing MS13 leading to Alain Pitiot improve the sensitivity of diagnosing MS.4 Yet misdi- increased use of phase-sensitive imaging at the time Laboratory of Image and Data Analysis, Ilixa Ltd, agnosis is still common,5,6 especially when the MRI of first clinical presentation.14,15 London, UK criteria are incorrectly applied outside of a typical Menno M Schoonheim clinical presentation or when there is incorrect inter- Some chronic MS lesions have persistent active Department of Anatomy and Neurosciences, Amsterdam pretation of MRI findings. It has been suggested that demyelination, the products of which are engulfed Neuroscience, Amsterdam 2212 journals.sagepub.com/home/msj I Meaton, A Altokhis et al. UMC, Vrije Universiteit within activated microglia/macrophages on the Insufficient co-registration resulted in exclusion from Amsterdam, Amsterdam, the periphery of the lesion. One such product is ferrous analysis. The registered images were then sectioned Netherlands iron released into the extracellular space during the into eight equal-sized 3D blocks to ensure blinding of Friedemann Paul Neurocure Clinical destruction of oligodendrocytes.16,17 This can be assessors to the patients’ diagnosis.14 Research Center, Charité- detected in vivo with phase-sensitive imaging where Universitätsmedizin Berlin, Corporate Member of it presents as a paramagnetic rim (PR).16,17 Freie Universität Berlin, Paramagnetic rim lesions (PRLs) appear as a hypoin- Image analysis Humboldt-Universität zu Berlin and Berlin Institute of tense, ring-like structures that surround WML on All image analysis was performed by two trained Health, Berlin, Germany phase-sensitive MRI sequences. PRLs may increase investigators (A.A. and I.M.) using 3D Slicer (version Mikolaj A Pawlak in size whereas non-PRLs decrease in size or remain 4.11.2). Each 3D block was reviewed by A.A. or I.M. Department of Neurology and Cerebrovascular unchanged.18,19 and results were collated after all image analysis was Disorders, Poznan University performed to avoid lesion classification in one part of of Medical Sciences, Poznan, Poland This imaging marker has been studied in detail using a brain influencing assessment of other regions of the Reinhold Schmidt 7 Tesla (T) MRI.19–22 Importantly, 3T MRI studies same brain. The supratentorial regions of the FLAIR Department of Neurology, have also detected PRLs in MS23–25 and corroborated MRI scans were analysed for WMLs with a long Medical University of Graz, Graz, Austria the possible diagnostic and prognostic value.26,27 axis ⩾ 3 mm. Lesions were classified based on their Cristina Granziera However, there are reservations on the clinical utility location as cortical/juxtacortical (in direct contact Ludwig Kappos Research Center for as PRLs are only seen in a minority of WMLs.13,28 with the cerebral cortex), periventricular (in direct Clinical Neuroimmunology contact with the lateral/third ventricles), deep WML and Neuroscience Basel (RC2NB), Departments This retrospective international, multicentre study (not in direct contact with the cortex or ventricles) or of Head, Spine and within the Magnetic Resonance Imaging in MS in direct contact with deep grey matter structures.29 Neuromedicine, Clinical Research and Biomedical (MAGNIMS) Study Group aimed to test the potential Engineering, University for PRLs in clinical practice. MRIs for patients with The SWI scans were then analysed for the presence of Hospital, University of Basel, Basel, Switzerland MS and MS mimics (including cerebral small vessel PRLs. A PRL was defined as a hypointense, ring-like Xavier Montalban disease, migraine and NMOSD) were compared. This structure on phase-sensitive imaging. The rim had to Centre d’Esclerosi Multiple dataset was originally collected by Sinnecker et al.14 correspond to the WML edge on the FLAIR scan, encir- de Catalunya (Cemcat), Department of Neurology/ to evaluate the value of CVS in MS. cle it fully or partially and must be visible on at least two Neuroimmunology, Hospital consecutive image slices (Figure 1). As part of this Universitari Vall d’Hebron, Universitat Autonoma de study, CVS was also analysed using the North American Barcelona, Barcelona, Spain Methods Imaging in MS Cooperative (NAIMS) criteria.30 Àlex Rovira Section of Neuroradiology, Department of Radiology, Participants Hospital Universitari Vall The study included 562 participants scanned at 7 MS Quality assessment d’Hebron, Universitat Autònoma de Barcelona, centres across Europe between 2010 and 2016. The Each block was assessed for artefacts and co-registra- Barcelona, Spain participants were enrolled in ongoing observational tion quality of FLAIR and SWI before the detection Jens Wuerfel studies or included in neuroimaging research data- of WMLs. A total of 18 out of 5196 blocks failed this Medical Image Analysis Center AG and Department bases, all of which were approved by the institutional quality test and were excluded from the analysis. of Biomedical Engineering, review board at each centre. All patients provided Once image analysis was completed, the blocks were University Basel, Basel, Switzerland/Neurocure written informed consent prior to MRI. The inclusion de-anonymised and matched to patient data. Clinical Research Center, criteria, diagnostic criteria and patient demographics Charité-Universitätsmedizin Berlin, Corporate Member have been reported previously.16 All patients with of Freie Universität Berlin, NMOSD had antibodies against aquaporin 4.14 Statistical analysis Humboldt-Universität zu Berlin and Berlin Institute of Susceptibility-weighted imaging (SWI) and three- The statistical analysis was performed using IBM Health, Berlin, Germany dimensional (3D) fluid-attenuated inversion recovery SPSS statistics, version 20 (IBM). Sensitivity and *These authors contributed equally to this manuscript. (FLAIR) scans acquired at 3T of sufficient quality specificity were calculated for having at least one were analysed. Scan acquisition details for each cen- PRL per complete scan and presented with 95% con- tre can be found in the supplementary materials of fidence intervals (CIs). A secondary analysis was per- Sinneker et al.16 formed, only considering lesions that demonstrated both the PRL and CVS; this was also presented as sensitivity and specificity with 95% CI. Then, a Image post-processing sequential analysis was performed that first checks FLAIR images from each participant were co-regis- for the PRL and then, if no PRL is detected, checks for tered to the SWI using the ITK registration library CVS across the entire scan. A chi-square test was per- (Insight Software Consortium), which was imple- formed to investigate the location of PRL and WMLs mented in 3D Slicer, version 4.6.2 (Slicer Community). (deep white matter vs all other locations). journals.sagepub.com/home/msj 2213 Multiple Sclerosis Journal 28(14) Figure 1. Consecutive slices of a paramagnetic rim lesion (with a central vein) detected using the fluid-attenuated inversion recovery (a.i. and b.i.) and phase-sensitive imaging (a.ii. and b.ii.), at 3T. As per the study protocol, the lesions demonstrate a hypointense, ring-like structure corresponding to the lesion edge which is present on at least two consecutive slices. The logistic regression was used to produce receiver- and 47 without) containing MS and non-MS lesions. operating characteristic (ROC) curves. In the logistic Reliability was calculated using Cohen’s Kappa. model, diagnosis (MS vs MS-mimics) was set as a dependent variable and PRL (or CVS) as an inde- pendent variable. The ROC curve is a plot of sensitiv- Results ity against 1 − specificity. The sensitivity and specificity values were obtained by varying the cut- Cohort description off to dichotomise PRL (or CVS) (S1). The demographics of the 562 participants (182 males and 380 females) are shown in Table 1. Interrater reliability Interrater and intra-rater reliability for lesion identifica- Lesion count and distribution tion and PRL detection was assessed in a randomly A total of 6017 WMLs were analysed, with 3987 in selected enriched data set of 100 blocks (53 with a PRL MS or clinically isolated syndrome (CIS) patients. The 2214 journals.sagepub.com/home/msj journals.sagepub.com/home/msj Table 1. Overview of participants and analysed lesions. MS CIS Cluster Migraine SLE NMOSD Diabetes Ageing healthy Total headache mellitus controls Number of patients (female) 166 (110) 88 (59) 3 (1) 21 (18) 19 (16) 30 (26) 18 (10) 217 (140) 562 (380) Mean age [SD] 37.3 [7.4] 32.6 [7.7] 49.3 [12.6] 40.8 [8.7] 32.5 [9.5] 46.5 [11.8] 68.3 [14.0] 64.1 [18.1] 48.4 [17.8] Symptom duration, mean (range) 6.5 (0–30.5) 0.3 (0–2.5) – – – 2.9 (0–9.6) – – – White matter lesions analysed No. of patients with ⩾1 WML, (%) 164 (98.8) 84 (95.4) 2 (66.7) 21 (100.0) 15 (78.9) 18 (60.0) 18 (100) 163 (75.1) 485 (86.3) No. of lesions 3065 922 19 266 77 171 355 1142 6017 Median no. per patient (range) 17 (0–61) 7 (0–54) 9.5 (4–15) 10 (1–38) 4 (0–12) 1 (0–38) 17.5 (1–66) 3 (0–30) 7 (0–66) Paramagnetic rim lesions No. of patients with ⩾1 PRL, (%) 38 (22.9) 23 (26.1) 0 0 0 0 1 (5.6) 0 62 (11.3) No. of lesions (%) 72 (2.3) 57 (6.2) 0 0 0 0 1 (0.3) 0 130 (2.2) Median no. per patient (range) 0 (0–6) 0 (0–8) 0 0 0 0 0 (0–1) 0 0 (0–8) MS: multiple sclerosis; CIS: clinically isolated syndrome; NMOSD: neuromyelitis optica disorder; SD: standard deviation; SLE: systematic lupus erythematosus; WML: white matter lesion; PRL: paramagnetic rim lesions. 2215 I Meaton, A Altokhis et al. Multiple Sclerosis Journal 28(14) Figure 2. A box and whisker plot showing the mean and interquartile ranges of the number of white matter lesions per patient in each condition analysed. NMOSD: neuromyelitis optica spectrum disorder; Control: healthy control; CIS: clinically isolated syndrome; MS: multiple sclerosis. mean and interquartile ranges of WMLs per patient 100% (CI = 98.8%–100.0%) and a PPV of 100. The found in each condition are represented in Figure 2. sensitivity of PRL with a CVS for MS was 20.5% Across the analysis, inter-rater reliability for lesion (CI = 12.9%–25.4%) and 22.7% (CI = 14.5%–32.3%) and PRL detection between investigators showed a in the CIS patients. In all MS/CIS patients displaying substantial agreement with a Cohen’s Kappa value of a PRL, 88.5% had a lesion displaying both PRL and 0.640 and 0.696, respectively. Furthermore, intra-rater CVS. The single PRL detected in the patient with dia- reliability had a Cohen’s Kappa value of 0.827. betes did not display the CVS. The identification of ⩾ 1 PRL (optimal cut-off) was associated with high specificity of 99.7%, but low Paramagnetic rim lesions sensitivity of 24.0%, and overall accuracy: area under PRLs were detected in 130 lesions across 62 patients. the curve (AUC) = 0.71, 95% CI = 0.64–0.78. CVS Within the MS cohort, 38 patients (22.9% (CI = 16.7%– detection alone (optimal cut-off of ⩾ 4 CVS) had 30.0%)) had at least one PRL. In the CIS cohort, the specificity of 88.3%, sensitivity of 56.7% and accu- proportion of individuals with at least one PRL was racy: AUC = 0.82, 95% CI = 0.79–0.86. 26.1% (CI = 17.3%–36.6%), or 23 patients. Half of PRL positive scans had a single PRL (Figure 3). A The combination of the two biomarkers (fulfilment of single PRL was found in the scan of a diabetic patient, either ⩾ 1 PRL or ⩾ 4 CVS) further improved the and this was the only PRL detected outside of the MS/ specificity (90.6%), and a relative increase in the sen- CIS cohorts. Although the combined sensitivity of sitivity (57.9%). The overall accuracy: AUC = 0.83 PRL for MS/CIS was 24.0% (CI = 18.9%–29.8%), (95% CI = 0.79–0.87). PRLs had a very high specificity of 99.7% (CI = 98.2%–99.99%) and a PPV (positive predictive We also performed sequential analysis of the two signs: value) of 98.39. identification of any PRLs first, and if no PRL was identified followed by assessment of the presence of All patients with a PRL showing a CVS in the same ⩾4 CVS. The sensitivity of this two-stage analysis was lesion (n = 54) had MS or CIS, giving a specificity of 79.55% (CI = 74.6–83.9) and 70.9% (CI = 64.8–76.4). 2216 journals.sagepub.com/home/msj I Meaton, A Altokhis et al. Figure 3. The number of paramagnetic rim lesions per patient for each cohort with ⩾1 paramagnetic rim lesion. CIS: clinically isolated syndrome; MS: multiple sclerosis. Across the cohort, 73.1% of WMLs were found in we examined a higher number of patients with MS the deep white matter, 19.4% in the periventricular mimics and ageing controls with brain scans showing region, 7.2% were juxtacortical and only 0.3% adja- WMLs, which more commonly cause diagnostic dif- cent to deep grey matter structures. In the MS and ficulties for MS clinicians. CIS cohorts, 70.1% and 66.5% of WMLs were located in the deep white matter, respectively. Yet in In addition to the analysis of the value of PRL and the MS cohort, 84.2% of PRLs were identified in CVS, we performed a sequential analysis (first look- the deep white matter. The chi-square test investi- ing for the presence of PRL, and in the absence of any gating the location of the PRLs and WMLs found PRL assessing for ⩾ 4 CVS). Although this did not PRLs to be more common in the deep white matter lead to improvement of sensitivity and specificity of (p = 0.003). the diagnosis of MS, it may prove popular with MS clinicians as it is time efficient while reviewing MRI scans with WMLs. This sequential analysis of course Discussion needs to be tested in a prospective study. The PRLs detected in phase-sensitive imaging have potential to aid MS diagnosis. In this article, we In both Sinnecker et al.14 and this analysis, we have expand beyond our original MAGNIMS study of the recognised that the special distribution of the MS CVS14 to evaluate PRLs using clinically determined lesions and lesion characteristics may inadvertently 3T MRI protocols. We found the presence of any PRL un-blind the observer to the diagnosis and influence highly specific for MS/CIS. Furthermore, the combi- subsequent lesion characterisation on the same scan. nation of PRL with CVS was found only in patients For that reason, we have tried to improve the blinding with MS or CIS and not in any other diseases studied by parcellating the brain into eight blocks and ran- with WMLs. Maggi et al.26 also reported in their large domising the order of blocks analysis. In this way, we study low sensitivity and high specificity of PRL in are certain that the investigators assessed individual MS. Our study, conducted in different centres, sup- lesions without influence of other brain/lesion ports their findings. It further adds value as characteristics. journals.sagepub.com/home/msj 2217 Multiple Sclerosis Journal 28(14) Although CVS is sensitive to MS, it was found to be lesions smaller than 3 mm in their longest axis. The less specific than PRL. Both of these imaging bio- parcellation method used to truly blind the investiga- markers are acquired on the same MRI sequence tors might have resulted in the moderate reproducibil- and may reduce the need for oligoclonal band test- ity we report. We suspect that in clinical practice, ing which many patients find unpleasant. Our study clinicians will be influenced also by other MRI diag- strengthens the evidence for the role of phase-sensi- nostic features of the WM lesions. Using automated tive imaging in the diagnostic pathway of MS. techniques may prove to be beneficial in improving the accuracy of PRL detection by eliminating human Our study is pragmatic, with clinical scans acquired errors. Once again only prospective studies can assess by many centres, resulting in variability of scan qual- the true diagnostic value of a test. ity, sequences and operators. The results are therefore representative of the performance of this radiological biomarker in clinical practice. The patient-level prev- Conclusion alence of PRLs is within the range previously Paramagnetic rims are a potential imaging biomarker, described by K.C. Ng Kee Kwong et al.31 In this with high diagnostic specificity for MS. They have a cross-sectional study, we did not aim to report on the clinical role to play in decreasing the diagnostic natural history of PRLs but found that the percentage uncertainty in MS. In this large study, a quarter of of lesions with an iron rim was higher in CIS com- MS/CIS patients had at least one PRL. Furthermore, pared to MS. This corroborates previous longitudinal 3T phase-sensitive MRI is widely available and has studies which have suggested that PRLs may eventu- already been proven to reliably identify the CVS. The ally dissipate as neuroinflammation is replaced by combination of these radiological markers detected neurodegenerative pathology.19,32 with the same MRI sequence shows great promise and requires further prospective evaluation, perhaps While in our study we examined the role of PRLs in with added improvements to sequence optimisation. the diagnosis of MS, the debate continues whether most smouldering lesions produce a visible PR. Declaration of Conflicting Interests Expanding lesion volume is important since it may be The author(s) declared no potential conflicts of inter- predictive of long-term clinical disability.18 Our results est with respect to the research, authorship and/or suggest that one PRL is enough to help the diagnosis publication of this article. of MS, but counting the number of PRL might be important as a prognostic factor for long-term disabil- Funding ity. Studies suggest that some PRLs shrink after The author(s) received no financial support for the 7 years, at which point the iron rim has faded along research, authorship and/or publication of this with the diffuse hyperintensity outside the rim.13,30 It article. would be useful to examine the effect of disease-mod- ifying treatments on PRLs. Unfortunately, data about ORCID iDs the multiple disease-modifying treatments used in our Margareta A Clarke https://orcid.org/0000-0001- cohort is unavailable. Furthermore, the scans available 5531-796X were not taken at the time of diagnosis of MS, thus we Nicola De Stefano https://orcid.org/0000-0003- have been unable to determine at what point in the dis- 4930-7639 ease progression PRLs may be most prevalent. Alain Pitiot https://orcid.org/0000-0003-0146- 8572 As this cohort was previously reported, the limita- Mikolaj A Pawlak https://orcid.org/0000-0002- tions are similar.14 This study relied on the investiga- 8607-0194 tors’ clinical diagnosis and we did not independently Cristina Granziera https://orcid.org/0000-0002- assess the accuracy of the MS diagnosis or MS by 4917-8761 subtype. Some publications suggest that relapsing- Ludwig Kappos https://orcid.org/0000-0003- remitting and secondary progressive MS have a dif- 4175-5509 fering prevalence of PRLs.33,34 The parcellated nature Àlex Rovira https://orcid.org/0000-0002-2132- of the blocks, although essential for blinding, may 6750 also have resulted in lesions not being counted if they were dissected by the border of the blocks. This may Supplemental Material account for why not all the MS patients had lesions Supplemental material for this article is available found on their scans, although we also excluded online. 2218 journals.sagepub.com/home/msj I Meaton, A Altokhis et al. References 13. AlTokhis AI, AlOtaibi AM, Felmban GA, et al. Iron 1. Solomon AJ and Corboy JR. The tension between rims as an imaging biomarker in MS: A systematic early diagnosis and misdiagnosis of multiple mapping review. Diagnostics 2020; 10(11): 968. sclerosis. Nat Rev Neurol 2017; 13(9): 567–572. 14. Sinnecker T, Clarke MA, Meier D, et al. Evaluation 2. Thompson AJ, Banwell BL, Barkhof F, et al. of the central vein sign as a diagnostic imaging Diagnosis of multiple sclerosis: 2017 revisions of biomarker in multiple sclerosis. JAMA Neurol 2019; the McDonald criteria. Lancet Neurol 2018; 17(2): 76: 1446–1456. 162–173. 15. Maggi P, Absinta M, Grammatico M, et al. Central 3. Gaetani L, Prosperini L, Mancini A, et al. 2017 vein sign differentiates multiple sclerosis from central revisions of McDonald criteria shorten the time to nervous system inflammatory vasculopathies. Ann diagnosis of multiple sclerosis in clinically isolated Neurol 2018; 83(2): 283–294. syndromes. J Neurol 2018; 265(11): 2684–2687. 16. Dal-Bianco A, Grabner G, Kronnerwetter C, et al. 4. Beesley R, Anderson V, Harding KE, et al. Impact Slow expansion of multiple sclerosis iron rim lesions: of the 2017 revisions to McDonald criteria on the Pathology and 7T magnetic resonance imaging. Acta diagnosis of multiple sclerosis. Mult Scler 2018; Neuropathol 2017; 133(1): 25–42. 24(13): 1786–1787. 17. Hametner S, Wimmer I, Haider L, et al. Iron and 5. Solomon AJ, Bourdette DN, Cross AH, et al. neurodegeneration in the multiple sclerosis brain. Ann The contemporary spectrum of multiple sclerosis Neurol 2013; 74(6): 848–861. misdiagnosis: A multicenter study. Neurology 2016; 18. Hametner S, Dal Bianco A, Trattnig S, et al. Iron 87(13): 1393–1399. related changes in MS lesions and their validity to 6. Geraldes R, Ciccarelli O, Barkhof F, et al. The current characterize MS lesion types and dynamics with ultra- role of MRI in differentiating multiple sclerosis from high field magnetic resonance imaging. Brain Pathol its imaging mimics. Nat Rev Neurol 2018; 14(4): 2018; 28(5): 743–749. 199–213. 19. Absinta M, Sati P, Masuzzo F, et al. Association 7. Wong YYM, de Mol CL, van der Vuurst de of chronic active multiple sclerosis lesions with Vries RM, et al. Real-world validation of the disability in vivo. JAMA Neurol 2019; 76: 1474– 2017 McDonald criteria for pediatric MS. Neurol 1483. Neuroimmunol Neuroinflamm 2019; 6(2): e528. 20. Suthiphosuwan S, Sati P, Guenette M, et al. 8. Gobbin F, Zanoni M, Marangi A, et al. 2017 The central vein sign in radiologically isolated McDonald criteria for multiple sclerosis: Earlier syndrome. AJNR: Am J Neuroradiol 2019; 40(5): diagnosis with reduced specificity. Mult Scler Relat 776–783. Disord 2019; 29: 23–25. 21. Kilsdonk ID, Lopez-Soriano A, Kuijer JP, et al. 9. Van der Vuurst de Vries RM, Mescheriakova JY, Morphological features of MS lesions on FLAIR* Wong YYM, et al. Application of the 2017 revised at 7T and their relation to patient characteristics. J McDonald criteria for multiple sclerosis to patients Neurol 2014; 261(7): 1356–1364. with a typical clinically isolated syndrome. JAMA Neurol 2018; 75(11): 1392–1398. 22. Sinnecker T, Dörr J, Pfueller CF, et al. Distinct lesion morphology at 7-T MRI differentiates neuromyelitis 10. Liu S, Kullnat J, Bourdette D, et al. Prevalence of optica from multiple sclerosis. Neurology 2012; brain magnetic resonance imaging meeting Barkhof 79(7): 708–714. and McDonald criteria for dissemination in space among headache patients. Mult Scler 2013; 19(8): 23. Tillema JM, Weigand SD, Dayan M, et al. Dark rims: 1101–1105. Novel sequence enhances diagnostic specificity in multiple sclerosis. AJNR: Am J Neuroradiol 2018; 11. Huh SY, Min JH, Kim W, et al. The usefulness of 39(6): 1052–1058. brain MRI at onset in the differentiation of multiple sclerosis and seropositive neuromyelitis optica 24. Absinta M, Sati P, Fechner A, et al. Identification of spectrum disorders. Mult Scler 2014; 20(6): 695–704. chronic active multiple sclerosis lesions on 3T MRI. AJNR: Am J Neuroradiol 2018; 39: 1233–1238. 12. Kim SS, Richman DP, Johnson WO, et al. Limited utility of current MRI criteria for distinguishing 25. Kaunzner UW, Kang Y, Zhang S, et al. Quantitative multiple sclerosis from common mimickers: Primary susceptibility mapping identifies inflammation in a and secondary CNS vasculitis, lupus and Sjogren’s subset of chronic multiple sclerosis lesions. Brain syndrome. Mult Scler 2014; 20(1): 57–63. 2019; 142(1): 133–145. journals.sagepub.com/home/msj 2219 Multiple Sclerosis Journal 28(14) 26. Maggi P, Sati P, Nair G, et al. Paramagnetic the North American Imaging in Multiple Sclerosis rim lesions are specific to multiple sclerosis: An Cooperative. Nat Rev Neurol 2016; 12(12): 714–722. international multicenter 3T MRI study. Ann Neurol 31. Ng Kee Kwong KC, Mollison D, Meijboom R, et 2020; 88(5): 1034–1042. al. The prevalence of paramagnetic rim lesions in 27. Clarke MA, Pareto D, Pessini-Ferreira L, et al. multiple sclerosis: A systematic review and meta- Value of 3T susceptibility-weighted imaging in analysis. PLoS ONE 2021; 16(9): e0256845. the diagnosis of multiple sclerosis. AJNR: Am J 32. Absinta M, Sati P, Schindler M, et al. Persistent Neuroradiol 2020; 41(6): 1001–1008. 7-Tesla phase rim predicts poor outcome in new 28. Clarke M, Pessini Ferreira LM, Pareto D, et al. The multiple sclerosis patient lesions. J Clin Invest 2016; central vein sign and iron rings: Insights from a 126(7): 2597–2609. large cohort of patients with multiple sclerosis and 33. Dal-Bianco A, Grabner G, Kronnerwetter C, et al. mimicking disorders. ECTRIMS 2019; 25: 38–39. Slow expansion of multiple sclerosis iron rim lesions: 29. Mittal S, Wu Z, Neelavalli J, et al. Susceptibility- Pathology and 7T magnetic resonance imaging. Acta weighted imaging: Technical aspects and clinical Neuropathol 2017; 133(1): 25–42. applications, part 2. AJNR: Am J Neuroradiol 2009; 34. Yao B, Ikonomidou VN, Cantor FK, et al. 30(2): 232–252. Visit SAGE journals online Heterogeneity of multiple sclerosis white matter journals.sagepub.com/ 30. Sati P, Oh J, Constable RT, et al. The central vein lesions detected with T2*-weighted imaging at home/msj sign and its clinical evaluation for the diagnosis 7.0 Tesla. J Neuroimaging 2015; 25(5): 799– SAGE journals of multiple sclerosis: A consensus statement from 806. 2220 journals.sagepub.com/home/msj