Abstract
The new McDonald 2010 criteria have been recommended in paediatric multiple sclerosis (PMS). We aimed to assess the utility of McDonald 2010 criteria in comparison with 2007 International Paediatric Multiple Sclerosis Study Group (IPMSSG)-recommended criteria for PMS diagnosis. Retrospective analysis of 38 PMS cases from three UK demyelination clinics was conducted. Dissemination in space (DIS) and time (DIT) for both McDonald and IPMSSG criteria were noted on initial and follow-up magnetic resonance imaging (MRI). At first MRI scan, IPMSSG DIS criteria were fulfilled in 68% of scans and McDonald DIS criteria in 84%. In total, 11/18 children given gadolinium contrast fulfilled both McDonald DIS and DIT criteria on initial scan. The 2010 McDonald criteria appear more sensitive than IPMSSG and may allow PMS diagnosis at first presentation of CIS in at least a half of cases.
Introduction
Paediatric multiple sclerosis (MS) is becoming increasingly recognized, 1 with up to 10% of adults having their first symptoms in childhood.2,3 Magnetic resonance imaging (MRI) is increasingly utilized in aiding the prompt diagnosis of MS in children. 4 Currently the International Paediatric MS Study Group (IPMSSG) recommends that paediatric MS diagnosis is based on McDonald 2001 MRI criteria5,6 whilst eliminating any lower age limit. In addition, the IPMSSG state that an episode of acute disseminated encephalomyelitis (ADEM) cannot be considered as the first event of MS, unless a second non-ADEM demyelinating event is accompanied by further evidence of dissemination in time (DIT) (either with new MRI T2 lesions ≥ 3 months from the second event, or a new [third] clinical event developing ≥ 3 months subsequent to the second event). Recently the new McDonald 2010 diagnostic criteria, for the first time, have been recommended for use in children with acute demyelination presenting with a clinically isolated syndrome (CIS). 7 Common and important to both the McDonald 2001 and McDonald 2010 MRI criteria is the requirement of lesion dissemination in space (DIS) and time (DIT). The McDonald 2001 criteria state DIS can be fulfilled if three out of four features are satisfied: (1) nine or more T2 white matter lesions or one gadolinium-enhanced lesion; (2) one juxtacortical lesion; (3) three periventricular lesions; (4) one infratentorial lesion. The McDonald 2001 DIT criteria are fulfilled if new T2 lesions or gadolinium-enhanced lesions develop after 3 months following the initial attack. 5 McDonald 2010 criteria state that DIS criteria are fulfilled if one or more T2 lesion is present in two out of four areas: (1) juxtacortical; (2) periventricular; (3) infratentorial; (4) spinal cord (if the patient has a brainstem or spinal cord syndrome, symptomatic lesions are excluded from the criteria and do not contribute to the count). McDonald 2010 DIT criteria can be fulfilled if a new T2 and or gadolinium-enhanced lesion is present on follow-up MRI irrespective of the timing of a baseline MRI, or if there is simultaneous presence of gadolinium-enhanced and non-enhanced lesions at any time (including at time of first scan).
Previous retrospective MRI paediatric MS studies have shown McDonald 2001 criteria lack sensitivity for diagnosis of paediatric MS.4,8 The same study groups have also shown that McDonald 2001 MRI criteria have high specificity and differentiated children with MS from those with relapsing non-demyelinating disorders with central nervous system (CNS) involvement, such as migraine and small vessel vasculitis. 8 We hence aimed to assess the utility of the new McDonald 2010 criteria in comparison with the 2007 IPMSSG criteria for the early diagnosis of paediatric MS, and also compare these criteria in a cohort of children with relapsing non-demyelinating neurological disorders. To our knowledge this is the first study examining the McDonald 2010 criteria in a paediatric cohort.
Methods
Cases with clinically definite relapsing–remitting MS 9 from three tertiary UK paediatric demyelination clinics were retrospectively identified using hospital databases for MS. The demyelination clinics are led by paediatric neurologists with a specialist interest in MS and allow comprehensive and longitudinal evaluation of children with CNS inflammatory demyelination. Consecutive cases of children younger than 16 years old with a first demyelinating episode (post publication of McDonald 2001 criteria; 2001–2009) and with both initial (within 3 months of clinical onset) and first follow-up scan (routine or relapse associated) available for review were included. All MRI scans were performed on 1.5T scanners. DICOM viewing software (IMPAX version; © 2007 Agfa HealthCare, Mortsel, Belgium) was used to view scans. Sequences used included, T1, T2, T2 FLAIR ± gadolinium-enhanced. A standardized MRI protocol was not used as this was a retrospective study based on collecting data from established clinical practice. MRI slice thickness varied from 3–5 mm. MRI scans were examined for the presence of DIS and DIT criteria which allowed diagnosis via the McDonald 2001 criteria and McDonald 2010 criteria, respectively. Two assessors reviewed scans and reports, with consensus agreement with regards lesion number and location, and then inputted anonymized data into a standardized Excel spreadsheet. Migraine patients (International Headache Society Criteria 10 ) with MRI brain T2 white matter signal changes with or without other CNS manifestations were used as controls (age and sex matched to patients with MS). Statistical analysis was performed using PASW Statistics for Windows version 18.0 (© SPSS, Inc., 2009, Chicago, IL, www.spss.com). Descriptive and non-parametric statistics were used (binomial, Fishers exact, Kruskal–Wallis, and McNemar tests) to summarize key aspects of the dataset. Ethical approval was deemed not necessary for this retrospective evaluation of criteria for routine clinical practice (National Research Ethics Service and National Patient Safety Queries Line 17/06/2011) and confidentiality of patients was maintained (data anonymized and cases cannot be identified).
Results
The initial search identified 49 patients, 11 of whom were excluded due to: lack of availability of initial or first follow-up scan for review (two patients); MS diagnosis not clinically confirmed (two patients); monophasic ADEM (three patients); multiphasic ADEM (two patients); or neuromyelitis optica (two patients). In total, 38 children with MS (Table 1) were included, 28 female and 10 male (p = 0.005). The median age of onset was 12.9 years (25–75th centile = 10.5–14.3 years) and duration of follow-up was median 3.5 years (25–75th centile = 2.3–4.6 years). Clinical features at first presentation of the MS cohort included: optic neuritis (16 patients); cerebellar (12 patients); pyramidal (11 patients); brainstem (10 patients); headache (seven patients); hemisensory (five patients) and myelitis (four patients). One patient presented with encephalopathy on initial attack but had two subsequent CIS relapses consistent with MS. Time to first relapse was median 0.6 years (25–75th centile = 0.3–0.8 years) and median relapse rate was 0.8 relapses/year (25–75th centile = 0.4–1.5 relapses/year). Cerebrospinal fluid oligoclonal bands were positive in 23/31 (74%) and aquaporin-4 antibody was negative in 6/6 patients tested. Twenty-nine children were given disease modifying treatment (DMT) in the form of interferon 1a (seven patients), interferon 1b (17 patients) or glatiramer acetate (five patients) at a median age of 13.7 years (25–75th centile = 12.4–15.9 years).
Demographics of the paediatric multiple sclerosis cohort (n = 38).
All patients had T2 white matter brain lesions at first MRI (Table 2). The McDonald 2001 DIS criteria were fulfilled in 26/38 (68%) at first scan, whilst McDonald 2010 DIS criteria were fulfilled in 32/38 (84% p = 0.03). Only 18 patients had gadolinium contrast administered at first scan, of which 11 showed lesion enhancement (61%) hence fulfilling McDonald 2010 DIT criteria. The 11 children who had gadolinium contrast upon their initial scan also fulfilled McDonald 2010 DIS criteria, and hence fulfilled a diagnosis of MS at first presentation. On follow-up MRI, 3/38 (8%) children failed to meet the McDonald 2001 DIS and DIT criteria, whilst all children fulfilled McDonald 2010 DIS and DIT criteria.
Initial MRI lesion distribution and fulfilment of McDonald 2001 and 2010 criteria on initial and follow-up scan.
Eighteen control cases were matched for age and sex to MS cases. The median age was 10.9 years for both groups (MS cases 25–75th centile = 8.1–14.1; control 25–75th centile = 7.7–14.2, p = not significant). Six control cases had follow-up imaging at a median of 0.7 years (range: 0.3–2.1 years). There was no significant difference between the number of total T2 white matter lesions found in MS and controls (Table 3). Of those control cases given contrast (n = 2) at first scan, none showed enhancement. Only 1/18 control case fulfilled both the McDonald 2001 and McDonald 2010 DIS criteria at first scan. This case did not fulfil the McDonald 2001 or McDonald DIT criteria on follow-up scan (was never given contrast). Of the five other control cases that had a follow-up scan performed, all failed to meet DIS and DIT according to both the McDonald 2001 criteria and McDonald 2010 criteria.
Lesion distribution and McDonald 2010 criteria eligibility of matched control and MS cases.
p < 0.001 Fisher and McNemar tests
Discussion
To our knowledge this is the first study examining the utility of the new McDonald 2010 criteria in a paediatric MS cohort. The 2010 McDonald criteria allow MS diagnosis on the first scan and may allow diagnosis at first presentation of CIS in at least a half of paediatric cases. In addition, our study confirms that on follow-up imaging, the McDonald 2001 MS criteria (currently recommended by IPMSSG) were met by fewer patients in contrast to McDonald 2010 criteria. This study also supports the hypothesis that children presenting with typical clinical features of a CIS consistent with CNS inflammatory demyelination 7 should be given gadolinium to allow for earlier diagnosis. The abolition of specific time limits, as those of the older criteria, 6 may make it easier for paediatric neurologists with an interest in MS to decide on early treatment of their patients and allow for prompt entry in to future paediatric clinical trials. From the control group, only one control case fulfilled the McDonald 2001 and McDonald 2010 DIS criteria but did not fulfil DIT criteria on follow-up. This is potentially reassuring, given that the new simpler criteria did not perform differently when compared with the more stringent established criteria.
Limitations of this study include: its retrospective nature; the fact that it contains cases with confirmed clinically definite MS as opposed to being a prospective follow-up cohort study at first CIS presentation; that not all children had spinal imaging and contrast administered; the relatively small numbers precludes testing subgroups; and that MRI sequence variation could affect the sensitivity of results. Nevertheless the study provides useful data towards assessing the utility of the new McDonald 2010 criteria in children. Future longitudinal prospective studies are required to further evaluate these criteria in children presenting with a first episode CIS.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflicts of interest statement
MJL, CH, EW, and MA are contributors to a prospective study of childhood CNS inflammatory demyelination funded by the MS Society and Action Medical Research (UK).