- Email: [email protected]
Comparison of the 2010 and 2005 versions of the McDonald MRI criteria for dissemination-in-time in Taiwanese patients with classic multiple sclerosis
Journal of the Neurological Sciences 329 (2013) 51–54
Contents lists available at SciVerse ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Comparison of the 2010 and 2005 versions of the McDonald MRI criteria for dissemination-in-time in Taiwanese patients with classic multiple sclerosis Chun-Jen Hsueh a, Hung-Wen Kao a, Shao-Yuan Chen b, c, Chung-Ping Lo d, e,⁎, Chia-Chun Hsu d, e, Dai-Wei Liu e, Wen-Lin Hsu e a
Department of Radiology, Tri-Service General Hospital and National Defense Medical Center, Taipei, Taiwan Department of Neurology and Hyperbaric Medicine, Cardinal Tien Hospital, New Taipei City, Taiwan School of Medicine, Fu Jen Catholic University, New Taipei City, Taiwan d Department of Radiology, Buddhist Tzu Chi General Hospital, Taichung Branch, Taichung, Taiwan e School of Medicine, Tzu Chi University, Hualien, Taiwan b c
a r t i c l e
i n f o
Article history: Received 3 February 2013 Received in revised form 10 March 2013 Accepted 25 March 2013 Available online 15 April 2013 Keywords: Multiple sclerosis Magnetic resonance imaging Brain Demyelinating disease Clinically isolated syndrome Diagnostic imaging
a b s t r a c t In 2010, the International Panel on the Diagnosis of Multiple Sclerosis revised the 2005 version of the McDonald criteria. The revisions to MRI dissemination-in-time criteria include adoption of a new criterion by demonstration of simultaneous asymptomatic gadolinium-enhancing and nonenhancing lesions on baseline MRI scans. The purpose of this study was to demonstrate the diagnostic validity of the modified MRI dissemination-in-time criteria. We collected 80 patients with an initial clinical attack suggestive of an acute central nervous system demyelinating disease. The patients were followed for at least two years or until the development of definite multiple sclerosis. The nonconverters were taken as negative cases. Their baseline and follow-up brain MRI studies were retrospectively reviewed by two neuroradiologists. The 2010 version had higher sensitivity (68.2% vs. 45.5%), slightly lower specificity (80.6% vs. 83.3%), and higher accuracy (73.8% vs. 62.5%) than the 2005 version, but the differences were without statistical significance. The new criteria are more sensitive and accurate and specific just as the old criteria. They allow the diagnosis of definite multiple sclerosis in 34.1% patients at first presentation of the clinically isolated syndrome. © 2013 Elsevier B.V. All rights reserved.
1. Introduction According to McDonald criteria, multiple sclerosis (MS) is diagnosed based on the objective demonstration of central nervous system (CNS) demyelinating lesions disseminated in space (DIS) and time (DIT) after exclusion of alternative diagnoses that might mimic the disease [1–3]. Magnetic resonance imaging (MRI) findings can substitute for clinical evidence of one lesion and one clinical attack if they fulfill the MRI DIS and DIT criteria, respectively [1–3]. In 2010, the International Panel on the Diagnosis of Multiple Sclerosis adopted a new MRI DIS criterion proposed by Swanton et al. (i.e., at least one T2-lesion in two or more of the following CNS regions: periventricular, juxtacortical, infratentorial, and spinal cord) [4,5]. Our previous study found that the specificity and sensitivity of this criterion were comparable to those of the MRI DIS criterion included in the 2005 version of the McDonald criteria when applied to Taiwanese patients with classic MS [3–6]. In addition, the panel revised the 2005 McDonald criteria by adopting a new criterion proposed by Rovira et al. [3,7]. The new MRI
DIT criteria emphasized the early diagnosis of MS based on a single baseline MRI demonstrating simultaneously the presence of asymptomatic gadolinium-enhancing and nonenhancing lesions indicative of demyelinating plaques in different stages of evolution [7,8]. The two sets of MRI DIT criteria (from 2005 to 2010) are summarized in Table 1. Clinically isolated syndrome (CIS) refers to a single clinical attack of CNS inflammatory demyelinating symptoms suggestive of MS. CIS presentations can be monofocal or multifocal, and typically involve the optic nerve, brainstem, cerebellum, spinal cord, or cerebral hemispheres. Although CIS may represent the first manifestation of definite multiple sclerosis (DMS), some patients (i.e., nonconverters) may not develop a second clinical relapse. The purpose of this study is to validate the diagnostic ability of the two sets of MRI DIT criteria in DMS patients and nonconverter patients. 2. Methods 2.1. Patient selection
⁎ Corresponding author at: Department of Radiology, Buddhist Tzu Chi General Hospital, Taichung Branch, No. 66, Sec. 1, Fongsing Road, Tanzih District, Taichung 427, Taiwan. Tel.: + 886 2 26479608. E-mail address: [email protected] (C.-P. Lo). 0022-510X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jns.2013.03.018
The study protocol was approved by the local institutional review board (TSGH IRB 1-101-05-004). From January 2001 to June 2010, we collected 80 patients with an initial clinical attack suggestive of an
52
C.-J. Hsueh et al. / Journal of the Neurological Sciences 329 (2013) 51–54
2.3. Statistical analysis
Table 1 The 2005 and 2010 revisions to the McDonald criteria for MRI DIT. 2005 revisions
2010 revisions
Detection of gadolinium enhancement at least 3 months after the onset of the initial clinical event, if not at the site corresponding to the initial event or detection of a new T2 lesion if it appears at any time compared with a reference scan done at least 30 days after the onset of the initial clinical event A new T2 and/or gadolinium-enhancing lesion(s) on follow-up MRI, with reference to a baseline scan, irrespective of the timing of the baseline MRI or simultaneous presence of asymptomatic gadolinium-enhancing and nonenhancing lesions at any time
3. Results
MRI = magnetic resonance imaging; DIT = dissemination-in-time.
Table 2 Comparison of MRI findings in DMS patients and nonconverters. New T2Simultaneous Gd+ and Gd− lesion on F-U MRI lesions on baseline MRI DMS (n = 44) 17 (38.6%) Nonconverter 1 (2.8%) (n = 36)
New Gd+ lesion on F-U MRI
Fulfill 2005 MRI
Patients who finally converted to DMS were considered positive cases and the nonconverters after at least two years of follow-up were considered negative cases. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of the two sets of MRI DIT criteria were calculated. The significance of differences was determined with 95% confidence intervals (95% CI), and the absence of overlapping was considered an indicator of difference.
Fulfill 2010 MRI DIT
DIT
20 (45.5%) 5 (11.4%) 20 (45.5%) 30 (68.2%) 6 (16.7%) 0 (0%) 6 (16.7%) 7 (19.4%)
MRI = magnetic resonance imaging; DMS = definite multiple sclerosis; Gd+ = gadolinium-enhancing; Gd− = gadolinium-nonenhancing; F-U = follow-up; DIT = dissemination-in-time.
acute CNS demyelinating disease (i.e., CIS). They had clinical follow-up of at least two years or until the development of DMS. Some of the subjects were in the cohort we studied previously [6,9]. All 80 patients met the following inclusion criteria: (1) age of 15 to 50 years; (2) baseline brain MRI obtained within three months of symptom onset; (3) follow-up brain MRI available; (4) MRI sequences containing gadolinium enhancement; and (5) no use of disease-modifying agents before the second clinical episodes (except for corticosteroid therapy). The exclusion criteria were: (1) clinical suspicion of acute disseminated encephalomyelitis (ADEM); and (2) symptoms and MRI abnormalities confined to optic nerves and spinal cord fulfilling the diagnostic criteria of neuromyelitis optica proposed by Wingerchuk et al. [10]. (3) other CNS white matter diseases were excluded through clinical work-ups.
2.2. MRI sequences and imaging analysis The MRI studies were performed on 1.5-T MR scanners. The brain MRI sequences included spin echo (SE) or fast spin echo (FSE) T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), and fluid-attenuated inversion recovery (FLAIR) in the axial plane, T2WI or T2-FLAIR in the sagittal plane, and contrast-enhanced T1WI in axial, coronal, and sagittal planes. Follow-up MRI scans were compared with the baseline scans. Spinal cord MRI studies were not included in this analysis because most patients received no or only one study and therefore DIT could not be demonstrated. Two neuroradiologists who were blind to the final diagnosis retrospectively reviewed the MRI scans and reached a consensus agreement on whether the two sets of MRI DIT criteria were fulfilled.
The recruited 80 patients consisted of 28 men and 52 women (mean age, 33.4 years; range, 15–50 years). Finally, 44 (55%) patients converted to DMS with a mean interval of 9.6 months after the onset of initial clinical events (range, 2–24 months) and 36 patients did not convert. The diagnosis of DMS in 44 patients was made by experienced neurologists and based on clinical grounds (two or more clinical attacks of CNS demyelinating events and objective clinical evidence of two or more lesions as defined in McDonald criteria). The mean clinical follow-up duration of the whole cohort was 66.4 months (range, 24–130 months). Clinical features at first presentation among the 80 patients included: unilateral optic neuritis (n = 24 [30.0%]), brainstem symptoms (n = 18 [22.5%]), multifocal symptoms (n = 23 [28.8%]), and spinal cord symptoms (n = 22 [27.5%]). All of the baseline brain MRI studies were performed within three months after initial symptom onset. Twenty-seven patients had more than one follow-up brain MRI scan (range, 1–4 scans). The mean interval between the baseline and follow-up MRI scans was 6.7 months (range, 3–28 months). Table 2 summarizes the MRI findings of the DMS patients and nonconverters. Among the 44 patients with DMS, 17 had simultaneous asymptomatic gadolinium-enhancing and nonenhancing lesions on baseline brain MRI; 20 had new T2-lesions, and five had new gadolinium-enhancing lesions on follow-up MRI scans. Some patients had more than one of these findings on MRI. Overall, 30 patients fulfilled the 2010 version of the McDonald criteria for MRI DIT and 20 patients fulfilled the 2005 version of these criteria. Among the 17 patients who had simultaneous asymptomatic gadolinium-enhancing and nonenhancing lesions on baseline brain MRI (with a mean interval of 15 days from symptom onset), 15 patients also fulfilled the MRI DIS criteria of 2010, which means that the diagnosis of DMS could be established in 34.1% (15/44) patients based on early clinical events and baseline brain MRI studies without the need for a second clinical attack or follow-up MRI according to the new criteria. The other 13 patients fulfilled the 2010 MRI DIT criteria based on follow-up MRI scans (with a mean interval of 168 days from symptom onset). Four of them already had a second clinical attack before follow-up MRI scans and thus had been diagnosed as DMS. Among the 36 nonconverters, one had simultaneous asymptomatic gadolinium-enhancing and nonenhancing lesions on baseline brain MRI; six had new T2-lesions and none had new gadoliniumenhancing lesions on follow-up scans (i.e., six nonconverters fulfilled the MRI DIT criteria of 2005 and seven fulfilled those of 2010).
Table 3 Performance of the 2005 and 2010 revisions to the McDonald criteria for MRI DIT.
2005 revisions 95% CI 2010 revisions 95% CI
TP
FP
TN
FN
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
Accuracy (%)
20
6
30
24
30
7
29
14
45.5 30.7–60.2 68.2 55.4–81.9
83.3 71.2–95.5 80.6 67.6–93.5
76.9 60.7–93.1 81.1 68.5–93.7
55.6 42.3–68.8 67.4 53.4–81.5
62.5 51.9–73.1 73.8 64.1–83.4
MRI = magnetic resonance imaging; DIT = dissemination-in-time; TP = true positive; FP = false positive; TN = true negative; FN = false negative; PPV = positive predictive value; NPV = negative predictive value; 95% CI = 95% confidence interval.
C.-J. Hsueh et al. / Journal of the Neurological Sciences 329 (2013) 51–54
53
The performance of the two sets of MRI DIT criteria is summarized in Table 3. The 2010 version had higher sensitivity (68.2% vs. 45.5%), slightly lower specificity (80.6% vs. 83.3%), and higher accuracy (73.8% vs. 62.5%) compared to the 2005 version. However, there was no significant statistical difference in the sensitivity, specificity, PPV, NPV, or accuracy among the two sets of criteria. 4. Discussion Our study shows that (compared to the 2005 version) the new version of the McDonald criteria is more accurate. The new version appears to increase sensitivity while maintaining the specificity attributable to the previous version. Recently a Spanish cohort study revealed a 52.63% sensitivity, a 75.00% specificity and a 59.26% accuracy for the 2010 MRI DIT criteria applying to patients with CIS [11]. However, their study was based on a single baseline brain MRI scan. If we exclude the follow-up brain MRI scans in our study, it would be a 38.6% sensitivity, a 97.2% specificity and a 65.0% accuracy, reflecting the relatively fewer enhancing brain lesions on baseline MRI in our study population. Six cases were false positives (without conversion to DMS) because new T2-lesions were detected on follow-up MRI. Positive MRI findings tend to occur more frequently than clinical findings in MS. New T2-lesions and gadolinium-enhancing lesions indicative of increasing axon pathology can be five to ten times more common than clinical signs of disease progression [12–14]. The involvement of non-eloquent areas of the brain or the masking effect of cortical adaptation may explain why MRI-detectable “subclinical” lesions may be present without clinically apparent disease [15]. Recent studies have also found individuals with incidentally-identified MRI abnormalities in the CNS highly suggestive of MS (fulfilling Barkhof–Tintoré criteria for DIS), but without clinical symptoms for demyelinating disease. The condition was defined as radiologically isolated syndrome (RIS), unexpected multiple sclerosis, or radiologically uncovered asymptomatic possible inflammatorydemyelinating disease (RAPIDD) [16–19]. About 22.7–36.7% patients with RIS have been reported to develop at least one clinical attack of CNS demyelinating symptoms during follow-up periods ranging from 2.4 to 12.5 years, and the initial MRI abnormalities may represent “preclinical” lesions in those patients [16–19]. In our study, the baseline MRI scans in two of our patients with DMS did not fulfill the MRI DIT criteria of 2010 because the only enhancing lesions were symptomatic and had to be excluded by definition. Enhancing lesions (particularly the nodular enhancing pattern) are usually encountered in the acute demyelinating stage and are more frequently symptomatic, while nonenhancing lesions usually indicate inactive or old plaques and are often asymptomatic. From an imaging perspective, the simultaneous presence of enhancing and nonenhancing lesions may suggest different stages of demyelination or dissemination-in-time (i.e., DIT) regardless of whether the lesions are symptomatic. Whether symptomatic enhancing lesions have to be excluded from the lesion count may need further discussion. The major concern of differential diagnosis is ADEM, which is typically a monophasic disease. ADEM often shows enhancement of all the lesions in the acute stage but a mixture of enhancing and nonenhancing lesions during recovery [8]. Patients' younger age, clinical history of prior infection or vaccination, presence of encephalopathy, frequent gray matter involvement on brain MRI, and less frequent oligoclonal bands in cerebrospinal fluid are characteristic features for ADEM [20,21]. The average duration of gadolinium enhancement of a demyelinating plaque is about one to two months [22,23], and temporal variations in the enhancement duration may potentially cause false positive and false negative results for DIT, depending on the timing of MRI examinations. For example, two plaques in the same stage but having different enhancement durations may lead to false positive results for DIT on baseline MRI (Fig. 1), whereas two plaques in different stages but
Fig. 1. Two plaques of the same stage but having different durations of contrast enhancement. If MRI study is performed at the time point marked by the dotted line, it may lead to a false positive result (simultaneous presence of enhancing and nonenhancing lesions).
having overlapping enhancement or nonenhancement periods may lead to false negative results (Fig. 2). The use of corticosteroid may also affect lesion enhancement. Consequently, follow-up MRI may still have an important role in the determination of DIT. The limitations of this study are as follows: (1) The minimum clinical follow-up duration was set at two years, and some patients may develop later conversion to DMS. (2) Spinal cord MRI studies were only obtained in patients with symptoms of spinal cord involvement, and most of them were performed once. As a result, the numbers of patients fulfilling McDonald MRI criteria might be underestimated. A previous study has shown limited impact of spinal cord MRI on the evidence of DIT [24]. However, other two studies have shown that the presence of spinal cord lesion was associated with a higher risk and shorter time to conversion to CDMS [25,26]. (3) The case number is relatively small and it may affect the statistical significance. (4) Because the study is a retrospective one, the time points for clinical evaluation and MRI follow-up were without a fixed schedule. The results may be influenced by this heterogeneity. In conclusion, our study shows that the MRI DIT criteria of 2010 increase sensitivity and accuracy without significantly compromising the specificity of the MRI DIT criteria of 2005. The new criteria allowed the diagnosis of DMS to be made in 34.1% patients at the CIS stage. It may help to select candidates for disease-modifying treatments from patients with CIS to delay the time to a second demyelinating event and to prevent future disability. We suggest that patients
Fig. 2. Two plaques of different stages and having overlapping durations of contrast enhancement and nonenhancement. The MRI findings may lead to false negative results if performed at the time points marked by the dotted lines.
54
C.-J. Hsueh et al. / Journal of the Neurological Sciences 329 (2013) 51–54
with clinical features of CIS should receive contrast-enhanced MRI for a possible earlier diagnosis of DMS, and follow-up MRI is still important in the demonstration of DIT. The current study of MRI DIT criteria along with our previous study of MRI DIS criteria [6] suggests that the 2010 version of McDonald MRI criteria is simple to use and as accurate as the 2005 version with possible early diagnosis of MS at CIS stage. Funding None. Conflict of interest None. Acknowledgments None. References [1] McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on diagnosis of multiple sclerosis. Ann Neurol 2001;50:121–7. [2] Polman CH, Reingold SC, Edan G, Fillipi M, Hartung HP, Kappos L, et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald Criteria”. Ann Neurol 2005;58:840–6. [3] Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Fillipi M, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 2011;69:292–302. [4] Swanton JK, Fernando K, Dalton CM, Miszkiel KA, Thompson AJ, Plant GT, et al. Modification of MRI criteria for multiple sclerosis in patients with clinically isolated syndromes. J Neurol Neurosurg Psychiatry 2006;77:830–3. [5] Swanton JK, Rovira A, Tintoré M, Altmann DR, Barkhof F, Fillipi M, et al. MRI criteria for multiple sclerosis in patients presenting with clinically isolated syndromes: a multicentre retrospective study. Lancet Neurol 2007;6:677–86. [6] Lo CP, Kao HW, Chen SY, Hsueh CJ, Lin WC, Hsu WL, et al. Prediction of conversion from clinically isolated syndrome to clinically definite multiple sclerosis according to baseline MRI findings: comparison of revised McDonald criteria and Swanton modified criteria. J Neurol Neurosurg Psychiatry 2009;80:1107–9. [7] Rovira A, Swanton J, Tintoré M, Huerga E, Barkhof F, Fillipi M, et al. A single, early magnetic resonance imaging study in the diagnosis of multiple sclerosis. Arch Neurol 2009;66:587–92. [8] Montalban X, Tintoré M, Swanton J, Barkhof F, Fazekas F, Fillipi M, et al. MRI criteria for MS in patients with clinically isolated syndromes. Neurology 2010;74:427–34.
[9] Chen SY, Lo CP, Hsu WL, Kao HW, Hsueh CJ, Lin WC, et al. Modifications to the McDonald MRI dissemination in space criteria for use in Asians with classic multiple sclerosis: the Taiwanese experience. Mult Scler 2010;16:1213–9. [10] Wingerchuk DM, Lennon VA, Pittock SJ, Lucchinetti CF, Scottsdale AZ. Revised diagnostic criteria for neuromyelitis optica. Neurology 2006;66:1485–9. [11] Gómez-Moreno M, Díaz-Sánchez M, Ramos-González. Application of the 2010. McDonald criteria for the diagnosis of multiple sclerosis in a Spanish cohort of patients with clinically isolated syndromes. Mult Scler 2012;18:39–44. [12] Thrower BW. Clinically isolated syndromes: predicting and delaying multiple sclerosis. Neurology 2007;68(Suppl. 4):S12–5. [13] Miller DH, Albert PS, Barkhof F, Francis G, Frank JA, Hodgkinson S, et al. Guidelines for the use of magnetic resonance techniques in monitoring the treatment of multiple sclerosis. Ann Neurol 1996;39:6–16. [14] Barkhof F, Scheltens P, Frequin ST, Nauta JJ, Tas MW, Valk J, et al. Relapsing–remitting multiple sclerosis in determining disease activity. AJR Am J Roentgenol 1992;158:1041–7. [15] Barkhof F. The clinico-radiological paradox in multiple sclerosis revisited. Curr Opin Neurol 2002;15:239–45. [16] Okuda DT, Mowry EM, Beheshtian A, Waubant E, Baranzini SE, Goodin DS, et al. Incidental MRI anomalies suggestive of multiple sclerosis: the radiologically isolated syndrome. Neurology 2009;72:800–5. [17] Lebrun C, Bensa C, Debouverie M, de Seze J, Wiertlievski S, Brochet B, et al. Unexpected multiple sclerosis: follow-up for 30 patients with magnetic resonance imaging and clinical conversion profile. J Neurol Neurosurg Psychiatry 2008;79:195–8. [18] Lebrun C, Bensa C, Debouverie M, Wiertlevski S, Brassat D, de Seze J, et al. Association between clinical conversion to multiple sclerosis in radiologically isolated syndrome and magnetic resonance imaging, cerebrospinal fluid, and visual evoked potential. Arch Neurol 2009;66:841–6. [19] Siva A, Saip S, Altintas A, Jacob A, Keegan BM, Kantarci OH. Multiple sclerosis risk in radiologically uncovered asymptomatic possible inflammatory-demyelinating disease. Mult Scler 2009;15:918–27. [20] Krupp LB, Banwell B, Tenembaum S. International Pediatric MS Study Group. Consensus definitions proposed for pediatric multiple sclerosis and related disorders.Neurology 2007;68:S7–S12 [Suppl.]. [21] de Seze J, Debouverie M, Zephir H, Lebrun C, Blanc F, Bourg V, et al. Acute fulminant demyelinating disease: a descriptive study of 60 patients. Arch Neurol 2007;64: 1426–32. [22] Guttmann CRG, Ahn SS, Hsu L, Kikinis R, Jolesz FA. The evolution of multiple sclerosis lesions on serial MR. Am J Neuroradiol 1995;16:1481–91. [23] He J, Grossman RI, Ge Y, Mannon LJ. Enhancing patterns in multiple sclerosis: evolution and persistence. AJNR Am J Neuroradiol 2001;22:664–9. [24] Jacobi C, Hähnel S, Martinez-Torres F, Rieger S, Jüttler E, Heiland S, et al. Prospective combined brain and spinal cord MRI in clinically isolated syndromes and possible early multiple sclerosis: impact on dissemination in space and time. Eur J Neurol 2008;15:1359–64. [25] Sombekke MH, Wattjes MP, Balk LJ, Nielsen JM, Vrenken H, Uitdehaag BM, et al. Spinal cord lesions in patients with clinically isolated syndrome: a powerful tool in diagnosis and prognosis. Neurology 2013;80:69–75. [26] Patrucco L, Rojas JI, Cristiano E. Assessing the value of spinal cord lesions in predicting development of multiple sclerosis in patients in patients with clinically isolated syndromes. J Neurol 2012;259:1317–20.
Contents lists available at SciVerse ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Comparison of the 2010 and 2005 versions of the McDonald MRI criteria for dissemination-in-time in Taiwanese patients with classic multiple sclerosis Chun-Jen Hsueh a, Hung-Wen Kao a, Shao-Yuan Chen b, c, Chung-Ping Lo d, e,⁎, Chia-Chun Hsu d, e, Dai-Wei Liu e, Wen-Lin Hsu e a
Department of Radiology, Tri-Service General Hospital and National Defense Medical Center, Taipei, Taiwan Department of Neurology and Hyperbaric Medicine, Cardinal Tien Hospital, New Taipei City, Taiwan School of Medicine, Fu Jen Catholic University, New Taipei City, Taiwan d Department of Radiology, Buddhist Tzu Chi General Hospital, Taichung Branch, Taichung, Taiwan e School of Medicine, Tzu Chi University, Hualien, Taiwan b c
a r t i c l e
i n f o
Article history: Received 3 February 2013 Received in revised form 10 March 2013 Accepted 25 March 2013 Available online 15 April 2013 Keywords: Multiple sclerosis Magnetic resonance imaging Brain Demyelinating disease Clinically isolated syndrome Diagnostic imaging
a b s t r a c t In 2010, the International Panel on the Diagnosis of Multiple Sclerosis revised the 2005 version of the McDonald criteria. The revisions to MRI dissemination-in-time criteria include adoption of a new criterion by demonstration of simultaneous asymptomatic gadolinium-enhancing and nonenhancing lesions on baseline MRI scans. The purpose of this study was to demonstrate the diagnostic validity of the modified MRI dissemination-in-time criteria. We collected 80 patients with an initial clinical attack suggestive of an acute central nervous system demyelinating disease. The patients were followed for at least two years or until the development of definite multiple sclerosis. The nonconverters were taken as negative cases. Their baseline and follow-up brain MRI studies were retrospectively reviewed by two neuroradiologists. The 2010 version had higher sensitivity (68.2% vs. 45.5%), slightly lower specificity (80.6% vs. 83.3%), and higher accuracy (73.8% vs. 62.5%) than the 2005 version, but the differences were without statistical significance. The new criteria are more sensitive and accurate and specific just as the old criteria. They allow the diagnosis of definite multiple sclerosis in 34.1% patients at first presentation of the clinically isolated syndrome. © 2013 Elsevier B.V. All rights reserved.
1. Introduction According to McDonald criteria, multiple sclerosis (MS) is diagnosed based on the objective demonstration of central nervous system (CNS) demyelinating lesions disseminated in space (DIS) and time (DIT) after exclusion of alternative diagnoses that might mimic the disease [1–3]. Magnetic resonance imaging (MRI) findings can substitute for clinical evidence of one lesion and one clinical attack if they fulfill the MRI DIS and DIT criteria, respectively [1–3]. In 2010, the International Panel on the Diagnosis of Multiple Sclerosis adopted a new MRI DIS criterion proposed by Swanton et al. (i.e., at least one T2-lesion in two or more of the following CNS regions: periventricular, juxtacortical, infratentorial, and spinal cord) [4,5]. Our previous study found that the specificity and sensitivity of this criterion were comparable to those of the MRI DIS criterion included in the 2005 version of the McDonald criteria when applied to Taiwanese patients with classic MS [3–6]. In addition, the panel revised the 2005 McDonald criteria by adopting a new criterion proposed by Rovira et al. [3,7]. The new MRI
DIT criteria emphasized the early diagnosis of MS based on a single baseline MRI demonstrating simultaneously the presence of asymptomatic gadolinium-enhancing and nonenhancing lesions indicative of demyelinating plaques in different stages of evolution [7,8]. The two sets of MRI DIT criteria (from 2005 to 2010) are summarized in Table 1. Clinically isolated syndrome (CIS) refers to a single clinical attack of CNS inflammatory demyelinating symptoms suggestive of MS. CIS presentations can be monofocal or multifocal, and typically involve the optic nerve, brainstem, cerebellum, spinal cord, or cerebral hemispheres. Although CIS may represent the first manifestation of definite multiple sclerosis (DMS), some patients (i.e., nonconverters) may not develop a second clinical relapse. The purpose of this study is to validate the diagnostic ability of the two sets of MRI DIT criteria in DMS patients and nonconverter patients. 2. Methods 2.1. Patient selection
⁎ Corresponding author at: Department of Radiology, Buddhist Tzu Chi General Hospital, Taichung Branch, No. 66, Sec. 1, Fongsing Road, Tanzih District, Taichung 427, Taiwan. Tel.: + 886 2 26479608. E-mail address: [email protected] (C.-P. Lo). 0022-510X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jns.2013.03.018
The study protocol was approved by the local institutional review board (TSGH IRB 1-101-05-004). From January 2001 to June 2010, we collected 80 patients with an initial clinical attack suggestive of an
52
C.-J. Hsueh et al. / Journal of the Neurological Sciences 329 (2013) 51–54
2.3. Statistical analysis
Table 1 The 2005 and 2010 revisions to the McDonald criteria for MRI DIT. 2005 revisions
2010 revisions
Detection of gadolinium enhancement at least 3 months after the onset of the initial clinical event, if not at the site corresponding to the initial event or detection of a new T2 lesion if it appears at any time compared with a reference scan done at least 30 days after the onset of the initial clinical event A new T2 and/or gadolinium-enhancing lesion(s) on follow-up MRI, with reference to a baseline scan, irrespective of the timing of the baseline MRI or simultaneous presence of asymptomatic gadolinium-enhancing and nonenhancing lesions at any time
3. Results
MRI = magnetic resonance imaging; DIT = dissemination-in-time.
Table 2 Comparison of MRI findings in DMS patients and nonconverters. New T2Simultaneous Gd+ and Gd− lesion on F-U MRI lesions on baseline MRI DMS (n = 44) 17 (38.6%) Nonconverter 1 (2.8%) (n = 36)
New Gd+ lesion on F-U MRI
Fulfill 2005 MRI
Patients who finally converted to DMS were considered positive cases and the nonconverters after at least two years of follow-up were considered negative cases. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of the two sets of MRI DIT criteria were calculated. The significance of differences was determined with 95% confidence intervals (95% CI), and the absence of overlapping was considered an indicator of difference.
Fulfill 2010 MRI DIT
DIT
20 (45.5%) 5 (11.4%) 20 (45.5%) 30 (68.2%) 6 (16.7%) 0 (0%) 6 (16.7%) 7 (19.4%)
MRI = magnetic resonance imaging; DMS = definite multiple sclerosis; Gd+ = gadolinium-enhancing; Gd− = gadolinium-nonenhancing; F-U = follow-up; DIT = dissemination-in-time.
acute CNS demyelinating disease (i.e., CIS). They had clinical follow-up of at least two years or until the development of DMS. Some of the subjects were in the cohort we studied previously [6,9]. All 80 patients met the following inclusion criteria: (1) age of 15 to 50 years; (2) baseline brain MRI obtained within three months of symptom onset; (3) follow-up brain MRI available; (4) MRI sequences containing gadolinium enhancement; and (5) no use of disease-modifying agents before the second clinical episodes (except for corticosteroid therapy). The exclusion criteria were: (1) clinical suspicion of acute disseminated encephalomyelitis (ADEM); and (2) symptoms and MRI abnormalities confined to optic nerves and spinal cord fulfilling the diagnostic criteria of neuromyelitis optica proposed by Wingerchuk et al. [10]. (3) other CNS white matter diseases were excluded through clinical work-ups.
2.2. MRI sequences and imaging analysis The MRI studies were performed on 1.5-T MR scanners. The brain MRI sequences included spin echo (SE) or fast spin echo (FSE) T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), and fluid-attenuated inversion recovery (FLAIR) in the axial plane, T2WI or T2-FLAIR in the sagittal plane, and contrast-enhanced T1WI in axial, coronal, and sagittal planes. Follow-up MRI scans were compared with the baseline scans. Spinal cord MRI studies were not included in this analysis because most patients received no or only one study and therefore DIT could not be demonstrated. Two neuroradiologists who were blind to the final diagnosis retrospectively reviewed the MRI scans and reached a consensus agreement on whether the two sets of MRI DIT criteria were fulfilled.
The recruited 80 patients consisted of 28 men and 52 women (mean age, 33.4 years; range, 15–50 years). Finally, 44 (55%) patients converted to DMS with a mean interval of 9.6 months after the onset of initial clinical events (range, 2–24 months) and 36 patients did not convert. The diagnosis of DMS in 44 patients was made by experienced neurologists and based on clinical grounds (two or more clinical attacks of CNS demyelinating events and objective clinical evidence of two or more lesions as defined in McDonald criteria). The mean clinical follow-up duration of the whole cohort was 66.4 months (range, 24–130 months). Clinical features at first presentation among the 80 patients included: unilateral optic neuritis (n = 24 [30.0%]), brainstem symptoms (n = 18 [22.5%]), multifocal symptoms (n = 23 [28.8%]), and spinal cord symptoms (n = 22 [27.5%]). All of the baseline brain MRI studies were performed within three months after initial symptom onset. Twenty-seven patients had more than one follow-up brain MRI scan (range, 1–4 scans). The mean interval between the baseline and follow-up MRI scans was 6.7 months (range, 3–28 months). Table 2 summarizes the MRI findings of the DMS patients and nonconverters. Among the 44 patients with DMS, 17 had simultaneous asymptomatic gadolinium-enhancing and nonenhancing lesions on baseline brain MRI; 20 had new T2-lesions, and five had new gadolinium-enhancing lesions on follow-up MRI scans. Some patients had more than one of these findings on MRI. Overall, 30 patients fulfilled the 2010 version of the McDonald criteria for MRI DIT and 20 patients fulfilled the 2005 version of these criteria. Among the 17 patients who had simultaneous asymptomatic gadolinium-enhancing and nonenhancing lesions on baseline brain MRI (with a mean interval of 15 days from symptom onset), 15 patients also fulfilled the MRI DIS criteria of 2010, which means that the diagnosis of DMS could be established in 34.1% (15/44) patients based on early clinical events and baseline brain MRI studies without the need for a second clinical attack or follow-up MRI according to the new criteria. The other 13 patients fulfilled the 2010 MRI DIT criteria based on follow-up MRI scans (with a mean interval of 168 days from symptom onset). Four of them already had a second clinical attack before follow-up MRI scans and thus had been diagnosed as DMS. Among the 36 nonconverters, one had simultaneous asymptomatic gadolinium-enhancing and nonenhancing lesions on baseline brain MRI; six had new T2-lesions and none had new gadoliniumenhancing lesions on follow-up scans (i.e., six nonconverters fulfilled the MRI DIT criteria of 2005 and seven fulfilled those of 2010).
Table 3 Performance of the 2005 and 2010 revisions to the McDonald criteria for MRI DIT.
2005 revisions 95% CI 2010 revisions 95% CI
TP
FP
TN
FN
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
Accuracy (%)
20
6
30
24
30
7
29
14
45.5 30.7–60.2 68.2 55.4–81.9
83.3 71.2–95.5 80.6 67.6–93.5
76.9 60.7–93.1 81.1 68.5–93.7
55.6 42.3–68.8 67.4 53.4–81.5
62.5 51.9–73.1 73.8 64.1–83.4
MRI = magnetic resonance imaging; DIT = dissemination-in-time; TP = true positive; FP = false positive; TN = true negative; FN = false negative; PPV = positive predictive value; NPV = negative predictive value; 95% CI = 95% confidence interval.
C.-J. Hsueh et al. / Journal of the Neurological Sciences 329 (2013) 51–54
53
The performance of the two sets of MRI DIT criteria is summarized in Table 3. The 2010 version had higher sensitivity (68.2% vs. 45.5%), slightly lower specificity (80.6% vs. 83.3%), and higher accuracy (73.8% vs. 62.5%) compared to the 2005 version. However, there was no significant statistical difference in the sensitivity, specificity, PPV, NPV, or accuracy among the two sets of criteria. 4. Discussion Our study shows that (compared to the 2005 version) the new version of the McDonald criteria is more accurate. The new version appears to increase sensitivity while maintaining the specificity attributable to the previous version. Recently a Spanish cohort study revealed a 52.63% sensitivity, a 75.00% specificity and a 59.26% accuracy for the 2010 MRI DIT criteria applying to patients with CIS [11]. However, their study was based on a single baseline brain MRI scan. If we exclude the follow-up brain MRI scans in our study, it would be a 38.6% sensitivity, a 97.2% specificity and a 65.0% accuracy, reflecting the relatively fewer enhancing brain lesions on baseline MRI in our study population. Six cases were false positives (without conversion to DMS) because new T2-lesions were detected on follow-up MRI. Positive MRI findings tend to occur more frequently than clinical findings in MS. New T2-lesions and gadolinium-enhancing lesions indicative of increasing axon pathology can be five to ten times more common than clinical signs of disease progression [12–14]. The involvement of non-eloquent areas of the brain or the masking effect of cortical adaptation may explain why MRI-detectable “subclinical” lesions may be present without clinically apparent disease [15]. Recent studies have also found individuals with incidentally-identified MRI abnormalities in the CNS highly suggestive of MS (fulfilling Barkhof–Tintoré criteria for DIS), but without clinical symptoms for demyelinating disease. The condition was defined as radiologically isolated syndrome (RIS), unexpected multiple sclerosis, or radiologically uncovered asymptomatic possible inflammatorydemyelinating disease (RAPIDD) [16–19]. About 22.7–36.7% patients with RIS have been reported to develop at least one clinical attack of CNS demyelinating symptoms during follow-up periods ranging from 2.4 to 12.5 years, and the initial MRI abnormalities may represent “preclinical” lesions in those patients [16–19]. In our study, the baseline MRI scans in two of our patients with DMS did not fulfill the MRI DIT criteria of 2010 because the only enhancing lesions were symptomatic and had to be excluded by definition. Enhancing lesions (particularly the nodular enhancing pattern) are usually encountered in the acute demyelinating stage and are more frequently symptomatic, while nonenhancing lesions usually indicate inactive or old plaques and are often asymptomatic. From an imaging perspective, the simultaneous presence of enhancing and nonenhancing lesions may suggest different stages of demyelination or dissemination-in-time (i.e., DIT) regardless of whether the lesions are symptomatic. Whether symptomatic enhancing lesions have to be excluded from the lesion count may need further discussion. The major concern of differential diagnosis is ADEM, which is typically a monophasic disease. ADEM often shows enhancement of all the lesions in the acute stage but a mixture of enhancing and nonenhancing lesions during recovery [8]. Patients' younger age, clinical history of prior infection or vaccination, presence of encephalopathy, frequent gray matter involvement on brain MRI, and less frequent oligoclonal bands in cerebrospinal fluid are characteristic features for ADEM [20,21]. The average duration of gadolinium enhancement of a demyelinating plaque is about one to two months [22,23], and temporal variations in the enhancement duration may potentially cause false positive and false negative results for DIT, depending on the timing of MRI examinations. For example, two plaques in the same stage but having different enhancement durations may lead to false positive results for DIT on baseline MRI (Fig. 1), whereas two plaques in different stages but
Fig. 1. Two plaques of the same stage but having different durations of contrast enhancement. If MRI study is performed at the time point marked by the dotted line, it may lead to a false positive result (simultaneous presence of enhancing and nonenhancing lesions).
having overlapping enhancement or nonenhancement periods may lead to false negative results (Fig. 2). The use of corticosteroid may also affect lesion enhancement. Consequently, follow-up MRI may still have an important role in the determination of DIT. The limitations of this study are as follows: (1) The minimum clinical follow-up duration was set at two years, and some patients may develop later conversion to DMS. (2) Spinal cord MRI studies were only obtained in patients with symptoms of spinal cord involvement, and most of them were performed once. As a result, the numbers of patients fulfilling McDonald MRI criteria might be underestimated. A previous study has shown limited impact of spinal cord MRI on the evidence of DIT [24]. However, other two studies have shown that the presence of spinal cord lesion was associated with a higher risk and shorter time to conversion to CDMS [25,26]. (3) The case number is relatively small and it may affect the statistical significance. (4) Because the study is a retrospective one, the time points for clinical evaluation and MRI follow-up were without a fixed schedule. The results may be influenced by this heterogeneity. In conclusion, our study shows that the MRI DIT criteria of 2010 increase sensitivity and accuracy without significantly compromising the specificity of the MRI DIT criteria of 2005. The new criteria allowed the diagnosis of DMS to be made in 34.1% patients at the CIS stage. It may help to select candidates for disease-modifying treatments from patients with CIS to delay the time to a second demyelinating event and to prevent future disability. We suggest that patients
Fig. 2. Two plaques of different stages and having overlapping durations of contrast enhancement and nonenhancement. The MRI findings may lead to false negative results if performed at the time points marked by the dotted lines.
54
C.-J. Hsueh et al. / Journal of the Neurological Sciences 329 (2013) 51–54
with clinical features of CIS should receive contrast-enhanced MRI for a possible earlier diagnosis of DMS, and follow-up MRI is still important in the demonstration of DIT. The current study of MRI DIT criteria along with our previous study of MRI DIS criteria [6] suggests that the 2010 version of McDonald MRI criteria is simple to use and as accurate as the 2005 version with possible early diagnosis of MS at CIS stage. Funding None. Conflict of interest None. Acknowledgments None. References [1] McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on diagnosis of multiple sclerosis. Ann Neurol 2001;50:121–7. [2] Polman CH, Reingold SC, Edan G, Fillipi M, Hartung HP, Kappos L, et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald Criteria”. Ann Neurol 2005;58:840–6. [3] Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Fillipi M, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 2011;69:292–302. [4] Swanton JK, Fernando K, Dalton CM, Miszkiel KA, Thompson AJ, Plant GT, et al. Modification of MRI criteria for multiple sclerosis in patients with clinically isolated syndromes. J Neurol Neurosurg Psychiatry 2006;77:830–3. [5] Swanton JK, Rovira A, Tintoré M, Altmann DR, Barkhof F, Fillipi M, et al. MRI criteria for multiple sclerosis in patients presenting with clinically isolated syndromes: a multicentre retrospective study. Lancet Neurol 2007;6:677–86. [6] Lo CP, Kao HW, Chen SY, Hsueh CJ, Lin WC, Hsu WL, et al. Prediction of conversion from clinically isolated syndrome to clinically definite multiple sclerosis according to baseline MRI findings: comparison of revised McDonald criteria and Swanton modified criteria. J Neurol Neurosurg Psychiatry 2009;80:1107–9. [7] Rovira A, Swanton J, Tintoré M, Huerga E, Barkhof F, Fillipi M, et al. A single, early magnetic resonance imaging study in the diagnosis of multiple sclerosis. Arch Neurol 2009;66:587–92. [8] Montalban X, Tintoré M, Swanton J, Barkhof F, Fazekas F, Fillipi M, et al. MRI criteria for MS in patients with clinically isolated syndromes. Neurology 2010;74:427–34.
[9] Chen SY, Lo CP, Hsu WL, Kao HW, Hsueh CJ, Lin WC, et al. Modifications to the McDonald MRI dissemination in space criteria for use in Asians with classic multiple sclerosis: the Taiwanese experience. Mult Scler 2010;16:1213–9. [10] Wingerchuk DM, Lennon VA, Pittock SJ, Lucchinetti CF, Scottsdale AZ. Revised diagnostic criteria for neuromyelitis optica. Neurology 2006;66:1485–9. [11] Gómez-Moreno M, Díaz-Sánchez M, Ramos-González. Application of the 2010. McDonald criteria for the diagnosis of multiple sclerosis in a Spanish cohort of patients with clinically isolated syndromes. Mult Scler 2012;18:39–44. [12] Thrower BW. Clinically isolated syndromes: predicting and delaying multiple sclerosis. Neurology 2007;68(Suppl. 4):S12–5. [13] Miller DH, Albert PS, Barkhof F, Francis G, Frank JA, Hodgkinson S, et al. Guidelines for the use of magnetic resonance techniques in monitoring the treatment of multiple sclerosis. Ann Neurol 1996;39:6–16. [14] Barkhof F, Scheltens P, Frequin ST, Nauta JJ, Tas MW, Valk J, et al. Relapsing–remitting multiple sclerosis in determining disease activity. AJR Am J Roentgenol 1992;158:1041–7. [15] Barkhof F. The clinico-radiological paradox in multiple sclerosis revisited. Curr Opin Neurol 2002;15:239–45. [16] Okuda DT, Mowry EM, Beheshtian A, Waubant E, Baranzini SE, Goodin DS, et al. Incidental MRI anomalies suggestive of multiple sclerosis: the radiologically isolated syndrome. Neurology 2009;72:800–5. [17] Lebrun C, Bensa C, Debouverie M, de Seze J, Wiertlievski S, Brochet B, et al. Unexpected multiple sclerosis: follow-up for 30 patients with magnetic resonance imaging and clinical conversion profile. J Neurol Neurosurg Psychiatry 2008;79:195–8. [18] Lebrun C, Bensa C, Debouverie M, Wiertlevski S, Brassat D, de Seze J, et al. Association between clinical conversion to multiple sclerosis in radiologically isolated syndrome and magnetic resonance imaging, cerebrospinal fluid, and visual evoked potential. Arch Neurol 2009;66:841–6. [19] Siva A, Saip S, Altintas A, Jacob A, Keegan BM, Kantarci OH. Multiple sclerosis risk in radiologically uncovered asymptomatic possible inflammatory-demyelinating disease. Mult Scler 2009;15:918–27. [20] Krupp LB, Banwell B, Tenembaum S. International Pediatric MS Study Group. Consensus definitions proposed for pediatric multiple sclerosis and related disorders.Neurology 2007;68:S7–S12 [Suppl.]. [21] de Seze J, Debouverie M, Zephir H, Lebrun C, Blanc F, Bourg V, et al. Acute fulminant demyelinating disease: a descriptive study of 60 patients. Arch Neurol 2007;64: 1426–32. [22] Guttmann CRG, Ahn SS, Hsu L, Kikinis R, Jolesz FA. The evolution of multiple sclerosis lesions on serial MR. Am J Neuroradiol 1995;16:1481–91. [23] He J, Grossman RI, Ge Y, Mannon LJ. Enhancing patterns in multiple sclerosis: evolution and persistence. AJNR Am J Neuroradiol 2001;22:664–9. [24] Jacobi C, Hähnel S, Martinez-Torres F, Rieger S, Jüttler E, Heiland S, et al. Prospective combined brain and spinal cord MRI in clinically isolated syndromes and possible early multiple sclerosis: impact on dissemination in space and time. Eur J Neurol 2008;15:1359–64. [25] Sombekke MH, Wattjes MP, Balk LJ, Nielsen JM, Vrenken H, Uitdehaag BM, et al. Spinal cord lesions in patients with clinically isolated syndrome: a powerful tool in diagnosis and prognosis. Neurology 2013;80:69–75. [26] Patrucco L, Rojas JI, Cristiano E. Assessing the value of spinal cord lesions in predicting development of multiple sclerosis in patients in patients with clinically isolated syndromes. J Neurol 2012;259:1317–20.