- Email: [email protected]
Clinical, MRI, CSF and electrophysiological findings in different stages of multiple sclerosis
Clinical Neurology and Neurosurgery 108 (2006) 271–274
Clinical, MRI, CSF and electrophysiological findings in different stages of multiple sclerosis Uroˇs Rot ∗ , Anton Mesec Department of Neurology, Medical Centre, Zaloˇska 7, 1525 Ljubljana, Slovenia
Abstract Effective therapy in the earliest stages of multiple sclerosis (MS) demands early correct diagnosis. Retrospective analysis included 130 patients (90 women) with a median age of 35.5 years, median duration of the disease of 2 years and median EDSS score of 3.0. Twenty-seven patients had clinically isolated syndrome (CIS) suggestive of MS, 66 relapsing-remitting (RR) MS, 19 secondary progressive (SP) MS and 18 primary progressive (PP) MS. The predominant symptoms were sensory in 52% of the patients with CIS compared to 27% in patients with RRMS, whereas they were more often motor in patients with PPMS. Patients with CIS had higher CSF cell counts than patients diagnosed in later stages of the disease and oligoclonal bands were found in 89% of all patients without statistically significant differences between the subgroups. Prolonged latencies of visual evoked potentials (VEP) were found in only 29% of patients with CIS compared to 66% in RRMS, 75% in SPMS and 65% of PPMS patients. Fifty-six percent of patients with CIS, 88% with RRMS, 74% with SPMS and 78% of patients with PPMS fulfilled modified the Barkhof et al. MRI criteria at the time of diagnosis. Patients in early MS often present with sensory symptoms. Brain MRI can be inconclusive in over 40% of patients with CIS but the elevated CSF cell count and positive oligoclonal bands are helpful in establishing the diagnosis of CIS suggestive of MS. In later stages of the disease the combination of clinical features, MRI, prolonged VEP latencies and positive CSF oligoclonal bands secures the correct diagnosis. © 2005 Elsevier B.V. All rights reserved. Keywords: Multiple sclerosis; Clinically isolated syndrome; Clinical features; Cerebrospinal fluid; MRI; Evoked potentials
1. Introduction In order to initiate effective therapy in the early stages of MS the correct diagnosis must be established as soon as possible. We reviewed the differences in the clinical, MRI, CSF and electrophysiological characteristics of 130 patients with clinically isolated syndrome (CIS) and MS, with special emphasis on relapsing-remitting (RR) MS patients, in order to facilitate the diagnostic process in the early stages of the disease.
2. Patients and methods A neurologist with a special interest in MS (UR) analyzed the hospital records of patients admitted to the Department of Neurology, Medical Centre, Ljubljana between January ∗
Corresponding author. Tel.: +386 1 522 43 20. E-mail address: [email protected] (U. Rot).
0303-8467/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.clineuro.2005.11.021
1, 2000 and December 31, 2003 and discharged with the diagnoses MS or CIS. All patients who fulfilled McDonald criteria for MS were included in the study [1]. Patients with CIS suggestive of MS had either typical brain/spinal cord MRI lesions or positive oligoclonal bands in CSF [2]. Tests for serum electrolytes, urea, creatinine, full blood count, erythrocyte sedimentation rate, proteinogram, anticardiolipin antibodies, antinuclear antibodies, serum and CSF Borrelia and lues, urinanalysis and chest X-ray were performed to exclude other diseases. Vitamin B12 and folic acid were also measured in patients with progressive MS. One-hundred and thirty patients, 90 women and 40 men were included in the study. Twenty-seven patients had CIS, 66 RRMS, 19 secondary-progressive (SP) MS and 18 primaryprogressive (PP) MS. The median age of patients was 35.5 years (range 18–68 years) and the median duration of the disease was 2 years (range 0–30 years). Nineteen of the CIS patients had spinal cord syndromes, five brainstem syndromes, one optic neuritis, one a cerebellar syndrome and
272
U. Rot, A. Mesec / Clinical Neurology and Neurosurgery 108 (2006) 271–274
one a hemispheric syndrome. Fifteen PPMS patients had a progressive myelopathy, two a progressive cerebellar syndrome and one had progressive cognitive deterioration. The predominant neurological symptom, number of symptoms and disability using EDSS score were determined in every patient [3]. CSF white cells were counted in Fuchs-Rosenthal chambers (normal <3 cells/mm3 ). The total CSF protein and IgG concentrations were measured by automatic immunoprecipitation nephelometry (DOSACAT nephelometer, DOSATEC, Munich, Germany). CSF IgG level was expressed as a percentage of total protein concentration (normal <11%). The total CSF protein concentration between 15 and 45 mg% was considered to be the normal value. Isoelectric focusing on polyacrylamide gels of concentrated CSF and undiluted serum was performed for the detection of oligoclonal bands (OB). Four or more OB in alkaline region were deemed positive. MR imaging was performed on 1.5 T imager (Siemens Magnetom 63 SP, Munich, Germany) using conventional T1, T2 and proton density-weighted sequences with or without gadolinium injection. Lesions were considered MS-like if they were of high intensity on T2 or proton density scans, located in the white matter and greater than 3 mm in diameter. The following semi-quantitative classification scheme was used for analysis: normal MRI, 1 or 2 MS-like lesions, >3 MS-like lesions without Barkhof criteria, and fulfilled modified Barkhof criteria [4,5]. Three of four of the following are required to fulfill the modified Barkhof MRI criteria: (1) one gadolinium enhancing lesion or 9 T2 hyperintense lesions; (2) at least one infratentorial lesion; (3) at least one juxtacortical lesion; (4) at least three periventricular lesions. Visual (VEP) and somatosensory (SEP) evoked potentials were measured by standard methods [6,7]. Statistical analysis was performed using the statistical program GraphPad Prism Version 3.03 (GraphPad Software, San Diego, CA). Mann–Whitney U-test and Fisher exact test were used as appropriate for comparison of clinical and paraclinical findings in patients with CIS and RRMS. Spearman correlation coefficient was used for correlation analyses.
3. Results 3.1. Clinical features The median duration of disease at diagnosis was 0 years (range 0–1) in patients with CIS, 2.5 years (range 0–21) in RRMS, 10 years (range 1–30) in SPMS and 2 years (range 0.5–25) in PPMS. The median EDSS score was 3 (range 0–6.5), 2.5 (range 0–3) in CIS, 2.75 (range 0–6) in RRMS, 4 (range 2.5–6) in SPMS and 3.5 (range 2–6.5) in PPMS. The various clinical symptoms in the subgroups of patients are shown in Table 1.
4. CSF findings The median CSF leukocyte number in all the patients included in the study was 5 mm−3 (range 0–60), protein concentration 38 mg% (range 15–110) and IgG level 12.83% (range 5.03–34.47). Thirty-seven percent of all patients had normal CSF cell counts, 72% normal protein concentrations and 36% normal IgG levels. OB were present in 103/116 (89%) patients. The median CSF leukocyte number in CIS was 11 mm−3 (range 1–30) compared to 5 mm−3 (range 0–60) in RRMS, 2 mm−3 (range 1–18) in SPMS, and 5 mm−3 (range 1–34) in PPMS. Eighty-one percent of CIS patients, 67% of RRMS, 42% of SPMS and 56% of PPMS patients had elevated CSF cell counts. The median protein concentration was 34 mg% (range 15–67) in CIS, 38 mg% (range 19–66) in RRMS, 43 mg% (range 21–110) in SPMS and 40 mg% (range 31–69) in PPMS. The median IgG level was 14.92% (range 7.14–27.75) in CIS, 12.46% (range 6.37–30.69) in RRMS, 10.12% (range 5.03–33.33) in SPMS and 14.29% (range 9.02–34.47) in patients with PPMS. Seventy-nine percent of patients with CIS, 63% with RRMS, 44% of SPMS and 71% of PPMS had elevated IgG levels. OB were found in 20/24 (83%) of CIS, 56/60 (93%) of RRMS, 11/15 (73%) of SPMS and 16/17 (94%) of patients with PPMS.
Table 1 Clinical features of patients at different stages of multiple sclerosis Symptoms
Motor Sensory Brainstem Cerebellar Visual Sphincter Cognitive
Clinically isolated syndrome (n = 27)
Relapsing-remitting MS (n = 66)
Secondary-progressive MS (n = 19)
Primary-progressive MS (n = 18)
Predominant
Present
Predominant
Present
Predominant
Present
Predominant
Present
6/27 (22) 14/27 (52) 5/27 (19) 0/27 2/27 (7) 0/27 0/27
13/27 (48) 18/27 (67) 5/27 (19) 1/27 (4) 4/27 (15) 4/27 (15) 0/27
22/66 (33) 18/66 (27) 10/66 (15) 7/66 (11) 9/66 (14) 0/66 0/66
41/66 (62) 36/66 (55) 15/66 (23) 19/66 (29) 16/66 (24) 3/66 (5) 0/66
12/19 (63) 1/19 (5) 0/19 5/19 (26) 0/19 0/19 1/19 (5)
18/19 (95) 17/19 (89) 3/19 (16) 8/19 (42) 3/19 (16) 6/19 (33) 8/19 (42)
13/18 (72) 1/18 (6) 1/18 (6) 2/18 (11) 0/18 0/18 1/18 (6)
17/18 (94) 11/18 (61) 1/18 (6) 5/18 (28) 2/18 (11) 6/18 (33) 2/18 (11)
Percentages are shown in brackets.
U. Rot, A. Mesec / Clinical Neurology and Neurosurgery 108 (2006) 271–274
273
Table 2 MRI findings in subgroups of patients
Brain MRI MS-like lesions Normal MRI 1–2 lesions >3 (no Barkhof criteria) Fulfilled modified Barkhof criteria Gd enhancement Spinal cord MRI
Clinically isolated syndrome (n = 27)
Relapsing-remitting MS (n = 66)
Secondary-progressive MS (n = 19)
Primary-progressive MS (n = 18)
20/25 (80) 5/25 (20) 4/25 (16) 2/25 (8) 14/25 (56) 8/18 (44) 8/11 (73)
65/65 (100) 0/65 2/65 (3) 6/65 (9) 57/65 (88) 21/40 (53) 9/12 (75)
17/19 (89) 2/19 (11) 2/19 (11) 1/19 (5) 14/19 (74) 3/15 (20) 3/4 (75)
18/18 (100) 0/18 2/18 (11) 2/18 (11) 14/18 (78) 4/9 (44) 7/9 (78)
Percentages are shown in brackets.
5. Correlation between clinical and CSF findings There was a positive correlation between CSF leukocyte numbers and IgG level (r = 0.5201; p < 0.0001), but negative correlations between CSF cell numbers and patients age, r = −0.3371; p = 0.0001, CSF cell number and duration of the disease (r = −0.3322; p = 0.0002) and CSF cell number and EDSS score (r = −0.2854; p = 0.0014). A modest negative correlation also existed between IgG level and duration of disease (r = −0.2215; p = 0.03), but none between IgG level and EDSS score (p = 0.44) and IgG level and patients age (p = 0.44).
patients. MRI results in the subgroups of patients are shown in Table 2.
7. Evoked potentials Prolonged latencies on VEP were found in 48/81 (59%) of our patients, 4/14 (29%) of patients with CIS, 29/44 (66%) of RRMS, 6/8 (75%) of SPMS and 9/14 (65%) of PPMS. Prolonged latencies of SEP were found in 28/52 (54%) of all patients, 7/14 (50%) of CIS, 13/22 (59%) of RRMS, 3/8 (38%) of SPMS, and 5/8 (63%) of PPMS.
8. Clinical and paraclinical findings in CIS and RRMS
6. MRI results MS-like lesions on brain MRI were seen in 121/127 (95%) patients. Among them nine (7%) had one or two lesions, 11 (9%) had >3 but did not fulfill the Barkhof et al. criteria but 101 (80%) did. Gadolinium enhancement occurred in 36/82 (44%) patients. Spinal cord lesions existed in 27/36 (75%)
Patients with CIS had more sensory symptoms (p = 0.03). They were younger (p = 0.02) and had shorter duration of disease at the time of diagnosis (p < 0.0001). Their CSF cell count was higher (p = 0.04), VEP latencies less often pro-
Table 3 Comparison of clinical and paraclinical characteristics in patients with CIS and RRMS
Sex ratio (F:M) Median age (years) Duration of the disease (years) Predominant sensory symptoms Median EDSS score Median CSF protein (g/l) Frequency of abnormal CSF protein (%) Median CSF cells/mm3 Frequency of abnormal CSF cells (%) Median IgG level (%) Frequency of abnormal IgG (%) Oligoclonal bands Frequency of abnormal VEP (%) Frequency of abnormal SEP (%) Brain MRI MS-like lesions Normal brain MRI 1–2 lesions ≥3 (no-Barkhof) Barkhof Gadolinium enhancement Spinal cord MRI
CIS (n = 27)
RRMS (n = 66)
p-value
16:11 29 (range 21–48) 0 (range 0–1) 14/27 (52%) 2.5 (range 0–3) 0.34 (range 0.15–0.67) 7/26 (27%) 11 (range 1–30) 21/26 (81%) 1492 (range 714–2775) 15/19 (79%) 20/24 (83%) 4/14 (29%) 7/14 (50%) 20/25 (80%) 5/25 (20%) 4/25 (16%) 2/25 (8%) 14/25 (56%) 8/18 (44%) 8/11 (73%)
50:16 34 (range 16–59) 2.5 (range 0–21) 18/66 (27%) 2.75 (range 0–6) 0.38 (range 0.19–0.66) 18/63 (29%) 5 (range 0–60) 42/63 (67%) 1246 (range 6.37–30.69) 33/52 (63%) 56/60 (93%) 29/44 (66%) 13/22 (59%) 65/65 (100%) 0/65 2/65 (5%) 6/65 (9%) 57/65 (88%) 21/40 (53%) 9/12 (75%)
0.11 0.02 <0.0001 0.03 0.07 0.53 0.88 0.04 0.21 0.15 0.26 0.22 0.03 0.73 0.0012 0.0012 0.05 1 0.0027 0.78 1
274
U. Rot, A. Mesec / Clinical Neurology and Neurosurgery 108 (2006) 271–274
longed (p = 0.03) and they were less likely to fulfill the MRI criteria (p = 0.02). Comparisons of clinical and paraclinical findings in CIS and RRMS are shown in Table 3.
9. Discussion We retrospectively analyzed the clinical and paraclinical characteristics of 130 patients with either MS or CIS. The most common symptoms in patients with CIS were sensory (52%), followed by motor (22%) and brainstem (19%) symptoms, whereas motor and sensory symptoms were more equally distributed in RRMS. Motor symptoms predominated in PPMS and were found in over 90% of patients. For comparison, in a Danish study 24% of patients started with motor symptoms, 32% with sensory and 20% with visual symptoms [8]. Bauer found that 43% of patients had motor, 41% sensory and 36% visual symptoms at the onset of the disease [9]. Only 7% of our CIS patients had visual symptoms reflecting the fact that optic neuritis patients are often referred to ophthalmologists rather than to neurologists. CSF OB were present in over 80% of patients with CIS and in more than 90% of all patients with MS. Sastre-Garriga et al. reported CSF OB in 71% of patients with isolated brainstem syndrome [10], whereas they were seen in 72% of optic neuritis patients in a Swedish study [2]. Interestingly, patients with CIS had higher CSF cell counts than RRMS and patients in more advanced stages of the disease. We also found a moderate negative correlation between CSF cells and patient age, duration of disease and EDSS. These results confirm our previous observation that patients with early MS have elevated CSF cell count [11], additional helpful information for the diagnosis of MS. Walker et al. also found a negative correlation between CSF cell number and duration of disease in their series of 120 MS patients[12]. On the other hand S¨oderstr¨om found no differences in CSF findings between CIS and MS [13]. Over 95% of our patients had MS-like lesions on brain MRI, confirming its high sensitivity for the diagnosis of MS. The modified Barkhof MRI criteria for dissemination in space provide an acceptable degree of sensitivity, specificity and accuracy and are included in current diagnostic criteria [1]. Fifty-six percent of our CIS and approximately 80% of our MS patients fulfilled them; as did 50% of Spanish patients with isolated brainstem syndrome [10]. Our results, and those of others, suggest that additional investigations should be performed in approximately half of CIS patients. Spinal cord MRI was performed in only one third of our patients and showed typical lesions in approximately 75%, underlining its importance in the diagnosis of MS [14]. The vast majority of our CIS patients had isolated spinal cord syndromes, thus accounting for the high sensitivity of spinal cord MRI in our study, paralleling Bot et al.’s report that 83% of 104 patients with recently diagnosed MS had spinal cord lesions [15].
The McDonald et al. criteria include only VEP [1]. They were obtained in 60% of our patients, but in addition, 40% also had SEP. Prolonged VEP latencies were found in only 30% of patients with CIS compared to over 60% of MS patients. SEP were more often prolonged in patients with CIS, again because most had isolated spinal cord syndrome. In Sastre-Garriga et al.’s Spanish study VEP latencies were prolonged in approximately 30% of patients and SEP in 35%, suggesting that evoked potential studies are not very helpful in confirming clinically silent lesions in CIS patients [10]. In conclusion, patients in the earliest stages of MS often present with sensory symptoms. Brain MRI can be inconclusive in over 40% of patients with CIS but the CSF profile of elevated cell counts, IgG ratio and positive OB bands is helpful in establishing the diagnosis. In later stages of the disease, the combination of clinical features, MRI, prolonged VEP latencies and positive CSF OB lead to the correct diagnosis.
References [1] McDonald W, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis. Ann Neurol 2001;50:121–7. [2] S¨oderstr¨om M, Ya-Ping J, Hillert J, et al. Optic neuritis. Prognosis for multiple sclerosis from MRI, CSF and HLA findings. Neurology 1998;50:708–14. [3] Kurtzke J. Rating neurological impairment in multiple sclerosis: an expanded disability rating scale. Neurology 1988;38:180–5. [4] Barkhof F, Filippi M, Miller D, et al. Comparison of MRI criteria at first presentation to predict conversion to clinically definite multiple sclerosis. Brain 1997;120:2059–69. [5] Tintor´e M, Rovira A, Martinez M, et al. Isolated demyelinating syndromes: comparison of different MR imaging criteria to predict conversion to clinically definite multiple sclerosis. AJNR 2000;21:702–6. [6] Brecelj J. Visual evoked potentials and electrophysiological evaluation of the visual pathway. Med Razgl 1994;33:339–59. [7] Deniˇsliˇc M, Prevec T. Analysis of somatosensory evoked potentials in multiple sclerosis. Zdrav Vestn 1987;56:7–10. [8] Bronnum-Hansen H, Koch-Henriksen N, Hyllested K. Survival of patients with multiple sclerosis in Denmark: a nationwide long-term epidemiologic survey. Neurology 1994;44:1901–7. [9] Bauer H. Problems of symptomatic therapy in multiple sclerosis. Neurology 1978;28(Suppl. 2):8–20. [10] Sastre-Garriga J, Tintor´e M, Rovira A, et al. Conversion to multiple sclerosis after a clinically isolated syndrome of the brainstem: cranial magnetic resonance imaging, cerebrospinal fluid and neurophysiological findings. Multiple Sclerosis 2003;9:39–43. [11] Rot U, Mesec A, Pogaˇcnik T. Multiple sclerosis with uncommon cerebrospinal fluid findings. Croat Med J 2003;44:697–701. [12] Walker R, Thompson E, McDonald W. Cerebrospinal fluid in multiple sclerosis: relationships between immunoglobulins, leukocytes and clinical features. J Neurol 1985;232:250–9. [13] S¨oderstr¨om M. Clues to the immunopathogenesis of multiple sclerosis by investigating untreated patients during the very early stage of disease. Neurol Sci 2001;22:145–9. [14] Lycklama G, Thompson A, Filippi M, et al. Spinal-cord MRI in multiple sclerosis. Lancet Neurol 2003;2:555–62. [15] Bot J, Barkhof F, Polman CH, et al. Spinal cord abnormalities in recently diagnosed MS patients: added value of spinal MRI examinations. Neurology 2004;62:226–33.
Clinical, MRI, CSF and electrophysiological findings in different stages of multiple sclerosis Uroˇs Rot ∗ , Anton Mesec Department of Neurology, Medical Centre, Zaloˇska 7, 1525 Ljubljana, Slovenia
Abstract Effective therapy in the earliest stages of multiple sclerosis (MS) demands early correct diagnosis. Retrospective analysis included 130 patients (90 women) with a median age of 35.5 years, median duration of the disease of 2 years and median EDSS score of 3.0. Twenty-seven patients had clinically isolated syndrome (CIS) suggestive of MS, 66 relapsing-remitting (RR) MS, 19 secondary progressive (SP) MS and 18 primary progressive (PP) MS. The predominant symptoms were sensory in 52% of the patients with CIS compared to 27% in patients with RRMS, whereas they were more often motor in patients with PPMS. Patients with CIS had higher CSF cell counts than patients diagnosed in later stages of the disease and oligoclonal bands were found in 89% of all patients without statistically significant differences between the subgroups. Prolonged latencies of visual evoked potentials (VEP) were found in only 29% of patients with CIS compared to 66% in RRMS, 75% in SPMS and 65% of PPMS patients. Fifty-six percent of patients with CIS, 88% with RRMS, 74% with SPMS and 78% of patients with PPMS fulfilled modified the Barkhof et al. MRI criteria at the time of diagnosis. Patients in early MS often present with sensory symptoms. Brain MRI can be inconclusive in over 40% of patients with CIS but the elevated CSF cell count and positive oligoclonal bands are helpful in establishing the diagnosis of CIS suggestive of MS. In later stages of the disease the combination of clinical features, MRI, prolonged VEP latencies and positive CSF oligoclonal bands secures the correct diagnosis. © 2005 Elsevier B.V. All rights reserved. Keywords: Multiple sclerosis; Clinically isolated syndrome; Clinical features; Cerebrospinal fluid; MRI; Evoked potentials
1. Introduction In order to initiate effective therapy in the early stages of MS the correct diagnosis must be established as soon as possible. We reviewed the differences in the clinical, MRI, CSF and electrophysiological characteristics of 130 patients with clinically isolated syndrome (CIS) and MS, with special emphasis on relapsing-remitting (RR) MS patients, in order to facilitate the diagnostic process in the early stages of the disease.
2. Patients and methods A neurologist with a special interest in MS (UR) analyzed the hospital records of patients admitted to the Department of Neurology, Medical Centre, Ljubljana between January ∗
Corresponding author. Tel.: +386 1 522 43 20. E-mail address: [email protected] (U. Rot).
0303-8467/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.clineuro.2005.11.021
1, 2000 and December 31, 2003 and discharged with the diagnoses MS or CIS. All patients who fulfilled McDonald criteria for MS were included in the study [1]. Patients with CIS suggestive of MS had either typical brain/spinal cord MRI lesions or positive oligoclonal bands in CSF [2]. Tests for serum electrolytes, urea, creatinine, full blood count, erythrocyte sedimentation rate, proteinogram, anticardiolipin antibodies, antinuclear antibodies, serum and CSF Borrelia and lues, urinanalysis and chest X-ray were performed to exclude other diseases. Vitamin B12 and folic acid were also measured in patients with progressive MS. One-hundred and thirty patients, 90 women and 40 men were included in the study. Twenty-seven patients had CIS, 66 RRMS, 19 secondary-progressive (SP) MS and 18 primaryprogressive (PP) MS. The median age of patients was 35.5 years (range 18–68 years) and the median duration of the disease was 2 years (range 0–30 years). Nineteen of the CIS patients had spinal cord syndromes, five brainstem syndromes, one optic neuritis, one a cerebellar syndrome and
272
U. Rot, A. Mesec / Clinical Neurology and Neurosurgery 108 (2006) 271–274
one a hemispheric syndrome. Fifteen PPMS patients had a progressive myelopathy, two a progressive cerebellar syndrome and one had progressive cognitive deterioration. The predominant neurological symptom, number of symptoms and disability using EDSS score were determined in every patient [3]. CSF white cells were counted in Fuchs-Rosenthal chambers (normal <3 cells/mm3 ). The total CSF protein and IgG concentrations were measured by automatic immunoprecipitation nephelometry (DOSACAT nephelometer, DOSATEC, Munich, Germany). CSF IgG level was expressed as a percentage of total protein concentration (normal <11%). The total CSF protein concentration between 15 and 45 mg% was considered to be the normal value. Isoelectric focusing on polyacrylamide gels of concentrated CSF and undiluted serum was performed for the detection of oligoclonal bands (OB). Four or more OB in alkaline region were deemed positive. MR imaging was performed on 1.5 T imager (Siemens Magnetom 63 SP, Munich, Germany) using conventional T1, T2 and proton density-weighted sequences with or without gadolinium injection. Lesions were considered MS-like if they were of high intensity on T2 or proton density scans, located in the white matter and greater than 3 mm in diameter. The following semi-quantitative classification scheme was used for analysis: normal MRI, 1 or 2 MS-like lesions, >3 MS-like lesions without Barkhof criteria, and fulfilled modified Barkhof criteria [4,5]. Three of four of the following are required to fulfill the modified Barkhof MRI criteria: (1) one gadolinium enhancing lesion or 9 T2 hyperintense lesions; (2) at least one infratentorial lesion; (3) at least one juxtacortical lesion; (4) at least three periventricular lesions. Visual (VEP) and somatosensory (SEP) evoked potentials were measured by standard methods [6,7]. Statistical analysis was performed using the statistical program GraphPad Prism Version 3.03 (GraphPad Software, San Diego, CA). Mann–Whitney U-test and Fisher exact test were used as appropriate for comparison of clinical and paraclinical findings in patients with CIS and RRMS. Spearman correlation coefficient was used for correlation analyses.
3. Results 3.1. Clinical features The median duration of disease at diagnosis was 0 years (range 0–1) in patients with CIS, 2.5 years (range 0–21) in RRMS, 10 years (range 1–30) in SPMS and 2 years (range 0.5–25) in PPMS. The median EDSS score was 3 (range 0–6.5), 2.5 (range 0–3) in CIS, 2.75 (range 0–6) in RRMS, 4 (range 2.5–6) in SPMS and 3.5 (range 2–6.5) in PPMS. The various clinical symptoms in the subgroups of patients are shown in Table 1.
4. CSF findings The median CSF leukocyte number in all the patients included in the study was 5 mm−3 (range 0–60), protein concentration 38 mg% (range 15–110) and IgG level 12.83% (range 5.03–34.47). Thirty-seven percent of all patients had normal CSF cell counts, 72% normal protein concentrations and 36% normal IgG levels. OB were present in 103/116 (89%) patients. The median CSF leukocyte number in CIS was 11 mm−3 (range 1–30) compared to 5 mm−3 (range 0–60) in RRMS, 2 mm−3 (range 1–18) in SPMS, and 5 mm−3 (range 1–34) in PPMS. Eighty-one percent of CIS patients, 67% of RRMS, 42% of SPMS and 56% of PPMS patients had elevated CSF cell counts. The median protein concentration was 34 mg% (range 15–67) in CIS, 38 mg% (range 19–66) in RRMS, 43 mg% (range 21–110) in SPMS and 40 mg% (range 31–69) in PPMS. The median IgG level was 14.92% (range 7.14–27.75) in CIS, 12.46% (range 6.37–30.69) in RRMS, 10.12% (range 5.03–33.33) in SPMS and 14.29% (range 9.02–34.47) in patients with PPMS. Seventy-nine percent of patients with CIS, 63% with RRMS, 44% of SPMS and 71% of PPMS had elevated IgG levels. OB were found in 20/24 (83%) of CIS, 56/60 (93%) of RRMS, 11/15 (73%) of SPMS and 16/17 (94%) of patients with PPMS.
Table 1 Clinical features of patients at different stages of multiple sclerosis Symptoms
Motor Sensory Brainstem Cerebellar Visual Sphincter Cognitive
Clinically isolated syndrome (n = 27)
Relapsing-remitting MS (n = 66)
Secondary-progressive MS (n = 19)
Primary-progressive MS (n = 18)
Predominant
Present
Predominant
Present
Predominant
Present
Predominant
Present
6/27 (22) 14/27 (52) 5/27 (19) 0/27 2/27 (7) 0/27 0/27
13/27 (48) 18/27 (67) 5/27 (19) 1/27 (4) 4/27 (15) 4/27 (15) 0/27
22/66 (33) 18/66 (27) 10/66 (15) 7/66 (11) 9/66 (14) 0/66 0/66
41/66 (62) 36/66 (55) 15/66 (23) 19/66 (29) 16/66 (24) 3/66 (5) 0/66
12/19 (63) 1/19 (5) 0/19 5/19 (26) 0/19 0/19 1/19 (5)
18/19 (95) 17/19 (89) 3/19 (16) 8/19 (42) 3/19 (16) 6/19 (33) 8/19 (42)
13/18 (72) 1/18 (6) 1/18 (6) 2/18 (11) 0/18 0/18 1/18 (6)
17/18 (94) 11/18 (61) 1/18 (6) 5/18 (28) 2/18 (11) 6/18 (33) 2/18 (11)
Percentages are shown in brackets.
U. Rot, A. Mesec / Clinical Neurology and Neurosurgery 108 (2006) 271–274
273
Table 2 MRI findings in subgroups of patients
Brain MRI MS-like lesions Normal MRI 1–2 lesions >3 (no Barkhof criteria) Fulfilled modified Barkhof criteria Gd enhancement Spinal cord MRI
Clinically isolated syndrome (n = 27)
Relapsing-remitting MS (n = 66)
Secondary-progressive MS (n = 19)
Primary-progressive MS (n = 18)
20/25 (80) 5/25 (20) 4/25 (16) 2/25 (8) 14/25 (56) 8/18 (44) 8/11 (73)
65/65 (100) 0/65 2/65 (3) 6/65 (9) 57/65 (88) 21/40 (53) 9/12 (75)
17/19 (89) 2/19 (11) 2/19 (11) 1/19 (5) 14/19 (74) 3/15 (20) 3/4 (75)
18/18 (100) 0/18 2/18 (11) 2/18 (11) 14/18 (78) 4/9 (44) 7/9 (78)
Percentages are shown in brackets.
5. Correlation between clinical and CSF findings There was a positive correlation between CSF leukocyte numbers and IgG level (r = 0.5201; p < 0.0001), but negative correlations between CSF cell numbers and patients age, r = −0.3371; p = 0.0001, CSF cell number and duration of the disease (r = −0.3322; p = 0.0002) and CSF cell number and EDSS score (r = −0.2854; p = 0.0014). A modest negative correlation also existed between IgG level and duration of disease (r = −0.2215; p = 0.03), but none between IgG level and EDSS score (p = 0.44) and IgG level and patients age (p = 0.44).
patients. MRI results in the subgroups of patients are shown in Table 2.
7. Evoked potentials Prolonged latencies on VEP were found in 48/81 (59%) of our patients, 4/14 (29%) of patients with CIS, 29/44 (66%) of RRMS, 6/8 (75%) of SPMS and 9/14 (65%) of PPMS. Prolonged latencies of SEP were found in 28/52 (54%) of all patients, 7/14 (50%) of CIS, 13/22 (59%) of RRMS, 3/8 (38%) of SPMS, and 5/8 (63%) of PPMS.
8. Clinical and paraclinical findings in CIS and RRMS
6. MRI results MS-like lesions on brain MRI were seen in 121/127 (95%) patients. Among them nine (7%) had one or two lesions, 11 (9%) had >3 but did not fulfill the Barkhof et al. criteria but 101 (80%) did. Gadolinium enhancement occurred in 36/82 (44%) patients. Spinal cord lesions existed in 27/36 (75%)
Patients with CIS had more sensory symptoms (p = 0.03). They were younger (p = 0.02) and had shorter duration of disease at the time of diagnosis (p < 0.0001). Their CSF cell count was higher (p = 0.04), VEP latencies less often pro-
Table 3 Comparison of clinical and paraclinical characteristics in patients with CIS and RRMS
Sex ratio (F:M) Median age (years) Duration of the disease (years) Predominant sensory symptoms Median EDSS score Median CSF protein (g/l) Frequency of abnormal CSF protein (%) Median CSF cells/mm3 Frequency of abnormal CSF cells (%) Median IgG level (%) Frequency of abnormal IgG (%) Oligoclonal bands Frequency of abnormal VEP (%) Frequency of abnormal SEP (%) Brain MRI MS-like lesions Normal brain MRI 1–2 lesions ≥3 (no-Barkhof) Barkhof Gadolinium enhancement Spinal cord MRI
CIS (n = 27)
RRMS (n = 66)
p-value
16:11 29 (range 21–48) 0 (range 0–1) 14/27 (52%) 2.5 (range 0–3) 0.34 (range 0.15–0.67) 7/26 (27%) 11 (range 1–30) 21/26 (81%) 1492 (range 714–2775) 15/19 (79%) 20/24 (83%) 4/14 (29%) 7/14 (50%) 20/25 (80%) 5/25 (20%) 4/25 (16%) 2/25 (8%) 14/25 (56%) 8/18 (44%) 8/11 (73%)
50:16 34 (range 16–59) 2.5 (range 0–21) 18/66 (27%) 2.75 (range 0–6) 0.38 (range 0.19–0.66) 18/63 (29%) 5 (range 0–60) 42/63 (67%) 1246 (range 6.37–30.69) 33/52 (63%) 56/60 (93%) 29/44 (66%) 13/22 (59%) 65/65 (100%) 0/65 2/65 (5%) 6/65 (9%) 57/65 (88%) 21/40 (53%) 9/12 (75%)
0.11 0.02 <0.0001 0.03 0.07 0.53 0.88 0.04 0.21 0.15 0.26 0.22 0.03 0.73 0.0012 0.0012 0.05 1 0.0027 0.78 1
274
U. Rot, A. Mesec / Clinical Neurology and Neurosurgery 108 (2006) 271–274
longed (p = 0.03) and they were less likely to fulfill the MRI criteria (p = 0.02). Comparisons of clinical and paraclinical findings in CIS and RRMS are shown in Table 3.
9. Discussion We retrospectively analyzed the clinical and paraclinical characteristics of 130 patients with either MS or CIS. The most common symptoms in patients with CIS were sensory (52%), followed by motor (22%) and brainstem (19%) symptoms, whereas motor and sensory symptoms were more equally distributed in RRMS. Motor symptoms predominated in PPMS and were found in over 90% of patients. For comparison, in a Danish study 24% of patients started with motor symptoms, 32% with sensory and 20% with visual symptoms [8]. Bauer found that 43% of patients had motor, 41% sensory and 36% visual symptoms at the onset of the disease [9]. Only 7% of our CIS patients had visual symptoms reflecting the fact that optic neuritis patients are often referred to ophthalmologists rather than to neurologists. CSF OB were present in over 80% of patients with CIS and in more than 90% of all patients with MS. Sastre-Garriga et al. reported CSF OB in 71% of patients with isolated brainstem syndrome [10], whereas they were seen in 72% of optic neuritis patients in a Swedish study [2]. Interestingly, patients with CIS had higher CSF cell counts than RRMS and patients in more advanced stages of the disease. We also found a moderate negative correlation between CSF cells and patient age, duration of disease and EDSS. These results confirm our previous observation that patients with early MS have elevated CSF cell count [11], additional helpful information for the diagnosis of MS. Walker et al. also found a negative correlation between CSF cell number and duration of disease in their series of 120 MS patients[12]. On the other hand S¨oderstr¨om found no differences in CSF findings between CIS and MS [13]. Over 95% of our patients had MS-like lesions on brain MRI, confirming its high sensitivity for the diagnosis of MS. The modified Barkhof MRI criteria for dissemination in space provide an acceptable degree of sensitivity, specificity and accuracy and are included in current diagnostic criteria [1]. Fifty-six percent of our CIS and approximately 80% of our MS patients fulfilled them; as did 50% of Spanish patients with isolated brainstem syndrome [10]. Our results, and those of others, suggest that additional investigations should be performed in approximately half of CIS patients. Spinal cord MRI was performed in only one third of our patients and showed typical lesions in approximately 75%, underlining its importance in the diagnosis of MS [14]. The vast majority of our CIS patients had isolated spinal cord syndromes, thus accounting for the high sensitivity of spinal cord MRI in our study, paralleling Bot et al.’s report that 83% of 104 patients with recently diagnosed MS had spinal cord lesions [15].
The McDonald et al. criteria include only VEP [1]. They were obtained in 60% of our patients, but in addition, 40% also had SEP. Prolonged VEP latencies were found in only 30% of patients with CIS compared to over 60% of MS patients. SEP were more often prolonged in patients with CIS, again because most had isolated spinal cord syndrome. In Sastre-Garriga et al.’s Spanish study VEP latencies were prolonged in approximately 30% of patients and SEP in 35%, suggesting that evoked potential studies are not very helpful in confirming clinically silent lesions in CIS patients [10]. In conclusion, patients in the earliest stages of MS often present with sensory symptoms. Brain MRI can be inconclusive in over 40% of patients with CIS but the CSF profile of elevated cell counts, IgG ratio and positive OB bands is helpful in establishing the diagnosis. In later stages of the disease, the combination of clinical features, MRI, prolonged VEP latencies and positive CSF OB lead to the correct diagnosis.
References [1] McDonald W, Compston A, Edan G, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis. Ann Neurol 2001;50:121–7. [2] S¨oderstr¨om M, Ya-Ping J, Hillert J, et al. Optic neuritis. Prognosis for multiple sclerosis from MRI, CSF and HLA findings. Neurology 1998;50:708–14. [3] Kurtzke J. Rating neurological impairment in multiple sclerosis: an expanded disability rating scale. Neurology 1988;38:180–5. [4] Barkhof F, Filippi M, Miller D, et al. Comparison of MRI criteria at first presentation to predict conversion to clinically definite multiple sclerosis. Brain 1997;120:2059–69. [5] Tintor´e M, Rovira A, Martinez M, et al. Isolated demyelinating syndromes: comparison of different MR imaging criteria to predict conversion to clinically definite multiple sclerosis. AJNR 2000;21:702–6. [6] Brecelj J. Visual evoked potentials and electrophysiological evaluation of the visual pathway. Med Razgl 1994;33:339–59. [7] Deniˇsliˇc M, Prevec T. Analysis of somatosensory evoked potentials in multiple sclerosis. Zdrav Vestn 1987;56:7–10. [8] Bronnum-Hansen H, Koch-Henriksen N, Hyllested K. Survival of patients with multiple sclerosis in Denmark: a nationwide long-term epidemiologic survey. Neurology 1994;44:1901–7. [9] Bauer H. Problems of symptomatic therapy in multiple sclerosis. Neurology 1978;28(Suppl. 2):8–20. [10] Sastre-Garriga J, Tintor´e M, Rovira A, et al. Conversion to multiple sclerosis after a clinically isolated syndrome of the brainstem: cranial magnetic resonance imaging, cerebrospinal fluid and neurophysiological findings. Multiple Sclerosis 2003;9:39–43. [11] Rot U, Mesec A, Pogaˇcnik T. Multiple sclerosis with uncommon cerebrospinal fluid findings. Croat Med J 2003;44:697–701. [12] Walker R, Thompson E, McDonald W. Cerebrospinal fluid in multiple sclerosis: relationships between immunoglobulins, leukocytes and clinical features. J Neurol 1985;232:250–9. [13] S¨oderstr¨om M. Clues to the immunopathogenesis of multiple sclerosis by investigating untreated patients during the very early stage of disease. Neurol Sci 2001;22:145–9. [14] Lycklama G, Thompson A, Filippi M, et al. Spinal-cord MRI in multiple sclerosis. Lancet Neurol 2003;2:555–62. [15] Bot J, Barkhof F, Polman CH, et al. Spinal cord abnormalities in recently diagnosed MS patients: added value of spinal MRI examinations. Neurology 2004;62:226–33.