New diagnostic criteria for multiple sclerosis

Patients and methods.

The current study is based on longitudinal clinical, CSF, and MRI data prospectively acquired from a cohort of patients with CIS recruited since 1995. The McDonald criteria were retrospectively applied to this prospectively acquired data set. Patients presenting for the first time with monophasic neurologic symptoms of the type seen in MS were recruited at Hospital Universitari Vall d’Hebron in Barcelona. Inclusion criteria were as follows: 1) a CIS suggestive of CNS demyelination involving the optic nerve, brainstem, or spinal cord, not attributable to other diseases; 2) aged <50 years; 3) onset of symptoms within 3 months of both clinical and MRI examinations; and 4) clinical follow-up of 12 months or longer. All patients were specifically asked about any history of neurologic disturbances.

One hundred and thirty-nine patients were included. The mean follow-up was 39 ± 17.2 months (range 12 to 77 months). Detailed history and physical examination at baseline and follow-up were performed by one of four neurologists. Patients were seen every 3 to 6 months and were instructed to report any new or worsening of preexisting symptoms. A diagnosis of conversion to CDMS was made when new symptoms occurred after an interval of at least 1 month and only when other diagnoses had been excluded. CDMS was diagnosed when there was a second attack with a new neurologic abnormality that was confirmed by examination.

To detect IgG OB, unconcentrated CSF together with appropriately diluted plasma were examined by agarose isoelectric focusing combined with immunoblotting and avidin–biotin-amplified double-antibody peroxidase staining.6,7⇓ OB were considered positive when two or more bands were demonstrated in CSF only.

Brain MRI was performed after the first demyelinating event and repeated after 12 months of follow-up. MRI was performed on a 1.0- or 1.5-T machine with a standard head coil. MRI included the following pulse sequences: transverse proton-density and T2-weighted conventional spin-echo (2200/20–90/1 [repetition time/echo time/excitations]), or fast spin-echo (3000/14–85/2), and in some patients contrast-enhanced (0.1 mmol/kg; scan delay, 5 minutes) T1-weighted spin-echo (600/15/2). For both sequences, we used a section thickness of 5 mm, a pixel size of approximately 1 × 1 mm, and an interleaved scan mode with an intersection gap of either 1.5 mm when using conventional spin-echo sequences or 5 mm when using fast spin-echo sequences, resulting in 26 contiguous sections covering the whole brain. The MRI scans were assessed by two neuroradiologists who were blinded to clinical follow-up. In case of discrepancy, a final decision was reached by consensus after a third analysis of the images. The proton-density and T2-weighted images were scored according to the following items: total number of lesions, number of periventricular lesions, presence of juxtacortical lesions, and presence of infratentorial lesions. In the baseline scan, we applied the Barkhof four-criteria dichotomized MRI model that included the following: 1) presence of at least one gadolinium-enhancing lesion (or the presence of at least nine lesions on the T2-weighted images); 2) presence of at least three periventricular lesions; 3) presence of at least one juxtacortical lesion; and 4) presence of at least one infratentorial lesion (see table E-1 on the Neurology Web site; go to www.neurology.org).3 In patients in whom a contrast-enhanced, T1-weighted sequence was obtained (n = 33), the presence of at least one enhancing area related to a lesion seen on T2-weighted images was scored. In those patients who did not receive contrast-enhanced MRI (n = 106), the alternative option proposed by Barkhof (nine T2-weighted lesions) when gadolinium is lacking was applied.3

According to the MRI component of the new McDonald criteria, evidence of dissemination in space was provided in two ways: 1) presence of at least three of four MRI Barkhof criteria; and 2) presence of at least two T2 lesions plus OB. Dissemination in time was fulfilled when at least one new T2 lesion had appeared in the first-year MRI. The MRI criteria were met when patients fulfilled the MRI definitions for dissemination in time and space. In addition, patients with a second clinical attack also fulfilled the new criteria.

Based on clinical follow-up (conversion or nonconversion to CDMS), we calculated sensitivity, specificity, and accuracy of the two MRI definitions for dissemination in space. True-positive is defined as abnormal MRI results and conversion to CDMS, false-positive as positive MRI results in the absence of conversion to CDMS, false-negative as normal MRI results but conversion to CDMS, and true-negative as normal MRI results and no conversion to CDMS. In the subset of patients with at least 3 years of follow-up, we also calculated sensitivity, specificity, and accuracy of the McDonald MRI criteria (the McDonald MRI criteria were met when patients fulfilled the MRI definitions for dissemination in time and space).

At 1 year, a diagnosis of MS was determined using both Poser and McDonald criteria.

Results.

The 139 patients in this study included 97 women and 42 men. Their ages ranged from 13 to 49 years, with a mean age of 30 years. Fifty-eight patients (41.5%) presented with optic neuritis, 34 (24.5%) with brainstem syndromes, 39 (28%) with spinal cord syndromes, and eight (6%) patients had a polysymptomatic presentation. The mean clinical follow-up was 39 ± 17.2 months (range 12 to 77 months, median 36.7) (see table E-2 on the Neurology Web site). All patients had at least 12 months of follow-up, 122 patients had at least 2 years of follow-up, and 86 had been followed up for at least 3 years. CDMS was diagnosed in 38 patients (27.3%) (mean conversion time 463 days, median 400 days) (table 1).

Accuracy of the “dissemination in space” criteria for the development of CDMS.

Fulfilling Barkhof criteria (definition 1) and conversion to CDMS showed a sensitivity of 71%, specificity of 69%, and accuracy of 70%. Fulfilling “two lesions plus positive OB” (definition 2) and conversion to CDMS, we obtained a sensitivity of 78%, specificity of 63%, and accuracy of 67%. However, using definition 1 or 2 and conversion to CDMS, we obtained a higher sensitivity of 87%,but a lower specificity of 56% and an accuracy of 65% (table 3 and figure 1).

View this table:

• View inline
• View popup

Table 3 Accuracy of dissemination in space definitions (Barkhof or two lesions and positive OB) and both combined in determining MS according to the Poser criteria

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Sensitivity, specificity and accuracy of the two MRI definitions for dissemination in space: definition 1 (3 or 4 Barkhof criteria), definition 2 (at least two lesions plus oligoclonal bands [OB]), and the combination of definitions (definition 1 or 2). Solid line = sensitivity; short dashed line = accuracy; long dashed line = specificity.

Accuracy of the MRI McDonald criteria.

The MRI McDonald criteria were met when patients fulfilled the MRI definitions for dissemination in time and space.

In the subset of patients that had at least 3 years of follow-up (n = 86), the MRI McDonald criteria at 12 months showed a sensitivity of 74%, a specificity of 86%, andan accuracy of 80% for the development of CDMS (table 4 and figure 2).

View this table:

• View inline
• View popup

Table 4 Accuracy of the MRI McDonald criteria at 12 months in predicting clinically definite MS (CDMS) in the cohort of patients with at least 3 years of follow-up (n = 86)

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Conversion to clinically definite MS (CDMS) in patients fulfilling the MRI McDonald criteria (MRI McDonald positive; gray bars) and not fulfilling the MRI McDonald criteria (MRI McDonald negative; white bars) at 12 months.

Discussion.

The International Panel on the Diagnosis of MS have developed new diagnostic criteria that integrate MRI into the diagnostic scheme and hence facilitate the diagnosis of MS in patients with a first demyelinating episode.1 In this group of patients, establishing a diagnosis of MS requires the demonstration of dissemination in time and space. The Panel recommended that evidence of dissemination in space can be met when MRI abnormalities fulfill three of four MRI criteria (first definition).3,4⇓ In our series, 58 patients (42%) fulfilled this condition, which is similar to other studies (34%, 36%, and 41%).3,4,8⇓⇓ In patients lacking such robust evidence, at least two brain lesions plus positive OB may also be used to demonstrate dissemination in space (second definition). In our cohort, 19 patients (14%) fulfilled this second condition. Thus, in total, 77 patients (56%) fulfilled criteria for dissemination in space using baseline MRI and, when required, CSF examination as diagnostic tools. As in previous studies, we have shown that fulfilling at least three of four Barkhof criteria is highly specific for predicting conversion to CDMS.4,7⇓ The Barkhof criteria were chosen by the panel of experts because they agreed that stringent criteria for MRI abnormalities should be followed in making a diagnosis of MS. The alternative criteria proposed by the panel (at least two lesions plus OB) also showed reasonably good accuracy with good balance between sensitivity and specificity. Nevertheless, when applying both criteria for dissemination in space (definition 1 or definition 2), we increase sensitivity but we lose specificity. The combination of definition 1 or definition 2 should be used with caution because the high specificity of both criteria individually may be compromised (see table 3 and figure 1).

According to the new diagnostic criteria, the demonstration of dissemination in time in patients with CIS may require three MRI scans if the first scan is performed less than 3 months after the onset of the clinical event.1 In this situation, a second scan is recommended 3 or more months after the clinical event, and if no gadolinium enhancement is seen or used at this second scan, a third scan will be necessary to confirm the presence of new T2 lesions. In our cohort, two scans were performed 1 year apart and a third was not done. For this reason, the presence of new T2 lesions could, in theory, be somewhat overinterpreted because some of these new T2 lesions may have been related to the first episode. In this regard, it is perhaps reassuring that 45% of new T2 lesions found in our cohort is very similar to the 43% of new T2 lesions found when brain MRI was repeated 3 months after a first event and is also similar to the 53% found when three scans are performed within 1 year.8,9⇓ Clearly, when considering dissemination in time, the Panel took a rigorous approach requiring three MRI scans to confirm the presence of a new T2 lesion. This degree of rigor may not be necessary and preliminary results suggest that when evaluating dissemination in time, the appearance of new T2 lesions at least 3 months after a CIS is a powerful predictor for CDMS and does not result in reduced specificity.10

In the current study, after a follow-up of 1 year, 40% of the patients fulfilled the clinical and MRI criteria for dissemination in time and dissemination in space necessary for a diagnosis of MS according to the McDonald criteria, compared to only 11% according to the Poser criteria. Therefore, in just 12 months the new criteria more than tripled the frequency of diagnosis of MS. The English cohort resulted in 48% of the patients fulfilling McDonald criteria at 1 year.8 Early diagnosis of MS is now of great relevance, as some currently available treatments have demonstrated a delay in the occurrence of a second attack11,12⇓ and some neurologists consider treating patients early in the course of this condition, although the long-term clinical impact of these treatments remains to be demonstrated. It was reassuring to see that the majority (80%) of patients fulfilling the MRI McDonald criteria at 12 months developed a second clinical event during follow-up, demonstrating the usefulness of these criteria in predicting a diagnosis of MS (see table 1).