Subclinical demyelinating lesions may occur in the brain of asymptomatic individuals. A few case reports have been published on first degree relatives of patients with multiple sclerosis (MS) who are at a particular risk, demonstrating 10% subclinical demyelination in asymptomatic siblings.1 2 We report a descriptive retrospective study of the clinical and MRI findings after a 5 year follow-up in patients with subclinical demyelinating lesions fulfilling Barkhof criteria3 with a normal neurological examination.

PATIENTS

Nine neurologists were asked for medical advice concerning asymptomatic patients with possible MS suspected on brain MRI. All patients underwent brain MRI (1.5 T, axial fluid attenuated inversion recovery and fast spin echo T2 and T1 weighted sequences without and with gadolinium) for various medical problems not suggestive of MS. When demyelinating lesions fulfilled Barkhof–Tintoré MRI criteria,3 4 patients were recommended extensive neurological examinations, paraclinical studies (blood samples, spinal fluid analysis, evoked potentials) and MRI follow-up. All patients had a normal neurological examination. Differential diagnoses for multiple T2 hypersignals with or without gadolinium enhancement were screened. For each examination, brain MRI was performed with and without gadolinium injection.

RESULTS

Thirty patients were identified (26 women, four men; mean age 29 years (range 16–48)). Five had a familial history (one MS and four migraine). Twelve patients had a medical background: migraine (n = 5), depression (n = 4), atopic eczema (n = 2) and breast cancer (n = 1). The first brain MRI was performed for various medical events: headaches (n = 14), migraine with (n = 2) or without (n = 4) aura, craniocerebral trauma (n = 3), depression (n = 3), dysmenorrhoea (n = 2), epilepsy (n = 1) and cognitive changes (n = 1). Mean time between consultation and the first brain MRI was 6.4 months (1–48). Nine patients had gadolinium enhancement and 10 had infratentorial lesions (table 1).

Twenty-five patients underwent paraclinical studies and five decided to have only MRI follow-up. All patients had a normal examination and biological screening (serological panel, antinuclear and antiphospholipid antibodies). CSF characteristics were: mean cell count 2 (0–9 with three patients >4 cells/mm3), nine with oligoclonal bands and all had increased Ig levels. Eight patients had abnormal asymptomatic visual evoked potentials (VEPs). For patients with less than eight T2 hypersignals on brain MRI, CSF abnormalities were never associated with abnormal VEPs. Among the 21 patients with more than nine T2 hypersignals, 6/21 had an abnormal VEP, all associated with positive CSF. Mean time for the second brain MRI was 6 months (range 3–30); 14 patients 3–6 months, three patients 6–12 months, 12 patients 12–24 months and one patient 30 months. Twenty-five patients had temporospatial dissemination (nine with gadolinium enhancement, with five patients who already had active lesions) (fig 1). All 17 patients who had their second MRI scan during the first year had MRI dissemination, and 8/12 for the 12–24 month period. Eleven patients had clinical conversion: optic neuritis in five, brainstem (diplopia or internuclear ophthalmoplegia) in three, sensitive symptoms (paresthesias in the lower limbs) in two and cognitive deterioration in one. Four patients had the first clinical event during the first year after the MRI, three patients during the second year, three during the third year and one at 5 years. Eight (72%) already had criteria of dissemination to space and time before the clinical event. Mean time between the first brain MRI and clinically isolated syndrome (CIS) was 2.3 years (0.8–5 years) (table 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 (A) T2 weighted fluid attenuated inversion recovery (FLAIR) brain MRI showing hypersignals suggestive of multiple sclerosis. (B) Gadolinium enhancement of two lesions. (C) T2 weighted FLAIR brain MRI showing hypersignals with (D) dissemination in time on the MRI 6 months later.

No particular event was found in the 3 months before the CIS. Three patients were treated immediately with interferon. Three patients had two relapses, with a mean time of 9 months. To date, four patients have received immunomodulatory treatment, three with interferon and one with copaxone.

DISCUSSION

Subclinical MS has been incidentally identified at autopsy or in familial MS studies.1 2 A few case reports have been published on the first degree relatives of patients with MS who are at particular risk, demonstrating 10% subclinical demyelination in asymptomatic siblings. The concept of subclinical MS is now well recognised.5 Extrapolating results found in the literature to the general population would give rise to the possibility that undiagnosed MS exists at a similar prevalence to the diagnosed disease.6 Our pre-MS patients who developed CIS fitted with the literature but already had the MS dissemination criteria on MRI. In our study, all 17 patients who had their second MRI scan during the first year had MRI dissemination. These results fit with the CIS patients study with a placebo control group where serial brain MRI showed 25% of new activity at 3 months, 50% at 6 months and 85% at 2 years.

Review of a large database of patients with CIS found that 21% presented with optic neuritis, 46% with long tract symptoms and signs, 10% with brainstem syndrome and 23% with multifocal abnormalities.7 Therefore, the syndrome was isolated in space in only 77% of presentations, although all cases were isolated in time. Considering the three studies of CIS patients which had placebo arms, the second attack occurred within 2 years for 45% of patients and for 45% within the 3 years for the last study. In our study, 6/11 patients had their first clinical event within the 2 years after their first MRI.

There is evidence that the presence and number of lesions on MRI has an effect on the course of the disorder; the risk of having a diagnosis defining second episode after a CIS is higher in those with an abnormal scan.8 The presenting symptoms did not seem to affect the rate of conversion to definite MS, but lesions on MRI at baseline were strongly predictive of diagnosis, and the number of lesions was related to time to relapse and subsequent disability.9 All patients who had their clinical event during the first year fulfilled the Barkhof criteria on the first MRI scan. Our study confirms the burden of the preclinical phase with an increased ratio of MRI activity to clinical relapse. In 1997, Barkhof et al published criteria for MRI assessment to improve the prediction of CIS developing into clinically definite MS.3 These criteria were later modified by Tintoré et al and were also predictive in patients with CIS.4 Finally, the McDonald guidelines accepted the modified Barkhof–Tintoré criteria for evidence of dissemination in space. The MRI criteria for dissemination in time require the presence of gadolinium enhanced lesions on an MRI performed at least 3 months after the CIS.10 Applied to a large cohort of patients with CIS, fulfilment of the criteria at baseline showed after a survival time of 2 years, a conversion rate of approximately 45% and a median survival time of 11 months when fulfilling at least three Barkhof–Tintoré criteria, as in our pre-MS patients.11 Recent revised criteria accepted the dissemination in time if the baseline MRI performed more than 1 month after the clinical event shows active lesions.12

Even if incidental brain T2 lesions are frequents, patients with hypersignals and/or gadolinium enhancement fulfilling Barkhof–Tintoré MRI criteria and dissemination to time and space are more rare; 80% of our presymptomatic patients already had theoretical MRI-MS diagnostic criteria before the CIS. In longitudinal studies, it has been shown that 88% of patients with brain abnormalities at CIS developed MS,8 and even if new lesions are clinically silent, the ratio of MRI activity to clinical relapse is approximately 7–10%. Nevertheless, before the clinical threshold, other inflammatory diseases can often be suspected. Certain inflammatory or infections disorders (eg, systemic lupus erythematosus and neuroboreliosis) that are commonly part of the differential diagnosis can be investigated by blood and CSF analysis. At this step of the diagnosis, MRI is a crucial investigation for differentiation of inflammatory demyelination from other diseases.13

The question of treatment with immunomodulatory agents (IMA) should be raised. In 1998, the National Multiple Sclerosis Society issued a consensus statement recommending that IMA therapy should be initiated immediately on establishing a diagnosis of definite relapsing remitting MS.14 In recent years, investigators have addressed a related but different question: the effectiveness of the IMAs for reducing the risk of developing MS after a single demyelinating episode. The CHAMPS study, designed to evaluate the effect of interferon β1aIM in lowering the rate of developing MS after a single demyelinating event,15 and later the BENEFIT study16 with interferon β1b confirmed evidence of diminished disease activity.

For the pre-MS patient, without clinical events but with evidence of dissemination to time and space, medical discussion on the therapeutic options of IMAs for reducing the risk of MS should be conducted, with the knowledge that subclinical MS can evolve into relapsing remitting MS in the majority of cases17 or more rarely into progressive MS.18

To our knowledge, after many published case reports, this is the first cohort of CIS with preclinical follow-up.5 17^–20 All patients fulfilled the MS diagnosis according to space and time before the first clinical event. Discussion about early treatment has to be conducted. Because many patients with CIS and MRI lesions at presentation develop MS but have a benign course,8 prospective follow-up studies are needed to address these issues.