Article Text

Short report
Isolated cognitive relapses in multiple sclerosis
  1. Matteo Pardini1,2,3,
  2. Antonio Uccelli1,
  3. Jordan Grafman4,
  4. Özgür Yaldizli3,5,
  5. Gianluigi Mancardi1,2,
  6. Luca Roccatagliata2,6,7
  1. 1Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health, University of Genoa, Genoa, Italy
  2. 2Magnetic Resonance Research Centre on Nervous System Diseases, University of Genoa, Genoa, Italy
  3. 3MS Centre, University College London Institute of Neurology, London, UK
  4. 4Brain Injury Research, Rehabilitation Institute of Chicago, Chicago, Illinois, USA
  5. 5Department of Neurology, University Hospital Basel, Basel, Switzerland
  6. 6Department of Diagnostic and Interventional Neuroradiology, San Martino University Hospital, Genoa, Italy
  7. 7Department of Health Sciences, University of Genoa, Genoa, Italy
  1. Correspondence to Dr Matteo Pardini, Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics and Maternal and Child Health, University of Genoa, Largo Daneo 3, Genoa 16143, Italy;


Objective While cognition can be affected during sensorimotor multiple sclerosis (MS) relapses, the relevance of isolated cognitive relapses (ICRs ie, those occurring in absence of new sensorimotor symptoms) remain poorly characterised. Here, we decided to explore the relationship between ICR, subjective evaluation of cognitive performance and long-term cognitive decline in a group of subjects with relapsing-remitting MS.

Methods We analysed the cognitive performance of 99 clinically stable relapsing-remitting MS for whom data from four consequent clinical and cognitive evaluations were available, that is, a baseline evaluation (t0), followed in the subsequent 6 months by a second evaluation performed not later than 2 weeks after a routine brain scan positive for at least one area of gadolinium enhancement (t1) and two gadolinium enhancement-negative follow-up evaluations after 6 months (t2) and 1 year (t3) from t1. Based on published literature, we defined as a meaningful change in cognition a transient reduction of Symbol Digit Modalities Test score of at least four points at t1 compared with t0 and t2.

Results ICRs were found in 17 patients and were not associated with subjective cognitive deficits or depression. Subjects who presented with an ICR at t1 presented with a significantly reduced cognitive performance at the follow-up evaluations compared with patients without ICR.

Conclusions and relevance We showed that ICRs were not associated with changes in mood, fatigue levels or cognitive performance self-evaluations. Our study introduces an operational definition of ICRs and suggests to their role as a factor for cognitive decline in MS.

Statistics from


Relapses, one of the key features of relapsing-remitting multiple sclerosis (RRMS), are defined as episodes of transient neurological disturbance due to disease activity, which last for at least 24 h and are not better accounted for by other clinical conditions.1 Relapses recovery can be full or with residual stable deficits.2

Relapses are protean in their clinical manifestations, ranging from optic neuritis to motor or sensory disturbances, making diagnosis challenging, especially for the less common presentations.

Accurate and sensitive identification of relapses is important because they represent a major element of RRMS diagnostic criteria and because they are a widely used measure to assess treatment success.1 ,3

Among the different possible presentations of a multiple sclerosis (MS) relapse, cognitive relapses lack a clear operational definition applicable in clinical practice, and so their possible relevance has tended to be overlooked. Indeed, while transient cognitive impairments have been described in association with other symptomatic neurological deficits during MS disease activity,4 the construct of isolated cognitive relapse (ICR), that is, a transient reduction in cognitive functioning not associated with other subjective or objective neurological symptomatology, albeit described in different single case reports,5 ,6 remains to date poorly characterised.

Our main research questions were as follows: (1) Is it possible to evaluate ICR in the clinical setting? (2) What is the relationship between ICR and subjective cognitive evaluation, taking into account that transient changes in cognition are not a frequent complaint in MS? (3) What is the relationship between ICR and stable cognitive impairment?


ICR definition

We applied the definition of relapse to the cognitive domain to operationally define ICR as (1) a transient significant (see below) cognitive decline in objective neuropsychological performance, (2) without clinical or subjective evidence of other new neurological signs and symptoms (3) associated with brain disease activity defined as a positive gadolinium enhancing scan (Gd+). We would like to point out that while the last criterion is not necessary for a clinical relapse per se, we still decided to apply it in this proof-of-concept study to ensure that enrolled subjects presented with current disease activity.

Patients’ population

Clinical and cognitive data were retrospectively collected from a group of patients with MS who underwent longitudinal cognitive and behavioural evaluations at our MS centre between 2008 and 2012. Our inclusion criteria were as follows: (1) a diagnosis of RRMS, (2) age range: 18–50 years, (3) EDSS score lower than 6, (4) at least 12 years of formal education and (5) no significant comorbidities or psychoactive drug use.

For all time points described below we collected the following clinical and cognitive parameters: EDSS score, Symbol Digit Performance Test (SDMT) score,7 Hospital Anxiety and Depression Scale, depression score,8 Modified Fatigue Impact Factor Scale total score9 and MS Neuropsychological Screening Questionnaire10 self-report score. Cognitive reserve was evaluated with the Cognitive Leisure Activity Questionnaire, completed at the end of the observation period.11 Descriptions of all tests and baseline cut-off values for inclusion in the study are described in the supplementary online methods.

To comply with our proposed operational definition of ICR, we focused on those subjects for whom data from four consequent clinical and cognitive evaluations were available, that is, a baseline evaluation (t0), followed in the subsequent 6 months by a second evaluation performed not later than 2 weeks after a routine brain scan positive for at least one Gd+ area (t1) and two gadolinium enhancement-negative follow-up evaluations after 6 months (t2) and 1 year (t3) from t1. We excluded from the analysis those subjects who presented with a change in Expanded Disability Status Scale (EDSS) score, novel subjective non-cognitive complaints or steroid use during the entire study period.

Subjects consented to all procedures, which were performed according to the Helsinki declaration. All procedures were approved by the ethics committee.

SDMT cut-off

We focused on the SDMT, a widely used screening test for cognitive impairment in MS12 to evaluate the possibility of diagnosing ICR in the clinical setting. Based on published literature,4 we defined as a meaningful change in cognition a transient reduction of SDMT score of at least four points at t1 compared with t0 and t2. This cut-off was previously used to explore transient cognitive changes in MS in a seminal study on the association of cognitive impairment during sensorimotor relapses,4 and it has been linked with an ecologically valid negative outcome.13 Patients presenting with a transient SDMT reduction of at least four points between t0 and t1 were thus enrolled in the ICR group, while all other patients were enrolled in the no-ICR group. Statistical methods are reported in the online supplementary methods.


Group composition

We included in our analysis 99 patients with RRMS (50 women, age: 41.2±5.6 years, EDSS: 3.5±2.2) of which 17 (9 women, age: 39.8±5.0 years, EDSS: 3.3±2.5) fulfilled our criteria for the ICR group. Demographic data are reported in online supplementary table S1, while cognitive and clinical data are reported in table 1. At baseline, we did not observe any significant differences in clinical or cognitive parameters between the two groups. Moreover the two groups were matched on Cognitive Leisure Activity Questionnaire scores and other measures of resilience (see online supplementary materials). Gd+ lesion location is reported in online supplementary table S2. The distribution of Gd+ lesions was different between the two groups with ICR subjects showing only frontoparietal Gd+ lesions (see online supplementary results).

Table 1

Cognitive data for the isolated cognitive relapse (ICR) and no-ICR groups

Differences in cognitive performance over time

Our model revealed a significant Time×Group interaction effect (F=45.116, p<0.001, partial η2 0.344) while all other effects were non-significant. Analysis of between-subjects effects inside the aforementioned model revealed an overall difference in SDMT scores (F=14.7, p<0.001), with the ICR group showing significantly reduced SDMT scores at t1,t2 and t3 (t1: p<0.001, t=8.76, partial η2=0.47; t2: p=0.003, t=5.7, partial η2=0.35; t3: p=0.004,t=3.7, partial η2=0.30), while there was no statistically significant difference in performance at baseline between the two groups (t0: t=0.461, p=0.64).

Longitudinal cognitive changes: t0-to-t1

Between t0 and t1, subjects in the ICR group presented with a reduction in SDMT performance (figure 1, p<0.001, t=20.0), while they did not present with any other statistically significant changes in other tests, including the MS Neuropsychological Screening Questionnaire and Hospital Anxiety and Depression Scale-D scales. All comparisons in the control group were not statistically significant.

Figure 1

Symbol Digit Modalities Test (SDMT) data for isolated cognitive relapse (ICR) and no-ICR groups. Data are reported as mean±SDs.

Longitudinal cognitive changes: t1-to-t2

Between t1 and t2, subjects in the ICR group presented with a significant increase in SDMT (figure 1, p<0.001, t=15.1). The range of SDMT increase was four to seven points. All other comparisons were not statistically significant.


Here, we showed that ICRs were neither associated with subjective changes in mood or fatigue levels nor with a significant alteration in cognitive abilities insights. Moreover we showed that compared with ICR-free patients, patients with ICR presented with significantly lower cognitive performances at follow-up evaluations after 6 months and 12 months.

Our study builds on previous descriptions of effects of inflammatory activity on cognitive performances. Indeed, in RRMS the presence of transient cognitive impairment during symptomatic sensorimotor relapses is well described,4 while an increase in cognitive performance was previously observed when comparing baseline Gd+ with a gadolinium negative (Gd−) follow-up evaluation.14 Compared with these data our results suggest the relevance of transitory changes in cognitive performance in MS, and of their conceptualisation as ICR. Our data suggest that ICR are associated with a persistent reduction of cognitive performance over time; by analogy these findings are consistent with the known relationship between optical neuritis and long-term visual deficits in demyelinating disorders.15 Future studies are needed to verify the potential of treatment of ICR as a possible strategy to reduce the cognitive disability burden in this population.

From a methodological point of view, our data highlight the value of longitudinal neuropsychological testing in subjects with good baseline cognitive performances and lack of subjective complaints. Indeed, the partial awareness for cognitive performance changes in MS is well known11 and suggests a possible explanation why acute cognitive impairment is rarely reported by patients with MS.

Regarding the composition of the cognitive protocol, we acknowledge that our examination included only the SDMT, thus possibly leading to an underestimation of the frequency of ICR. While this observation, together with the retrospective nature of the study, does not allow quantifying the prevalence of IC, our positive results based on this simple test suggest the feasibility of the identification of transient cognitive changes in clinical practice.

This retrospective work presents with some limitations. First, to control for confounding effects due to chronic cognitive deficits, we focused on a carefully selected population with MS presenting with very good baseline cognitive performances and few confounding factors. Moreover the heterogeneity of our MRI data regarding the radiological protocol did not allow evaluating the lesion loads to carry on any quantitative analyses on the relationship between ICR and MRI metrics. Lastly, as the timing of our follow-ups and the limited time frame did not allow us to explore the frequency of ICR over time; future studies are warranted to clarify this issue.

Despite these limitations, however, we believe that our data reveal that ICRs are a clinically relevant facet of cognitive function in MS and suggest that heightened attention to transient cognitive status changes in MS could improve patient management and help in the development of individualised pathways to care.


MP thanks the non-profit patient association AKWO for their unrestricted support. The AKWO association had no role in the study including study design and conduction, data management, analysis, data interpretation and manuscript preparation. MP also thanks Novartis for unrestricted research support for this project. Novartis had no role in the study including study design and conduction, data management, analysis, data interpretation and manuscript preparation. The authors thank Dr Declan Chard, UCL Institute of Neurology, London, for the very useful insights on this manuscript.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • Contributors MP, AU, GLM and LR conceived the study. MP supervised the clinical evaluations. All authors collaborated to the statistical analysis, the interpretation of the data and the draft of the manuscript. All authors approved the final version of the manuscript. MP had access to all data and takes full responsibility for data analysis.

  • Competing interests MP: received research support from Novartis. AU received consulting fees from Biogen Idec, Merck Serono, Allergan and Roche; payment for lectures from Merck Serono, Biogen Idec, Teva, Sanofi-Aventis, Novartis, Roche. JG: serves as co-editor of Cortex. ÖY has received lecture fees from Teva, Novartis and Bayer Schering which was exclusively used for funding of research and continuous medical education in the Department of Neurology at the University Hospital Basel. GLM received honoraria for lecturing, travel expenses for attending meetings and financial support for research from Bayer Schering, Biogen Idec, Sanofi-Aventis, Novartis and Merck Serono. LR: nothing to disclose.

  • Ethics approval University of Genoa.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Cognitive data are available for collaborative studies upon request by no-profit researchers.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.