Practice Points
- Serum neurofilament light chain (sNfL) measurement can help identify individuals with radiologically isolated syndrome (RIS) who are at high risk for multiple sclerosis (MS) conversion.
- Despite evidence supporting early treatment in RIS, many neurologists hesitate to initiate disease-modifying therapy, revealing a limited
perception of the benefits of biomarkers like sNfL. - Neurologists not fully dedicated to MS are more likely to defer treatment initiation, emphasizing the need for specialized MS care.
- Standardizing sNfL protocols, defining reference ranges, and raising awareness of clinical utility are key to improving RIS management.
The rising adoption of advanced brain imaging in medicine has led to the detection of incidental brain and spinal cord abnormalities, even in the absence of clinical symptoms.1 Radiologically isolated syndrome (RIS) describes individuals who exhibit focal white matter T2-hyperintense MRI lesions in the central nervous system that are highly suggestive of multiple sclerosis (MS), despite showing no symptoms and having a normal neurological examination.2-4 Although RIS is relatively rare, with a cumulative incidence of 0.1%-0.7% according to Okuda criteria, between 30.7% and 51.2% of individuals develop MS within 3 years and 10 years following the initial detection of these incidental lesions, respectively.5-7 Young age; male sex; presence of oligoclonal bands (OCBs) in cerebrospinal fluid; and infratentorial, spinal cord, or gadolinium-enhancing lesions on the index MRI scan are associated with an increased risk of developing MS in individuals with RIS.6-10
Serum neurofilament light chain (sNfL) concentrations serve as a biomarker of neuroaxonal damage in different inflammatory and neurodegenerative disorders and are increasingly used in MS care to predict clinical outcomes and monitor treatment response.11-14 Elevated sNfL levels have been identified as a new independent risk factor of conversion to MS among individuals with RIS, with higher values correlating with a shorter time to first symptom.4,15 Testing of sNfL helps detect spinal cord lesions and/or OCBs and aids in discriminating individuals who have a high risk for clinical conversion from individuals with stable RIS.4
The findings of 2 recent studies demonstrated that early treatment with disease-modifying therapies (DMT), such as dimethyl fumarate or teriflunomide, in patients with RIS significantly reduces the risk of developing the first clinical symptoms associated with MS.16,17 However, there is still uncertainty about whether all individuals should receive treatment.4,7,18 In this context, having a reliable and noninvasive biomarker of inflammatory activity like sNfL could aid neurologists in making prompt therapeutic decisions.4,19 The aim of this study was to assess how sNfL testing influences neurologists’ decisions regarding the initiation of DMT in the management of RIS.
Methods
We conducted a noninterventional, cross-sectional, web-based study in collaboration with the Spanish Society of Neurology (SEN) as part of the NewFeeLs-MS project.20,21 This study aimed to evaluate neurologists’ perspectives on the utility of sNfL measurement in guiding treatment decisions in MS. Specifically, the present analysis examines choices regarding the initiation of DMT in individuals with RIS.
Neurologists actively involved in MS care completed a questionnaire and provided details about their demographic characteristics, professional background, clinical setting, and behavioral characteristics, including attitude toward evidence-based innovation, medical empathy, care-related regret, risk attitude at work, and occupational burnout. Participants were then provided with a simulated case scenario reflecting a common encounter with an individual with RIS to assess treatment initiation supported by sNfL levels. This case scenario was developed by the research team based on clinical practice and literature review.2,3,6,14,21
Participants were invited to participate via e-mail from SEN. The study was approved by the Hospital Universitario Clínico San Carlos, Madrid, Spain, ethical review board (reference 23/471-E). All participants provided written informed consent and were recruited from August 28, 2023, to December 12, 2023.
RIS Experiment
The simulated case scenario described a 24-year-old woman diagnosed with RIS based on an incidental brain MRI performed for a headache episode that showed 1 juxtacortical and 8 periventricular white matter T2-hyperintense lesions highly suggestive of MS. The patient had OCBs in her cerebrospinal fluid and sNfL levels of 24 pg/mL. Participants were informed that there was no other neurological or nonneurological condition that could explain the radiological findings and sNfL levels. They then had to decide whether to recommend initiating treatment with teriflunomide or dimethyl fumarate, or to schedule the patient for a reassessment in 6 to 12 months.
Before starting the experiment, participants were shown a screen with the following instructions: “We invite you to participate in a short survey to understand how sNfL testing influences decisions regarding the initiation of disease-modifying treatment in the management of RIS. We will present a case as a hypothetical situation in which the decision to treat should not be conditioned by hospital protocols, drug access, or costs. Responses should not be conditioned by reasons other than clinical judgment. The pathological cutoff point for sNfL levels will be 10 pg/mL.”22
Outcome Measures
The primary outcome was decisions misaligned with emerging evidence (DMEE), defined as not recommending the initiation of DMT despite an elevated sNfL level.
Participants’ perceptions regarding the benefits of sNfL measurement were assessed by level of agreement with the following statement: “On a scale of 1% to 100%, what is your expectation of the usefulness of the sNfL determination in defining treatment for this patient?” The Evidence-Based Practice Attitude Scale (EBPAS) is a validated 15-item instrument designed to evaluate health care professionals’ openness to embracing new evidence-based treatments, interventions, and practices.23 Scores on the EBPAS range from 0 to 4, with higher scores reflecting a more favorable attitude toward health care innovations. Empathy levels among health care professionals were measured using the 20-item Jefferson Scale of Empathy-Health Professionals.24 This scale produces a continuous variable score ranging from 20 to 140, with higher scores indicating a greater degree of empathy toward patients. A previous experience involving regret related to caring for a patient was evaluated using the 10-item Regret Intensity Scale.25 A cutoff score of greater than 3 indicates the presence of care-related regret. The risk attitude at work was assessed using a single question from the German Socio-Economic Panel, a validated survey that measures willingness to take risks across various domains.26 Respondents were asked to rate their inclination to take risks in their work on a scale of 0 (not at all) to 10 (extremely). Occupational burnout was assessed using a single-item measure derived from the Physician Worklife Study.27 This measure utilizes a 5-category ordinal scale, where a cutoff score of greater than or equal to 3 indicates the presence of burnout.
Statistical Analysis
Descriptive statistics were calculated for all variables. Continuous variables were summarized using measures of central tendency (mean and median) and variability/dispersion (SD and IQR). Categorical variables were described using absolute and relative frequencies (percentages).
Logistic regression models were used to evaluate the relationship between participants’ characteristics and the primary binary outcome of interest. The explanatory variables included in the model were predefined based on conceptual and clinical relevance, ensuring a balance between empirical and conceptual justification: age, sex, years of professional practice, years caring for patients with MS, practice setting (academic vs nonacademic), full or partial dedication to MS care, number of people with MS seen weekly, access to sNfL testing, participation in MS clinical trials (yes or no), burnout (yes or no), empathy, attitude toward innovation, care-related regret, and risk attitude at work. All variables were included simultaneously in the multivariable model to assess their independent associations with DMEE. Responses were mandatory, resulting in no missing data. Model fit was evaluated using the Hosmer-Lemeshow test and the C statistic.
Results
121 completed the survey (completion rate, 89.6%). After excluding 5 participants who did not meet the inclusion criteria, the final analysis included 116 neurologists. The participants had a mean age of 41.9 years (SD, 10.1), and 53.4% were men. Seventy-eight (67.2%) neurologists were fully dedicated to the care of demyelinating disorders and managed a median of 16.0 (IQR, 10.0-25.0) patients with MS per week. Although 94.8% (n = 110) worked in academic hospitals, only 34.5% (n = 40) had access to sNfL testing at their hospitals. The Table shows participants’ characteristics.
Overall, 58.6% of participants (n = 68) showed DMEE. Participants’ overall expectations regarding the benefits of sNfL testing in this case were moderate, with a median of 60.5 (IQR, 40.0, 77.8). Participants with DMEE cared for fewer patients with MS per week (P = .01), participated less frequently in clinical trials (P = .01), and reported a lower willingness to adopt innovations compared with their counterparts (P = .03). The multivariable analysis, which was adjusted for age, years of experience, number of patients per week, EBPAS score, and global perception of sNfL value (as continuous variables), as well as sex, participation in clinical trials, and full or partial dedication to MS care (as dichotomous variables), indicated 2 significant associations with DMEE. Being a neurologist not exclusively dedicated to MS care (OR, 2.52; 95% CI, 1.02-6.20; P = .04) and having a low perception of the clinical utility of sNfL (OR, 1.03; 95% CI, 1.01-1.05; P < .001) were associated with DMEE. Participants’ demographic and behavioral characteristics, type of hospital, number of patients with MS, or access to sNfL tests were not associated with DMEE. There was no evidence of collinearity. The models demonstrated good fit, as indicated by the Hosmer-Lemeshow test (P = .11) and a C statistic of 0.75.
Discussion
Longitudinal studies assessing individuals with RIS found that over 50% will experience the first clinical event of MS within 10 years after the detection of incidental lesions.4 Elevated levels of sNfL have been identified as an independent risk factor for clinical conversion within 3 years of RIS diagnosis, with a hazard ratio of 2.8 (95% CI, 1.2-6.6; P = .02).15 Further, randomized, double-blind clinical trials showed that dimethyl fumarate and teriflunomide significantly delay the time to the first clinical event in individuals with RIS, with a risk reduction of 80% and 72%, respectively, vs placebo.16,17 In this context, RIS still presents a clinical dilemma for neurologists, balancing the potential benefits of early initiation of DMT to prevent conversion to MS against the risk of adverse events in asymptomatic individuals.4,19,28
In our study, nearly 60% of participants in a survey organized in collaboration with the SEN chose not to initiate any DMT in a simulated case of RIS with an elevated sNfL level, despite the high risk of MS conversion. The lack of full dedication to MS care and a limited perception of sNfL benefits were associated with suboptimal decisions in this scenario. Notably, neurologists who managed neurological disorders other than MS were 2.5 times more likely to refrain from recommending the initiation of treatment. These findings suggest that although sNfL is a valuable prognostic biomarker, there may be areas where neurologists’ confidence or perception of its clinical utility in guiding therapeutic decisions for high-risk RIS patients could be further developed.
Neurologists’ decision-making process regarding whether to treat individuals with RIS has been explored in a few studies, all predating the outcomes of clinical trials involving dimethyl fumarate and teriflunomide and the availability of sNfL testing.29-31 In general, although there were variations among neurologists in Europe, the United States, and Latin America, the prevailing tendency was not to initiate DMT in cases of RIS.29-31 Follow-up MRI played a crucial role in this decision, particularly if gadolinium-enhanced lesions and/or new or enlarging T2 lesions were detected.
The likelihood of experiencing a first clinical event increases gradually as the number of risk factors rises.10 The emergence of ultrasensitive single-molecule detection technologies has enabled the measurement of sNfL as a marker of ongoing inflammatory disease activity in different neurological disorders.11 Measuring sNfL appears to be a reliable and easy-to-implement additional tool for identifying high-risk individuals with RIS in the context of the treatment decision-making process.4 Rival et al found that 84% of a sample of 57 individuals with RIS and OCBs, spinal cord lesions, and sNfL levels greater than 5 pg/mL at diagnosis experienced clinical conversion within 3 years, compared with 34% in the remaining cases where no association of these factors was observed.15 In our simulated case scenario, there were 3 risk factors associated with MS: being younger than 37 years, the presence of OCBs, and an elevated sNfL determination.
Physicians’ decisions to adopt or reject a new diagnostic tool are perceptual phenomena influenced by a combination of objective knowledge and emotional factors.32-34 These factors include demographic elements (eg, age, sex, and practice setting) and behavioral elements (eg, risk preferences and personal approach). In addition, considerations related to the innovation itself, such as ease of use, perceived usefulness, potential risks, and financial implications, also play a role. The moderated recognition of the benefits of sNfL in our study can be attributed to several factors. These may include its lack of specificity; susceptibility to confounding factors such as age, sex, renal function, or body mass index; and the absence of standardized measurement techniques and reference ranges across different neurological populations and diseases.14,35,36
Our study’s findings align with previous research, showing that general neurologists or those not exclusively focused on managing demyelinating diseases are more likely to make suboptimal medical decisions when compared with MS specialists.37 Multidisciplinary MS care units provide a comprehensive and specialized approach to managing demyelinating diseases, ultimately leading to improved patient outcomes and enhanced quality of life, as well as cost-effective care.38 In today’s MS care landscape, management has become increasingly complex due to the diverse array of diagnostic and follow-up techniques available, as well as the varying efficacy and safety profiles of treatment options. This dynamic environment demands a comprehensive approach that integrates evolving treatments, outcomes assessment, and biomarker utilization to optimize care delivery and patient outcomes. Consequently, neurologists who are not exclusively focused on managing demyelinating diseases may find it challenging to fully integrate the wide variety of factors that affect current MS care. Given the complexity of RIS management, we suggest that patients with RIS be treated by neurologists with greater clinical and radiological expertise to optimize outcomes. Future research could explore the impact of specialist care on RIS outcomes. 4,28,39,40
The emergence of sNfL testing as a promising biomarker in the context of RIS management presents both opportunities and challenges. Although sNfL measurement holds potential for identifying high-risk individuals and informing treatment decisions, its integration into clinical practice requires careful consideration of various factors, including standardization of measurement techniques, establishment of reference ranges, and enhancement of neurologists’ awareness and understanding of its clinical implications.
Our study has some limitations. First, we included 3 risk factors associated with conversion to MS.4 This suggests that our results might have been more pronounced (meaning that a larger number of participants might have deferred treatment initiation) if we had included a simulation of elevated sNfL levels in isolation. Second, the study did not collect the participants’ level of knowledge about sNfL in demyelinating disorders, a factor potentially impacting their perception of the utility of this assessment. Third, our exclusive focus on practicing neurologists from Spain might limit the generalizability of our findings to other cultural contexts and health care systems. Fourth, this study did not investigate the financial implications or the practical ease of obtaining sNfL testing, nor the current lack of funding of DMT in the RIS indication, for health care systems in Spain, factors that could impact real-world treatment decisions despite being explicitly removed from our simulated scenario. Fifth, although our study demonstrated that elevated sNfL levels did not uniformly lead to treatment initiation in our simulated scenario, a limitation is that the study design did not include a direct comparison with a hypothetical case with low sNfL levels. Such a comparison could further elucidate the precise weight neurologists attribute to this biomarker in guiding therapeutic decisions, providing a clearer understanding of whether a low sNfL would indeed lead to a different recommendation. Finally, our study was conducted before the release of the 2024 revision of the McDonald MS diagnostic criteria, which introduced a major change by reclassifying RIS as MS for individuals meeting specific criteria, such as the presence of dissemination in space and time and OCBs in the cerebrospinal fluid.40,41 This update is expected to significantly influence clinical decision-making, as it expands the definition of MS to include a subset of individuals previously categorized as having RIS. Consequently, neurologists may now be more likely to initiate treatment in such cases. As a result, our findings, which are based on the prior classification system, may not fully reflect current therapeutic strategies, highlighting the need for future studies to explore the impact of these updated criteria on clinical practice. Despite these limitations, a strength of our study is its representation of neurologists managing MS and other demyelinating disorders in Spain, with our sample comprising approximately 58% of this group, according to SEN data.42
Conclusions
Despite the recognized association between elevated sNfL levels and an increased risk of conversion to MS, more than 50% of neurologists made decisions misaligned with emerging evidence by not recommending the initiation of DMT in a simulated case of RIS with high sNfL levels. Neurologists not exclusively dedicated to MS care and those with a limited perception of sNfL benefits were more likely to defer treatment initiation, highlighting potential gaps in knowledge and practice patterns. Efforts led by scientific societies are essential to develop evidence-based guidelines and educational initiatives that enhance neurologists’ understanding of sNfL testing and other risk factors relevant to caring for patients with RIS.
