Coordination between neuroradiology and neurology departments in the care of patients with multiple sclerosis: Recommendations for optimization

Magnetic resonance imaging (MRI) plays an essential role in the diagnosis and follow-up of patients with multiple sclerosis (MS).¹ The diagnostic criteria for MS (McDonald, 2017) position MRI as a support to the clinical findings when already in the early stages of the disease, emphasising its contribution to early diagnosis and facilitating the start of disease-modifying treatment (DMT). Once the DMT has started, MRI in combination with clinical assessments also helps identify the effectiveness of the DMT and predict the risk of disease progression.

The importance of neuroradiology and neurology departments throughout the journey of patients with MS is clear. However, there are limitations in both the functioning of the departments separately and in coordination between the two which need to be overcome in clinical practice. On the one hand, different acquisition protocols and MRI equipment are used in neuroradiology departments, which, combined with the lack of uniform criteria among neuroradiologists, makes it difficult to compare results. The lack of specific knowledge about clinical aspects of MS on the part of neuroradiologists also limits the interpretation of these results. On the other hand, neurology departments can be guilty of requesting MRI scans without providing either the information the neuroradiologist requires to justify the test, the patient's clinical problems or the aspects they want the neuroradiologist to resolve. There is certainly room for improvement in the interaction between the neurology and neuroradiology departments, both in terms of coordination among the healthcare professionals and general communication between the two departments.²

With the aim of responding to these needs, a nine-recommendation proposal has recently been presented to optimise coordination between the two departments in the care of patients with MS.³ The aim of these recommendations, drawn up by a panel of 17 neurologists and neuroradiologists from eight Spanish hospitals, is to improve aspects detected as deficient in clinical practice, whether in one or the other department, or in the interaction between the two. With this article, we present our opinion on the most important aspects of these recommendations.

First of all, we consider it essential to standardise neurology department requests for MRI scans. The more complete and precise the neurologist's request for the tests to be performed by the neuroradiologist, the more complete and precise the subsequent radiological report will be. The minimum information neurologists should include in the request for MRI tests is shown in Table 1. Having clinical information describing the patient's context and the reason for referral to neuroradiology will positively impact the performing and interpretation of the MRI. The neuroradiologist should therefore be sent this information, but if not, should request it. Also helpful would be if the neuroradiology department standardised the structure of the radiology reports. Previous studies have shown that both clinicians and radiologists consider structured reports to have better content and clarity than unstructured reports.⁴ Having a structured report benefits the inclusion of information the neurologist needs for decision making⁵ and the standardised recording of data, in turn facilitating the use of these data in Real-World Evidence studies. Table 2 shows the recommendations on the structure that radiological reports should follow, both for diagnosis and for patient follow-up. Another aspect that requires standardisation is the scheduling of MRI scans. This could be done by implementing a system for scheduling routine follow-up tests in advance, reducing the number of unscheduled examinations, prioritising urgent tests, and reminding patients of upcoming appointments.

Table 1.

Minimum information to include in the request for magnetic resonance imaging tests.

Diagnosis	Follow-up
• Date of clinically isolated syndrome (CIS)	• Clinical status: symptoms, stability
• Clinical symptoms and how they evolved over time	• Diagnosis: MS phenotype
• Suspected diagnosis	• Reason for performing the test: suspected flare-up, change in treatment, routine control in a stable patient, risk or suspicion of progressive multifocal leukoencephalopathy
• Corticosteroid therapy (start and end dates)	• Clinical signs of disease activity and/or progression
• Parts of the central nervous system that should be examined (with clinical justification)	• Degree of disability
• Relevant comorbidities	• Current treatment and start date (corticosteroids, DMT)
• Existence of severe renal failure	• Parts of the central nervous system that should be examined (with clinical justification)
• Known allergy to contrast media	• Relevant comorbidities
• Special needs (degree of autonomy, claustrophobia, pregnancy and lactation, presence of devices, etc.)	• Existence of severe renal failure
• Priority	• Known allergy to contrast media
	• Special needs (degree of autonomy, claustrophobia, pregnancy and lactation, presence of devices, etc.)
	• Priority

Note: clinical priority could be categorised into four types: immediate urgent (on the same day), deferrable urgent (maximum of 72 h), preferential (maximum of 15 days) and ordinary (before the next appointment with neurology). Table reproduced with permission of the authors.³

Table 2.

Radiological report proposal for diagnosis and follow-up.

Diagnosis	Follow-up
Study technique: brain and/or spinal cord magnetic resonance imaging (MRI) (type of sequence) with or without gadolinium (dose)	Study technique: brain and/or spinal cord MRI (type of sequence) with or without gadolinium (dose)
Date of the previous study for comparison (if applicable)	Comparative analysis with previous studies:
Comparative study with previous MRI to indicate the number of new/increased lesions on T2	• Date of previous study
Detailed description of the findings:	• Indicate whether or not the comparative analysis is technically possible
• Overall assessment of the number of lesions on T2 and T2-FLAIR (minimal, low, moderate, high, very high)	• Number of lesions on T2
• Presence of lesions:	• Number and topography of new or enlarged lesions on T2
• Periventricular: yes/no	• Number and topography of gadolinium-enhancing lesions
• Juxtacortical: yes/no	• Presence of leptomeningeal enhancement (T2-FLAIR)
• Brainstem: yes/no	• Detection of focal lesions with severe tissue damage (hypointense on T1 SE or T2-FLAIR) Progression in number and size
• Cerebellum: yes/no	• Presence and progression of brain atrophy
• Corpus callosum: yes/no	• Relevant incidental findings
• Spinal cord: yes/no
• Description of the size and shape of the lesions
• Number and topography of gadolinium-enhancing lesions
• Presence of leptomeningeal enhancement (T2-FLAIR)
• Detection of focal demyelinating lesions with severe tissue damage (hypointense on T1 SE or T2-FLAIR)
• Presence of brain atrophy: no, mild, moderate, severe
• Relevant incidental findings
• Compliance with McDonald radiological criteria:
– Dissemination in time: yes/no
– Dissemination in space: yes/no
• Interpretation of findings and differential diagnosis

Table reproduced with permission of the authors.³

Secondly, we believe protocols should be designed for MRI tests, using the different published national and international guidelines as reference. These need to be written protocols, so that all staff involved have easy access to them. They need to specify when and how to perform the MRI scans, including the essential sequences in specific brain and spinal cord MRI studies for the initial diagnosis and follow-up.

Thirdly, multidisciplinary committees and coordination sessions between healthcare professionals from the two departments should be set up. Multidisciplinary committees would enable an overall approach to the patient from different specialist areas, which would have a favourable impact on patient management. Table 3 shows the recommendations on the members who should form part of these committees, their functions and responsibilities, and the aspects that need to be defined within each committee. The purpose of the coordination sessions is to make up for the present lack of a management structure to formalise the coordination of the two departments. These sessions would promote the sharing of information between the different healthcare professionals and help detect any deficiencies in the implementation of the rest of the recommendations (Table 4).

Table 3.

Recommendations on the multidisciplinary committee.

Composition

Stable team: neurologist, neuroradiologist, registered nurse

Support team: neurorehabilitator, speech therapist, neurophysiologist, urologist, neuro-ophthalmologist, psychologist, psychiatrist and pharmacologist

Roles and responsibilities

Prior identification of complex cases

Case presentation

Discussion of diagnosis

Discussion of therapeutic management

Minimisation of risks

Discussion on patient involvement and shared decisions

Aspects to be established after being made official within each centre

Assumed competences, adapted to the needs of each centre

Composition and roles of its members

Coordination of the meetings, suggesting that they be managed by both neurologists and neuroradiologists of reference

Making decisions official through the creation of official clinical committees

Meetings at predetermined intervals, at least monthly, even if they are of short duration. Contents will vary depending on the clinical workload, and it may be cancelled a certain time in advance if it is deemed not necessary.

Table reproduced with permission of the authors.³.

Table 4.

Recommendations for carrying out coordination sessions between departments.

Regularity of at least two meetings a year

Outline of the subject areas to be discussed in the sessions some time in advance to allow prior reflection by those involved

Making the sessions official with drafting and signing of minutes

Communication and channelling of decisions involving significant changes to the management of the centre.

Healthcare professionals involved in the sessions: neuroradiologists, neurologists and registered nurses specialising in neurology. Also involve specialised radiology technicians and other specialists according to the areas to be discussed in the sessions.

Table reproduced with permission of the authors.³.

Last of all, the creation of formal communication channels between the healthcare professionals of the two departments would help achieve more efficient communication. Communication could be carried out using registered alert systems or a mobile application developed for this purpose, depending on the resources of each centre. This would prevent important information for neuroradiologists and the other healthcare professionals involved in the care of the MS patient from being lost or not arriving in time due to the use of informal communication channels.

Future in the acquisition and analysis of magnetic resonance imaging studies

In clinical practice, radiologists interpret MRI scans by visual analysis of the images obtained. However, this may not be the most appropriate strategy, particularly if the aim is to use MRI as a true biomarker in MS, as visual analysis of certain measurements (lesion volume, degree of atrophy, active lesions on T2) is subject to great variability and inadequate sensitivity, which limits their usefulness, especially if they are needed to create prognostic models for the disease and to predict and monitor the efficacy of DMT.

Hence the need to implement automatic tools, many based on artificial intelligence (AI), to aid the objective and robust quantification of MRI measurements important for patient management.

Research in the field of AI applied to the diagnosis, prognosis and monitoring of MS with MRI provides tangible results, with the development and implementation of automatic programs for clinical use, which enable precise and reproducible quantitative analyses to be obtained, such as automatic detection of active lesions on T2, lesion volume, overall and regional brain volume and tissue properties (T1 and T2 relaxometry).

Synthetic MRI acquisition techniques are also being developed (MRI fingerprinting, Synthetic MRI), which generate images in less time, and which, due to the design for obtaining them, are reproducible regardless of the MRI equipment used. This should facilitate the harmonisation of MRI studies, an essential factor for obtaining reproducible quantitative measurements, which are particularly necessary in longitudinal studies.

With the progressive implementation in clinical practice of standardised and harmonised sequences, and with the application of automatic quantification programs, in the very near future, radiologists are going to have to adapt the way they obtain and analyse diagnostic and follow-up MRI studies in MS. These new strategies are certain to give more value to MRI in patient diagnosis, prognosis and follow-up and to the work of the radiologist who, in addition to continuing to visually interpret MRI images based on their knowledge and experience, will also have to validate and interpret the quantitative measurements obtained. The ultimate goal is to provide the neurologist with the necessary information for optimal and individualised management of patients with MS.

Funding

Support in writing and editing the manuscript was funded by Novartis Farmacéutica S.A.

Conflicts of interest

A.R. is or has been involved in scientific advisory boards for Novartis, Sanofi-Genzyme, SyntheticMR, Bayer, Roche, Biogen, Tensor Medical, Icometrix and OLEA Medical, and has received fees as a speaker from Sanofi-Genzyme, Merck-Serono, Teva, Novartis, Roche and Biogen.

S.L. has received fees for conducting training sessions and for consultancy from Novartis, Merck, Teva, Sanofi, Biogen, Roche and Genzyme.

Acknowledgements

We thank Laura Prieto del Val, from Dynamic Science, for her support in writing the article.