Ultrahigh field MRI in clinical neuroimmunology: a potential contribution to improved diagnostics and personalised disease management
Conventional magnetic resonance imaging (MRI) at 1.5 Tesla (T) is limited by modest spatial resolution and signal-to-noise ratio (SNR), impeding the identification and classification of inflammatory central nervous system changes in current clinical practice. Gaining from enhanced susceptibility effects and improved SNR, ultrahigh field MRI at 7 T depicts inflammatory brain lesions in great detail. This review summarises recent reports on 7 T MRI in neuroinflammatory diseases and addresses the question as to whether ultrahigh field MRI may eventually improve clinical decision-making and personalised disease management.
Keywords7 Tesla Ultrahigh field MRI Multiple sclerosis Neuromyelitis optica Susac syndrome Neuroimmunology Central vein sign Cortical lesions Predictive, Preventive and Personalised Medicine
clinically isolated syndrome
central nervous system
double inversion recovery
deoxyribonucleic acid double-strand breaks
fluid attenuated inversion recovery
fast low angle shot
magnetisation prepared rapid acquisition gradient echo
MPRAGE with multiple echoes
magnetic resonance imaging
optical coherence tomography
susceptibility weighted fluid attenuated inversion recovery
Magnetic resonance imaging (MRI) revolutionised clinical neuroimmunology since brain MRI depicted multiple sclerosis (MS) lesions already in early technical developmental stages at 0.1 Tesla (T) . During the past decade, MRI became a crucial tool to diagnose and monitor inflammatory central nervous system (CNS) alterations . Nonetheless, today’s physicians are faced with a key issue in clinical neurology: many distinct CNS diseases are characterised by nearly identically appearing white matter changes and brain lesions that are often unspecific in appearance, limiting the diagnostic value of conventional MRI.
Ultrahigh field (UHF) MRI at 7 T benefits from increased signal-to-noise ratio (SNR) and enhanced spatial resolution as good as 100 μm . Future studies will show whether these 7 T MRI advantages indeed improve diagnosis and our understanding of the underlying pathophysiology in inflammatory CNS diseases. Following the recommendations of the "EPMA White Paper" , this review summarises technical opportunities, challenges, and findings of recent clinical 7 T MRI studies on multiple sclerosis, neuromyelitis optica, and Susac syndrome.
Technical improvements and limitations
However, there are still few practical and technical considerations to be made when applying UHF MRI: Some patients may be excluded from an examination at 7 T due to an increased number of contraindications at UHF as compared to lower field strengths, such as tattoos, dental implants, metallic intrauterine devices, stents, surgical clips, and piercings. These may also include otherwise "MRI-safe" implants such as pacemakers or orthopaedic replacements.
Furthermore, there are technical challenges that deserve attention: Increased magnetic field inhomogeneity may impact post-processing procedures despite excellent gray to white matter contrast. Radiofrequency (RF) power deposition constitutes another practical challenge since it scales superlinearily with the magnetic field strength. Local RF coils that offer improved transmission efficiency versus large volume coils can be instrumental to offset this challenge [9, 10, 11].
When considering these constraints, UHF MRI is believed to be safe and it is well tolerated by the vast majority of patients [12, 13]. Nonetheless, temporary adverse events were reported during 7 T at higher frequency compared to 1.5 T MRI . In addition, 5 % of all subjects or patients reported vertigo during UHF MR exams . During scan with magnetic field gradients being rapidly switched, visual disturbances or temporary muscle contractions may occur [15, 16, 17]. Deteriorating vital signs or long-term effects have—to the best of our knowledge—not been described during or after 7 T MRI investigations [13, 18, 19], but the relevance of preliminary in vitro studies on potential deoxyribonucleic acid (DNA) damage caused by a static magnetic field of 1.5 T or by rapidly changing magnetic fields is still subject to discussion [20, 21]. A recent analysis of DNA double-strand breaks (DSB) in human peripheral blood mononuclear cells after exposure to 7 T did not show a significant increase in DSB levels compared to the unexposed control group .
Multiple sclerosis is an inflammatory and neurodegenerative autoimmune CNS disorder affecting white as well as gray matter of the brain and spinal cord [22, 23, 24]. The disease is characterised by a wide range of symptoms and a large heterogeneity in clinical presentation. Besides neurological impairment in visual, pyramidal, cerebellar, sensory, and vegetative functional systems, more global symptoms of CNS dysfunction such as fatigue and cognitive dysfunction may occur that negatively impact patients’ quality of life [23, 25, 26, 27, 28, 29, 30]. MRI and more recently optical coherence tomography (OCT) have emerged as valuable imaging tools for contributing to diagnosis, differential diagnosis, and disease monitoring [31, 32, 33, 34, 35, 36, 37, 38, 39]. These imaging techniques have shown that beyond focal lesions, diffuse and widespread tissue damage occurs in both the gray and the white matter already in early disease stages [40, 41, 42, 43] and more pronounced in progressive disease . However, diagnosis and treatment decisions in clinical routine are still widely based on the detection of focal cerebral white matter lesions hyperintense on T2 weighted (T2w) or fluid attenuated inversion recovery (FLAIR) images. An accurate diagnosis of MS remains challenging given the insufficient specificity of focal white matter lesions [45, 46]. In this regard, UHF MRI improves both the detection and morphological description of MS lesions and may thus be used in the future to distinguish MS from lesions of other origins and to improve our understanding of the disease. This is of high clinical relevance as the broadening MS treatment landscape will pave the way for an individualised and tailored MS therapy . However, with the increasing number of available efficacious immunosuppressive and immunomodulatory drugs for MS, a correct and timely diagnosis is a prerequisite for personalised medicine that weighs benefits and risks of these drugs in every individual patient [24, 48, 49, 50, 51, 52, 53, 54].
Cortical gray matter lesions
The detection of cortical lesions is greatly improved by 7 T MRI . Gray matter pathology accumulates during disease progression and may affect major areas of the cortex in long-standing multiple sclerosis [56, 57, 58]. Recent studies revealed that cortical lesions are associated with disease progression, disability, and cognitive dysfunction [59, 60, 61]. In conventional MRI, the vast majority of cortical lesions remain undetected even when applying double inversion recovery (DIR) techniques at 1.5 T [62, 63]. UHF MRI at 7 T improves the detection of cortical lesions and depicts up to 48 % of all cortical lesions revealed by ex vivo immunohistochemical staining for myelin . These results were confirmed by several in vivo studies. Magnetisation transfer imaging at 7 T was reported to detect roughly 25 % more cortical lesions than 3 T DIR in a recent study . Furthermore, 7 T 3D FLAIR is highly sensitive in detecting cortical lesions and detects 89 % more lesions than 7 T 3D DIR . A multi-contrast 3 T versus 7 T comparative study reported 7 T MRI to detect up to 238 % more cortical lesions than 3 T . In addition, it was shown that 7 T T1 weighted magnetisation prepared rapid acquisition gradient echo (MPRAGE) imaging increases the detection rate of cortical lesions by twofold in comparison to 1.5 T MPRAGE [65, 68].
In sum, there is increasing evidence that 7 T MRI detects significantly more (subpial) cortical lesions than 3 T, but the detection of some type III lesions still remains challenging .
Improved depiction of white matter lesions
Persisting T1 weighted (T1w) hypointense lesions—namely black holes—contribute to disability in MS in addition to cortical lesions [78, 79]. At UHF strength—and in contrast to conventional MRI at 1.5 T—virtually, every T2w hyperintense lesion is visible as a distinct hypointense plaque on 7 T T1w MPRAGE images as shown by our group and others [68, 80]. Contrarily, 1.5 T T1w MPRAGE delineated only 68 to 78 % of T2w lesions in the same study . Moreover, 7 T T1w MPRAGE is even more sensitive in detecting MS lesions than 1.5 T T2w (728 versus 545 lesions)  or 3 T FLAIR imaging (1043 versus 812 lesions) .
White matter lesion morphology
Gaining from increased susceptibility effects and spatial resolution, T2*w imaging at 7 T delineates distinct morphological features of MS lesions. Most importantly, a very small vein can be displayed within the center of the MS lesion on T2*w images, and the lesion often follows the course of the vessel (Fig. 3b) [71, 73, 74, 81, 82]. This feature is not only detectable in relapsing-remitting MS but also observable in primary progressive MS . In addition, a proportion of MS lesions is characterised by a T2*w hypointense rim surrounding the lesion (Fig. 3b) [71, 73, 74]. A comparative 7 T and histopathological study found that these rims correspond to iron-rich CD68-positive cells of the macrophage lineage . Hence, a positive rather thick rim-like phase shift is detectable around these lesions at 7 T . Contrarily, rather thin rim-like phase shifts around MS lesions without major T2*w hypointensity in these areas were associated with blood-brain barrier breakdown and inflammatory activity . In general, MRI phase imaging can provide additional information on the tissue microstructure that is not encoded in the magnitude of the MR signal. Thus, MRI phase imaging at 7 T depicts white matter lesions prior to conventional T2w imaging as revealed by a case series . Finally, susceptibility changes indicative of iron deposition within the center of a proportion of MS lesions can be found even in the earliest MS disease stages . The origin of these iron deposits, however, still remains unclear and highly speculative. Leakages of haemoglobin through a leaky blood vessel or dying iron-rich oligodendrocytes releasing iron into the extracellular matrix are only two hypotheses among many others [87, 88, 89, 90, 91].
Differential diagnosis by 7 T MRI
The detailed description of the lesion morphology facilitates the distinction of MS lesions versus brain lesions of other origin [92, 93, 94, 95]. A first study on 28 MS patients and 17 subjects with non-symptomatic lesions presumably caused by small vessel disease found that the “central vein sign” differentiates MS patients from these controls by using a central vein cutoff of 40 % . The same cutoff was reported to be beneficial in predicting MS conversion of clinically isolated syndrome (CIS) patients . In detail, each of 13 CIS patients with a positive central vein sign (>40 %) at baseline included in a prospective study developed MS, and all CIS patients (n = 9) with a negative central vein sign (<40 %) at baseline were ultimately diagnosed as not having MS . The median follow-up time in this study was 26 months (range, 4–37 months) . Although these initial results must be confirmed in a larger dataset with longer follow-up, this study illustrates the potential predictive capability of 7 T MRI.
Venous abnormalities in MS
The controversy on cerebrospinal venous insufficiency in MS [97, 98, 99, 100] revitalised a discussion on vascular abnormalities within MS lesions that were first described by Dawson et al. in early 1916 . Today, 7 T T2*w imaging can depict very small brain veins in vivo (Fig. 1) [71, 74, 82, 102]. The venous density is reduced in MS compared to healthy controls presumably as a consequence of hypometabolism, gliosis, and vascular damage . This reduction in (periventricular) venous density is already detectable in the earliest MS disease stages and patients with CIS . Furthermore, shrinkage of intra-lesional compared to extra-lesional veins was reported recently . Although the degree of intra-lesional venous shrinkage was smaller in another study , intra-lesional venous shrinking is a potential in vivo imaging marker of inflammation since it is hypothesised to be the consequence of thickened vein walls caused by inflammation leading to obstruction and reduced blood flow .
Structural damage and atrophy in MS
High-resolution 7 T T2*w imaging visualises strongly myelinated aligned structures such as the optic radiation (OR, Fig. 1). Furthermore, very small lesions can be displayed within the OR on 7 T images . The lesion volume affecting the optic radiation was reported to be associated with OR atrophy and retinal thinning as revealed by OCT . This association between OR damage and retinal atrophy may reflect retrograde transsynaptic degeneration, but independent mechanisms may play a role, too.
Quantifying the total volume of brain tissue and volumes of gray or white matter is impeded at 7 T by the local field inhomogeneity. This limitation may be overcome by a T1w MPRAGE sequence with two inversion pulses, e.g., MPRAGE with multiple echoes (MP2RAGE), a technique recently recommended for generating a homogenised T1w image free of proton density or T2w contrast . Indeed, the MP2RAGE approach yielded sufficient cortical surface reconstructions  and voxel-based morphometry (VBM) analyses estimating gray matter volume can be of good quality regarding superior cortical areas [109, 110].
In summary, these 7 T MRI imaging characteristics may be used in the future to improve the differentiation between NMO and MS, which is highly relevant for the individual patient since therapeutic approaches in MS and NMO differ considerably [123, 124, 125, 126]. The central vein sign is a potential future biomarker to distinguish MS from NMO patients. It is noteworthy that the sensitivity in detecting venous structures on 7 T gradient echo images largely relates to the imaging sequence, the post-processing, and the acquisition parameters such as the spatial resolution, flip angle, or echo time . Thus, a “central vein cutoff value” for the differentiation of MS versus NMO lesions may vary in relationship to these parameters. An important limitation of current studies on NMO and 7 T MRI is the absence of spinal cord imaging at 7 T and small sample sizes [92, 93].
An increasing number of 7 T MRI studies described unique features of MS lesions—most importantly, the central vein sign—that may be used in the future to differentiate MS lesions from brain lesions of other origin. Today there is, however, only limited evidence on these findings since many 7 T MRI studies comprise small patient cohorts or are hampered by a cross-sectional design. In addition, not all differential diagnoses of MS have been investigated at 7 T yet. From a more technical and practical perspective, technical limitations such as magnetic field inhomogeneity and economic as well as safety concerns have to be solved before widely applying 7 T in clinical practice. By then, we should aim to apply knowledge from these preliminary 7 T MRI studies to 3 T MRI platforms that are available for clinical imaging. Recently, different approaches to display venous structures within MS lesions at 3 T were published: FLAIR* combines FLAIR and T2*w images [145, 146], whereas susceptibility weighted FLAIR (sFLAIR) combines SWI and FLAIR images [102, 147]. In addition, optimised 3 T T2*w contrast may improve vessel detection at 3 T .
In the emerging field of personalised medicine, 7 T MRI may be used in patients with suspected neuroinflammatory disease such as MS, but conflicting clinical or paraclinical findings to support making the correct diagnosis early. Today, this should be done within the framework of clinical trials.
This work was supported by the German Research Foundation (DFG Exc 257 to FP) and by the German Ministry of Education and Research (Competence Network Multiple Sclerosis KKNMS to FP and JW) and a research grant from the Guthy Jackson Charitable Foundation/National Multiple Sclerosis Society of the USA.
- 5.Haacke E, Brown R, Thompson M, Venkatesan R. Magnetic resonance imaging: physical principles and sequence design. John Wiley & Sons (USA). 1999. p. 378.Google Scholar
- 14.Theysohn J. Subjective acceptance of 7T: initial experience in the first 210 subjects. Proc Intl Soc Mag Reson Med. 2008;16:1049.Google Scholar
- 16.Fatahi M, Reddig A, Friebe B, Reinhold D, Speck O. Analysis of DNA double-strand breaks in human peripheral blood mononuclear cells after exposure to 7T MRI. ISMRM Toronto, Canada. 2015;2015:0300.Google Scholar
- 31.Scheel M, Finke C, Oberwahrenbrock T, Freing A, Pech L, Schlichting J, et al. Retinal nerve fibre layer thickness correlates with brain white matter damage in multiple sclerosis: a combined optical coherence tomography and diffusion tensor imaging study. Mult Scler J. 2014;20:190–7.CrossRefGoogle Scholar
- 43.Bellmann-Strobl J, Stiepani H, Wuerfel J, Bohner G, Paul F, Warmuth C, et al. MR spectroscopy (MRS) and magnetisation transfer imaging (MTI), lesion load and clinical scores in early relapsing remitting multiple sclerosis: a combined cross-sectional and longitudinal study. Eur Radiol. 2009;19:2066–74.PubMedCrossRefGoogle Scholar
- 77.Harrison DM, Oh J, Roy S, Wood ET, Whetstone A, Seigo MA, et al. Thalamic lesions in multiple sclerosis by 7T MRI: clinical implications and relationship to cortical pathology. Mult Scler J. 2015. doi: 10.117/1352458514558134.
- 79.Sailer M, Losseff NA, Wang L, Gawne-Cain ML, Thompson AJ, Miller DH. T1 lesion load and cerebral atrophy as a marker for clinical progression in patients with multiple sclerosis. A prospective 18 months follow-up study. Eur J Neurol Off J Eur Fed Neurol Soc. 2001;8:37–42.Google Scholar
- 90.Bozin I, Ge Y, Kuchling J, Dusek P, Chawla S, Harms L, et al. Magnetic resonance phase alterations in multiple sclerosis patients with short and long disease duration. PLoS One. 2015;10(7):e0128386.Google Scholar
- 96.Mistry N, Dixon J, Tallantyre E, Tench C, Abdel-Fahim R, Jaspan T, et al. Central veins in brain lesions visualized with high-field magnetic resonance imaging: a pathologically specific diagnostic biomarker for inflammatory demyelination in the brain. JAMA Neurol. 2013;70:623–8.PubMedCrossRefGoogle Scholar
- 101.Dawson J. The histology of disseminated sclerosis. Trans Roy Soc Edin. 1916;50:517-740.Google Scholar
- 116.Jarius S, Paul F, Fechner K, Ruprecht K, Kleiter I, Franciotta D, et al. Aquaporin-4 antibody testing: direct comparison of M1-AQP4-DNA-transfected cells with leaky scanning versus M23-AQP4-DNA-transfected cells as antigenic substrate. J Neuroinflammation. 2014;11:129.PubMedCentralPubMedCrossRefGoogle Scholar
- 139.Ringelstein M, Albrecht P, Kleffner I, Bühn B, Harmel J, Müller A, et al. Retinal pathology in Susac syndrome detected by spectral-domain optical coherence tomography. Neurology. 2015;85(7):610-8.Google Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.