Книги по МРТ КТ на английском языке / MR Imaging in White Matter Diseases of the Brain and Spinal Cord - K Sartor Massimo Filippi Nicola De Stefano Vincent Dou

glia (Holtmannspötter et al. 2003), may also help to differentiate ADEM from MS. In individual cases, however, this may still be difficult, at least in the early stages (Dale et al. 2000; Tenenbaum et al. 2002). Contrast enhancement needs not to occur absolutely simultaneously in all lesions caused by ADEM.At the same time, contrast enhancement of both lesions in the brain and spinal cord is not uncommon for MS, as mentioned previously (Thorpe et al. 1996a). With more advanced MS, however, there is almost always a larger proportion of non-enhancing lesions, with some appearing as T1 hypointensities or socalled black holes, that strongly argue against ADEM. Nevertheless, in many instances clinical and imaging follow-up is necessary to make a final diagnosis from the absence of further disease activity, although it is known that a second bout of ADEM may also occur (Tenenbaum et al. 2002).

18.2.2 Neuromyelitis Optica

Neuromyelitis optica (NMO), or Devic’s syndrome, is a severe form of idiopathic inflammatory demyelinating disease and is increasingly considered as a separate entity (Weinshenker 2003). The specific features include its topographic restriction to the optic nerve and spinal cord, and a greater attack severity than MS (Wingerchuk et al. 1999). This is also reflected by much more extensive spinal cord lesions on MRI than those seen in MS (Fig. 18.6). These lesions usually involve large segments of the cord with diffuse swelling and extensive contrast enhancement. The necrotising character of this type of demyelination is also evidenced in the post-acute phase by the frequent occurrence of cystic areas within the cord together with severe focal or diffuse atrophy as

indicators of extensive tissue destruction (Figs. 18.6, 18.7).

In terms of disease evolution, it is important to note that NMO can have both a monophasic and a relapsing course. In a review of 70 patients with NMO at the Mayo Clinic, patients with a monophasic course usually presented with rapidly sequential index events with moderate recovery. Two thirds of patients, however, presented with an extended interval between index events followed within 3 years by clusters of severe relapses isolated to the optic nerves and spinal cord. In these patients severe disability developed in a stepwise manner; thus, a prolonged interval between damage to the spinal cord and optic nerves does not rule against a diagnosis of NMO (Wingerchuk et al. 1999).

Absence of brain MRI changes is a hallmark finding of NMO and part of the diagnostic criteria for this disorder [30]. Unaffected brain white matter has also been shown in one study using magnetisation

transfer (Filippi et al.1999).Non-specific abnormalities, however, do not preclude a diagnosis of NMO.

18.3

Transverse Myelitis

Acute transverse myelitis can be caused by a number of bacterial, viral, fungal or parasitic infections, by connective tissue diseases, such as sarcoidosis, Behçet’s disease, Sjögren’s syndrome, systemic lupus erythematosus, antiphospholipid syndrome, and mixed connective tissue disease, and can be the presenting feature of NMO and less likely also of MS (Transverse Myelitis Consortium Working Group 2002). In rare cases no aetiology can be defined, and it may therefore be suggested to group such disorders under the umbrella of the idiopathic inflammatory demyelinating disorders as well. Some

reports and our own experience suggest a similarity of spinal cord abnormalities with those observed in NMO. This has been especially true for patients with relapsing acute transverse myelitis (Fig. 7; T. Seifert et al., submitted). Support for this hypothesis comes from the fact that some of the patients with relapsing transverse myelitis subsequently also developed optic neuritis but otherwise showed a normal or at least non-specific MRI of the brain (Katz and Ropper 2000).Marked inflammatory CSF changes and the absence of oligoclonal bands are also more commonly found in this type of spinal cord disorder than with MS (Transverse Myelitis Consortium Working Group 2002).

18.3.1

MRI Evaluation of Spinal Cord Damage in Relation to Function

Spinal cord lesions do correlate with spinal cord symptoms as shown in cross-sectional and followup studies (Thorpe et al. 1996a; Lycklama et al. 1998); however, the relationship between a patient’s symptoms and spinal cord findings frequently remains limited. The reasons for this are manifold. As in the brain, conventional MRI cannot serve to grade the severity of axonal destruction within lesions or

display other factors which may impact on functioning also at non-morphological levels (Caramia et al. 2004). In addition, it is even more difficult to quantify the total extent of damage in the spinal cord than in the brain. This has also been seen in histopathological comparisons where the actual spinal cord damage was frequently much more pronounced than would have been expected from imaging alone (Bergers et al. 2002b). Finally, quantification of tissue loss, i.e. of spinal cord atrophy, is also not satisfactory by visual analysis. For this reason it has been attempted to use various MRI metrics for the global description of spinal cord damage from demyelinating diseases, especially in MS.

Much work has been done on the measurement and longitudinal follow-up of spinal cord atrophy (Zivadinov and Bakshi 2004). Measuring the crosssectional size of the spinal cord, however, always has to be performed at the exactly same cervical level, and various technical problems may cause significant variance in the measurements. With the appropriate techniques,however,longitudinal follow-up can show small but statistically significant decreases in spinal cord area (Stevenson et al. 1998) and a relatively strong relationship between degree of spinal cord atrophy and EDSS values has been reported (Losseff et al. 1996). Especially in primary progressive MS, atrophy measurements may be more important in

relation to function than those of lesion volumes (Ukkonenen et al. 2003).

In the brain,magnetisation transfer imaging (MTI) has shown the capability of detecting white matter abnormalities not accessible with conventional MRI (Filippi 2003). It was therefore speculated that this technique should also allow delineation of diffuse changes of the spinal cord from and independent of focal abnormalities. First studies in relatively small cohorts of patients confirmed that the cervical cord of MS patients had lower MTR values than those of controls (Inglese et al. 2002). Histogram analyses, however, showed that compared with control subjects patients with relapsing, -remitting MS had similar cervical cord MTR histogram-derived measures, whereas those with primary progressive and secondary progressive MS had abnormal MTR histograms (Filippi et al. 2000). More importantly, the cervical cord MTR histogram parameters were independent predictors of loco-motor disability. A further study confirmed that MTR metrics do not differ between primary and secondary progressive MS, and they also appear to be independent of the cerebral lesion load (Rovaris et al. 2000, 2001).

With technical advances it has also become possible to apply diffusion-weighted MRI (DWI) to the spinal cord (Bammer et al. 2002; Clark and Werring 2002). A preliminary study assessed water diffusion in seven cord lesions of three MS patients with loco-motor disability and found increased diffusivity compared with healthy volunteers (Weeler-Kingshott et al. 2002); however, for the individual patient no useful contribution can yet be expected from DWI, and especially in the early phase of MS only very subtle changes appear detectable (Mezzapesa et al. 2004). The development of techniques which allow to incorporate tractography into spinal cord MRI may be an important next step. Currently, DWI of the spinal cord still faces significant technical challenges that can limit the reproducibility of quantitative measurements. Specifically, motion of the patient and the cord itself, and the inhomogeneous magnetic environment within and around the spinal column, can cause deleterious artefacts.

Magnetization transfer imaging and DWI promise to provide complementary biophysical information about microstructural changes of the spinal cord. While MT is more targeted to the chemical–struc- tural environment of the myelin sheath, DWI probes for potential changes in the micro-environment that may impair or facilitate the mobility of protons. In combination or individually these two biophysical

mechanisms can help to provide insight and improve contrast beyond conventional MRI methods.

Recent technological developments in MR hardware, especially improved radio-frequency (RF) coils and higher field strengths, have furthered MRI’s capability to image the spinal cord at highest spatial resolution. At numerous centres, neuroimaging is currently migrating from 1.5 to 3 T or even higher, trying to capitalize on the increased signal strength afforded by the stronger magnetic field. Here, especially for the spinal cord, the excessive signal-to- noise ratio can be invested in greatly improved spatial resolution. In combination with dedicated spine arrays, this strategy allows imaging of the entire spinal cord at once and at a higher spatial resolution, which should definitely impact diagnostic sensitivity and also specificity.

One issue associated with higher field strength that has to be kept in mind is the increased energy depositioninthepatient’sbody,especiallywhentransmitting with the large-body-volume coil. This can have major implications for pulse sequences with relatively high RF duty cycles, such as MT sequences. Fortunately, if only the spine needs to be imaged, its unique location in the body allows for the use of transmit/receive coils which demonstrate much fewer problems with energy deposition. Thus far, only little has been reported on the overall benefits of using higher magnetic fields in demyelinating diseases, especially when focused on the spinal cord. One should also be aware that the T1 relaxation times of semi-solid tissues changes with field strength, while CSF remains almost unchanged; therefore, certain adaptations in pulse sequence are mandatory to maintain contrast properties, and the altered relaxation properties should be kept in mind when interpreting/comparing studies performed at different field strengths.

18.4 Conclusion

Magnetic resonance imaging of the spinal cord provides characteristic patterns of abnormalities in different demyelinating diseases. It has to be recognised, however, that other aetiologies may mimic these abnormalities and have to be considered in the differential diagnosis as described elsewhere (Fazekas and Kapeller 1999; Bot et al. 2002). On the other hand, the recognition of spinal cord abnormalities as a consequence of demyelinating diseases makes an important contribution to their diagnosis. Future de-

Demyelinating Diseases of the Spinal Cord

velopments also promise to generate insights beyond these diagnostic contributions in terms of information relevant for function and repair.

References

Bammer R, Augustin M, Prokesch RW et al (2002) Diffusionweighted imaging of the spinal cord: interleaved echoplanar imaging is superior to fast spin-echo. J Magn Reson Imaging 15:364–373

Bergers E, Bot JC, de Groot CJ et al (2002a) Axonal damage in the spinal cord of MS patients occurs largely independent of T2 MRI lesions. Neurology 59:1766–1771

Bergers E, Bot JC, van der Valk P et al (2002b) Diffuse signal abnormalities in the spinal cord in multiple sclerosis: direct postmortem in situ magnetic resonance imaging correlated with in vitro high-resolution magnetic resonance imaging and histopathology. Ann Neurol 51:652–656

Bot J, Barkhof F, Lycklama a Nijeholt G et al (2002) Differentiation of multiple sclerosis from other inflammatory disorders and cerebrovascular disease: value of spinal MR imaging. Radiology 223:46–56

Bot J, Barkhof F, Polman CH et al (2004) Spinal cord abnormalities in recently diagnosed MS patients. Added value of spinal MRI examination. Neurology 62:226–233

Brex P, O’Riordan JI, Miszkiel KA et al (1999) Multisequence MRI in clinically isolated syndromes and the early development of MS. Neurology 53:1184–1190

Caramia M, Palmieri MG, Desiato MT et al (2004) Brain excitability changes in the relapsing and remitting phase of multiple sclerosis: a study with transcranial magnetic stimulation. Clin Neurophysiol 115:956–965

Clark C, Werring D (2002) Diffusion tensor imaging in spinal cord: methods and applications – a review. NMR Biomed 15:578–586

Dale RC, de Sousa C, Chong WK et al (2000) Acute disseminated encephalomyelitis, multiphasic disseminated encephalomyelitis and multiple sclerosis in children. Brain 123:2407–2422

Fazekas F, Kapeller P (1999) Diseases of the spinal cord. In: Greenberg J (ed) Neuroimaging: a companion to Adam and Victor’s principles of neurology. McGraw-Hill, New York, pp 521–542

Fazekas F, Barkhof F, Filippi M et al (1999) The contribution of magnetic resonance imaging to the diagnosis of multiple sclerosis. Neurology 53:448–456

Filippi M (2003) Magnetization transfer MRI in multiple sclerosis and other central nervous system disorders. Eur J Neurol 10:3–10

Filippi M, Rocca MA, Moiola L et al (1999) MRI and magnetization transfer imaging changes in the brain and cervical cord of patients with Devic’s neuromyelitis optica. Neurology 53:1705–1710

Filippi M, Bozzali M, Horsfield MA et al (2000) A conventional and magnetization transfer MRI study of the cervical cord in patients with MS. Neurology 54:207–213

Holtmannspötter M, Inglese M, ROvaris M et al (2003) A diffusion tensor MRI study of basal ganglia from patients with ADEM. J Neurol Sci 15:27–30

Inglese M, Salvi F, Iannucci G et al (2002) Magnetization trans-

277

fer and diffusion tensor MR imaging of acute disseminated encephalomyelitis. Am J Neuroradiol 23:267–272

Katz J, Ropper A (2000) Progressive necrotic myelopathy: clinical course in 9 patients. Arch Neurol 57:355–361

Kidd D, Thorpe JW, Thompson AJ et al (1993) Spinal cord MRI using multi-array coils and fast spin echo. II. Findings in multiple sclerosis. Neurology 43:2632–2637

Losseff N, Webb SL, O’Riordan J et al (1996) Spinal cord atrophy and disability in multiple sclerosis. A new reproducible and sensitive MRI method with potential to monitor disease progression. Brain 119:701–708

Lycklama à Nijeholt G, Barkhof F, Scheltens P et al (1997) MR of the spinal cord in multiple sclerosis: relation to clinical subtype and disability. Am J Neuroradiol 18:1041–1048

Lycklama à Nijeholt G, van Walderveen MAA, Castelijns JA et al (1998) Brain and spinal cord abnormalities in multiple sclerosis. Correlation between MRI parameters, clinical subtypes and symptoms. Brain 121:687–697

Lycklama G, Thompson AJ, Filippi M et al (2003) Spinal-cord MRI in multiple sclerosis. Lancet Neurol 2:555–562

McDonald W, Compston A, Edan G et al (2001) Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the Diagnosis of Multiple Sclerosis. Ann Neurol 50:121–127

Mezzapesa D, Rocca MA, Falini A et al (2004) A preliminary diffusion tensor and magnetization transfer magnetic resonance imaging study of early-onset multiple sclerosis. Arch Neurol 61:366–368

Miller DH, Filippi M, Fazekas F et al (2004) The role of MRI within diagnostic criteria for multiple sclerosis: a critique. Ann Neurol (in press)

Nijeholt G, Bergers E, Kamphorst W et al (2001) Post-mortem high-resolution MRI of the spinal cord in multiple sclerosis: a correlative study with conventional MRI, histopathology and clinical phenotype. Brain 124:154–166

O’Riordan J, Losseff NA, Phatouros C et al (1998) Asymptomatic spinal cord lesions in clinically isolated optic nerve, brain stem, and spinal cord syndromes suggestive of demyelination. J Neurol Neurosurg Psychiatry 64:353–357

Rovaris M, Bozzali M, Santuccio G et al (2000) Relative contributions of brain and cervical cord pathology to multiple sclerosis disability: a study with magnetisation transfer ratio histogram analysis. J Neurol Neurosurg Psychiatry 69:723–727

Rovaris M, Bozzali M, Santuccio G et al (2001) In vivo assessment of the brain and cervical cord pathology of patients with primary progressive multiple sclerosis. Brain 124:2540–2549

Silver N, Good CD, Sormani MP et al (2001) A modified protocol to improve the detection of enhancing brain and spinal cord lesions in multiple sclerosis. J Neurol 248:215–224

Stevenson V, Leary SM, Losseff NA et al (1998) Spinal cord atrophy and disability in MS: a longitudinal study. Neurology 51:234–238

Tartaglino LM, Friedman DP, Flanders AE et al (1995) Multiple sclerosis in the spinal cord: MR appearance and correlation with clinical parameters. Radiology 195:725–732

Tenembaum S, Chamoles N, Fejerman N (2002) Acute disseminated encephalomyelitis.A long-term follow-up study of 84 pediatric patients. Neurology 59:1224–1231

Thielen K, Miller G (1996) Multiple sclerosis of the spinal cord: magnetic resonance appearance. J Comput Assist Tomogr 20:434–438

278

Thompson AJ, Montalban X, Barkhof F et al (2000) Diagnostic criteria for primary progressive multiple sclerosis: a position paper. Ann Neurol 47:831–835

Thorpe J, Kidd D, Kendall BE et al (1993) Spinal cord MRI using multi-array coils and fast spin echo. I. Technical aspects and findings in healthy adults. Neurology 43:2625–2631

Thorpe J, Kidd D, Moseley IF et al (1996a) Spinal MRI in patients with suspected multiple sclerosis and negative brain MRI. Brain 119:709–714

Thorpe JW, Kidd D, Moseley IF et al (1996b) Serial gadoliniumenhanced MRI of the brain and spinal cord in early relaps- ing-remitting multiple sclerosis. Neurology 46:373–378

Transverse Myelitis Consortium Working Group (2002) Proposed diagnostic criteria and nosology of acute transverse myelitis. Neurology 59:499–505

Ukkonen M, Dastidar P, Heinonen T et al (2003) Volumetric quantitation by MRI in primary progressive multiple scle-

R. Bammer et al.

rosis: volumes of plaques and atrophy correlated with neurological disability. Eur J Neurol 10:663–669

Weinshenker B (2003) Neuromyelitis optica: what it is and what it might be. Lancet 361:889–890

Weinshenker B, Miller DH (1998) MS: One disease or many? In: Siva A, Kesselring J, Thompson AJ (eds) Frontiers in multiple sclerosis. Dunitz, London

Wheeler-Kingshott C, Hickman SJ, Parker GJ et al (2002) Investigating cervical spinal cord structure using axial diffusion tensor imaging. Neuroimage 16:93–102

Wingerchuk DM (2003) Postinfectious encephalomyelitis. Curr Neurol Neurosci Rep 3:256–264

Wingerchuk D, Hogancamp WF, O’Brien PC et al (1999) The clinical course of neuromyelitis optica (Devic’s syndrome). Neurology 53:1107–1114

Zivadinov R, Bakshi R (2004) Role of MRI in multiple sclerosis II: brain and spinal cord atrophy. Front Biosci 9:647–664

CONTENTS

19.1Introduction 279

19.2.1Artefacts 280

19.4.1STIR Imaging 281

19.4.2 Fat Saturated Fast Spin Echo Imaging 282

19.2

Technical Aspects of Optic Nerve Imaging

Magnetic resonance imaging of the optic nerve presents many challenges (Barker 2000), principally due to:

1)Its small size and tortuosity

2)Its mobility

3)The surrounding cerebrospinal fluid (CSF) sheath

4)Orbital fat

19.5New Pulse Sequences and Methods of

Analysis in Optic Neuritis 282

19.5.1Gadolinium “Enhancement” 283

19.5.4Diffusion-Weighted Imaging 286

19.7Conclusion 287 References 288

19.1 Introduction

The principal disease that will be discussed here will be optic neuritis, which is the prototypical demyelinating disease of the optic nerve and has been studied in most detail.Other conditions such as neuromyelitis optica and sarcoidosis will be considered principally in terms of the optic nerve imaging features that help in diagnosis. Magnetic resonance imaging (MRI) of the optic nerves presents many technical difficulties which will be discussed.

S. J. Hickman, MB, MRCP

D. H. Miller, MD, FRCP

NMR Research Unit, Department of Neuroinflammation, Institute of Neurology, University College London, Queen Square, London, WC1N 3BG, UK

faces from the adjacent sphenoid and ethmoid sinuses

The optic nerves are approximately 40–50 mm long and 3–5 mm in diameter (Sadun 1998; Tamraz 1994; Williams et al. 1989). Resolution of the small structures or lesions contained within the nerves is dependent upon the voxel size. If the voxel size is too large then resolution will be impaired due to partial volume effects (i.e. mixing of signal from neighbouring structures). The voxel size can be reduced by increasing the matrix size; however, if the voxel size is too small then there may not be enough signal resolved to produce an image because the signal-to-noise ratio (SNR) is reduced.ResolutionandSNRcanbeincreasedbyincreasing the acquisition time or by increasing the magnetic field strength. Increasing the magnetic field strength, however, will increase the longitudinal relaxation time of the tissue being imaged, also potentially leading to increased acquisition time. There is also the possibility of more susceptibility artefacts (see later) from surrounding tissues at higher field strengths leading to worsening of the SNR (Horsfield 2000). Signal averaging can also improve the SNR and thereby improve the resolution, again at the expense of increased acquisition time (Barker 2000). Increasing acquisition time will lead to more movement artefact, particularly in the mobile optic nerve. Fast imaging techniques are therefore required to give high SNR with an acceptable acquisition time (Gass et al. 1995).

The optic nerves are surrounded by a CSF-filled nerve sheath and, in the orbit, by lipid (Sadun 1998;

Globe

Optic nerve

Orbital fat

Sphenoid sinus

Optic canal

Fig. 19.1. Axial PD-weighted fast spin echo image demonstrating optic nerve anatomy

Williams et al. 1989). Due to its high proton density (PD) lipid typically gives high signal intensity on both T1and PD/T2-weighted imaging (Fig. 19.1). CSF gives high signal intensity on PDand T2-weighted imaging. This may cause problems due to obscuring the edge of the optic nerves and hence CSF suppression may be desirable. In certain circumstances, however, the bright CSF may assist in identification of the nerve (Barker 2000).

19.2.1 Artefacts

There is also a considerable potential for artefact in optic nerve imaging, more so than in the brain. The most important sources of artefact with possible solutions are discussed below.

19.2.1.1

Chemical Shift Artefacts

As has been discussed, the optic nerves are surrounded by lipid in the orbits. Chemical shift artefacts can occur at lipid-water boundaries. Due to their differing chemical environments protons in lipid have a lower precession frequency than protons in water. The lipid protons are therefore misregistered relative to the water protons in the frequency encoding direction (Soila et al. 1984). This can lead to decreased signal where the lipid is displaced away from water and high signal where lipid and water overlap. This chemical shift artefact is more prominent at higher field strengths (Barker 2000). High signal and chemical shift artefacts from orbital fat can be suppressed using short tau inversion recovery (STIR) imaging (Johnson et al. 1987) or a frequencyselective fat saturation excitation pulse prior to imaging time (Gass et al. 1995).

19.2.1.2

Motion Artefacts

Motion in a structure causes image blurring in the direction of movement and “ghost” artefacts in the phase-encoding direction due to the signal being reconstructed over and over. The major motion problem in optic nerve imaging is random motion due to eye movements.On ocular movement the intra-orbital portion of the optic nerve shifts against the direction of gaze with an increasing extent from the relatively fixed posterior part at the orbital apex anteriorly (Liu et al. 1992).Motion artefacts can be reduced by using fast imaging techniques with multiple signal acquisitions and signal averaging or by using gradient echo (GE) rather than spin echo techniques because a 180° refocusing pulse is not required in GE imaging (mobile protons may have moved before the 180° pulse has been applied) (Barker 2000; Taber et al. 1998). Subjects undergoing optic nerve imaging can also be encouraged to relax, close their eyes and avoid deliberate eye movements.

19.2.1.3

Susceptibility Artefacts

Susceptibility artefacts occur at interfaces between different tissues, due to the different magnetic properties of the tissues. It is more apparent at higher field strengths and can cause rapid dephasing of spins with signal loss and mismapping artefacts (Hashemi and Bradley Jr 1997). It is a particular problem in the canalicular portion of the optic nerves due to the bony cavity of the optic canal and air from the adjacent sphenoid and ethmoid sinuses. Metals, particularly iron, can cause a lot of image distortion and subjects should be carefully screened for metal in clothing before imaging commences. Susceptibility artefacts can be reduced by using spin echo rather than GE or echoplanar imaging (EPI) because the spin echo sequence includes a refocusing pulse (Barker 2000).

Imaging of the optic nerves is therefore a tradeoff between imaging time, resolution, SNR and the amount of artefact induced.

19.3

MRI in the Diagnosis of Demyelinating Diseases

Optic neuritis can usually be diagnosed on its clinical features (Hickman et al. 2002b). The principal use of MRI is in assessing the brain for asymptomatic lesions which gives an indication of the risk of subse-

quent development of multiple sclerosis (MS) (Optic Neuritis Study Group 1997). If there is doubt about the diagnosis, in particular if a compressive lesion is suspected then optic nerve,MRI is very useful.In optic neuritis, it will usually demonstrate the lesion, either with conventional or gadolinium-enhanced imaging (Fig. 19.2). It will also demonstrate, in most cases, any compressive lesions (Fig. 19.3).

In neuromyelitis optica (Devic’s disease) high signal and enhancement may be seen in both optic nerves (Eggenberger 2001), usually in combination with a long confluent spinal cord lesion extending over a number of vertebral segments (Fardet et al.2003; Filippi et al. 1999). Optic nerve sheath enhancement may be seen in cases of sarcoidosis, optic perineuritis, or optic nerve sheath meningioma (Fig. 19.4) (Atlas and Galetta 1996; Carmody et al. 1994; Purvin et al. 2001). This feature has also been reported in demyelinating optic neuritis (Atlas and Galetta 1996) and we have witnessed it occurring in several of our own cases.

19.4

Conventional MRI Studies in Optic Neuritis

MRI provides a valuable opportunity for in vivo study of the pathological evolution of optic neuri-

Fig. 19.4. Optic nerve sheath enhancement (arrow)

tis and, by inference, other central nervous system inflammatory/demyelinating lesions with clinical and electrophysiological correlation. There follows a summary of the studies and pulse sequences used in optic neuritis.

19.4.1

STIR Imaging

The first imaging technique used to investigate optic neuritis was the STIR sequence. In a study of 37 patients following an episode of optic neuritis using STIR imaging, lesions were detected in 84% of symptomatic nerves and 20% of asymptomatic

nerves (Miller et al. 1988). Slow (>4 weeks for vision to improve to 6/9) or poor (<6/9 at 6 months) visual recovery was associated with longer lesions (mean number of abnormal 5-mm coronal slices 3.50 versus 1.84, p<0.01), although there was no correlation between lesion length and latency or amplitude of the P100 response of the visual evoked potential (VEP). Of 15 patients with lesions involving the optic canal, 11 (73%) also had a poor recovery which was thought to be due to compression of the swollen optic nerve within the tight confines of the bony canal. These findings were confirmed by Dunker and Wiegand (1996). They examined 22 patients acutely and again after a mean of 4.7 years (range 2.5–8 years) with STIR MRI of the optic nerves. Lesions less than 17.5 mm long either acutely or chronically were correlated with complete visual recovery (visual acuity >20/25, no visual field, colour vision or contrast sensitivity defect). Intracanalicular lesions correlated with incomplete visual recovery (visual acuity >20/25 but other deficits present).

Kakisu et al. (1991) demonstrated that, although the lesion length on STIR was not correlated with pat- tern-reversal VEP findings in the acute phase, there was a significant correlation in the chronic phase (r=0.68, p<0.005). There was, however, no correlation between lesion length and visual acuity.

High signal in the optic nerves following optic neuritis usually persists despite improvements in vision and VEP findings (Youl et al. 1996) and may be seen in MS in the absence of acute attacks of optic neuritis (Davies et al. 1998).

19.4.2

Fat Saturated Fast Spin Echo Imaging

For optic nerve imaging with fat saturation the fast spin echo (FSE) sequence allows for increased resolution and image quality with an acceptable acquisition time. In a study comparing STIR with fatsaturated FSE, lesions were detected in 18 out of 21 symptomatic optic nerves with acute optic neuritis using STIR compared with 20/21 using FSE (Gass et al. 1996). Mean lesion length was 15.4 mm on STIR images and 17.3 mm on FSE images. Using fatsaturated FSE dual echo imaging is possible giving both PDand T2-weighted images from the same acquisition if appropriate TEefs are chosen (Barker 2000).

19.4.3

Combined Fat and Water Suppression Imaging

On FSE and, particularly the lower resolution STIR, CSF gives high signal which may obscure the signal from the optic nerves (Gass et al. 1996). It is possible to suppress the signal from CSF using fluid-attenuated inversion recovery (FLAIR) techniques. The FLAIR sequence uses an inversion pulse with a long inversion time for CSF suppression which increases the acquisition time. For optic nerve imaging fat saturation and fast acquisitions are required, hence a combined selective partial inversion recovery pre-pulse with a FLAIR sequence and FSE acquisition (SPIR-FLAIR) has been developed to try to achieve these aims (Jackson et al. 1998). SPIR-FLAIR was compared with both SPIR and STIR imaging in the assessment of optic neuritis. Both SPIR-FLAIR and SPIR detected lesions in all 21 symptomatic optic nerves imaged whereas STIR imaging detected only 20/21. Lesion lengths were also longest on the SPIR-FLAIR images (17.2 mm on SPIR-FLAIR, 15.8 mm on SPIR and 13.8 mm on STIR, p<0.01) even though the in-plane resolution was less than with SPIR imaging.

19.5

New Pulse Sequences and Methods of Analysis in Optic Neuritis

Although the above conventional MRI techniques are sensitive in detecting the lesions of optic neuritis there is a lack of pathological specificity to allow prediction of the chronological age of these lesions. This is because oedema, demyelination, axonal loss and gliosis all cause high signal on T2-weighted images (Miller et al.1998).As has been discussed,imaging of the optic nerves is particularly limited by issues of poor resolution. Increasing the SNR requires very large increases in acquisition time. Using conventional imaging techniques at present is therefore limited by the technology available (Horsfield 2000). The only pointers to age the lesion have been a subjective impression of optic nerve swelling in the acute phase and atrophy in the chronic phase (Gass et al. 1996; Jackson et al. 1998; Kapoor et al. 1998; Miller et al. 1988). Other techniques have been developed which are able to study the optic nerves in order to understand better the pathophysiology of optic neuritis. These have principally been applied in brain imaging in MS but, with recent technological advances, optic nerve imaging using these techniques has also been possible.