Elsevier

NeuroImage

Volume 45, Issue 3, 15 April 2009, Pages 679-686
NeuroImage

Optic nerve diffusion changes and atrophy jointly predict visual dysfunction after optic neuritis

https://doi.org/10.1016/j.neuroimage.2008.12.047Get rights and content

Abstract

Recently, there has been strong interest in the development of imaging techniques to quantify axonal and myelin pathology in patients with multiple sclerosis (MS). Optic neuritis, a condition characterised by inflammatory demyelination of the optic nerve, is one of the commonest sites of MS relapse, and exhibits similar pathological alterations to MS lesions elsewhere in the central nervous system (CNS). Unlike other regions of the CNS, however, the function of the optic nerve can be accurately assessed using clinical measures, as well as electrophysiological techniques such as visual evoked potential recordings. Therefore, optic neuritis is useful for investigating the relationship between abnormalities in optic nerve structure, assessed using magnetic resonance imaging (MRI), and visual dysfunction, assessed clinically and electrophysiologically. The aims of the present study were to assess optic nerve structural abnormalities in patients with a history of unilateral optic neuritis using MRI, and then to identify correlations between abnormalities in optic nerve MRI and visual dysfunction. Ten controls and sixteen patients underwent high resolution optic nerve diffusion tensor imaging (DTI), T2- and T1-weighted MRI. In addition, Snellen visual acuity and the latency and amplitude of multifocal visual evoked potentials (mfVEP) were tested in all patients. Diffusion and volumetric MRI indices were correlated to mfVEP functional indices. Significant abnormalities were detected in MRI and mfVEP measures in patients' affected nerves compared to unaffected optic nerves or optic nerves from healthy controls. Reduced mfVEP amplitude in the affected side significantly correlated with both affected optic nerve atrophy (R = 0.58, p = 0.02) and reduced fractional anisotropy (FA) (R = 0.52, p = 0.04). However, atrophy and reduced FA did not correlate with each other. To further investigate this disassociation, we used linear regression analysis with optic nerve atrophy and optic nerve FA as independent variables and mfVEP amplitude as the dependent variable. The resulting linear regression model was highly significant (R = 0.819, p = 0.001). These results show that, 4 years after unilateral optic neuritis, MRI-based measures of optic nerve structural abnormalities (decreased anisotropy and volume) independently predict visual dysfunction.

Introduction

Multiple sclerosis (MS) is as an inflammatory demyelinating disease which also features significant axonal degeneration (Trapp et al., 1998). However, long term accumulation of irreversible neurological disability in MS most likely results from axonal degeneration (Lassmann, 2004). Axonal damage may be the direct result of inflammatory damage or secondary to persistent demyelination. Therefore therapeutic interventions that protect or regenerate oligodendrocytes and myelin or, alternatively, directly protect axons may improve clinical outcomes. Several therapeutic candidates with remyelinating or neuroprotective potential have been identified such as Leukemia Inhibitory Factor (Butzkueven et al., 2002) and inhibition of LINGO1 (Mi et al., 2007). However, the lack of sensitive, robust, and validated surrogate markers of axon and myelin pathology inhibits early phase human trials of these agents.

The visual system is potentially useful for developing surrogate markers of myelin and axonal damage, and repair. The optic nerve is anatomically distinct from the brain, can be imaged using magnetic resonance imaging (MRI) (Hickman, 2007), and can be functionally assessed using visually evoked potentials (VEP) (Halliday et al., 1972). Additionally, the optic nerve is a common site of localised inflammatory demyelination, known as optic neuritis. Although clinical visual dysfunction associated with optic neuritis can resolve within months, persistent subclinical changes can be detected using MRI and VEP. However, the specific relationship between visual dysfunction and underlying optic nerve pathology remains unclear.

Histo-pathological assessment of the optic nerve in MS is limited to one study, which confirmed significant axonal loss of around 40% (Evangelou et al., 2001). Therefore, the development of paraclinical assessment tools able to detect axonal degeneration in the MS-affected optic nerve might be feasible. Ideally, such a paraclinical tool should be altered by optic neuritis, have specific visual functional correlates, and have validated histo-pathological correlates in animal models. Optic nerve pathology associated with optic neuritis can be imaged using conventional T2-weighted MRI. Unfortunately, inflammation, demyelination, remyelination, gliosis and axonal degeneration all produce T2 hyperintensity, so T2 sequences are unlikely to fulfil a role as biomarkers of neurodegeneration. Persistent T1 hypointensity, on the other hand, is known to be associated with regions of severe axonal degeneration (Bruck et al., 1997, Fisher et al., 2007, Mottershead et al., 2003, van Waesberghe et al., 1999). White matter atrophy might also be a useful marker of gross axonal loss (Miller et al., 2002). Reduced optic nerve cross-sectional area has been reported in patients after optic neuritis (Hickman et al., 2001, Hickman et al., 2002, Hickman et al., 2004) and appears to correlate with retinal nerve fibre layer thinning indicative of axonal loss (Trip et al., 2006a). White matter tracts can also be characterised using diffusion weighted imaging (DWI), which has been demonstrated in animal studies to reveal subtle changes in the microscopic structure of axons and myelin (Beaulieu, 2002). DWI can be used to detect abnormalities in MS plaques and in normal-appearing white matter (NAWM) (Werring et al., 1999). When diffusion is measured in six or more directions, the diffusion tensor can be calculated (Basser et al., 1994) and, from it, measures of the diffusion parallel to (λ||) and perpendicular to (λ) the axonal tract can be derived. Animal model studies of axon and myelin pathology have demonstrated that reduced λ|| is associated with acute axonal pathology (Song et al., 2003, Wu et al., 2007), whereas increased λ is associated with demyelination (Song et al., 2002, Song et al., 2003, Song et al., 2005).

DWI of the human optic nerve is challenging given its size, mobility and geometric distortions resulting from the surrounding tissue interfaces. Nonetheless, novel MR sequences using either reduced field of view spin-echo echo planar imaging (EPI) (Wheeler-Kingshott et al., 2002, Wheeler-Kingshott et al., 2006, Xu et al., 2007) or fast spin-echo (Chabert et al., 2005) sequences have been used to successfully image the optic nerve. These studies have reported healthy optic nerve mean diffusivity (MD) values of between 1.0 and 1.3 × 10 3 mm2/s and fractional anisotropy (FA) of between 0.4 and 0.6 (Xu et al., 2008). The optic nerves of patients with optic neuritis have been studied using DWI (Hickman et al., 2005, Iwasawa et al., 1997) and DTI (Trip et al., 2006b). These studies have shown that chronically affected nerves consistently exhibit increased diffusion compared to unaffected or control optic nerves and DTI has shown that affected nerves exhibit increased λ||, also confirmed in spinal cord lesions (Clark et al., 2000).

After optic nerve inflammation resolves, prolonged latency and reduced amplitude of VEP are believed to be produced by persistent demyelination and axonal loss, respectively (Klistorner et al., 2008). Optic nerve atrophy (Hickman et al., 2002, Trip et al., 2006a) and diffusion changes (Hickman et al., 2005, Trip et al., 2006b) have been shown to correlate with changes in VEP. However, the correlation coefficients reported in these studies suggest that individually, atrophy or diffusivity changes only explain 30–40% of the variance in VEP measures. Such moderate correlations may reflect a more complex relationship between optic nerve structural abnormalities and visual dysfunction. Also, conventional VEP might be insensitive to some optic neuritis related visual field defects, because it principally assesses central vision. In order to overcome this deficiency of standard VEP, multifocal VEP (mfVEP) was developed. Multifocal VEP is a technique for recording VEP from multiple sectors of the visual field simultaneously. Compared to conventional VEP, mfVEP is more sensitive to smaller or peripheral field defects after optic neuritis (Klistorner et al., 2007a).

In this study we aimed to comprehensively assess patient optic nerve structural abnormalities after acute optic neuritis using high resolution T1- and T2-weighted imaging and DTI. Furthermore, we aimed to identify correlative relationships between optic nerve structural abnormalities and visual dysfunction, measured using mfVEP.

Section snippets

Subjects, clinical assessments and disease classification

In order to assess the long term outcomes after acute optic neuritis, sixteen patients were recruited randomly from all the patients diagnosed with acute unilateral optic neuritis at the Royal Victorian Eye and Ear Hospital between July 2002 and August 2003. All patients were imaged a mean of 4 years (range = 3.4 to 4.8 years) after acute optic neuritis presentation. The demographic and disease information for each patient recorded at the time of MRI assessment is contained in Table 1. Patient

Results

Optic nerve DTI and volumetric measurements and mfVEP measurements for each clinical group were not found to be significantly different from the normal probability distribution when compared using the K–S statistic, allowing the use of parametric statistical tools to compare groups. Intra-rater and inter-rater reliability coefficients for FA (0.98/0.94) and optic nerve volume (0.98/0.96) were high.

Discussion

Given the lack of ‘gold standard’ markers for myelin and axonal pathologies, a convergence of evidence approach using measures from various imaging techniques can be used to infer the nature of underlying pathologies. We aimed to characterise any structural abnormalities in optic nerves after optic neuritis using MRI, and identify correlative relationships between optic nerve MRI and visual dysfunction measured using mfVEP. Significant abnormalities were observed in conventional MRI contrast,

Acknowledgments

This work was supported by the National Health and Medical Research Council (NHMRC) Principal Research Fellowship (Grant 400317, G. F. Egan), Australia; the joint MS Research Australia/Trish MS Foundation/NHMRC Betty Cuthbert Fellowship (Grant 400476, H. Butzkueven); the joint MS Research Australia/Australian Financial Advisors Association Postgraduate Scholarship (S. Kolbe); Charityworks for MS, Melbourne; and the Neurosciences Victoria Neuro-Informatics Platform. We thank all the participants

References (44)

  • WuQ. et al.

    MR diffusion changes correlate with ultra-structurally defined axonal degeneration in murine optic nerve

    Neuroimage

    (2007)
  • BeaulieuC.

    The basis of anisotropic water diffusion in the nervous system — a technical review

    NMR Biomed.

    (2002)
  • BruckW. et al.

    Inflammatory central nervous system demyelination: correlation of magnetic resonance imaging findings with lesion pathology

    Ann. Neurol.

    (1997)
  • ButzkuevenH. et al.

    LIF receptor signaling limits immune-mediated demyelination by enhancing oligodendrocyte survival

    Nat. Med.

    (2002)
  • ChabertS. et al.

    Diffusion tensor imaging of the human optic nerve using a non-CPMG fast spin echo sequence

    J. Magn. Reson. Imaging

    (2005)
  • ClarkC.A. et al.

    Diffusion imaging of the spinal cord in vivo: estimation of the principal diffusivities and application to multiple sclerosis

    Magn. Reson. Med.

    (2000)
  • EvangelouN. et al.

    Size-selective neuronal changes in the anterior optic pathways suggest a differential susceptibility to injury in multiple sclerosis

    Brain

    (2001)
  • FisherE. et al.

    Imaging correlates of axonal swelling in chronic multiple sclerosis brains

    Ann. Neurol.

    (2007)
  • FraserC. et al.

    Multifocal visual evoked potential latency analysis: predicting progression to multiple sclerosis

    Arch. Neurol.

    (2006)
  • HasanK.M. et al.

    Computation of the fractional anisotropy and mean diffusivity maps without tensor decoding and diagonalization: theoretical analysis and validation

    Magn. Reson. Med.

    (2003)
  • HickmanS.J.

    Optic nerve imaging in multiple sclerosis

    J. Neuroimaging

    (2007)
  • HickmanS.J. et al.

    Detection of optic nerve atrophy following a single episode of unilateral optic neuritis by MRI using a fat-saturated short-echo fast FLAIR sequence

    Neuroradiology

    (2001)
  • Cited by (74)

    • Incorporating non-linear alignment and multi-compartmental modeling for improved human optic nerve diffusion imaging

      2019, NeuroImage
      Citation Excerpt :

      Nevertheless, like any other dMRI studies, motion-induced artifactual signal dropout did occasionally occur in our optic nerve dMRI data, which would require manual identification and removal of these outlier volumes. Besides the rFOV or IVI approach for high resolution optic nerve dMRI acquisition, other researchers have adapted axial or coronal whole brain dMRI protocols, using parallel imaging techniques, with high in-plane resolution (e.g., 1.2–1.5 mm), but thick (e.g., 3–4 mm) slices to study the optic nerve in clinical populations (Gerlach et al., 2017; Kolbe et al., 2009; Li et al., 2014; Smith et al., 2011; Techavipoo et al., 2009; Walt et al., 2013). We do not expect our proposed scheme to be applicable to facilitate the volume alignment of optic nerve dMRI images acquired from this type of optic nerve dMRI protocols, because our algorithm performs 3D non-linear registration, which assumes continuous identifiable optic nerve voxels between slices, while the relatively thick slices, as compared to the small optic nerve structure, from this type of protocol typically results in discontinuous optic nerve voxels between slices.

    • Sequential phases of RGC axonal and somatic injury in EAE mice examined using DTI and OCT

      2019, Multiple Sclerosis and Related Disorders
      Citation Excerpt :

      From the ON, DTI measures paralleled findings from human MS in several respects. The findings of increased RD with concomitant decreases in FA after 4 and 8 weeks are consistent with long-term findings from MS patients, which show similar DTI changes in ON even years after symptoms have resolved (Kolbe et al., 2009; Naismith et al., 2009). These changes in FA and RD have also been shown to correspond with visual acuity (Naismith et al., 2010), and discriminate between visual recovery groups after optic neuritis (Naismith et al., 2010).

    • Microstructural MR Imaging Techniques in Multiple Sclerosis

      2017, Neuroimaging Clinics of North America
    • Utility of coronal contrast-enhanced fat-suppressed FLAIR in the evaluation of optic neuropathy and atrophy

      2017, European Journal of Radiology Open
      Citation Excerpt :

      Although comprehensive neuro-ophthalmic exams were commonly documented in the acute optic neuritis phase, commentary on optic nerve atrophy at the time of MRI in this patient cohort was inconsistent. In this regard, other studies have evaluated the degree of optic atrophy via cross-sectional measurements and segmentation, based on DTI, FS-short echo FLAIR, and magnetization transfer imaging [12,14–16]. Also, alternative imaging modalities such as optical coherence tomography have directly measured optic nerve atrophy via quantification of the microscopic retinal nerve fiber layer [16,30].

    View all citing articles on Scopus
    View full text