Download Neuroimaging Biomarkers in Multiple Sclerosis (MS)

REVIEW
BIOMARKERS.IMAGILYS.COM
Neuroimaging Biomarkers
in Multiple Sclerosis (MS)
State-of-the-art MR imaging in MS patients: conventional MRI, quantitative
neuroimaging biomarkers, and advanced brain imaging techniques.
By Dr. Laurent Hermoye, CEO
About Multiple Sclerosis
Multiple sclerosis (MS) is a chronic inflammatory
demyelinating disease affecting the central nervous system
(Compston and Coles, 2008). The disease usually affects
young adults, who suffer from a variety of neurological
symptoms, with periods of relapse and remission (relapsingremitting MS), or progressively leading to irreversible
disabilities (primary or secondary progressive MS).
FIGURE 1: FLAIR images of a patient with multiple sclerosis, loaded in
BrainMagix’s longitudinal follow-up module. An increased lesion load,
and a large atrophy of the brain can be observed, 20 years after the
onset of the disease.
The pathophysiology of the disease is complex, and not
perfectly understood at this time. An immune-mediated
inflammation process is associated with demyelination
and subsequent axonal damage. Although multiple
sclerosis is primarily a white matter disease, gray
matter is also affected. Some studies suggest that gray
matter atrophy could be linked with an independent
degenerative process of the disease (Steenwijk et al., 2014),
in addition to the well-known demyelinating one.
At the diagnostic step, MR imaging is included in the diagnostic
work-up, alongside clinical assessment, analysis of the cerebrospinal fluid (CSF), and evoked potentials. The MR-based
diagnosis relies on the demonstration of lesion dissemination
in space and time, as defined by consensus criteria, such as
the revised McDonald criteria (Polman et al., 2011). MRI is also
useful for the differential diagnosis with other brain diseases.
MRI in Multiple Sclerosis
In the brain, different MRI sequences can reflect
different aspects of the disease’s complex
pathophysiology (Filippi and Rocca, 2011):
Magnetic resonance imaging (MRI) is useful:
• at the diagnostic step, in patients with a clinically
isolated syndrome (CIS), or MS. A variant of the CIS is
the radiologically isolated syndrome (RIS), in which the
discovery of lesions suggestive of MS is an incidental
finding in a routine MRI examination (Granberg et al. 2013).
• at the follow-up step, in patients with established MS.
• in clinical trials, in order to assess the
therapeutic effect of a new drug.
• to monitor the potential adverse effects of MS-targeted
drugs and, especially, to exclude the rare onset of
progressive multifocal leukoencephalopathy (PML) in
patients treated with Natalizumab (Yousry et al., 2012).
EDU-001
Revision #0
• Fluid-attenuated inversion-recovery (FLAIR) images, in
which the white matter lesions appear as bright spots,
reflecting different levels of myelin loss, inflammatory
activity, and gliosis (figure 1). The sensitivity of the
sequence is high but its specificity is low, as other
lesions (e.g., vascular lesions) can mimic MS plaques.
• Dual-echo T2- and proton density-weighted images, in
which the white matter lesions appear hyperintense,
such as on FLAIR images. They can be particularly
useful in the posterior fossa, where FLAIR images have
limited sensitivity, and are more prone to artifacts.
© Imagilys 2014
Page 1
REVIEW
BIOMARKERS.IMAGILYS.COM
is necessary, in most of the cases, in order to improve the
matching. A transparent fusion tool, within (intra-) or between
(inter-) the time points, is useful for the characterization (on
multiple image weighting) and the follow-up of each lesion.
Image subtraction techniques can also be used in order to
quickly identify new or growing lesions (Moraal et al., 2009).
FIGURE 2: Volumetric segmentation of the brain structures, in
BrainMagix’s SurferMagix module.
• 3D high-resolution T1-weighted images, useful for
brain volumetry and for detecting the black holes, i.e.
persistently dark lesions, associated with severe tissue
damage (i.e., both demyelination and axonal loss).
• Post-contrast T1-weighted images, acquired after the
injection of a gadolinium-based contrast agent. Active
lesions, corresponding to areas of ongoing inflammation,
show a signal enhancement on these images, as a result
of an increased blood-brain-barrier permeability.
• Double inversion recovery (DIR) images, in which two
inversion pulses are used, in order to suppress both the
signal of white matter and the one of the CSF. The sequence
is useful for depicting cortical gray matter lesions, which
are usually not well visible on the other series. However, the
sensitivity of the sequence remains limited (Sethi et al., 2012).
• Phase-sensitive inversion recovery (PSIR) images,
only available at 3 Tesla, which may be more
sensitive than DIR images for the detection of
cortical gray matter lesions (Sethi et al., 2012).
Imaging of the spinal cord and of the optic nerve,
which are also affected by multiple sclerosis, is
outside the scope of this review. Readers interested
in this topic can refer to (Filippi and Rocca, 2011).
Quantitative Imaging in MS Patients
Longitudinal MRI follow-up in MS patients can be optimized
by the use of dedicated tools, such as BrainMagix’s followup module (figure 1), in order to compare the examinations
and follow the disease’s progression. Although standardized
slice positioning, parallel to the sub-callosal plane, improves
the repeatability of the examinations, images registration
EDU-001
Revision #0
The counting and volume of the T2 (white matter) lesions,
and their progression, are useful biomarkers. These are usually
measured on FLAIR images. However, the sensitivity of this
contrast is poor in the posterior fossa (Stevenson et al., 1997),
and may require the segmentation of the T2 images in this
region. Total lesion volume increases by approx. 5%–10%
per year in untreated patients (Filippi and Rocca, 2011).
There is a substantial body of histopathological evidence
that supports chronic black holes (i.e. white matter
lesions with a persistently hypointense appearance
relative to normal-appearing white matter (NAWM) on
a T1-weighted MRI) as being indicative of irreversible
demyelination and axonal damage. As a result, the
progression of black holes is considered as a promising
imaging surrogate endpoint in MS (Tam et al., 2012).
Contrast-enhancing lesions can also be quantified, as
a marker of the disease’s inflammatory activity. At this
stage, the quantification of cortical lesions, as measured
in DIR or PSIR sequence, is still challenging because of
their limited sensitivity and of the lack of standardization
between centers (Filippi and Rocca, 2011).
Automated MS lesion segmentation algorithms have been
widely published in the literature (García-Lorenzo et al.,
2013). However, their use on a variety of patients, centers,
MRI scanners, and sequences is still challenging. Therefore,
BrainMagix’s longitudinal follow-up module implements
a semi-automated algorithm that, while reducing the
user’s interaction time, still requires his/her validation.
Technical parameters, such as slice thickness, can also affect the
measurement and its reproducibility. The reduction of the slice
thickness from 5 to 3 mm makes it possible to detect smaller
lesions, leading to an increase of the measured lesion volume
by approximately 8% (Molyneux et al., 1998) and a decrease of
the intra- and inter-observer variability (Filippi et al., 1998).
© Imagilys 2014
Page 2
REVIEW
BIOMARKERS.IMAGILYS.COM
to the brain tissue matrix, can be used to measure their
capacity to exchange magnetization with the surrounding
free water, i.e. the magnetization transfer ratio (MTR). This
ratio is reduced in MS lesions, correlated with the degree of
myelin loss and axonal damage (Schmierer et al., 2004), and its
reduction in NAWM and gray matter can even be predictive
of the development of new lesions (Filippi and Agosta, 2007)
and of the future clinical disability (Agosta et al., 2006).
Perfusion studies have identified both diffuse and focal
perfusion abnormalities in MS patients. Functional MRI
(fMRI) has been used to show functional reorganization in
the brain of MS patients. Dynamic changes of metabolites
concentrations can be observed, even preceding the lesion
formation, with MR spectroscopy. However, none of these
three modalities is currently included in routine MS protocols,
due to the difficulties in interpreting the results and their
meaning at the individual level (Filippi and Rocca, 2011).
FIGURE 3: Color-coded diffusion tensor imaging (DTI) map, overlaid on
a T1-weighted image. Degeneration of the white matter tracts can be
observed, especially in the forceps major.
The brain volume can be measured based on the
segmentation of a high resolution 3D T1-weighted image
(figure 2). In MS patients, this volume decreases by approx.
0.7%–1% per year, on average (Filippi and Rocca, 2011).
The brain parenchymal fraction (BPF), i.e. the ratio of brain
parenchymal volume to the total volume within the brain
surface contour (Rudick et al., 1999; Vågberg et al., 2013), is
also often used as an indicator of atrophy in MS patients. In
addition, some brain structures seem to be more affected
by the atrophy than others, depending on the phase of the
disease. Therefore, brain morphometry, implemented in
BrainMagix’s SurferMagix module, is a useful biomarker.
MS Imaging for Treatment Monitoring and in Clinical Trials
There is a poor correlation between lesion load and symptoms
(Messina and Patti, 2014). Brain atrophy seems to be better
correlated with disability progression (Sormani et al., 2014)
and especially with cognitive impairment (Messina and Patti,
2014), in line with the presence of a neurodegenerative
process in the disease (Steenwijk et al., 2014).
Advanced MRI in Multiple Sclerosis
Diffusion tensor imaging is useful in assessing the integrity
and myelination of white matter tracts (figure 3). Increased
apparent diffusion coefficient (ADC) and decreased
fractional anisotropy (FA) can be observed in T2 lesions,
and even in normal appearing white matter (NAWM).
Globally, abnormal mean diffusivity and FA generally
correlate with the progression of the disease, and with
cognitive impairment (Hulst et al., 2013). Locally, white
matter abnormalities can, in some cases, be correlated with
clinical disabilities, e.g., mean diffusivity of the cortico-spinal
tract with motor impairment (Filippi and Rocca, 2011).
Magnetization transfer MR imaging, which uses an offresonance pulse in order to saturate the protons bounds
EDU-001
Revision #0
The quantitative parameters described hereinabove, and
especially the number and volume of new and enlarged
lesions, can be used to monitor treatment, alongside
clinical assessment. Studies have shown that these
measurements are valid surrogate markers of clinical activity
(Sormani et al., 2009). These may become increasingly
useful in clinical practice, especially for monitoring
the new drugs, such as Natalizumab, which targets a
complete stabilization of the disease (zero tolerance).
At the individual level, the prediction of the efficacy of a
treatment would be a major step toward precision medicine.
Although some studies have shown that MR-based
measurement could be used as early predictors of treatment
response, there is still no consensus on the optimal criteria,
and MR imaging is not recommended as the sole predictor
of treatment response at this stage (Filippi and Rocca, 2011).
These neuroimaging biomarkers are also used in the
framework of clinical trials, testing new drugs. Follow-up of
the lesion load is mainly linked with the inflammatory aspect
of the disease. In addition, specific biomarkers, such as the
measurement of brain atrophy, of T1-hypointense black
holes, and of the magnetization transfer ratio, can be used
to monitor the neurodegenerative aspect of the disease
(Barkhof et al., 2009), especially when testing drugs with a
neuroprotective effect. However, standardization and validation
of the biomarkers, across clinical trials and centers, is still
missing and should be promoted by consensus meetings.
© Imagilys 2014
Page 3
REVIEW
BIOMARKERS.IMAGILYS.COM
References
• Agosta et al. Magnetization transfer MRI metrics predict
the accumulation of disability 8 years later in patients with
multiple sclerosis. Brain J. Neurol. 2006, 129: 2620–2627.
• Barkhof et al. Imaging outcomes for neuroprotection and repair
in multiple sclerosis trials. Nat. Rev. Neurol. 2009, 5: 256–266.
• Compston and Coles. Multiple sclerosis.
Lancet 2008, 372: 1502–1517.
• Filippi and Agosta. Magnetization transfer MRI in
multiple sclerosis. J. Neuroimaging Off. J. Am. Soc.
Neuroimaging 2007, 17 Suppl 1: 22S – 26S.
• Filippi and Rocca. MR imaging of multiple
sclerosis. Radiology 2011, 259: 659–681.
• Filippi et al. Intraobserver and interobserver
variability in measuring changes in lesion volume
on serial brain MR images in multiple sclerosis.
AJNR Am. J. Neuroradiol. 1998, 19: 685–687.
• García-Lorenzo et al. Review of automatic
segmentation methods of multiple sclerosis white
matter lesions on conventional magnetic resonance
imaging. Med. Image Anal. 2013, 17: 1–18.
• Granberg et al. Radiologically isolated syndrome--incidental
magnetic resonance imaging findings suggestive of multiple
sclerosis, a systematic review. Mult. Scler. 2013, 19: 271–280.
• Hulst et al. Cognitive impairment in MS: Impact of
white matter integrity, gray matter volume, and
lesions. Neurology 2013, 80: 1025–1032.
• Messina and Patti. Gray matters in multiple
sclerosis: cognitive impairment and structural
MRI. Mult. Scler. Int. 2014, 2014: 609694.
• Molyneux et al. The effect of section thickness on MR
lesion detection and quantification in multiple sclerosis.
AJNR Am. J. Neuroradiol. 1998, 19: 1715–1720.
• Moraal et al. Subtraction MR images in a multiple sclerosis
multicenter clinical trial setting. Radiology 2009, 250: 506.
• Polman et al. Diagnostic criteria for multiple
sclerosis: 2010 Revisions to the McDonald
criteria. Ann. Neurol. 2011, 69: 292–302.
• Rudick et al. Use of the brain parenchymal fraction
to measure whole brain atrophy in relapsingremitting MS. Multiple Sclerosis Collaborative
Research Group. Neurology 1999, 53: 1698–1704.
• Schmierer et al. Magnetization transfer ratio
and myelin in postmortem multiple sclerosis
brain. Ann. Neurol. 2004, 56: 407–415.
• Sethi et al. Improved detection of cortical MS lesions
with phase-sensitive inversion recovery MRI. J. Neurol.
Neurosurg. Psychiatry 2012, 83: 877–882.
EDU-001
Revision #0
• Sormani et al. Magnetic resonance imaging as a potential
surrogate for relapses in multiple sclerosis: a metaanalytic approach. Ann. Neurol. 2009, 65: 268–275.
• Sormani et al. Treatment effect on brain atrophy
correlates with treatment effect on disability in
multiple sclerosis. Ann. Neurol. 2014, 75: 43–49.
• Steenwijk et al. What Explains Gray Matter Atrophy in Longstanding Multiple Sclerosis? Radiology 2014, 272: 832–842.
• Stevenson et al. Imaging of the spinal cord and brain
in multiple sclerosis: a comparative study between fast
FLAIR and fast spin echo. J. Neurol. 1997, 244: 119–124.
• Tam et al. Improving the clinical correlation of multiple
sclerosis black hole volume change by paired-scan
analysis. NeuroImage Clin. 2012, 1: 29–36.
• Vågberg et al. Automated determination of brain
parenchymal fraction in multiple sclerosis. AJNR
Am. J. Neuroradiol. 2013, 34: 498–504.
• Yousry et al. Magnetic resonance imaging pattern
in natalizumab-associated progressive multifocal
leukoencephalopathy. Ann. Neurol. 2012, 72: 779–787.
About the Author
After graduating as an electrical engineer and doctor of
medical sciences, Laurent Hermoye founded Imagilys.
He endeavors to bring the most advanced brain imaging
techniques from bench to bedside. His efforts have resulted in
multiple publications in medical journals and at international
conferences, but also in a neuroimaging software, BrainMagix,
which has been used on thousands of patients. Through
continuing medical education at Harvard Medical School, he
is continually at the leading edge of the newest techniques.
About Imagilys
Imagilys develops and sells a neuroimaging software,
BrainMagix. We help our customers, throughout the world,
to diagnose and to treat patients with severe neurological,
neurosurgical, and psychiatric diseases. We help pharmaceutical
companies and CROs to use neuroimaging biomarkers for
the discovery of promising drugs. Finally, we train doctors
on advanced neuroimaging techniques, in the framework
of our Advanced Clinical Neuroimaging Courses.
© Imagilys 2014
Page 4