Chronic active lesions: a new MRI biomarker to monitor treatment effect in multiple sclerosis?

KEYWORDS:

• Chronic active lesions
• MRI
• multiple sclerosis
• slowly-evolving lesions
• susceptibility-weighted imaging

1. Introduction

Multiple sclerosis (MS) is a chronic inflammatory, demyelinating and neurodegenerative disease of the central nervous system (CNS) characterized by the accumulation of focal white matter (WM) lesions that can be classified as active, chronic active (also referred to as mixed active/inactive lesions), inactive, and remyelinated, according to the presence and topography of ongoing demyelination and active inflammation or remyelination [1–3].

In pathological studies, chronic active lesions represent up to 57% of all focal WM lesions, being more frequently described in patients with progressive (P) MS [1–3]. Chronic active lesions typically show a peripheral ‘rim’ of iron-laden activated microglia/macrophages associated with ongoing demyelination and axonal loss, around an inactive core without blood-brain barrier damage, thus reflecting a compartmentalized chronic inflammatory pathological process [1–4].

Pathology represents the gold standard to identify chronic active lesions [5]. However, promising neuroimaging strategies have been proposed to detect in vivo these lesions. First, chronic active lesions have been evaluated by looking at susceptibility-based magnetic resonance imaging (MRI) scans at high- and ultra-high field. On these sequences, chronic active lesions show a paramagnetic hypointense rim (i.e. ‘iron rim lesions’) that corresponds pathologically to peripheral iron-laden microglia [3,4,6] (Figure 1(a)). Moreover, by combining susceptibility-based MRI scans and serial post-contrast T1-weighted sequences rapidly acquired after gadolinium injection (i.e. dynamic contrast enhancement [DCE]), chronic active lesions, ~9.4% of all chronic lesions, showed no contrast-enhancement, thus they lack substantially abnormal blood-brain barrier permeability that typically characterizes acute lesions [7]. Iron rim lesions are also characterized by a more limited lesional repair and have been described in all MS phenotypes although with heterogeneous reported rates: clinically isolated syndrome/relapsing-remitting (RR) MS from 6% to 53%, PMS from 7% to 62% [3,4,8,9]. By evaluating the longitudinal evolution of iron rim lesions over up to 7 years, a recent study with the longest follow-up currently available showed that such lesions are characterized by a slow rate of increase in size in the first years after their formation and then stabilize. Moreover iron rim lesions persist for years, although they gradually diminish over time, likely reflecting a reduction of peripheral chronic activity [4]. Quantitative susceptibility mapping (QSM) has also been recently applied to quantify and localize brain iron more accurately, with iron rim lesions being reported with rates from 4.2% to 10.6% [10–12].

Figure 1. Example of chronic active lesion visualization using different neuroimaging modalities. (A) On 3D axial fluid-attenuated inversion recovery sequence, two hyperintense white matter lesions are visible. Both lesions (orange arrows) show a hypointense rim on phase image derived from a multi echo gradient-echo T2* sequence, thus they represent ‘iron rim lesions.’ (B) White matter lesions showing a linear expansion on serial MRI scans (i.e. ‘slowly expanding lesions’ [SELs]) can be identified by linearly fitting the Jacobian of the non-linear deformation field between timepoints obtained using merged T1- and T2-weighted images scans. In particular, T2-hyperintense lesions containing one cluster showing a linear annual increase ≥12.5% and embedding neighboring voxels with an annual increase of at least 4% are classified as SELs. (C) Demonstration of a chronic active lesion showing a peripheral rim on quantitative susceptibility mapping (QSM) (red circle) and ¹¹C-PK11195 uptake, a marker of activated microglia/macrophages, on positron emission tomography (red circle). Figure 1(c) adapted from [11] by permission of Oxford University Press on behalf of the Guarantors of Brain

Second, since chronic active lesions slowly increase in size over time (i.e. slowly evolving lesions [SEL]), such lesions have been identified among those WM lesions showing a linear and progressive longitudinal expansion over long-enough periods of time on conventional T1- and T2-weighted sequences [13–15] (Figure 1(b)).

Third, chronic active lesions have been studied with positron emission tomography (PET) using radiotracers specific to microglia/macrophages [16–18] (Figure 1(c)).

More recently, sodium (²³Na) MRI has been also proposed to identify chronic active lesions [19]; however it has not yet been evaluated to study treatment effects.

Chronic active lesions seem to represent one of the most relevant pathological substrates associated with more severe clinical disability, progressive disease course, and brain atrophy and contributing to disability progression in MS, also in the absence of overt inflammatory activity [6,9].

Accordingly, the investigation of the potential beneficial effects of disease-modifying treatments (DMTs) on the occurrence, accumulation and microstructural features of chronic active lesions may represent a rewarding strategy to understand whether specific therapeutic interventions may limit this compartmentalized chronic inflammation in MS.

2. Treatment effects on chronic active lesions

2.1. Susceptibility weighted imaging

The influence of DMTs on the occurrence and evolution of iron rim lesions has not been evaluated yet. Indirectly, a recent study suggested that iron rim lesions were present in patients treated with most of the currently available DMTs, including first and second line therapies (proportion of MS patients with ≥1 iron rim lesion: untreated = 28/61 [45.9%], interferon/glatiramer acetate = 28/51 [54.9%], dimethyl fumarate = 19/29 [65.5%], natalizumab = 8/18 [44.4%], fingolimod = 5/10 [50.0%], rituximab/ocrelizumab = 7/10 [70.0%]) [6]. However, the lack of a longitudinal evaluation and the limited sample size did not allow to compare the effects of different DMTs in limiting this chronic inflammation.

An ongoing phase II randomized controlled trial (RCT) (NCT04742400) evaluating the efficacy and safety of tolebrutinib, an oral, brain-penetrant, Bruton’s tyrosine kinase (BTK) inhibitor, included the assessment of iron rim lesions and changes in size and in T1 relaxation time of paramagnetic rim lesions at the end of 96 weeks compared with non-rim lesions as secondary endpoints [20].

At present no study evaluated the effect of DMTs on iron rim positive lesions on QSM.

2.2. Conventional T1- and T2-weighted sequences

The direct evaluation of gradual linear expansion of WM lesions on conventional T1- and T2-weighted sequences, acquired at different timepoints and commonly available in RCTs and clinical setting, has been recently proposed to identify SELs [14].

By evaluating the pooled relapsing MS (RMS) population (n = 1334) of two phase III RCTs (OPERA I [NCT01247324] and OPERA II [NCT01412333]) and of primary progressive (PP) MS patients (n = 555) of the phase III ORATORIO RCT (NCT01194570), a recent study assessed the burden of lesions showing a gradual linear expansion over 96 to 120 weeks (i.e. SELs) [14].

The prevalence of SELs was similar in PPMS and RMS patients (71.9% and 68.2%). However, PPMS patients had a higher mean number of SELs (6.3 vs 4.6, p = 0.002), a higher mean T2-hyperintense volume defined as SELs (1838 vs 1223 mm³, p < 0.001), and a higher mean proportion of T2-hyperintense lesion volume identified as SELs (11.3% vs 8.6%, p < 0.001) [14]. SELs were also characterized by a lower T1 signal intensity compared with non-SELs, and a significant longitudinal T1 signal intensity decline that was not found in non-SEL [14], suggesting the occurrence of progressive demyelination and axonal damage.

By evaluating 555 PPMS patients of phase III ORATORIO RCT (NCT01194570), another study showed that SELs were characterized by a larger decrease in mean normalized T1 signal intensity and a greater relative accumulation of T1-hypointense volume that predicted 12-week confirmed composite disability progression (measured by Expanded Disability Status Scale worsening, ≥20% increase in Timed 25-Foot Walk time or in 9-Hole Peg test time) from baseline to Week 120 [13].

Of note, although the overall prevalence of SELs was not significantly influenced by treatment (proportion of PPMS patients with ≥1 SELs = 73.2% in ocrelizumab group vs 69.0%, in placebo group), ocrelizumab-treated patients had a lower proportion of total preexisting T2-hyperintense lesions identified as SEL (median = 2.5 vs 3.4, p = 0.044) from baseline to week 120 [13]. Moreover, compared with placebo, ocrelizumab, was associated with a lower T1-hypointense lesion volume increase from baseline to Week 120 (p < 0.001), not only in acute new T2-hyperintense lesions (p < 0.001), but also in preexisting T2-hyperintense lesions that were classified as SELs (+27% vs +40%, p < 0.001) and non-SELs (+15% vs +18%, p = 0.005) [13]. Finally, ocrelizumab was associated with a significant reduced decrease in normalized T1 signal intensity in SELs from baseline to Week 120 compared with placebo (−0.24 vs −0.28, p = 0.013) [13].

Since the accumulation of chronic active lesions may represent one of the contributors to clinical disability in PMS [9], the demonstration of a beneficial effect of ocrelizumab in limiting chronic inflammation, demyelination, and axonal loss could further support the use of this drug in limiting clinical worsening in these patients.

By applying a similar approach to identify SELs, another recent study evaluated the burden and microstructural features of SELs over two years in RRMS patients starting natalizumab (n = 28) or fingolimod (n = 24) [15]. Although natalizumab-group showed, compared to fingolimod-group, a lower proportion of patients with ≥1 SEL (46% vs 75%), lower number (median = 0 vs 2) and volume (median = 0.0 vs 1.7 ml) of T2-hyperintense lesions defined as SELs, the results did not survive correction for multiple comparisons [15]. Conversely, the volume change of SELs in natalizumab-group was significantly lower compared with fingolimod-group (median = 0.01 vs 0.10 ml, p = 0.002) [15]. Consistently with previous studies [13,14], SELs were also characterized by lower magnetization transfer ratio (MTR) and T1 signal intensity compared with non-SELs, and a significant longitudinal T1 signal intensity decline that was not found in non-SEL [15].

The results of this study suggested a limited effect of natalizumab and fingolimod in preventing SEL occurrence, possibly due to their specific mechanisms of action. Both drugs limit lymphocyte migration into the CNS and promote an excellent control of disease activity. Despite this, their efficacy could be more limited on smoldering inflammation that is more hardly targeted by currently available DMTs.

Clearly, further studies including larger cohorts of MS patients with different DMTs are still necessary to better investigate the beneficial effects of DMTs on SELs evaluated using conventional T1- and T2-weighted sequences.

2.3. Positron emission tomography

Thanks to the application of radiotracers specific for the different pathophysiological substrates of MS, such as inflammation and myelin dynamics, PET is another emerging technique for monitoring therapeutic targets [21].

PET studies showed a significant reduction of ¹¹C-(R)-PK11195, a radioligand for translocator protein specific for activated microglia, in T2-hyperintense WM lesions after 24 weeks of treatment with fingolimod [17], and in T2-hyperintense WM lesions, and the rim of chronic lesions in natalizumab-treated patients after 6–12 months [16,18].

Although these are encouraging findings, several factors should be considered before using PET for monitoring chronic active lesions. Specificity of PET radioligands to different therapeutic targets still needs to be fully validated, PET acquisition procedures and analyses in both cross-sectionally and longitudinally should be optimized and standardized, and further studies should explore the possible risks from serial PET imaging.

3. Expert Opinion

Chronic inflammation represents a relevant pathological process contributing to disease progression in MS. Accordingly, the in vivo evaluation using MRI of how specific treatments may influence the occurrence and microstructural features of chronic active lesions may contribute to determine the potential effects of different therapeutic strategies on a more compartmentalized inflammation that seems clinically relevant but not yet included in treatment monitoring.

The application of different approaches, including susceptibility-based imaging, longitudinal evaluation of T1- and T2-weighted sequences and PET, is suggesting that neuroimaging techniques are reliable to identify chronic inflammation in vivo, and, possibly, to monitor treatment response. These approaches, combined with other quantitative measures, such as T1 signal intensity, MTR, and diffusion tensor MRI [13–15,22], can better disentangle the pathological processes, such as ongoing demyelination and neuro-axonal loss, contributing to the progression of irreversible clinical disability.

Further effort is still necessary to make these biomarkers feasible in the clinical setting, defining and validating standardized MRI protocols for their assessment. Moreover, longitudinal studies with large cohorts of MS patients being treated with the different DMTs are needed to evaluate the effects of specific drugs in limiting such a chronic inflammation more strictly limited within the CNS.

Finally, it would be of interest to study how treatment effects on chronic active lesions may positively influence not only locomotor functions but also other disability outcomes such as cognitive performances.

Declaration of interest

P Preziosa received speaker honoraria from Biogen, Novartis, Merck Serono, Bristol Myers Squibb and ExceMED. He is supported by a senior research fellowship FISM – Fondazione Italiana Sclerosi Multipla - cod. 2019/BS/009 and financed or co-financed with the ‘5 per mille’ public funding. M. Filippi is Editor-in-Chief of the Journal of Neurology and Associate Editor of Radiology, Human Brain Mapping and Neurological Sciences, received compensation for consulting services and/or speaking activities from Almiral, Alexion, Bayer, Biogen, Celgene, Eli Lilly, Genzyme, Merck-Serono, Novartis, Roche, Sanofi, Takeda, and Teva Pharmaceutical Industries, and receives research support from Biogen Idec, Merck-Serono, Novartis and Roche. MA Rocca received speaker honoraria from Bayer, Biogen, Bristol Myers Squibb, Celgene, Genzyme, Merck Serono, Novartis, Roche, and Teva, and receives research support from the MS Society of Canada and Fondazione Italiana Sclerosi Multipla. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or conflict with the subject matter or materials discussed in this manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was not funded.