Аккумуляция железа в множестве склерозов и последствия для когнитивной и клинической прогрессии: систематический обзор передовых исследований в области нейронизации.

Introduction

Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system (CNS) characterized by demyelination, axonal damage, and progressive neurodegeneration. The clinical manifestations include autonomic dysfunction, visual, and motor and sensory deficits, which are strictly related to the localization of lesions within the CNS. Along with these neurological signs and symptoms, cognitive impairment is highly prevalent in MS, even from the early stages of the disease, with estimates ranging from 40 to 65%. Cognitive impairment can also be associated with depressive symptoms, poor long-term prognosis, and increased disability, thus being responsible for poor quality of life [ 1 , 2 ]. Cognitive domains typically affected in MS include information processing speed, attention, memory, and executive functions, reflecting widespread network disruption rather than focal lesion burden [ 3 ]. Although inflammatory demyelination is a pathological hallmark, growing evidence indicates that neurodegeneration constitutes a core process that may occur early and independently of focal lesions [ 4 ]. Accordingly, cognitive impairment may not be the sole consequence of demyelination but may also reflect diffuse neurodegenerative patterns that disrupt large-scale brain networks [ 5 , 6 ]. Neurodegeneration is thought to arise from interacting pathophysiological mechanisms, such as chronic inflammation, oxidative stress, metabolic dysfunction, and iron dysregulation [ 7 ]. Neuropathological studies have identified iron-enriched, immunologically activated microglia and macrophages at the edges of a subset of chronically active MS lesions [ 8 , 9 ]. These lesions, commonly referred to as paramagnetic rim lesions (PRLs), can be detected in vivo using susceptibility-based MRI techniques, such as susceptibility-weighted imaging (SWI), R 1 and R 2* relaxometry, and quantitative susceptibility mapping (QSM), at both 3 T [ 10 ] and 7 T MRI [ 11 – 17 ]. PRLs have been associated with greater neurological disability, reduced brain volumes, and worse longitudinal outcomes, supporting their proposed role as in vivo biomarkers of chronic active inflammation and ongoing tissue damage [ 11 , 12 , 15 , 18 – 20 ]. Iron-related alterations in MS are also reported in deep gray matter (DGM) structures and may occur even before atrophy and demyelination, potentially contributing to early pathological changes and cognitive decline [ 21 – 25 ]. However, findings on the direction and magnitude of iron-related alterations in MS remain inconsistent, and their relationship with cognitive impairment is still underexplored. Against this background, the present systematic review aims to synthesize the available evidence on cerebral iron deposition assessed using susceptibility-based MRI techniques, with a specific focus on DGM structures in people with MS (pwMS). Specifically, we seek to clarify the association between iron accumulation and cognitive impairment and to evaluate the potential role of iron-sensitive MRI metrics as diagnostic and prognostic biomarkers of cognitive decline in MS.

Materials and methods

Search strategy

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [ 26 ]. Search keywords included: multiple sclerosis, magnetic resonance imaging, susceptibility, iron, deep grey matter, subcortical, thalamus, basal ganglia, caudate, and putamen. The following databases were searched in October 2025 to identify studies investigating the correlation between iron homeostasis and cognitive impairment in subjects with MS: PubMed, Scopus, Web of Science and Embase; no dedicated preprint servers were searched. The search string included combinations of the following terms: (“multiple sclerosis”[All Fields] OR “MS”[All Fields]) AND (“heavy Metal”[All Fields] OR (“iron”[Supplementary Concept] OR “iron”[All Fields] OR “iron”[MeSH Terms])) AND (“magnetic resonance imaging”[All Fields] OR “MRI”[All Fields] OR “SWI”[All Fields] OR “Susceptibility Weighted Imaging”[All Fields] OR “QSM”[All Fields] OR “Quantitative Susceptibility Mapping”[All Fields]) AND (“cognition”[MeSH Terms] OR “cognition”[All Fields] OR “cognitions”[All Fields] OR “cognitive”[All Fields] OR “cognitively”[All Fields] OR “cognitives”[All Fields] OR “cognitive impairment”[All Fields]). The search strategy was adapted as appropriate for each database. This research was not restricted by the year of publication of the articles considered. This study is registered with the International Prospective Register of Systematic Reviews (PROSPERO) under record number CRD420261281845.

Study selection

Study selection was based on the following inclusion criteria: (i) original or protocol studies of any type; (ii) adult patients with MS; (iii) patients undergoing MRI, including one of the following: R2, SWI, and QSM; (iv) microstructural changes involving iron deposition. We excluded studies that met any of the following criteria: (i) systematic, integrated, or narrative reviews, conferences, and editorials; (ii) duplicated studies; (iii) all articles written in other languages than English; (iv) other neurodegenerative conditions; (v) other different MRI parameters; (vi) histological-only studies of iron deposition; (vii) absence of cognitive assessment. Furthermore, studies that used other advanced MRI methods and did not specifically mention QSM values in the DGM were also excluded.

Data extraction

Following the full-text selection, data were extracted independently by two reviewers (C.S. and G.R.) and reported in a table using Microsoft Excel. For each eligible study, they collected information on: (i) general study characteristics (authors, year of publication, sample size, and follow-up duration); (ii) population characteristics (multiple sclerosis phenotype when available); (iii) MRI ( R 2*, SWI, QSM sequences); (iv) cognitive outcomes (neuropsychological tests or batteries employed, definition of cognitive impairment when available); and (v) main findings regarding the association between iron homeostasis and cognitive performance or clinical progression. Any doubt throughout the data extraction process between the two reviewers was resolved through discussion and, when required, consultation with a third reviewer (V.L.B.).

PECO evaluation

We employed a PECO (Population, Exposure, Comparison, Outcome) framework to structure our research question. The population consisted of adult patients (≥ 18 years) diagnosed with MS. The exposure was defined as iron dyshomeostasis reflecting cognitive decline and clinical progression, and investigated through MRI parameters involving QSM, R 1/ R 2*, and SWI-based sequences. The comparison groups included age- and sex-matched healthy controls (HC) with no history of neurological or psychiatric disorders or patients with different forms of multiple sclerosis. The outcome was the correlation between iron deposition and cognitive performance, evaluated using a standardized battery of neuropsychological tests. This approach enabled us to explore the role of iron deposition in cognitive decline and clinical progression of MS patients.

Assessing the quality of included studies: risk of bias

C.S. and G.R. independently evaluated the risk of bias in the results of the included studies. Any disagreements throughout this process, as well as in earlier stages, were settled by consulting V.L.B., a third reviewer, who provided the final decision. All eligible studies were assessed using ROBINS-E (Risk Of Bias In Non-randomized Studies of Exposures), a tool to assess risk of bias in non-randomized follow-up studies of exposure effects [ 27 ]. ROBINS-E evaluates bias across seven domains: (i) bias due to confounding (ii) bias arising from measurement of the exposure, (iii) bias in selection of participants into the study (or into the analysis), (iv) bias due to post-exposure interventions, (v) bias due to missing data, (vi) bias arising from measurement of the outcome, and (vii) bias in selection of reported result. Each domain was rated as “low risk of bias”, “some concerns”, “high risk of bias”, or “very high risk of bias”, and an overall judgement was derived accordingly. To graphically visualize the results obtained using the tool, a table was produced using Risk-of-bias VISualization ( robvis), a web app designed for visualizing risk-of-bias assessments performed as part of a systematic review [ 28 ].

Results

Study selection

A total number of 600 search entries were generated by searching on PubMed, EmBase, Scopus, and Web of Science databases (Fig. 1 ). These were reduced to 397 studies following deduplication. After the removal of duplicates, another 154 were excluded because of reviews, conferences, and editorials, and 204 after screening for title and abstract. Following the screening, 34 full-text articles were assessed for eligibility. After excluding additional 22 papers that did not meet inclusion criteria or provided insufficient data, 12 studies were suitable for analysis.

Study characteristics

Twelve studies were included, with sample sizes ranging from 29 to 640 pwMS. Most of the included studies ( n = 9) adopted a cross-sectional design, whereas the remaining implemented a longitudinal component ( n = 3). Study cohorts predominantly comprised individuals with relapsing–remitting MS (RRMS) followed by secondary progressive MS (SPMS), primary progressive MS (PPMS), and clinically isolated syndrome (CIS). A healthy control group was included in 7/12 studies, whereas the remaining studies relied on within-MS comparisons without external controls. Studies were conducted across Europe, North America, and the United Kingdom: 7/12 cohorts were recruited in European countries (Austria, Belgium, Germany, Italy, and Switzerland), 4/12 were conducted in North America (United States and Canada), and 1/12 in the United Kingdom. Austria and the United States were the most represented settings, contributing three studies each. Table 1 summarizes the clinical and demographic characteristics of the included studies [ 18 , 22 , 24 , 29 – 37 ].

MRI acquisition and iron-sensitive metrics

The scanner field strength was 3 T in 9 studies, 4.7 T in 1 study, and 7 T in 2 studies. Conventional MRI sequences (T1 and T2-weighted and FLAIR) were also included. Across the 12 included studies, QSM was the most frequently adopted technique ( n = 8), followed by R 2* relaxometry ( n = 5) and phase/SWI measures ( n = 2). However, several studies implemented combined or multiparametric approaches. Regarding structural brain features, the most frequently investigated targets were DGM structures, principally basal ganglia (putamen, caudate, and globus pallidus) and thalamus; furthermore, in studies on chronic active lesions, PRLs were assessed with SWI/QSM.

Cognitive assessment and clinical outcomes

While a few studies adopted a comprehensive neuropsychological battery, such as the Brief Repeatable Battery of Neuropsychological Tests (BRB-N), most relied on domain-specific measures. Processing speed was the primary cognitive outcome, mainly assessed with the Symbol Digit Modalities Test (SDMT). Several studies also evaluated executive functions (e.g., Flanker task and Stroop test), verbal (e.g., California Verbal Learning Test-II [CVLT-II] or Verbal Learning List Test [VLL]), and visuospatial memory (e.g. Brief Visuospatial Memory Test-Revised [BVMT-R]) and global cognition in selected protocols.

Clinical disability was assessed in all the included studies using the Expanded Disability Status Scale (EDSS) and many studies included motor/functional measures, such as the Nine-Hole Peg Test (9HPT) for manual dexterity, the 25-foot walk test (T25FW) for motor performance. Fatigue scale for motor and cognitive functions (FSCF) and Sloan low-contrast visual acuity (LCVA) were also administered to examine fatigue and visual function.

Adjustment for brain or regional atrophy was heterogeneous across the included studies. Six studies explicitly accounted for volumetric measures or atrophy in their statistical models, and iron-outcome model, whereas the remaining studies either analyzed atrophy separately or did not clearly adjust iron-related associations for atrophy.

MRI acquisition parameters, iron-sensitive metrics, neuropsychological assessments, and main findings are summarized in Table 2 .

Synthesis of evidence

Basal ganglia susceptibility/iron and cognition

Multiple studies reported associations between higher basal ganglia iron-related metrics and poorer cognitive performance. Increased iron-related measures in the putamen and caudate were associated with worse SDMT performance [ 30 ]. Schmalbrock and colleagues [ 35 ] coherently reported that higher susceptibility in the caudate and anterior putamen was associated with poorer inhibitory control. Higher susceptibility was also reported in the globus pallidus in association with a lower cognitive composite score (memory and processing speed) and independent of atrophy [ 22 ]. In a multiparametric framework, basal ganglia quantitative markers contributed to cognitive efficiency alongside conventional MRI variables [ 33 ]. In a large 3 T QSM cohort, higher basal ganglia susceptibility was associated with higher lesion burden and lower SDMT scores after adjustment for relevant covariates [ 36 ].

Thalamic susceptibility: directionality and clinical meaning

Across included studies and metrics, findings regarding the thalamus were heterogeneous. Thalamic atrophy emerged as a dominant predictor of cognitive performance in some cohorts, with thalamic QSM/ R 2* providing limited additional modulation [ 22 ], while in a susceptibility-myelin framework, R 1 alterations and atrophy showed stronger relationships with SDMT than susceptibility [ 34 ]. In early MS, thalamic QSM associations with lesion burden were modest and baseline QSM did not predict short-term clinical outcomes, suggesting a predominantly cross-sectional signal in that context [ 36 ]. In SPMS, thalamic susceptibility differed from controls and was related to clinical and cognitive disability [ 37 ]. Similarly, a thalamus-focused multiparametric MRI work integrating susceptibility with complementary microstructural metrics reported that thalamic quantitative markers were significant predictors of clinical and cognitive disability, even when accounting for lesion load and atrophy [ 29 ].

Memory and broader cognitive domains

Evidence linking DGM iron to memory outcomes was more heterogeneous than processing speed outcomes. Phase/SWI-based estimates in the caudate, putamen, globus pallidus, and pulvinar nucleus showed moderate associations with specific memory measures, although volumetric measures often explained a larger share of variance [ 24 ]. Fujiwara et al. [ 22 ] observed that iron levels across DGM structures correlated with deficits in both memory and processing speed. Overall, across studies incorporating verbal learning measures (e.g., CVLT-II) and broader batteries, memory performance frequently tracked thalamic/hippocampal and global neurodegenerative markers more robustly than iron-only indices [ 22 , 33 ].

Chronic active lesions, neurodegeneration, and cognitive performance

Two studies directly examined chronic active inflammation using rim lesion markers. In a 7 T framework targeting iron rim lesions (IRLs), patients with rim lesions exhibited smaller thalamic volume and slightly worse SDMT performance compared with non-IRL patients, consistent with a more aggressive pathological profile [ 18 ]. In a large 3 T QSM cohort, the presence of at least one rim + lesion was associated with worse cognitive performance (including SDMT and verbal learning measures) and greater structural neurodegeneration, particularly involving thalamic and cortical volumes [ 32 ]. Taken together, these findings support lesions as markers of a more severe disease phenotype with measurable cognitive and neurodegenerative consequences [ 18 , 32 ]. Moreover, in a multiparametric analysis of normal-appearing tissue, quantitative microstructural parameters (including R 2* within an MPM protocol) were associated with clinical and cognitive measures, with abnormalities detected across normal-appearing gray and white matter compartments [ 31 ].

Quality assessment: risk of bias

As presented in Fig. 2 , the ROBINS-E assessment showed an overall risk of bias ranging from some concerns to very high risk. Most study results were judged as presenting some concerns, whereas one [ 29 ] was assessed as being at very high risk of bias, while the remaining were judged at high risk of bias.

Across domains, confounding (D1) emerged as the primary source of bias. Key confounding factors were not fully addressed in any of the study results. This was mainly related to the observational design and the limited availability of key prognostic variables, influencing both exposure and outcomes. Consequently, bias ranged from some concerns to high risk. Bias arising from measurement of the exposure (D2) was generally judged to be at low risk of bias; some concerns were identified in three study results [ 18 , 29 , 32 ], mainly due to the use of indirect exposure measures and limited information on their validity and reliability. Selection of participants into the study (D3) was most often judged as presenting some concerns. Although study participants were generally well described, several studies did not clearly report inclusion and exclusion criteria and rarely recruited consecutive or population-based samples, limiting the ability to fully exclude selection mechanisms related to exposure and outcome. Post-exposure interventions (D4) were consistently judged at low risk of bias, as no relevant interventions influenced by prior exposure were identified or were unlikely to meaningfully affect the outcomes of interest. Missing data (D5) were identified as a source of concern in several studies, mainly due to limited or absent reporting of missing data and information on how missingness was handled in the analyses. Outcome measurement (D6) was judged at low risk of bias in most study results; some concerns were identified in two study results [ 22 , 32 ] mainly due to limited information on whether outcome assessment could have differed by exposure status or assessor blinding. Notably, more than half of the included studies involved two or more raters in the visual assessment of rim lesions, but intra-rater and inter-rater reliability was rarely reported, although many studies provided information on investigators’ level of experience. Selection of the reported result (D7) was mostly judged as presenting some concerns, mainly due to the absence of pre-specified analysis plans and the potential for selective reporting among multiple analyses, outcomes, or exposure definitions.

Discussion

This systematic review synthesizes evidence from 12 studies applying iron-sensitive quantitative MRI to examine the relationship between DGM susceptibility/iron-related metrics and cognitive performance in MS. Across heterogeneous cohorts and acquisition approaches, two converging patterns emerged: (i) higher basal ganglia susceptibility/iron-related metrics are consistently linked to worse cognitive efficiency, particularly information processing speed and executive control; and (ii) thalamic findings are more variable in direction and clinical meaning, with susceptibility often capturing a mixed signal that overlaps with (and is frequently outperformed by) thalamic atrophy.

From a network perspective, these regional effects are plausible. The basal ganglia and thalamus are central nodes within cortico-subcortical loops supporting processing speed, cognitive control, and behavioral inhibition [ 38 – 40 ]. Iron dyshomeostasis in these structures may contribute to a neurotoxic environment sustained by chronic microglial activation, mechanisms repeatedly implicated in MS-related neurodegeneration [ 41 , 42 ]. Basal ganglia measures showed the most reproducible cognitive associations. Across QSM/ R 2* studies, higher putaminal and caudate susceptibility was related to poorer SDMT performance and/or weaker inhibitory control, sometimes persisting after accounting for lesion burden and regional atrophy [ 22 , 30 , 35 , 36 ]. Importantly, multiparametric models indicate that iron-related indices can add explanatory value beyond conventional MRI, suggesting that basal ganglia susceptibility may capture microstructural injury not fully indexed by lesion load or volumetry [ 22 , 33 ]. Earlier work using T2 hypointensity as a proxy of iron similarly linked subcortical signal changes to cognitive dysfunction, providing convergent support for a basal ganglia contribution to cognitive impairment [ 43 ]. On the other hand, thalamic susceptibility findings were heterogeneous. Several studies emphasized that thalamic volume and/or myelin-sensitive metrics were stronger correlates of processing speed than susceptibility, particularly in earlier disease stages or mixed phenotypes [ 22 , 34 , 36 ]. Nevertheless, in progressive MS or in thalamus-focused multiparametric protocols, thalamic susceptibility and complementary quantitative measures were associated with both clinical and cognitive disability [ 29 , 37 ]. This apparent inconsistency likely reflects the fact that thalamic susceptibility is influenced by multiple components and may be phenotype-dependent. In fact, mechanistic models propose that thalamic susceptibility can decrease with longer disease duration, potentially reflecting inflammatory/oligodendroglial disruption and reduced iron storage capacity, while other regions, such as the basal ganglia, may demonstrate net iron accumulation [ 19 ].

In DGM structures, particularly in the thalamus, susceptibility may reflect a composite signal influenced not only by iron content, but also by tissue loss, myelin changes, and atrophy-related effects [ 19 , 34 ]. By contrast, susceptibility changes associated with PRLs likely reflect a distinct pathological substrate, namely chronic active inflammation sustained by iron-laden microglia and macrophages at the lesion edge [ 8 , 9 , 11 ]. Therefore, PRL-related susceptibility and DGM susceptibility should be considered related but not interchangeable markers of MS pathology [ 19 , 34 , 37 ]. A related issue is the linkage between iron-related metrics and macroscopic neurodegeneration. The included evidence suggests that susceptibility abnormalities can be detected even when volumetric loss is limited, supporting the view that iron-sensitive MRI may act as an early indicator of tissue distress. However, when entered in regression models, thalamic atrophy often remains the most robust independent predictor of cognitive performance, with susceptibility providing incremental explanatory value [ 22 , 33 , 34 ]. Longitudinal datasets further indicate that DGM susceptibility evolves over time and relates to disability, reinforcing its potential utility for tracking disease-related microstructural change together with morphometric measures [ 23 , 25 ]. Beyond processing speed, evidence for memory and other cognitive domains was less consistent, partly because fewer studies implemented comprehensive cognitive batteries. While some reports linked DGM susceptibility (including pulvinar or globus pallidus measures) to memory-related outcomes, other analyses suggested that learning and memory are more strongly associated with broader neurodegenerative markers (e.g., thalamic, hippocampal, and cortical atrophy) than with solely iron-sensitive metrics [ 22 , 24 , 33 ]. This divergence in the literature highlights a methodological gap: domain-specific conclusions about iron and memory/executive functions remain provisional without harmonized testing and consistent covariate control. Two studies also linked chronic active lesion biology to cognition. At ultra-high field, iron rim lesions were associated with smaller thalamic volume and modestly worse processing speed, while at 3 T, the presence of rim-positive lesions related to worse cognitive performance and more pronounced subcortical/cortical neurodegeneration [ 18 , 32 ]. These findings align with the concept that iron-laden microglia at lesion edges sustain compartmentalized inflammation and may contribute to network-level degeneration of connected GM, offering a plausible bridge between focal lesion pathology and diffuse cognitive decline [ 11 , 44 ]. Consistently, quantitative microstructural alterations in normal-appearing tissue (including susceptibility-related parameters within multiparametric protocols) have also been associated with clinical and cognitive measures, suggesting a broader substrate than focal lesions alone [ 31 ].

Taken together, the reviewed evidence supports iron-sensitive MRI as a clinically informative marker of MS-related brain injury with regional specificity. Basal ganglia susceptibility appears most consistently linked to processing speed and executive control, whereas thalamic susceptibility captures a more complex, stage-dependent signal that frequently interacts with myelin changes and atrophy. Clinically, higher susceptibility in DGM and the presence of rim lesions tend to co-occur with higher disability and greater neurodegeneration, supporting an iron-associated phenotype of more severe disease burden [ 18 , 29 , 30 , 32 , 37 ].

Limitations and future directions

Given the several methodological limitations, the interpretation of associations between cerebral iron-related MRI metrics, cognitive impairment, and clinical outcomes in MS should be made with caution. According to ROBINS-E assessments, residual confounding (D1) represents the main source of uncertainty. Incomplete adjustment for key prognostic factors, such as disease duration, phenotype, lesion burden, treatment exposure and baseline disability, may have biased effect estimates or likely contributed to heterogeneity across studies. Bias related to exposure and outcome measurement (D2, D6) was generally low, supporting the reliability of QSM, R 2*, and SWI metrics and of the cognitive/clinical assessments. However, other methodological constraints, including non-consecutive sampling, incomplete eligibility reporting (D3), insufficient handling of missing data (D5), and lack of pre-specified analytical strategies (D7), may affect generalizability and precision, while raising the possibility of selective reporting.

Overall, the literature on cerebral iron in MS is highly heterogeneous. Variability arises from differences in MRI techniques, reconstruction methods, and processing pipelines. Even within QSM studies, direct comparisons are challenging due to differences in susceptibility referencing, phase processing techniques [ 45 , 46 ], and algorithms [ 47 ]. Comparisons with non-QSM methods, such as R 2* mapping, are further complicated by their differential sensitivity to iron microdistribution, chemical form, and tissue microstructure [ 48 ]. Additional variability may reflect a non-linear aging trajectory of thalamic iron concentration [ 49 ], as well as clinical heterogeneity across MS cohorts.

The causes of iron-related signal variation in MS are still poorly understood. Iron-sensitive MRI metrics may not exclusively reflect static iron deposition, but also dynamic pathological processes, such as chronic pro-inflammatory activity within WM lesions [ 50 ] and early stages of structural atrophy [ 51 – 53 ]. Current evidence suggests a close association between structural atrophy and increased iron density, highlighting that these processes are likely intertwined rather than independent.

Finally, heterogeneity of study populations, the lack of standardized protocols, and the inconsistent adjustment for relevant covariates, in particular brain and regional atrophy, highlight the need for investigations in more homogeneous MS cohorts at well-defined disease stages, ideally using longitudinal designs with harmonized MRI protocols and rigorous confounder control. Because the available literature is still largely cross-sectional, future studies should address these limitations to improve comparability and clarify causal and prognostic relationships before iron-sensitive MRI metrics can be established as reliable prognostic biomarkers of cognitive impairment in MS.

Author contributions

CS and VLB conceptualized the paper. CS and GR conducted a literature search and performed study selection, data extraction, and quality assessment. GT and CS designed and conducted data analysis. GM informed the interpretation of the findings and their implications. CS, GR, and VLB prepared the first draft of the manuscript. GD, SC, ES, and AQ contributed to visualization. All authors checked the study results and revised the manuscript to the final version.

Funding

This work was supported by Current Research Fund 2025, Ministry of Health, Italy.

Data availability

The present article is a systematic review of already published literature, thus no new data was analyzed.

Declarations

Conflict of interest

The authors declare no conflicts of interest.

5 References

1. Chiaravalloti ND, DeLuca J (2008) Cognitive impairment in multiple sclerosis. Lancet Neurol 7:1139–1151. 10.1016/S1474-4422(08)70259-X

2. Spence H, McNeil CJ, Waiter GD (2020) The impact of brain iron accumulation on cognition: a systematic review. PLoS ONE 15:e0240697. 10.1371/journal.pone.0240697

3. Mirabile A, Susinna C, Cipriano GL, D’Aleo G, Rifici C, Sessa E et al (2025) Correlation of cognitive dysfunctions and diffusion tensor MRI measures in subjects with RRMS. Front Aging Neurosci 17:1661821. 10.3389/fnagi.2025.1661821

4. Mey GM, Mahajan KR, DeSilva TM (2023) Neurodegeneration in multiple sclerosis. WIREs Mech Dis 15:e1583. 10.1002/wsbm.1583

5. Colato E, Stutters J, Tur C, Sridar N, Arnold D, Kingshott CW et al (2021) Predicting disability progression and cognitive worsening in multiple sclerosis using gray matter network measures. Neurology. 10.1101/2021.03.23.21253388

6. Rimkus CdeM, Nucci MP, Avolio IB, Apóstolos-Pereira SL, Callegaro D, Wagner MB et al (2024) Atrophy patterns in patients with multiple sclerosis with cognitive impairment, fatigue, and mood disorders. Neurology 103:e210080. 10.1212/WNL.0000000000210080. ( Wolters Kluwer )

7. Reich DS, Lucchinetti CF, Calabresi PA (2018) Multiple sclerosis. N Engl J Med 378:169–180. 10.1056/NEJMra1401483

8. Frischer JM, Weigand SD, Guo Y, Kale N, Parisi JE, Pirko I et al (2015) Clinical and pathological insights into the dynamic nature of the white matter multiple sclerosis plaque. Ann Neurol 78:710–721. 10.1002/ana.24497

9. Hametner S, Wimmer I, Haider L, Pfeifenbring S, Brück W, Lassmann H (2013) Iron and neurodegeneration in the multiple sclerosis brain. Ann Neurol 74:848–861. 10.1002/ana.23974

10. Maggi P, Sati P, Nair G, Cortese ICM, Jacobson S, Smith BR et al (2020) Paramagnetic rim lesions are specific to multiple sclerosis: an international multicenter 3T MRI study. Ann Neurol 88:1034–1042. 10.1002/ana.25877

11. Absinta M, Sati P, Schindler M, Leibovitch EC, Ohayon J, Wu T et al (2016) Persistent 7-tesla phase rim predicts poor outcome in new multiple sclerosis patient lesions. J Clin Invest 126:2597–2609. 10.1172/JCI86198

12. Absinta M, Sati P, Masuzzo F, Nair G, Sethi V, Kolb H et al (2019) Association of chronic active multiple sclerosis lesions with disability in vivo. JAMA Neurol 76:1474. 10.1001/jamaneurol.2019.2399

13. Blindenbacher N, Brunner E, Asseyer S, Scheel M, Siebert N, Rasche L et al (2020) Evaluation of the ‘ring sign’ and the ‘core sign’ as a magnetic resonance imaging marker of disease activity and progression in clinically isolated syndrome and early multiple sclerosis. Mult Scler J Exp Transl Clin 6:2055217320915480. 10.1177/2055217320915480

14. Chawla S, Kister I, Sinnecker T, Wuerfel J, Brisset J-C, Paul F et al (2018) Longitudinal study of multiple sclerosis lesions using ultra-high field (7T) multiparametric MR imaging. PLoS ONE 13:e0202918. 10.1371/journal.pone.0202918

15. Dal-Bianco A, Grabner G, Kronnerwetter C, Weber M, Höftberger R, Berger T et al (2017) Slow expansion of multiple sclerosis iron rim lesions: pathology and 7 T magnetic resonance imaging. Acta Neuropathol 133:25–42. 10.1007/s00401-016-1636-z

16. Hammond KE, Metcalf M, Carvajal L, Okuda DT, Srinivasan R, Vigneron D et al (2008) Quantitative in vivo magnetic resonance imaging of multiple sclerosis at 7 Tesla with sensitivity to iron. Ann Neurol 64:707–713. 10.1002/ana.21582

17. Pitt D, Boster A, Pei W, Wohleb E, Jasne A, Zachariah CR et al (2010) Imaging cortical lesions in multiple sclerosis with ultra-high-field magnetic resonance imaging. Arch Neurol. 10.1001/archneurol.2010.148

18. Dal-Bianco A, Grabner G, Kronnerwetter C, Weber M, Kornek B, Kasprian G et al (2021) Long-term evolution of Multiple Sclerosis iron rim lesions in 7 T MRI. Brain 144:833–847. 10.1093/brain/awaa436

19. Schweser F, Raffaini Duarte Martins AL, Hagemeier J, Lin F, Hanspach J, Weinstock-Guttman B et al (2018) Mapping of thalamic magnetic susceptibility in multiple sclerosis indicates decreasing iron with disease duration: a proposed mechanistic relationship between inflammation and oligodendrocyte vitality. Neuroimage 167:438–452. 10.1016/j.neuroimage.2017.10.063

20. Ward RJ, Dexter DT, Crichton RR (2022) Iron, neuroinflammation and neurodegeneration. Int J Mol Sci 23:7267. 10.3390/ijms23137267

21. De Lury AD, Bisulca JA, Lee JS, Altaf MD, Coyle PK, Duong TQ (2023) Magnetic resonance imaging detection of deep gray matter iron deposition in multiple sclerosis: a systematic review. J Neurol Sci 453:120816. 10.1016/j.jns.2023.120816

22. Fujiwara E, Kmech JA, Cobzas D, Sun H, Seres P, Blevins G et al (2017) Cognitive implications of deep gray matter iron in multiple sclerosis. AJNR Am J Neuroradiol 38:942–948. 10.3174/ajnr.A5109

23. Hagemeier J, Zivadinov R, Dwyer MG, Polak P, Bergsland N, Weinstock-Guttman B et al (2018) Changes of deep gray matter magnetic susceptibility over 2 years in multiple sclerosis and healthy control brain. Neuroimage Clin 18:1007–1016. 10.1016/j.nicl.2017.04.008

24. Modica CM, Zivadinov R, Dwyer MG, Bergsland N, Weeks AR, Benedict RHB (2015) Iron and volume in the deep gray matter: association with cognitive impairment in multiple sclerosis. AJNR Am J Neuroradiol 36:57–62. 10.3174/ajnr.A3998

25. Zivadinov R, Tavazzi E, Bergsland N, Hagemeier J, Lin F, Dwyer MG et al (2018) Brain iron at quantitative MRI is associated with disability in multiple sclerosis. Radiology 289:487–496. 10.1148/radiol.2018180136

26. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD et al (2021) The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 372:n71. 10.1136/bmj.n71

27. Higgins JPT, Morgan RL, Rooney AA, Taylor KW, Thayer KA, Silva RA et al (2024) A tool to assess risk of bias in non-randomized follow-up studies of exposure effects (ROBINS-E). Environ Int 186:108602. 10.1016/j.envint.2024.108602

28. McGuinness LA, Higgins JPT (2021) Risk-of-bias VISualization (robvis): An R package and Shiny web app for visualizing risk-of-bias assessments. Res Synth Methods 12:55–61. 10.1002/jrsm.1411

29. Cagol A, Ocampo-Pineda M, Lu P-J, Weigel M, Barakovic M, Melie-Garcia L et al (2024) Advanced quantitative MRI unveils microstructural thalamic changes reflecting disease progression in multiple sclerosis. Neurol Neuroimmunol Neuroinflamm 11:e200299. 10.1212/NXI.0000000000200299

30. Khalil M, Langkammer C, Ropele S, Petrovic K, Wallner-Blazek M, Loitfelder M et al (2011) Determinants of brain iron in multiple sclerosis: a quantitative 3T MRI study. Neurology 77:1691–1697. 10.1212/WNL.0b013e318236ef0e

31. Lommers E, Simon J, Reuter G, Delrue G, Dive D, Degueldre C et al (2019) Multiparameter MRI quantification of microstructural tissue alterations in multiple sclerosis. Neuroimage Clin 23:101879. 10.1016/j.nicl.2019.101879

32. Marcille M, Hurtado Rúa S, Tyshkov C, Jaywant A, Comunale J, Kaunzner UW et al (2022) Disease correlates of rim lesions on quantitative susceptibility mapping in multiple sclerosis. Sci Rep 12:4411. 10.1038/s41598-022-08477-6

33. Pinter D, Khalil M, Pichler A, Langkammer C, Ropele S, Marschik PB et al (2015) Predictive value of different conventional and non-conventional MRI-parameters for specific domains of cognitive function in multiple sclerosis. Neuroimage Clin 7:715–720. 10.1016/j.nicl.2015.02.023

34. Pontillo G, Petracca M, Monti S, Quarantelli M, Lanzillo R, Costabile T et al (2022) Clinical correlates of R1 relaxometry and magnetic susceptibility changes in multiple sclerosis: a multi-parameter quantitative MRI study of brain iron and myelin. Eur Radiol 33:2185–2194. 10.1007/s00330-022-09154-y

35. Schmalbrock P, Prakash RS, Schirda B, Janssen A, Yang GK, Russell M et al (2016) Basal ganglia iron in patients with multiple sclerosis measured with 7T quantitative susceptibility mapping correlates with inhibitory control. AJNR Am J Neuroradiol 37:439–446. 10.3174/ajnr.A4599

36. Voon CC, Meineke J, Wiltgen T, McGinnis J, Berg R, Preibisch C et al (2025) Quantitative susceptibility mapping of deep grey matter in MS: association with clinical scores and brain volume measures. Brain Behav 15:e70988. 10.1002/brb3.70988

37. Williams T, John N, Calvi A, Bianchi A, De Angelis F, Doshi A et al (2024) Investigating the relationship between thalamic iron concentration and disease severity in secondary progressive multiple sclerosis using quantitative susceptibility mapping: cross-sectional analysis from the MS-STAT2 randomised controlled trial. Neuroimage Rep 4:100216. 10.1016/j.ynirp.2024.100216

38. Halassa MM, Sherman SM (2019) Thalamocortical circuit motifs: a general framework. Neuron 103:762–770. 10.1016/j.neuron.2019.06.005

39. Hanakawa T, Goldfine AM, Hallett M (2017) A common function of basal ganglia-cortical circuits subserving speed in both motor and cognitive domains. eNeuro 4:ENEURO.0200-17.2017. 10.1523/ENEURO.0200-17.2017

40. Lanciego JL, Luquin N, Obeso JA (2012) Functional neuroanatomy of the basal ganglia. Cold Spring Harb Perspect Med 2:a009621–a009621. 10.1101/cshperspect.a009621

41. Mahad DH, Trapp BD, Lassmann H (2015) Pathological mechanisms in progressive multiple sclerosis. Lancet Neurol 14:183–193. 10.1016/S1474-4422(14)70256-X

42. Stephenson E, Nathoo N, Mahjoub Y, Dunn JF, Yong VW (2014) Iron in multiple sclerosis: roles in neurodegeneration and repair. Nat Rev Neurol 10:459–468. 10.1038/nrneurol.2014.118

43. Brass SD, Benedict RH, Weinstock-Guttman B, Munschauer F, Bakshi R (2006) Cognitive impairment is associated with subcortical magnetic resonance imaging grey matter T2 hypointensity in multiple sclerosis. Mult Scler 12:437–444. 10.1191/135248506ms1301oa

44. Treaba CA, Conti A, Klawiter EC, Barletta VT, Herranz E, Mehndiratta A et al (2021) Cortical and phase rim lesions on 7 T MRI as markers of multiple sclerosis disease progression. Brain Commun 3:fcab134. 10.1093/braincomms/fcab134

45. Özbay PS, Duerst Y, Wilm BJ, Pruessmann KP, Nanz D (2018) Enhanced quantitative susceptibility mapping (QSM) using real-time field control. Magn Reson Med 79:770–778. 10.1002/mrm.26735

46. Schweser F, Robinson SD, De Rochefort L, Li W, Bredies K (2017) An illustrated comparison of processing methods for phase MRI and QSM: Removal of background field contributions from sources outside the region of interest. NMR Biomed 30:e3604. 10.1002/nbm.3604

47. Wang Y, Liu T (2015) Quantitative susceptibility mapping (QSM): decoding mri data for a tissue magnetic biomarker. Magnetic Resonance in Med 73:82–101. 10.1002/mrm.25358

48. Dietrich O, Levin J, Ahmadi S-A, Plate A, Reiser MF, Bötzel K et al (2017) MR imaging differentiation of Fe2+ and Fe3+ based on relaxation and magnetic susceptibility properties. Neuroradiology 59:403–409. 10.1007/s00234-017-1813-3

49. Hallgren B, Sourander P (1958) The effect of age on the non‐haemin iron in the human brain. J Neurochem 3:41–51. 10.1111/j.1471-4159.1958.tb12607.x

50. Mehta V, Pei W, Yang G, Li S, Swamy E, Boster A et al (2013) Iron is a sensitive biomarker for inflammation in multiple sclerosis lesions. PLoS ONE 8:e57573. 10.1371/journal.pone.0057573

51. Blazejewska AI, Al‐Radaideh AM, Wharton S, Lim SY, Bowtell RW, Constantinescu CS et al (2015) Increase in the iron content of the substantia nigra and red nucleus in multiple sclerosis and clinically isolated syndrome: A 7 Tesla MRI study. Magn Reson Imaging 41:1065–1070. 10.1002/jmri.24644

52. Hagemeier J (2013) Gray matter SWI-filtered phase and atrophy are linked to disability in MS. Front Biosci E5:525–532. 10.2741/E634

53. Hagemeier J, Weinstock-Guttman B, Bergsland N, Heininen-Brown M, Carl E, Kennedy C et al (2012) Iron deposition on SWI-filtered phase in the subcortical deep gray matter of patients with clinically isolated syndrome may precede structure-specific atrophy. AJNR Am J Neuroradiol 33:1596–1601. 10.3174/ajnr.A3030

Associated Data

Data Availability Statement

The present article is a systematic review of already published literature, thus no new data was analyzed.