In this study, we aimed at assessing whether the iron deposition in the MS brain is a problem of quantity or distribution. The main findings were that the global iron load was not altered in MS compared to healthy controls, and the global iron concentrations did not significantly correlate with any of the clinical parameters. The iron concentrations in MS lesions and NAWM were altered considerably compared to controls, and the regional iron concentrations correlated with several clinical parameters, i.e., age, disease duration, EDSS, T25FW test, and DMTs duration.
In the previous literature, the vast majority of the studies focused on iron evaluation in the deep grey matter of MS patients, and a significant association between the iron load in these regions and the disability progression was reported [2, 9,10,11,12, 17, 18]. To the best of our knowledge, this is the first study to assess the global iron load in MS, not only the regional iron concentrations, and to study the correlation between the iron concentrations inside the MS lesions with the clinical profile.
The global iron load in our MS cohort was not significantly different from the global iron load in healthy controls even after adjustment for the loss of volume observed in MS patients. It seems that the noted aberrant iron deposition in MS is rather a problem of distribution, not overall global load. Additionally, no significant correlation was found between the global iron concentrations and any of the studied clinical parameters, which supports the idea that MS is not a disease of iron overload or deficiency. These findings are interesting as they might help explain why the previous therapeutic trials of chelation therapy and endovascular interventions failed to show a beneficial effect [31, 32]. Given these findings, the hypotheses that aberrant iron deposition in MS is due to altered iron influx or active clearance/elimination from the brain should be downweighed [6, 19]. Iron deposition in MS is likely a consequence of perturbed iron homeostasis, supported by the evidence of polymorphism of the genes encoding iron export from the cells, iron-binding, and iron transport in MS patients [43, 44].
In agreement with the previous literature findings, the regional iron concentrations in our MS cohort were different from the healthy controls. The methodology of measuring regional iron concentration in our study was different from what has been performed previously. We used fixed-size ROIs placed at selected areas rather than segmentation techniques. This allowed us to measure and compare the iron concentrations not only in the basal ganglia and thalami but also in the MS lesions and NAWM. In our cohort, the iron concentrations were high in the basal ganglia and low in thalami but not significantly different from the healthy control. The iron concentrations were significantly low in MS lesions and high in the NAWM compared to corresponding locations in healthy controls. High basal ganglia iron has been consistently reported in MS in the literature [37, 45,46,47].
In contrast, reports on thalamic iron concentrations were conflicting between the studies. Thalamic iron concentrations were reported to be lower , higher , or not significantly different from the healthy controls . The iron load was reported to be reduced inside the MS lesions in histological studies . It was reported to be shifted to the lesion periphery forming a rim in slowly expanding chronic active lesions [50, 51]. In our study, the iron load in NAWM was higher than the iron load in similar regions of healthy controls. In disagreement with this finding, Hametner et al.  reported low iron concentrations in their histopathological studies of the NAWM of four MS brains compared to three control brains. Similarly, their second study of formalin-fixed autopsies of 24 MS brains versus 18 controls revealed the same results . Their findings, however, were exclusive to the NAWM around the MS lesions edges, which is not the case in our study. Similar low iron concentrations were reported in NAWM of MS patients in a cross-sectional radiological study using R2* sequence for iron quantification . The patients recruited in their study had a longer disease duration (12.3 years versus 7.0 years in ours) . This might explain their difference from our results, given the accumulating evidence that the iron overload in different brain regions, even the basal ganglia, is reduced over time [6, 33, 53].
Taken altogether, the aberrant iron deposition could be concluded to be due to a process of shifting iron from certain regions to others inside the brain without affection of the overall global brain iron load. In a recently proposed explanation, iron was proposed to be reduced in brain regions where progressive damage to iron-containing cells (i.e., the oligodendrocytes and myelinated neurons) takes place, such as the MS lesions and thalami (the relay of several neurons where Wallerian degeneration is reflected) [19, 54] and increased in brain regions where chronic iron-rich microglia are activated such as the NAWM and basal ganglia [6, 14]. There seems to be a piece of evidence that the iron load is elevated in brain regions where chronic microglial activation and ongoing oligodendrocyte and myelin damage (the iron most rich structures in the brain) with subsequent iron deposition [6, 14, 55]. Over time, when the vast majority of myelin and oligodendrocytes are lost, the iron load is reduced [6, 19, 33]. This might explain the noted correlation between the regional iron concentrations and different clinical parameters in our cohort.
The patients’ disability (evaluated by the EDSS or T25FW test) was significantly correlated with high iron concentrations in the basal ganglia and lower concentrations in the thalami and MS lesions. Similarly, patients with aggressive MS had higher basal ganglia iron than benign patients reflecting more prominent pathology. This might also explain the differences in iron concentrations reported in the literature in different brain regions, as the concentrations largely depend on the disease duration [6, 19, 33, 49, 53]. In our cohort, MS patients on interferon-beta therapy had higher iron concentrations in MS lesions. The longer the interferon therapy was used, the less iron loss was observed in MS lesions. We propose that the interferon-beta therapy received might have reduced myelin damage and subsequent iron loss. However, there is no data in the literature that can substantiate such speculation; and the cross-sectional design of the study does not allow confirmation or negation of this explanation. Though we tried to adjust for the impact of disease duration on iron concentrations by selecting patients within a narrow range of disease duration (i.e., a range of 5 years), the duration was significantly associated with iron concentrations in the MS lesions in this relatively narrow range of years.
The main strength points of this study are that it is the first one assessing the global brain iron load in MS, to compare the iron concentrations inside the MS lesions as well as the deep grey matter and NAWM via fixed-sized ROIs to avoid brain volume issues and to measure the iron concentrations in corresponding locations to MS lesions in healthy controls.
However, this study has limitations. First, a selection bias could not be avoided to control for the confounding effect of several well-established clinical variables on iron concentration, such as the disease duration. Accordingly, the results should be cautiously interpreted and should not be generalised to patients outside the scope of the inclusion criteria. This makes the generalizability of the study results limited. Second, the ventricles and sulci could not be excluded during the measurement of global iron load. However, we do not expect the CSF iron to affect the global iron load measurement due to its dynamic nature. Finally, measurement of iron concentrations via fixed-sized ROIs in certain slices might be inaccurate if the iron concentrations were not evenly distributed inside the MS lesions, NAWM, or deep grey matter.
In conclusion, our results showed that the aberrant iron deposition in MS is likely a distribution problem rather than the overall iron load inside the brain. Iron global concentrations are comparable between MS patients and healthy controls, but the regional iron concentrations are significantly different with areas showing low iron concentrations (such as MS lesions and thalami) and others showing high MS concentrations (such as basal ganglia and NAWM).