Characterization of chronic active multiple sclerosis lesions with sodium (23 Na) magnetic resonance imaging-preliminary observations.

BACKGROUND AND PURPOSE: There has been an increasing interest in chronic active multiple sclerosis (MS) lesions as a new magnetic resonance imaging (MRI) marker of disease progression. Chronic active lesions are characterized by progressive tissue matrix damage, axonal loss and chronic inflammation. Sodium (23 Na) MRI provides a biochemical marker of cell integrity and tissue viability in a quantitative manner. The aim of this study was to investigate with 23 Na MRI tissue abnormalities in chronic active lesions as indicators of tissue destruction. METHODS: To identify chronic active lesions, two 3D magnetization-prepared rapid acquisition gradient-echo datasets obtained 12 months apart were processed using the voxel-guided morphometry algorithm. Cross-sectional 23 Na MRI was performed during the 12-month follow-up period. Total sodium concentration was calculated in chronic active lesions compared to shrinking, chronic stable and acute contrast-enhancing lesions. RESULTS: Overall, 70 MS lesions (21 chronic active, 10 shrinking, 29 chronic stable lesions, 10 acute contrast-enhancing lesions) in 12 patients were included. Total sodium concentration in chronic active lesions (49.57 ± 8.47 mM) was significantly higher than in shrinking (42.16 ± 3.9 mM; p = 0.03) and chronic stable lesions (39.92 ± 4.82 mM; p < 0.001). Chronic active lesions showed similar sodium values compared to acute contrast-enhancing lesions (48.06 ± 6.65 mM; p = 0.97). No differences between shrinking and chronic stable lesions were observed (p = 0.89). CONCLUSION: High sodium values in chronic active MS lesions may be an indicator of ongoing inflammation and tissue damage.

Individual Assessment of Brain Tissue Changes in MS and the Effect of Focal Lesions on Short-Term Focal Atrophy Development in MS: A Voxel-Guided Morphometry Study.

We performed voxel-guided morphometry (VGM) investigating the mechanisms of brain atrophy in multiple sclerosis (MS) related to focal lesions. VGM maps detect regional brain changes when comparing 2 time points on high resolution T1-weighted (T1w) magnetic resonace imaging (MRI). Two T1w MR datasets from 92 relapsing-remitting MS patients obtained 12 months apart were analysed with VGM. New lesions and volume changes of focal MS lesions as well as in the surrounding tissue were identified by visual inspection on colour coded VGM maps. Lesions were dichotomized in active and inactive lesions. Active lesions, defined by either new lesions (NL) (volume increase > 5% in VGM), chronic enlarging lesions (CEL) (pre-existent T1w lesions with volume increase > 5%), or chronic shrinking lesions (CSL) (pre-existent T1w lesions with volume reduction > 5%) in VGM, were accompanied by tissue shrinkage in surrounding and/or functionally related regions. Volume loss within the corpus callosum was highly correlated with the number of lesions in its close proximity. Volume loss in the lateral geniculate nucleus was correlated with lesions along the optic radiation. VGM analysis provides strong evidence that all active lesion types (NL, CEL, and CSL) contribute to brain volume reduction in the vicinity of lesions and/or in anatomically and functionally related areas of the brain.

A generalizable deep voxel-guided morphometry algorithm for the detection of subtle lesion dynamics in multiple sclerosis.

Introduction: Multiple sclerosis (MS) is a chronic neurological disorder characterized by the progressive loss of myelin and axonal structures in the central nervous system. Accurate detection and monitoring of MS-related changes in brain structures are crucial for disease management and treatment evaluation. We propose a deep learning algorithm for creating Voxel-Guided Morphometry (VGM) maps from longitudinal MRI brain volumes for analyzing MS disease activity. Our approach focuses on developing a generalizable model that can effectively be applied to unseen datasets. Methods: Longitudinal MS patient high-resolution 3D T1-weighted follow-up imaging from three different MRI systems were analyzed. We employed a 3D residual U-Net architecture with attention mechanisms. The U-Net serves as the backbone, enabling spatial feature extraction from MRI volumes. Attention mechanisms are integrated to enhance the model's ability to capture relevant information and highlight salient regions. Furthermore, we incorporate image normalization by histogram matching and resampling techniques to improve the networks' ability to generalize to unseen datasets from different MRI systems across imaging centers. This ensures robust performance across diverse data sources. Results: Numerous experiments were conducted using a dataset of 71 longitudinal MRI brain volumes of MS patients. Our approach demonstrated a significant improvement of 4.3% in mean absolute error (MAE) against the state-of-the-art (SOTA) method. Furthermore, the algorithm's generalizability was evaluated on two unseen datasets (n = 116) with an average improvement of 4.2% in MAE over the SOTA approach. Discussion: Results confirm that the proposed approach is fast and robust and has the potential for broader clinical applicability.

Long-term dynamics of multiple sclerosis iron rim lesions.

BACKGROUND: Several studies have pointed out that seemingly chronic multiple sclerosis (MS) lesions may also be in inflammatory states. In pathological studies, up to 40% of chronic MS lesions are characterized as "chronic active" or "smoldering" lesions that are characterized by a rim of iron-laden proinflammatory macrophages/microglial cells at the lesion edge with low-grade continuous myelin breakdown. In vivo, these lesions can be visualized as "iron rim lesions" (IRLs) on susceptibility-weighted imaging (SWI). The aim of this study was to investigate the long-term dynamics of IRLs in vivo for a more detailed evolution of dynamic lesion volume changes occurring over time. METHODS: We retrospectively identified patients with MS who were followed for at least 36 months (up to 72 months) and underwent at least an annual MRI on the same 3 Tsystem. Using Voxel-Guided Morphometry (VGM) we investigated regional volume changes within lesions and correlated these findings with SWI for the presence of a characteristic hypointense lesion rim. To estimate tissue damage, apparent diffusion coefficient (ADC) values for every lesion at baseline and follow-up MRIs were determined. RESULTS: Forty-three patients were included in the study. Overall, we identified 302 supratentorial non-confluent MS lesions (52 persistent IRLs, nine transient IRLs, 228 non-IRLs and 13 acute contrast-enhancing lesions). During follow-up, persistent IRLs significantly enlarged, whereas non-IRLs showed a tendency to shrink. At baseline MRI, ADC values were significantly higher in persistent IRLs (1.23 × 10-3 mm/s2) compared to non-IRLs (1.01 × 10-3 mm/s2; p < 0.001), but not compared to transient IRLs (1.06 × 10-3 mm/s2; p = 0.15) and contrast-enhancing lesions (1.15 × 10-3 mm/s2; p = 1.0). During follow-up, ADC values significantly increased more often in persistent IRLs compared to all other lesion types (p < 0.0001). CONCLUSIONS: Our long-term data demonstrate that persistent IRLs enlarge during disease duration, whereas non-IRLs show a tendency to shrink. Furthermore, IRLs are associated with sustained tissue damage, supporting the notion that IRLs could represent a new imaging biomarker in MS.

Quantitative MRI texture analysis in chronic active multiple sclerosis lesions.

OBJECTIVE: Recently, there has been an increasing interest in "chronic enlarging" or "chronic active" multiple sclerosis (MS) lesions that are associated with clinical disability. However, investigation of dynamic lesion volume changes requires longitudinal MRI data from two or more time points. The aim of this study was to investigate the application of texture analysis (TA) on baseline T1-weighted 3D magnetization-prepared rapid acquisition gradient-echo (MPRAGE) images to differentiate chronic active from chronic stable MS lesions. MATERIAL AND METHODS: To identify chronic active lesions as compared to non-enhancing stable lesions, two MPRAGE datasets acquired on a 3 T MRI at baseline and after 12 months follow-up were applied to the Voxel-Guided Morphometry (VGM) algorithm. TA was performed on the baseline MPRAGE images, 36 texture features were extracted for each lesion. RESULTS: Overall, 374 chronic MS lesions (155 chronic active and 219 chronic stable lesions) from 60 MS patients were included in the final analysis. Multiple texture features including "DISCRETIZED_HISTO_Energy", "GLCM_Energy", "GLCM_Contrast" and "GLCM_Dissimilarity" were significantly higher in chronic active as compared to chronic stable lesions. Partial least squares regression yielded an area under the curve of 0.7 to differentiate both lesion types. CONCLUSION: Our results suggest that multiple texture features extracted from MPRAGE images indicate higher intralesional heterogeneity, however they demonstrate only a fair accuracy to differentiate chronic active from chronic stable MS lesions.

Venous Diameter Changes in Chronic Active Multiple Sclerosis Lesions.

BACKGROUND AND PURPOSE: To investigate the temporal evolution of venous diameter in chronic active and nonenhancing shrinking multiple sclerosis (MS) lesions in a longitudinal magnetic resonance imaging (MRI) study including susceptibility-weighted images (SWI). METHODS: We compared the venous diameter in chronic active and nonenhancing shrinking lesions to the venous diameter in nonenhancing stable lesions on two 3 T MRI data sets obtained 12 months apart. Chronic active and nonenhancing shrinking lesions were identified by Voxel-Guided Morphometry. Coregistered, overlaid fluid-attenuated inversion recovery/SWI were analyzed for the presence of a central vein. Quantitative calculation of the venous diameter for each time point was performed on the reconstructed veins. RESULTS: Sixty-two relapsing-remitting MS patients (50 women; mean age: 36 ± 11 years; mean disease duration: 4 ± 7 years) were included in the study. Overall, we identified 222 chronic MS lesions (48 chronic active, 48 shrinking, 126 stable) with a corresponding intralesional central vein. On baseline MRI, the mean venous diameter did not statistically differ between all subgroups, whereas on follow-up MRI, the mean intralesional venous diameter was smaller in chronic active (0.92 ± 0.15 mm) and shrinking lesions (0.90 ± 0.19 mm) compared to stable lesions (1.10 ± 0.18 mm; P < .001). CONCLUSION: Our findings demonstrate venous narrowing in chronic active and nonenhancing shrinking MS lesions. The smaller diameter of intralesional veins during follow up in these lesions may reflect structural, degenerative, and metabolic changes due to chronic inflammation, (perivascular) fibrosis, collagenous thickening, and increased levels of oxygenated hemoglobin.

Mobilisation of hematopoietic CD34+ precursor cells in patients with acute stroke is safe--results of an open-labeled non randomized phase I/II trial.

BACKGROUND: Regenerative strategies in the treatment of acute stroke may have great potential. Hematopoietic growth factors mobilize hematopoietic stem cells and may convey neuroprotective effects. We examined the safety, potential functional and structural changes, and CD34(+) cell-mobilization characteristics of G-CSF treatment in patients with acute ischemic stroke. METHODS AND RESULTS: Three cohorts of patients (8, 6, and 6 patients per cohort) were treated subcutaneously with 2.5, 5, or 10 µg/kg body weight rhG-CSF for 5 consecutive days within 12 hrs of onset of acute stroke. Standard treatment included i.v. thrombolysis. Safety monitoring consisted of obtaining standardized clinical assessment scores, monitoring of CD34(+) stem cells, blood chemistry, serial neuroradiology, and neuropsychology. Voxel-guided morphometry (VGM) enabled an assessment of changes in the patients' structural parenchyma. 20 patients (mean age 55 yrs) were enrolled in this study, 5 of whom received routine thrombolytic therapy with r-tPA. G-CSF treatment was discontinued in 4 patients because of unrelated adverse events. Mobilization of CD34(+) cells was observed with no concomitant changes in blood chemistry, except for an increase in the leukocyte count up to 75,500/µl. Neuroradiological and neuropsychological follow-up studies did not disclose any specific G-CSF toxicity. VGM findings indicated substantial atrophy of related hemispheres, a substantial increase in the CSF space, and a localized increase in parenchyma within the ischemic area in 2 patients. CONCLUSIONS: We demonstrate a good safety profile for daily administration of G-CSF when begun within 12 hours after onset of ischemic stroke and, in part in combination with routine i.v. thrombolysis. Additional analyses using VGM and a battery of neuropsychological tests indicated a positive functional and potentially structural effect of G-CSF treatment in some of our patients. TRIAL REGISTRATION: German Clinical Trial Register DRKS 00000723.