Introduction. Smoldering lesions (also known as paramagnetic rim lesions, PRLs) are chronic active lesions that are associated with higher disability and faster progression to secondary progressive multiple sclerosis (MS). The factors contributing to their formation, and their association with the thickness of the peripapillary retinal nerve fiber layer (pRNFL), the ganglion cell-inner plexiform layer (GCIPL) and the inner nuclear layer (INL) remain to be elucidated.
Objectives. Our aim was to determine the role of haptoglobin gene or genotype and hemolysis parameters in the formation of PRLs, and their association with the retinal layer thickness (pRNFL, GCIPL, INL). In addition, we focused on the pathophysiological mechanisms of iron uptake into myeloid cells at the lesion edge, and the radiological correlate of faster disability progression in patients with PRLs.
Methods. We performed four independent retrospective studies including a total of 310 patients with relapsing-remitting and secondary progressive MS who underwent 3T brain MRI according to a standard protocol (T1-, FLAIR-, and SWI-weighted sequence, and in one of the studies also MDME-weighted sequence). Hemolysis parameters (erythrocyte count, reticulocyte count, hemoglobin concentration, hematocrit, potassium concentration, iron concentration, bilirubin concentration, free hemoglobin concentration, hemolysis index, lactate dehydrogenase concentration, fibrinogen concentration, aspartate transaminase concentration) were determined at the time of the blood draw. DNA was isolated from the blood of patients with MS, and the haptoglobin genotype was determined by allelic amplification by polymerase chain reaction. The concentration of soluble CD163 (sCD163) was determined in CSF samples of newly diagnosed patients. To determine the association between the number of PRLs and the retinal layer thickness, patients underwent optical coherence tomography (OCT). To determine the pathophysiological mechanisms of iron uptake into myeloid cells, brain tissue samples obtained at autopsy from 18 controls and 24 patients with MS were formalin-fixed, paraffin-embedded (FFPE). For iron detection, diaminobenzidine-enhanced Turnbull blue staining was applied. Iron, transferrin receptor (TfR), scavenger receptor class A member 5 (SCARA5), divalent metal transporter 1 (DMT1), natural resistance-associated macrophage protein 1 (NRAMP1), CD163, ferroportin, hephestin in hepcidin immunoreactivities were assessed in consecutive tissue sections. The following regions of interest were investigated: normal white matter of controls, MS normal-appearing white matter (NAWM) in at least 10,000 µm distance to any discernible lesion rim, white matter surrounding lesions, and early and late active areas. For chronic active lesions, we investigated the surrounding white matter, lesion rims and inactive lesion cores. MRI metrics (T1 and T2 relaxation times, proton density) were determined in PRLs, lesions with diffuse SWI-weighted hypointense signal (DSHLs), and SWI-isointense lesions (SILs), their surrounding periplaque area (PPA) and the NAWM.
Results. There was no significant difference in hemolysis parameters between patients with and without PRLs (regardless of gender and/or disease type), and controls, nor between hemolysis parameters and the number of PRLs. Hemolysis parameters remained stable during the observation period regardless of clinical and/or radiological activity. The distribution of haptoglobin genotypes was in Hardy-Weinberg equilibrium: Hp1-1 (n = 8; 8.2%), Hp2-2 (n = 39; 39.8%) in Hp2-1 (n = 51; 52.0%). Male sex (OR 3.83; 95% CI 1.44, 10.19; p = 0.007) and longer disease duration (OR 1.10; 95% CI 1.00, 1.22; p = 0.044) were both independently associated with higher risk for the PRL presence, whereas age, EDSS and haptoglobin genotype were not. Patients with ⡥ 4 PRLs had higher CSF sCD163 concentration than patients with ⡤ 3 PRLs (p = 0.009), and CSF sCD163 concentration correlated with the number of PRLs (r2 = 0.14, p = 0.023). Patients with ⡥ 4 PRLs had significantly lower pRNFL (85.8 [15.5] vs. 97.7 [12.0], p < 0.001) and GCIPL thickness (63.2 [9.1] vs. 67.7 [6.5], p = 0.039), and higher INL thickness (35.9 [2.3] vs. 34.3 [2.6], p = 0.027) than patients without PRLs. In a multivariate linear regression model, higher number of PRLs was associated with lower pRNFL (β = -0.18; 95% CI -0.98, -0.03; p = 0.038) and GCIPL (β = -0.21; 95% CI -0.58, -0.02; p = 0.039), but not INL thickness (β = 0.21; 95 % CI -0.01, 0.20; p = 0.073). In late active areas and at chronic lesion rims, we observed an accumulation of iron-positive myeloid cells. In late active areas, we also observed an increase in the number of myeloid cells positive for iron importers TfR, DMT1 (SLC11A2), NRAMP1 (SLC11A1), SCARA5 in CD163, whereas at chronic lesion rims, only DMT1 and CD163 expression was significantly increased. At chronic lesion rims, we also observed an increased number of CD68+ and CD163+ myeloid cells, which correlated with accumulated iron (CD163 [r2 = 0.0496, p = 0.033], CD68 [r2 = 0.3527, p < 0.001]). The number of CD163+ myeloid cells also correlated with HMOX1 expression (r2 = 0.1495, p = 0.003). In the whole cohort, PRLs had longer T1 relaxation times compared to DSHLs and SILs (2030.5 [1519–2540] vs. 1615.8 [1403.3–1953.5] vs. 1199.5 [1089.6–1334.6], both p < 0.001), and DSHLs had longer T1 relaxation times compared to SILs (p < 0.001). In all groups, T1 relaxation times in the PRL PPA were significantly longer compared to the SIL PPA and the NAWM but not the DSHL PPA. Similarly, in all groups, T2 relaxation times in the PRL PPA were significantly longer compared to the NAWM, and also compared to the DSHL PPA and the SIL PPA in patients with early and late RMS. Proton density in the PRL PPA was significantly higher compared to the SIL PPA and the NAWM in all groups, and also compared to the DSHL PPA in patients with SPMS.
Conclusions. We refuted the hypothesis that hemolysis parameters together with haptoglobin genotype play an important role in the development of PRLs. On the other hand, we have shown that the number of PRLs is directly associated with retinal layer thickness. In addition, we have shown increased expression of proteins involved in hem iron transport and metabolism in myeloid cells, supporting a peripheral origin of iron. Finally, we have shown that PRLs are associated with diffuse periplaque white matter damage, which could represent the driving mechanism of the so-called silent progression. In that way, we raised an important question whether PRLs could serve as a therapeutic target to prevent further neuroaxonal damage.
|
