ABSTRACT
Background and Purpose
Synthetic (Sy) MRI is a clinically approved technique providing quantitative MRI measures based on T1‐weighted, T2‐weighted, and proton density relaxometry. MRI sequences are often acquired after contrast injection with gadolinium (Gd) to assess active lesions in persons with multiple sclerosis (PwMS), affecting relaxation time. We aimed to assess the influence of Gd on the SyMRI‐based volumetrics in PwMS.
Methods
We enrolled 106 PwMS and 15 controls who performed pre‐/post‐contrast brain SyMRI on a 3T scanner. We evaluated mean change in brain parenchymal fraction (BPF), white matter (WM), grey matter (GM), myelin (Myl), non‐aqueous component (NAC), excess parenchymal water (EPW), and T1 enhancement (T1E) using paired sample t‐test for pre‐/post‐Gd volumes and independent sample t‐test for comparison between groups.
Results
The mean age was 40.9 and 39.9 years with 69% and 87% females in MS and controls, respectively. Compared to native volumetrics, Gd caused a significant observed volume increase (p < 0.001) in BPF 1.05 ± 0.3%, WM 2.8 ± 0.99%, Myl 1.42 ± 0.39%, NAC 1.04 ± 0.23%, and EPW 0.6 ± 0.4% and decrease in GM −3.05 ± 1.34% in MS. Similar change was seen in controls: BPF 0.99 ± 0.21%, WM 2.94 ± 0.93%, Myl 1.35 ± 0.37%, NAC 0.99 ± 0.22%, EPW 0.47 ± 0.29%, and GM −2.89 ± 1.18%. The change in T1E was 0.05 ± 0.12% in MS (p < 0.001) and 0.02 ± 0.25% (p = 0.76) in controls. The number of contrast‐enhancing lesions correlated with T1E (r = 0.348, p < 0.003).
Conclusion
There was a consistent pattern of volume changes in PwMS and controls, except for T1E, where the contrast could have affected the results in PwMS. Therefore, combining pre‐ and post‐contrast metrics in longitudinal studies should be interpreted with caution.
Keywords: gadolinium, multiple sclerosis, quantitative magnetic resonance imaging, volumetrics
1. Introduction
Synthetic (Sy) MRI produces contrast‐weighted images from quantitative maps based on simultaneous T1‐weighted (T1w), T2‐weighted (T2w), and proton density (PD) relaxometry and also provides quantitative MRI measures of estimated brain volumes including intracranial volume (ICV), brain parenchymal volume (BPV), white matter (WM), and grey matter (GM). Moreover, SyMRI can produce quantitative maps of myelin (Myl), T1 enhancement (T1E), non‐aqueous component (NAC), and excess parenchymal water (EPW) [1, 2, 3, 4]. The recently developed and validated SyMRI‐based Myl quantification tool, rapid estimation of Myl for diagnostics (REMyDI), has been approved for clinical use and validated to measure global cerebral demyelination [5, 6]. Clinically all of these measurements are provided to the neuro‐/radiologist in seconds alongside normative corrected Z‐scores. These quantitative MRI measures have significant potential in more specific and objective clinical monitoring for progression and discerning pathological changes.
It is routine in a clinical‐context for MRI acquisitions to be applied pre‐/post‐contrast (gadolinium based contrast agent, GBCA) to assess the specific features of various pathologies. For example, in persons with multiple sclerosis (PwMS), GBCAs are applied to determine the presence/absence of inflammatory activity in the form of contrast‐enhancing lesions (CELs). Moreover, to be more time‐efficient some MRI sequences are acquired post‐contrast. However, GBCAs significantly shorten T1 and T2 relaxation times, thereby increasing signal intensity on T1‐weighted images and decreasing it on T2‐weighted images, while having a minimal effect on PD‐weighted imaging [7]. Therefore, it is quite valuable to better understand the implications of the applying SyMRI post‐contrast, particularly in the context of MS where GBCAs are routinely applied.
It has previously been demonstrated in healthy volunteers that the SyMRI volumetric calculations are affected by introducing GBCAs, which can be explained by decreases in T1 and T2 relaxation time, and PD. The effect of GBCAs on the estimation of Myl has only been reported for brain metastases [7] that usually are considerably larger and have much higher GBCA uptake than lesions presented in multiple sclerosis (MS). Additionally, it is quite possible that degree of blood brain barrier (BBB) disruption can have a differential effect on the quantitative measures more drastically than for PwMS with and without CELs. In MS, both scans with GBCA and native scans are compared longitudinally to track the disease progression and assess treatment effects. Therefore, the effect of GBCA on Myl, EPW, and NAC needs to be evaluated in the context of PwMS.
Previous studies evaluating the effect of GBCAs on SyMRI quantitative metrics in MS have solely focused on volumes (including WM, GM, ICV, BPV, and brain parenchymal fraction [BPF]) and have only included patients with an advanced stage of the disease [7, 8]. Therefore, we aimed to investigate how applying GBCAs can alter SyMRI‐based quantitative metrics in a comprehensive cohort of PwMS and, secondarily, explore if CELs have any further impact.
2. Methods
2.1. Study Participants
We included 106 PwMS with known relapsing‐remitting MS, according to the McDonald criteria 2017 [9]. This was a subgroup from a larger prospective study (with inclusion between 2014 and 2016) performing MRI yearly [10], where patients were examined with an extra MRI sequence at one follow‐up between September 2020 and September 2021. We consecutively included patients from the prospective study during this one‐year time frame and performed a cross‐sectional study comparing pre‐/post‐contrast images. A total of 15 control subjects from the same prospective study [10] were invited for an MRI follow‐up during this time frame, as well. None of the controls had any history of neurological disease, and neurological status was normal in all of them. In a longitudinal part of this study, we analyzed all the images from PwMS (n = 106) previously completed during the prospective study. The study was approved by the Regional Ethical Review Board in Gothenburg (Dnr 895‐13) and by the Swedish Ethical Review Authority (Dnr 2020‐06766).
2.2. MR Acquisition and Image Analysis
All of the MRI examinations were performed on a 3T Philips machine (Achieva dStream, 16‐channel head coil). All study subjects underwent standardized conventional imaging and quantitative Sy imaging. First, pre‐contrast followed by post‐contrast quantitative Sy sequence brain was done. This is a multi‐dynamic multi‐echo (MDME) sequence covering the whole brain and quantifies the longitudinal relaxation rate (R1), transverse relaxation rate (R2), PD and local B1 field [1, 11]. The Myl is measured by its effect on the surrounding cellular water properties. Each voxel has four compartments (Myl partial volume cellular partial volume, free water partial volume, EPW partial volume) with their own R1, R2, and PD values. The Myl partial volume has lower values than other compartments and its detection was histologically validated [4, 5]. The post‐contrast MDME was performed directly after the contrast injection, estimated to 1 min. The acquisition time for pre‐ and post‐contrast MDME was identical, 5 min. The mean time difference between the start of pre‐ and post‐contrast MDME was 8:11 (SD 1:06) minutes. The automatic brain segmentation with quantitative values is done with SyMRI software (https://syntheticmr.com/products/symri‐neuro/ version 11.2; Sy MR, Linköping, Sweden) [1] with post‐processing time less than 1 min.
T1‐enhancement in SyMRI is quantified as an increase in the R1 = 1/T1 after gadolinium (Gd), since the contrast agent shortens T1 and thus elevates R1 in enhancing tissue. Unlike conventional subtraction techniques, SyMRI leverages its built‐in normative quantitative tissue maps (WM, GM, cerebrospinal fluid) so that post‐contrast voxels with R1 values falling significantly outside expected tissue ranges can be isolated as enhancement, allowing direct post‐contrast R1 quantification alone to reveal contrast uptake. The Warntjes et al. study [12] applied this principle in glioma patients where there is significant non‐uniform contrast enhancement, demonstrating that Sy R1‐based enhancement maps reliably depict tumor enhancement without needing pre‐contrast scans or image subtraction. This is particularly valuable in a clinical context saving time (≈5 min) in the scanner for both the patient and the institution.
The conventional imaging included the following sequences: three‐dimensional (3D) T1‐weighted (repetition time [TR] 8.2, echo time [TE] 3.8, inversion time [TI] 1650), T2‐weighted (TR 4000, [TE] 3.8, [TI] 1650) and 3D T2‐w fluid attenuated inversion recovery (FLAIR; [TR] 4600, [TE] 106, [TI] 1650). All images had a slice thickness of 3 mm and all conventional images were performed post‐contrast. The conventional images were evaluated by a trained neuroradiologist who counted the number of CELs. The T2‐FLAIR lesion volume segmentation was performed by Lesion Segmentation Toolbox, Lesion prediction algorithm (LPA, Munich and Jena, Germany; https://www.applied‐statistics.de/lst.html, Matlab 2019a, SPM12, LST v3.0.0). Dotarem (gadoterate meglumine) was the applied GBCA at a dosage of 0.2m/kg body weight.
2.3. Statistical Analysis
The Shapiro–Wilk test and the visual assessment of data distribution were used to assess normality. The paired sample t‐test was used for pre‐ and post‐Gd volumes and independent sample t‐test for comparison between groups. The p values are reported as two‐sided raw. After Bonferroni correction, the level of significance threshold was adjusted to α = 0.005, and values lower than α were considered significant. The Spearman rank correlation coefficient was used to investigate the association between the number of T1 CELs on conventional MRI and the volume of T1E on SyMRI in MS patients with CELs. All statistical analyses were performed using SPSS (version 28.0.0, IBM, Armonk, New York, https://www.ibm.com/products/spss‐statistics). The figures were created in GraphPad Prism (version 10.2.3 GraphPad Inc., California, USA, https://www.graphpad.com/).
3. Results
3.1. Cohort Characteristics
The characteristics of patients and controls are presented in Table 1. At the time point of the cross‐sectional study, the mean age of patients at the cross‐sectional study was 40.9 (SD 9.4) years, the mean disease duration was 9.5 (SD 5.1) years and median Expanded disability status scale (EDSS) was 1.5 (IQR 0–2). The median T2 lesion load was 27.1 (IQR 15–69) mL. Only three MS patients (2.8%) had CELs present on conventional, one patient had one CEL, one patient had two CELs, and one patient had three CELs. The mean age of controls was 39.9 (SD 9.2) years.
TABLE 1.
Characteristics.
| Multiple sclerosis | Controls | |
|---|---|---|
| Number of subjects | 106 | 15 |
| Age; years, mean ± SD | 40.9 ± 9.4 | 39.9 ± 9.2 |
| Gender; female, number (%) | 73 (68.9) | 13 (86.7) |
| Disease duration; years, mean ± SD | 9.5 ± 5.1 | NA |
| EDSS; median (IQR) | 1.5 (0–2) | NA |
| T2 lesion load; mL, median (IQR) | 27 (15–69) | NA |
| Number of patients with CELs (%) | 3 (2.8) | NA |
Abbreviations: CELs = contrast enhancing lesions, EDSS = expanded disability status scale, IQR = interquartile range, mL = milliliters, NA = not applicable, SD = standard deviation.
3.2. Volumetrics Pre‐ and Post‐Contrast
The volumetrics in MS patients and controls are presented in raw values (mL) in Table 2, including the results from paired t‐test, comparing the mean values. The corresponding values corrected for ICV (presented in %) are presented in Table 3. All raw and ICV normalized volumes differed significantly (p < 0.001) between pre‐ and post‐contrast in MS and controls, with the singular exception of raw T1E and BPV volume in controls that was unchanged.
TABLE 2.
Brain volumes (in mL) in patients with multiple sclerosis and controls.
| ICV | BPV | WM | GM | CSF | NON | Myl | NAC | EPW | T1E | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MS | Pre‐contrast | Mean | 1411 | 1200 | 481 | 649 | 211 | 70 | 149 | 217 | 37 | 3.6 |
| SD | 144 | 134 | 69 | 66 | 67 | 26 | 25 | 27 | 11 | 2.2 | ||
| Post‐contrast | Mean | 1399 | 1205 | 516 | 601 | 194 | 88 | 168 | 230 | 45 | 4.2 | |
| SD | 143 | 134 | 75 | 59 | 66 | 32 | 28 | 28 | 13 | 1.9 | ||
| Difference | Mean | −11.8 | 4.7 | 35.2 | −48.1 | −16.5 | 17.6 | 18.7 | 12.7 | 8.1 | 0.6 | |
| Pre/post‐contrast | SD | 8.1 | 8.5 | 14.2 | 18.9 | 4.1 | 14.2 | 6.1 | 3.0 | 5.8 | 1.6 | |
| p value | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | ||
| Controls | Pre‐contrast | Mean | 1392 | 1198 | 505 | 628 | 193 | 64 | 156 | 222 | 34 | 4.2 |
| SD | 104 | 109 | 48 | 63 | 65 | 21 | 16 | 19 | 9 | 4.7 | ||
| Post‐contrast | Mean | 1378 | 1200 | 541 | 583 | 178 | 76 | 173 | 233 | 40 | 4.3 | |
| SD | 99 | 105 | 52 | 61 | 65 | 16 | 17 | 20 | 9 | 2.0 | ||
| Difference | Mean | −13.4 | 2.1 | 35.5 | −45.2 | −15.3 | 11.7 | 16.9 | 11.4 | 6.0 | 0.0 | |
| Pre/post‐contrast | SD | 14.6 | 12.9 | 12.8 | 15.0 | 3.9 | 12.5 | 5.4 | 3.3 | 4.3 | 3.5 | |
| p value | 0.003 * | 0.544 | < 0.001 * | < 0.001 * | < 0.001 * | 0.003 * | < 0.001 * | < 0.001 * | < 0.001 * | 0.965 |
Abbreviations: BPV = brain parenchymal volume, CSF = cerebrospinal fluid, EPW = excess parenchymal water, GM = grey matter, ICV = intracranial volume, MS = multiple sclerosis, Myl = myelin, NAC = non‐aqueous component, NON = unclassified, SD = standard deviation, T1E = T1 enhancement, WM = white matter.
Significant p value.
TABLE 3.
Normalized intracranial volumes (%) defined as brain volumes divided by intracranial volume in patients with multiple sclerosis and controls.
| BPF | WM | GM | CSF | NON | Myl | NAC | EPW | T1E | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MS | Pre‐contrast | Mean | 85.1 | 34.1 | 46.1 | 14.9 | 4.9 | 10.6 | 15.4 | 2.6 | 0.3 | |
| SD | 4.5 | 3.3 | 3.2 | 4.5 | 1.5 | 1.1 | 1.0 | 0.6 | 0.2 | |||
| Post‐contrast | Mean | 86.1 | 36.9 | 43.1 | 13.9 | 6.2 | 12.0 | 16.4 | 3.2 | 0.3 | ||
| SD | 4.5 | 3.7 | 2.9 | 4.5 | 2.0 | 1.2 | 1.1 | 0.8 | 0.1 | |||
| Difference | Mean | 1.1 | 2.8 | −3.1 | −1.1 | 1.3 | 1.4 | 1.0 | 0.6 | 0.1 | ||
| Pre/post‐contrast | SD | 0.3 | 1.0 | 1.3 | 0.3 | 1.0 | 0.4 | 0.2 | 0.4 | 0.1 | ||
| p value | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | |||
| Controls | Pre‐contrast | Mean | 86.1 | 36.3 | 45.2 | 13.9 | 4.6 | 11.2 | 16.0 | 2.5 | 0.3 | |
| SD | 4.7 | 2.6 | 3.0 | 4.7 | 1.3 | 0.9 | 0.8 | 0.6 | 0.3 | |||
| Post‐contrast | Mean | 87.1 | 39.3 | 42.3 | 12.9 | 5.6 | 12.6 | 16.9 | 2.9 | 0.3 | ||
| SD | 4.7 | 3.0 | 2.8 | 4.7 | 1.2 | 1.0 | 0.9 | 0.7 | 0.2 | |||
| Difference | Mean | 1.0 | 2.9 | −2.9 | −1.0 | 0.9 | 1.4 | 1.0 | 0.5 | 0.0 | ||
| Pre/post‐contrast | SD | 0.2 | 0.9 | 1.2 | 0.2 | 0.9 | 0.4 | 0.2 | 0.3 | 0.3 | ||
| p value | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | < 0.001 * | 0.760 | |||
Abbreviations: BPF = brain parenchymal fraction, CSF = cerebrospinal fluid, EPW = excess parenchymal water, GM = grey matter, MS = multiple sclerosis, Myl = myelin, NAC = non‐aqueous component, NON = unclassified, SD = standard deviation, T1E = T1 enhancement, WM = white matter.
Significant p value.
Compared to native volumetrics, GBCAs caused a significant increase (p < 0.001) in BPF 1.05 ± 0.3%, ICV corrected WM 2.8 ± 0.99%, Myl 1.42 ± 0.39%, NAC 1.04 ± 0.23%, EPW 0.6 ± 0.4%, and decrease in GM −3.05 ± 1.34% in MS. Similar changes were observed in healthy volunteers: BPF 0.99 ± 0.21%, WM 2.94 ± 0.93%, Myl 1.35 ± 0.37%, NAC 0.99 ± 0.22%, EPW 0.47 ± 0.29%, and decrease in GM −2.89 ± 1.18%. The change in T1E was 0.05 ± 0.12% in MS (p < 0.001) and 0.02 ± 0.25% (p = 0.76) in controls.
Pearson correlations of pre‐ and post‐contrast volumetrics were consistently very high in MS (0.904–0.998, p < 0.001) and controls (0.875–0.999, p < 0.001), except for T1E 0.731 (p < 0.001) and 0.753 (p = 0.001), respectively (Table 4).
TABLE 4.
Paired samples correlations between pre‐ and post‐contrast values.
| MS | Controls | |||
|---|---|---|---|---|
| Measure | Correlation | p value | Correlation | p value |
| ICV | 0.998 | < 0.001 | 0.991 | < 0.001 |
| BPV | 0.998 | < 0.001 | 0.994 | < 0.001 |
| BPF | 0.998 | < 0.001 | 0.999 | < 0.001 |
| WM/ICV | 0.969 | < 0.001 | 0.955 | < 0.001 |
| GM/ICV | 0.906 | < 0.001 | 0.919 | < 0.001 |
| Myl/ICV | 0.95 | < 0.001 | 0.934 | < 0.001 |
| NAC/ICV | 0.977 | < 0.001 | 0.977 | < 0.001 |
| EPW/ICV | 0.869 | < 0.001 | 0.9 | < 0.001 |
| T1E/ICV | 0.622 | < 0.001 | 0.64 | 0.005 |
| WM | 0.984 | < 0.001 | 0.97 | < 0.001 |
| GM | 0.959 | < 0.001 | 0.972 | < 0.001 |
| Myl | 0.977 | < 0.001 | 0.947 | < 0.001 |
| NAC | 0.995 | < 0.001 | 0.986 | < 0.001 |
| EPW | 0.904 | < 0.001 | 0.875 | < 0.001 |
| T1E | 0.731 | < 0.001 | 0.753 | 0.001 |
Abbreviations: BPF = brain parenchymal fraction, BPV = brain parenchymal volume, EPW = excess parenchymal water, GM = grey matter, ICV = intracranial volume, MS = multiple sclerosis, Myl = myelin, NAC = non‐aqueous component, T1E = T1 enhancement, WM = white matter.
The Myl/ICV, BPF and T1E/ICV volumes are presented in Figure 1. The results were unchanged when excluding the patients with CELs. There was no correlation between number of CELs and T1E volumes. An example of PwMS with CEL is presented in Figure 2.
FIGURE 1.
SyMRI volumetrics pre‐ and post‐contrast. The figure shows myelin volume corrected for intracranial volume (A), brain parenchymal fraction (B), and T1 Enhancement volume corrected for intracranial volume (C) in patients with multiple sclerosis (MS) and controls. The mean values and standard deviations are shown pre‐contrast (●) and post‐contrast (■). The significant p values are marked *(p < 0.001), all other differences with p values (p < 0.05) are not marked.
FIGURE 2.
Example of a person with multiple sclerosis with contrast enhancing lesion. An exemplary depiction of a participant with MS (age: 35 years, female, relapsing‐remitting MS, disease duration: 7.8 years) who has a Gd enhancing lesion (dashed lines), which notably alters the myelin content in and surrounding the lesion in pre‐ versus post‐gadolinium injection. This is most well evidenced in contrast‐laden vessels (dotted lines), where the myelin signal is elevated relative to the non‐contrast acquisition. Gd = Gadolinium, T1w = T1‐weighted image.
3.3. Comparison of Volumetric Between MS and Controls
Differences between PwMS and controls are presented in Table 5, depicting that the pre‐ and post‐contrast raw and ICV corrected volumes did not differ between groups.
TABLE 5.
Differences between patients with multiple sclerosis and controls.
| MS versus controls | ICV | BPV | WM | GM | CSF | NON | Myl | NAC | EPW | T1E | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Pre‐contrast | Mean diff | 19.5 | 2.0 | −24.6 | 20.7 | 17.6 | 5.9 | −6.8 | −4.7 | 2.5 | −0.6 |
| SE diff | 38.6 | 36.1 | 18.3 | 18.0 | 18.5 | 7.1 | 6.7 | 7.2 | 2.9 | 0.7 | |
| p value | 0.614 | 0.955 | 0.183 | 0.252 | 0.344 | 0.406 | 0.307 | 0.514 | 0.392 | 0.420 | |
| Post‐contrast | Mean diff | 21.2 | 4.7 | −24.8 | 17.9 | 16.4 | 11.8 | −5.0 | −3.5 | 4.6 | 0.0 |
| SE diff | 38.3 | 36.2 | 20.1 | 16.4 | 18.2 | 8.5 | 7.3 | 7.5 | 3.5 | 0.5 | |
| p value | 0.582 | 0.897 | 0.218 | 0.279 | 0.371 | 0.168 | 0.492 | 0.646 | 0.195 | 0.968 | |
| BPF | WM/ICV | GM/ICV | CSF/ICV | NON/ICV | Myl/ICV | NAC/ICV | EPW/ICV | T1E/ICV | |||
| Pre‐contrast | Mean diff | −1.0 | −2.3 | 0.9 | 1.0 | 0.3 | −0.7 | −0.6 | 0.1 | 0.0 | |
| SE diff | 1.2 | 0.9 | 0.9 | 1.2 | 0.4 | 0.3 | 0.3 | 0.2 | 0.1 | ||
| p value | 0.405 | 0.011 | 0.278 | 0.405 | 0.477 | 0.028 | 0.038 | 0.504 | 0.389 | ||
| Post‐contrast | Mean diff | −1.0 | −2.4 | 0.8 | 1.0 | 0.7 | −0.6 | −0.5 | 0.2 | 0.0 | |
| SE diff | 1.3 | 1.0 | 0.8 | 1.3 | 0.5 | 0.3 | 0.3 | 0.2 | 0.0 | ||
| p value | 0.437 | 0.018 | 0.322 | 0.437 | 0.206 | 0.069 | 0.075 | 0.274 | 0.707 | ||
Abbreviations: BPV = brain parenchymal volume, BPF = brain parenchymal fraction, CSF = cerebrospinal fluid, Diff = difference, EPW = excess parenchymal water, GM = grey matter, ICV = intracranial volume, MS = multiple sclerosis, Myl = myelin, NAC = non‐aqueous component, NON = unclassified, SE = standard error, T1E = T1 enhancement, WM = white matter.
3.4. Correlation Between CELs and T1E in a Longitudinal Study
Due to the low number of pre‐ and post‐Gd images with CELs, we performed a longitudinal subanalysis on patients performing yearly assessments within a larger prospective study. Here, we identified 70 post‐Gd MRI examinations from PwMS with CELs. Number of CELs correlated with T1E volume and T1E/ICV (r = 0.383, p < 0.001 and 0.348, p < 0.003, respectively). Correlation between CELs and T1E volume and T1E/ICV was higher when excluding patients with only one CELs (r = 0.537, p < 0.001 and r = 0.525, p < 0.001, respectively). There was no correlation between CELs and Myl/ICV or BPF (r = 0.027, p = 0.827 and r = 0.030, p = 0.802, respectively).
4. Discussion
In this study, we investigated the difference in volume estimations by SyMRI in pre‐ and post‐contrast acquisition of the brain in PwMS and healthy brain tissue. Further, we compared the differences between MS patients with and without CELs. Compared to native volumetrics, GCBA caused a significant observed volume increase in volumetrics of BPF, WM, Myl, NAC, EPW, and a decrease of GM in MS and controls with the same proportional change in all tissue volumes. Moreover, the results remained unchanged in sensitivity analysis, when performed without PwMS who had CELs.
4.1. The Effects of GBCA on Volumetrics
There is only one previous report including patients with MS and studying the effects of GBCA on volumetrics. In that study, it was concluded that comparing images MS post‐Gd and HC without Gd is a valid approach [8], owing to a smaller effect of GBCA on tissue volumes than the volume changes due to the MS disease itself. They also suggested that further investigation is needed for patients with MS at earlier disease stages [8]. The effect of CELs on tissue volumes has not been previously investigated in MS.
The tissue volumes in our study differs due to different stages of the disease in patients included in the earlier studies. In the prior study by Wartnjes [8], included patients with progressive disease with higher age (mean 47 years), longer disease duration (mean 15 years) BPF in PwMS and controls was 82% and 88.8%, respectively, and had higher EDSS (mean 3.5). Interestingly, the controls with higher mean age (48 years) showed higher BPF than the younger controls in our study (BPF 86.1%, mean age 39.9 years). Though the direct comparison can be hampered by comparison across platform and software versions. In a study by Vågberg [13], PwMS had BPF 85.2%, mean age 43 years, disease duration 9 years, comparable to our cohort (85.1%, 40.9 years, 9.5 years) and slightly higher EDSS 2.5 than our patients (EDSS 1.5). The controls with BPF 89% were younger (mean age 35 years) compared to our controls. Therefore, the results of our study are contrariwise and the group differences between our patient and controls are small and differences within groups were significantly changed after GBCA.
4.2. The Presence of CELs and T1E Volume
It is known that GCBA decreases the T1 relaxation in tissue [14], but PwMS patients with CELs have not previously been included in previous studies and Myl volumes have also not previously been investigated [8]. A study on brain metastates [7] reported no correlation between the change in measured Myl volume and total volume of brain metastases. Our study therefore fills this knowledge gap and complements the two previous studies [7, 8]. In our study, we included a significantly larger amount of PwMS patients, and additionally those with shorter disease duration and included PwMS who had Gd lesions on MRI, which was lacking in only other GBCA in PwMS study [8]. We also examined the effect of Gd on Myl volumetrics in healthy tissue. The volumetrics were significantly changed with GBCA in both PwMS and healthy volunteers. Reflecting an effect that extends beyond that of a pathologically driven effect, and could possibly be due to the universal capillary GBCA uptake in the tissue. Suggesting that even a small or no contrast enhancement affects the results, thus in longitudinal studies including PwMS with larger volumes of CELs, meaning a greater disruption of BBB, could be precarious to interpret. In the subanalysis, there was correlation between T1E volume and the number of CELs reported by neuroradiologist. The correlation was stronger when excluding patients with only one CELs. This suggests that T1E values could reflect the CELs load. Interestingly, T1E values were the only values that did not change in healthy brain tissue after the application of GBCA, thus, T1E volume seems likely to indeed be reflecting specificity for macroscopic tissue enhancement, such as in CELs, but not for the microscopic global enhancement of tissues.
This apparent difference between the diffuse post‐contrast changes observed in the Myl SyMRI parameter versus the stability of T1E could potentially be explained by their distinct physiological and analytical approaches. The SyMRI Myl and volumetric parameters are derived from model‐based compartmental estimation using voxel‐wise R1, R2, and PD values. These metrics are sensitive to even small, globally distributed relaxometric shifts induced by low‐level Gd distribution within both the vascular and interstitial spaces, to produce diffuse apparent changes across both MS and control groups. In contrast, T1E specifically isolates voxels where the post‐contrast R1 exceeds expected (SyMRI‐defined) tissue‐specific thresholds, representing true focal accumulation of Gd and macroscopic BBB disruption. However, diffuse BBB permeability has previously been captured using the more sensitive technique of dynamic contrast‐enhanced MRI methods [15]. Therefore, the unchanged post‐contrast T1E remains unchanged in NAWM, aligning with the mechanistically lower sensitivity of SyMRI to subtle permeability alterations compared with the more sensitive dynamic contrast‐enhanced MRI. This indicates that T1E primarily reflects overt enhancement, while the diffuse changes in Myl‐related parameters reflect non‐pathological, global relaxation effects of contrast administration rather than true alterations in tissue integrity or BBB permeability.
The pathophysiology of MS involves immune attacks against the central nervous system, resulting in demyelinating lesions, irreversible axonal loss [16], and subsequent atrophy development of both WM and GM [17]. Although the degree of brain atrophy is partially due to focal tissue damage, there are other undefined factors contributing to degeneration in MS [18] that seem independent of demyelination [16]. Over recent years, brain atrophy has not only been used as a measure of global neurodegeneration in MS but also as a measure of treatment response and was recommended to be included in patient assessment [19, 20], but it is still not widely used in a clinical context. Conventional MRI is optimized to accentuate tissue contrast with a high sensitivity for detecting pathological changes in the brain, for example, T2‐FLAIR. It detects the WM lesions in MS, T2‐weighted hyperintense lesions and T1 CELs, reflecting disease activity [21]. However, it is a time‐consuming and therefore clinically costly approach necessitating multiple scans that relies on the visual assessment of the neuroradiologist. On the other hand, SyMRI uses a single scan to reproduce conventional images, but also quantitative maps with volumetric data, including Myl and also shows some utility in assessing CELs. Unlike other methods for determining Myl content in the brain [22], the Myl measurement by SyMRI is validated for clinical use providing robust measurements [6].
4.3. Limitations and Future Directions
Measuring change in the brain volumetrics including Myl content in different brain regions in vivo could provide new insights into neurodegeneration and the pathophysiology of MS, and the impact of disease‐modifying treatments in these processes that could impact clinical patient care. Thus, SyMRI could provide new quantitative imaging biomarkers for MS, with short automatic post‐processing, thus less time‐consuming, and more reliable than semi‐quantitative measurements used to track disease progression at present. SyMRI provides quantitative information about the brain tissue and showed a good performance and validity in cross‐sectional studies on BPF and global Myl content [6, 23]. However, larger longitudinal prospective studies on Myl are still lacking and how the Myl content in different brain regions changes over time is not known.
The limitation of the present study, similar to previous studies on Gd and SyMRI [7, 8] is that only single‐dose Gd and fixed time after Gd administration were investigated. Even if those parameters are fixed, the Gd concentration in the tissues still can be confounded by weight, blood volume, and perfusion rate. The capillary diffusion of the contrast can affect the measurement of enhancement volume and have an impact on the other volumes provided by SyMRI. Thus, the vascular permeability in the brain is another confounding factor. Even though the post‐GBCA imaging was performed directly after Gd administration, and the examination lasted approximately 5 min, this minimizes the effects of this confounding factor. Our results on volumes are consistent with previous studies, even if the acquisition of data was performed on a different scanner and we used a newer version of SyMRI software. Compared to previous studies [7, 8], supporting the usage of this method. However, there were limited observations of CELs, which hampers comparisons and therefore reduces the strength of CELs observations. Thus, further studies should specifically consider CELs and the added value of GBCA on SyMRI.
4.4. Conclusion
In conclusion, the SyMRI measures showed high correlation pre‐ and post‐contrast. There was a consistent pattern of change in brain volumes across the dataset in MS and controls, with the exception of T1E, where the presence of CELs in PwMS affected this value. Thus, our results indicate the same proportional increase and decrease in both groups. However, when comparing values in prospective studies, the usage of the same type of acquisition (pre‐ or post‐contrast only) would be preferred. To completely exclude confounding factors connected to GBCA usage, the acquisition without contrast would be the most suitable option.
Conflicts of Interest
Russell Ouellette has received spearkship honoraria from Siemens Healthineers, Novartis, and Sanofi, and has served on advisory boards for Sanofi.
Igal Rosenstein has received compensation for lectures from Merck, Biogen, Novartis, and Sanofi, and has served on advisory boards for Sanofi.
Markus Axelsson has received compensation for lectures and/or advisory boards from Biogen, Genzyme and Novartis.
Lenka Novakova has received compensation for lecture from Biogen, Novartis, Teva, Sanofi and Merck, has served on advisory boards for Merck, Janssen and Sanofi, has received an unconditional research grant from Novartis and Sanofi and participated in clinical trials as Principal Investigator sponsored by Amgen, Sanofi, ArgenX and Takeda.
Funding
The study was financed by grants from the Swedish state under the agreement between the Swedish government and the county councils, the ALF‐agreement (ALFGBG‐984087 and ALFGBG‐1005346), the Gothenburg Society of Medicine (GLS‐1001019), the Swedish MS research fund, Research Foundation of the Multiple Sclerosis Society of Gothenburg/NEURO Gothenburg, Edith Jacobson's Foundation, Amlöv's foundation, Göteborg Foundation for Neurological Research, the Swedish Society for Medical Research (Grant No. PG‐22‐0440), and the Region Stockholm Center for Innovative Medicine (Grant No. FoUI‐1025176).
Novakova L., Rosenstein I., Axelsson M., and Ouellette R., “The Effect of Gadolinium on Synthetic Magnetic Resonance Quantitative Imaging.” Journal of Neuroimaging 35, no. 6 (2025): e70104. 10.1111/jon.70104
References
- 1. Hagiwara A., Warntjes M., Hori M., et al., “SyMRI of the Brain: Rapid Quantification of Relaxation Rates and Proton Density, With Synthetic MRI, Automatic Brain Segmentation, and Myelin Measurement,” Investigative Radiology 52 (2017): 647–657, 10.1097/RLI.0000000000000365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Hagiwara A., Hori M., Yokoyama K., et al., “Synthetic MRI in the Detection of Multiple Sclerosis Plaques,” American Journal of Neuroradiology 38 (2017): 257–263, 10.3174/ajnr.A5012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Hagiwara A., Hori M., Kamagata K., et al., “Myelin Measurement: Comparison Between Simultaneous Tissue Relaxometry, Magnetization Transfer Saturation Index, and T1w/T2w Ratio Methods,” Scientific Reports 8 (2018): 10554, 10.1038/s41598-018-28852-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Warntjes M., Engstrom M., Tisell A., et al., “Modeling the Presence of Myelin and Edema in the Brain Based on Multi‐Parametric Quantitative MRI,” Frontiers in Neurology 7 (2016): 16, 10.3389/fneur.2016.00016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Warntjes J. B. M., Persson A., Berge J., et al., “Myelin Detection Using Rapid Quantitative MR Imaging Correlated to Macroscopically Registered Luxol Fast Blue‐Stained Brain Specimens,” American Journal of Neuroradiology 38 (2017): 1096–1102, 10.3174/ajnr.A5168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Ouellette R., Mangeat G., Polyak I., et al., “Validation of Rapid Magnetic Resonance Myelin Imaging in Multiple Sclerosis,” Annals of Neurology 87 (2020): 710–724, 10.1002/ana.25705. [DOI] [PubMed] [Google Scholar]
- 7. Maekawa T., Hagiwara A., Hori M., et al., “Effect of Gadolinium on the Estimation of Myelin and Brain Tissue Volumes Based on Quantitative Synthetic MRI,” American Journal of Neuroradiology 40 (2019): 231–237, 10.3174/ajnr.A5921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Warntjes J. B., Tisell A., Landtblom A. M., et al., “Effects of Gadolinium Contrast Agent Administration on Automatic Brain Tissue Classification of Patients With Multiple Sclerosis,” American Journal of Neuroradiology 35 (2014): 1330–1336, 10.3174/ajnr.A3890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Thompson A. J., Banwell B. L., Barkhof F., et al., “Diagnosis of Multiple Sclerosis: 2017 Revisions of the McDonald Criteria,” Lancet Neurology 17 (2018): 162–173, 10.1016/S1474-4422(17)30470-2. [DOI] [PubMed] [Google Scholar]
- 10. Novakova L., Axelsson M., Malmestrom C., et al., “Searching for Neurodegeneration in Multiple Sclerosis at Clinical Onset: Diagnostic Value of Biomarkers,” PLoS ONE 13 (2018): e0194828, 10.1371/journal.pone.0194828. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Warntjes J. B., Leinhard O. D., West J., et al., “Rapid Magnetic Resonance Quantification on the Brain: Optimization for Clinical Usage,” Magnetic Resonance in Medicine: Official Journal of the Society of Magnetic Resonance in Medicine /Society of Magnetic Resonance in Medicine 60 (2008): 320–329, 10.1002/mrm.21635. [DOI] [PubMed] [Google Scholar]
- 12. Warntjes M., Blystad I., Tisell A., et al., “Synthesizing a Contrast‐Enhancement Map in Patients With High‐Grade Gliomas Based on a Post‐Contrast MR Imaging Quantification Only,” American Journal of Neuroradiology 39 (2018): 2194–2199, 10.3174/ajnr.A5870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Vagberg M., Lindqvist T., Ambarki K., et al., “Automated Determination of Brain Parenchymal Fraction in Multiple Sclerosis,” American Journal of Neuroradiology 34 (2013): 498–504, 10.3174/ajnr.A3262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Costelloe C. M., Amini B., and Madewell J. E., “Risks and Benefits of Gadolinium‐Based Contrast‐Enhanced MRI,” Seminars in Ultrasound, CT, and MRI 41 (2020): 170–182, 10.1053/j.sult.2019.12.005. [DOI] [PubMed] [Google Scholar]
- 15. Cramer S. P., Simonsen H., Frederiksen J. L., et al., “Abnormal Blood‐Brain Barrier Permeability in Normal Appearing White Matter in Multiple Sclerosis Investigated by MRI,” NeuroImage: Clinical 4 (2014): 182–189, 10.1016/j.nicl.2013.12.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Trapp B. D. and Nave K. A., “Multiple Sclerosis: An Immune or Neurodegenerative Disorder?,” Annual Review of Neuroscience 31 (2008): 247–269, 10.1146/annurev.neuro.30.051606.094313. [DOI] [PubMed] [Google Scholar]
- 17. Trapp B. D., Vignos M., Dudman J., et al., “Cortical Neuronal Densities and Cerebral White Matter Demyelination in Multiple Sclerosis: A Retrospective Study,” Lancet Neurology 17 (2018): 870–884, 10.1016/S1474-4422(18)30245-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Fisher E., Rudick R. A., Simon J. H., et al., “Eight‐Year Follow‐Up Study of Brain Atrophy in Patients With MS,” Neurology 59 (2002): 1412–1420, 10.1212/01.wnl.0000036271.49066.06. [DOI] [PubMed] [Google Scholar]
- 19. Sastre‐Garriga J., Pareto D., Battaglini M., et al., “MAGNIMS Consensus Recommendations on the Use of Brain and Spinal Cord Atrophy Measures in Clinical Practice,” Nature Reviews Neurology 16 (2020): 171–182, 10.1038/s41582-020-0314-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. De Stefano N., Battaglini M., Pareto D., et al., “MAGNIMS Recommendations for Harmonization of MRI Data in MS Multicenter Studies,” NeuroImage: Clinical 34 (2022): 102972, 10.1016/j.nicl.2022.102972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Sormani M. P. and Bruzzi P., “MRI Lesions as a Surrogate for Relapses in Multiple Sclerosis: A Meta‐Analysis of Randomised Trials,” Lancet Neurology 12 (2013): 669–676, 10.1016/S1474-4422(13)70103-0. [DOI] [PubMed] [Google Scholar]
- 22. Alonso‐Ortiz E., Levesque I. R., and Pike G. B., “MRI‐Based Myelin Water Imaging: A Technical Review,” Magnetic Resonance in Medicine: Official Journal of the Society of Magnetic Resonance in Medicine /Society of Magnetic Resonance in Medicine 73 (2015): 70–81, 10.1002/mrm.25198. [DOI] [PubMed] [Google Scholar]
- 23. Granberg T., Uppman M., Hashim F., et al., “Clinical Feasibility of Synthetic MRI in Multiple Sclerosis: A Diagnostic and Volumetric Validation Study,” American Journal of Neuroradiology 37 (2016): 1023–1029, 10.3174/ajnr.A4665. [DOI] [PMC free article] [PubMed] [Google Scholar]