Quantication of Gadolinium Deposition in the Brain Using Quantitative MRI

Gadolinium depositions in the brain caused by Gadolinium Based Contrast Agents (GBCAs) represent a topical issue since depositions have been conrmed, but their potential effects remain unknown. The purpose of this study is to show that quantitative MRI provides an alternative to conventional MRI for detection of gadolinium depositions without the need for reference areas, while also avoiding methodological limitations associated with conventional MRI, facilitating future studies on the subject. This retrospective observational cohort study of longitudinally acquired data uses quantitative MRI to quantify gadolinium depositions in six areas of the brain; the dentate nuclei, globus pallidi, thalami, pons, white matter and frontal cortex grey matter of 43 patients who received two to ve doses of Magnevist (a linear GBCA) from 2009 to 2015. R 1 was measured at baseline and after four years in the six areas of the brain of each patient. The changes in R 1 were examined using a Paired-Samples T test. The correlation between doses of Magnevist and R 1 was analysed using an Independent-Samples T test. There was a signicant increase in R 1 (p<0.05), consistent with gadolinium depositions in the dentate nuclei, globus pallidi and grey matter. No signicant increase of R 1 was found in the thalami, white matter or pons. A correlation between doses of Magnevist and R 1 was found in the dentate nuclei and grey matter, but not in the globus pallidus. The results suggest that quantitative MRI is an alternative to conventional MRI for early detection of gadolinium depositions, while additionally providing quantitative measurements for comparison and follow-up.


Introduction
Gadolinium-based contrast agents (GBCAs) used for MRI are known to cause gadolinium depositions in several areas of the human body, including the brain, skin and bone (1). After the initial report in 2014 (2), gadolinium depositions in the brain have mainly been found in the dentate nuclei and globus pallidi, mostly associated with linear GBCAs (3,4). There have also been studies indicating gadolinium uptake in the grey matter, thalami, pons and white matter (1,5). Since 2014, studies on gadolinium depositions in the brain have focused on detection, distribution, potential toxicity and clinical consequences (1). This study focuses on the former, hopefully facilitating studies on the latter.
Various studies have shown that gadolinium depositions become detectable using conventional MRI after approximately four to six administrations of GBCAs, although most studies have been conducted on patients receiving far more than that (4). Previous studies on gadolinium depositions in the brain mainly used conventional T1-weighted MR images, where the affected areas were compared to reference areas using intensity ratios, and gadolinium depositions were detected by an increase in these intensity ratios. This method does not provide quantitative data on gadolinium depositions (T1 signal intensity depends on MRI parameters and does not provide linear associations) and the results cannot be compared over time or between different patients or populations. Furthermore, histopathological studies have shown that gadolinium retention is widespread and also occurs in brain regions that are often used as reference areas, such as the thalami (6).
In this study we use quantitative MRI to detect early gadolinium depositions after two to ve (mean 3.4) doses of linear GBCAs without the use of reference areas. To our knowledge, so far, there have been only a few crosssectional studies with quantitative MRI in the context of gadolinium retention (7,8), and no studies on longitudinal data.
Our hypothesis is that using quantitative MRI, we can con rm the gadolinium depositions after fewer doses of GBCAs than generally seen on conventional MRI, without the need for reference areas, providing quantitative data Page 3/10 making comparison and follow-up in future studies on gadolinium depositions easier.

Subjects and GBCA doses
Four different GBCAs were used in MRI exams before and during the study; the linear GBCAs Magnevist and Omniscan, and the macrocyclic GBCAs Dotarem and Gadovist. Only linear GBCAs were studied since macrocyclic GBCAs are not associated with signal intensity change on MRI (3,4). The mean dose of Magnevist received by the 43 patients before the study was 0.14 (range 0-3), and the mean dose received during the study was 3.4 (range 2-5). The mean dose of Omniscan received before the study was 0.42 (range 0-2). No patients received Omniscan (or any other linear GBCA than Magnevist) during the study period (Table 1).

Discussion
This study is to our knowledge the rst to use quantitative MRI (quantitatively measured longitudinal relaxation) in longitudinal data, measuring gadolinium depositions over time, in our case in six areas of the brain following 3.4 doses (mean) of linear GBCA administration.
Since the results are quantitative, the measured changes are directly proportional to the gadolinium concentration (12). In contrast, when using conventional T1-weighted MRI, the intensity changes due to gadolinium depositions will depend non-linearly on the change in R 1 (i.e. also non-linearly on the gadolinium concentration) (13).
Additionally, quantitative MRI has fewer methodological limitations than conventional MRI, such as varying MRI parameters, repetition time, echo time, spin echo/gradient echo or systematic imperfections of RF and B0 inhomogeneities. Furthermore, as mentioned earlier, there is indication of potential gadolinium depositions in areas commonly used as reference areas (6). Arguably, quantitative data in combination with directly measuring areas without using reference areas may provide earlier detection of gadolinium depositions, since there are fewer confounding factors present. Comparisons of conventional and quantitative MRI in this regard could be a potentially interesting subject for future studies.
R 1 was signi cantly higher in the dentate nuclei, globus pallidi and frontal cortex grey matter following GBCA administration, whereas there was no difference in R 1 in the pons, thalami or white matter. These ndings of an intensity increase in the dentate nuclei, globus pallidi and frontal cortex grey matter are consistent with previous ndings (4,6,14,15). They are also consistent with the ndings of a recently published cross-sectional study using quantitative MRI for comparison of gadolinium depositions in MS patients and controls (8).
Our study did not show any intensity change in the thalami or pons, which has previously been reported (6,16). However, GBCA doses in this study could potentially be too few to present as an intensity increase, since dosedependent change in T1-weighted intensity is smaller in the pons and thalami compared to the dentate nucleus (6). Although studies (including ours) have shown no change in signal intensity in white matter, autopsies have shown gadolinium in all areas of the brain, indicating that MRI probably has a limited sensitivity to detect low levels of gadolinium depositions in the brain (5).
Subgroup analyses regarding correlation of GBCA doses and R 1 increase showed a dose-dependent association of GBCA doses and R 1 increase in the dentate nuclei and frontal cortex grey matter. No statistical signi cance was obtained in the globus pallidi, despite a tendency for R 1 increase between 3 and 4 doses seen in Fig. 2. Outliers in this group are a possible explanation for this result.
To con rm these ndings, future studies including more patients and GBCA administrations should be performed, where using 3T MRI instead of 1.5T MRI can provide better image quality and accuracy of ROI placements.
Despite the numerous studies that suggest gadolinium deposition in the brain after multiple GBCA doses, few studies on the clinical manifestations correlated to each deposition area have been published, and those published have not been able to connect gadolinium depositions to signs of toxicity or clinical manifestations (17,18). Nor have gadolinium brain depositions been proven to have any toxic effects in animal models (19,20). Consequently, further studies are needed for evaluation of the clinical manifestations of gadolinium depositions. Our results suggest that the use of quantitative MRI can be bene cial for these future studies.
To conclude, quantitative MRI offers an alternative to conventional MRI for detecting and comparing gadolinium depositions in the brain, which can be useful in future studies regarding the distribution and toxicity of gadolinium depositions. This method provides potentially earlier detection (compared to conventional MRI) without using reference areas which may be subjected to gadolinium depositions, in addition to providing quantitative data with fewer methodological limitations than conventional MRI.

Methods
The study was approved by the local institutional review board of the Swedish Ethical Review Authority in Linköping, and all methods were performed in accordance with the relevant guidelines and regulations. Informed written consent was obtained from all patients.

Study population
Forty-three patients who were earlier enrolled in a prospective longitudinal cohort study of early Multiple Sclerosis (MS) at the Department of Neurology at the University Hospital in Linköping, Sweden, were included. Originally, forty-six patients with a clinical suspicion of MS were consecutively enrolled and examined between October 2009 and May 2015. Three patients were excluded, one due to withdrawal of consent, one due to the nding of a trigeminal schwannoma, which explained the patient's clinical ndings, and one due to missing quantitative MRI data. All patients had a normal Glomerular Filtration Rate during the study. Table 2 shows details of the patient demographics. MR imaging according to a standard clinical MS protocol with the addition of quantitative MRI (qMRI) before and after administration of Magnevist (gadopentate dimeglumine, 0.2 mL/kg body weight, Bayer HealthCare Pharmaceuticals) was performed at inclusion (baseline) and after one year, two years and four years.
In order to have as homogenous a group as possible, all studies used Magnevist as the contrast agent. Some of the patients had received other contrast agents before the study, and during the study when seeking emergency healthcare. However, none of the patients received another linear contrast agent during the study period (Table 1).
We have chosen six areas for our study; two areas that are most established as targets for gadolinium depositions (the dentate nuclei and globus pallidi), and additionally four areas that are often studied regarding gadolinium depositions (the thalami, pons, white matter and frontal cortex grey matter).

MR imaging acquisition
Images were acquired on a 1.5T MR imaging scanner (Achieva, Philips Healthcare, Best, the Netherlands) by using an eight-channel phased array head coil. The sequence parameters for conventional images have previously been described (9).
The quantitative sequence, QMAP (10, 11) is a multi-section, multi-echo, and multi-saturation delay qMRI technique, used with the following parameters in this study: All images had a section thickness of 3 mm without an intersection gap. Post-contrast qMRI images were acquired approximately 15 minutes after intravenous injection of Magnevist.

Post-processing
The qMRI sequence yields quantitative maps of longitudinal relaxation rate R 1 (R 1 = 1/T1), transverse relaxation rate R 2 , and proton density (PD), which are used for measurements and to create quantitative images matching the conventional ones. The post-processing time of the raw image dataset was approximately one minute on an ordinary PC by using SyMRI Diagnostic software (SyntheticMR, Linköping, Sweden) to create quantitative maps of R 1 , R 2 , PD and synthetic T1, T2 and T2 FLAIR images.

Radiologic evaluation
Region of Interests (ROIs) were placed in the quantitative images (Fig. 3) in the area of the dentate nuclei, globus pallidi, thalami, pons, normal-appearing white matter and frontal cortex grey matter (Fig. 4). There were no MS lesions present in the areas of R 1 measurements. The segmentations were performed in MeVisLab (MeVis Medical Solutions AG, Bremen, Germany) by a radiology resident (A.L.L.), supervised by an experienced neuro-radiologist (I.B.). Conventional MR images were available for reference to con rm correct ROI placement. R 1 of the areas was measured in each MR study (the baseline study and the four-year control).

Statistical analysis
All statistical calculations were performed in IBM SPSS Statistics for Macintosh, Version 27.0 (IBM Corp, Armonk NY). A p-value of ≤ 0.05 was considered signi cant.
The R 1 value from the four-year control was compared to the R 1 value of the baseline study for each subject using a Paired-Samples T test, which was performed for each of the above-mentioned six areas in the brain. For areas with a signi cant increase of R 1 , an Independent-Samples T test was performed to correlate GBCA doses to the increase in R 1 in these areas. The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.