The effect of gadolinium-based contrast-agents on automated brain atrophy measurements by FreeSurfer in patients with multiple sclerosis

Abstract

Objective To determine whether reliable brain atrophy measures can be obtained from post-contrast 3D T1-weighted images in patients with multiple sclerosis (MS) using FreeSurfer. Methods Twenty-two patients with MS were included, in which 3D T1-weighted MR images were obtained during the same scanner visit, with the same acquisition protocol, before and after administration of gadolinium-based contrast agents (GBCAs). Two FreeSurfer versions (v.6.0.1 and v.7.1.1.) were applied to calculate grey matter (GM) and white matter (WM) volumes and global and regional cortical thickness. The consistency between measures obtained in pre- and post-contrast images was assessed by intra-class correlation coefficient (ICC), the difference was investigated by paired t-tests, and the mean percentage increase or decrease was calculated for total WM and GM matter volume, total deep GM and thalamus volume, and mean cortical thickness. Results Good to excellent reliability was found between all investigated measures, with ICC ranging from 0.926 to 0.996, all p values < 0.001. GM volumes and cortical thickness measurements were significantly higher in post-contrast images by 3.1 to 17.4%, while total WM volume decreased significantly by 1.7% (all p values < 0.001). Conclusion The consistency between values obtained from pre- and post-contrast images was excellent, suggesting it may be possible to extract reliable brain atrophy measurements from T1-weighted images acquired after administration of GBCAs, using FreeSurfer. However, absolute values were systematically different between pre- and post-contrast images, meaning that such images should not be compared directly. Potential systematic effects, possibly dependent on GBCA dose or the delay time after contrast injection, should be investigated. Trial registration Clinical trials.gov. identifier: NCT00360906. Key Points • The influence of gadolinium-based contrast agents (GBCAs) on atrophy measurements is still largely unknown and challenges the use of a considerable source of historical and prospective real-world data. • In 22 patients with multiple sclerosis, the consistency between brain atrophy measurements obtained from pre- and post-contrast images was excellent, suggesting it may be possible to extract reliable atrophy measurements in T1-weighted images acquired after administration of GBCAs, using FreeSurfer. • Absolute values were systematically different between pre- and post-contrast images, meaning that such images should not be compared directly, and measurements extracted from certain regions (e.g., the temporal pole) should be interpreted with caution.

Serum neurofilament as a predictor of 10-year grey matter atrophy and clinical disability in multiple sclerosis: a longitudinal study

“…Lastly, atrophy measurements were obtained from postcontrast images, which is not the standard approach for FreeSurfer. However, recent work has shown excellent consistency between values obtained from precontrast and postcontrast images 24…”

“…However, recent work has shown excellent consistency between values obtained from precontrast and postcontrast images. 24 …”

“…A detailed description of these methods has recently been described 24 and is available in online supplemental appendix 1 . Briefly, on images obtained at the 10-year follow-up visit, lesion segmentation was done on FLAIR images using Lesion Segmentation Tool (V.2.0.15; http://applied-statistics.de/lst.html ), 25 and morphological reconstruction was performed with FreeSurfer (V.7.1.1; http://surfer.nmr.mgh.harvard.edu/ ) on T1-weighted images.…”

The Effect of Smoking on Long-term Gray Matter Atrophy and Clinical Disability in Patients with Relapsing-Remitting Multiple Sclerosis

“…Furthermore, total cerebral GM and white matter (WM) volume and total deep GM and thalamus volume (average of left and right hemisphere) were obtained. Because of frequent suboptimal segmentation of the temporal pole (previously found in a minor to moderate degree in almost 50% of postcontrast scans 28 ), this region was excluded when calculating the total GM volume.…”

Reliable brain morphometry from contrast‐enhanced T1w‐MRI in patients with multiple sclerosis

“…Pearson correlation coefficients ( r ) were calculated between image pairs, both across all pairs ( n = 454) as well as on the subset of pairs with identical sequence parameters (TI/TR, n = 213). To allow direct comparison with previous work (Lie et al, 2022 ), intra‐class correlation coefficients (ICC) were calculated based on mean‐rating ( k = 2), consistency, 2‐way mixed‐effects model (Koo & Li, 2016 ) with the R‐package irr (Gamer et al, 2012 ).…”