Authors: Sébastien Ballet, Jean-Sébastien Raynaud and Pierre Desché
Affiliation: Guerbet, Villepinte, France
Dear Editor in Chief,
In a recent article published in Insights into Imaging, Splendiani et al reported appearance of visible dentate nucleus (DN) T1-hyperintensity and quantifiable increases in DN/pons (DNP) signal intensity (SI) ratio in some patients suffering from relapsing-remitting multiple sclerosis (RR-MS) and exposed to multiple administrations of gadoterate meglumine and/or gadobutrol . While these results may be relevant for a better understanding of the cerebral behavior of gadolinium-based contrast agents (GBCAs) over time in these specific patients undergoing multiple contrast-enhanced MRI examinations, we believe that several shortcomings and limitations in the methods and results analyses may somewhat preclude or limit the conclusions of the study and deserve discussion.
As mentioned by the authors, the same clinical team has already reported in 2018 visible T1-hyperintensity in the DN in one third of patients with RR-MS who had received cumulative doses of the macrocyclic GBCAs, namely gadobutrol or gadoterate meglumine . It is noteworthy that Raynaud et al highlighted several inaccuracies and mistakes in a published Letter to Editor .
In this 2019 publication, in the same site (L’Aquila) and on the same population (RR-MS patients), a longer survey period from January 2005 to December 2017 (versus March 2007 to July 2016 in the 2018 publication) and a lower injection threshold of only 2 injections (versus 4 injections in the 2018 publication) were reported. Despite the inclusion of 29 patients receiving both gadoterate meglumine and gadobutrol (versus none in the 2018 publication), an unexpected lower number of patients was reported in the 2019 publication compared to the 2018 publication (149 versus 158 patients, respectively). The authors did not provide explanations that could exclude any bias in patient selection process.
Moreover, the image analysis methods are different between the two studies. Indeed, the identification of the DN and the central part of the pons, the method to place the region of interest, and the method to calculate the DNP ratio (mean change in DNP ratio in 2018; DNP ratio values at first and last MRI examinations in the present study) were different without any rationale.
In terms of results, we also detected some unexpected discrepancies between the 2018 publication and the present study. While the mean number of GBCA administrations are similar between the two studies, the mean change in DNP ratio between first and last MRI examinations are unexpectedly different. In 2018, mean change in DNP ratio was 0.0032 ± 0.0216 and 0.0019 ± 0.0346 after multiple administrations of gadoterate meglumine and gadobutrol, respectively. In the present study, we calculated values equal to 0.004 and 0.001 for the patients, injected with gadoterate meglumine and gadobutrol, respectively, without visible T1-hyperintensity. These values are consistent with those reported in 2018. However, calculated mean change in DNP ratio in patients with visible T1-hyperintensity in the DN drastically increased at similar order of magnitude compared to those reported in the literature for linear GBCAs: 0.107 and 0.155 after multiple administrations of gadoterate meglumine and gadobutrol, respectively. The latter result is unexpected since the part of patients with visible T1-hyperintensity in the DN are quite similar in both publications (one third in 2018, and 26% and 30% in the present study for gadoterate meglumine and gadobutrol, respectively). Moreover, the mean number of GBCA administrations at first visible T1-hyperintensity in the DN is around 4 for both studied GBCAs, which was never observed in previous publications after multiple administrations of macrocyclic GBCAs. The reason(s) why the mean change in DNP ratio increased so drastically and unexpectedly was not explained in the present study.
Taking into consideration the multiple differences and inconsistencies between the two studies from the same clinical team, we consider that the present findings do not fully support those from Splendiani et al (2018) as it is stated by the authors in the present publication and draw our attention on the methodology.
As acknowledged by the authors the absence of an age-matched control group of RR-MS patients, who had never been exposed to any GBCA, represent a main limitation of the present study. Because MS is a degenerative disease over time, progression can be a potential confounding variable which necessarily requires DNP ratio measurements in patients without GBCA administration. Moreover, the authors should report individual SI values in the DN and in the pons at the first and last MRI examination in both control and GBCA-administered patients to exclude any potential equivocal bias in the change in DNP ratio which might be due to either disease-related increase of SI value in the DN or disease-related decrease of SI value in the pons between the first and the last MRI examination. Previous studies [4, 5] confirmed that it is difficult to determine whether T1-hyperintensity in the DN of MS patients is due to the extent of the disease or to the type of GBCA administered. For this reason, any new study exploring MS patients should follow the recent E-GREC recommendations  stating that comparison between signal intensities should be done using age-matched patient group not exposed to GBCAs.
In summary, considering both the unexpected and unexplained differences between Splendiani et al (2018) and the present study, the absence of data from a control group (RR-MS patients without GBCA exposure), and lacking information on patient characteristics, we believe that several of the conclusions from the present study should be discussed by taking into consideration the abovementioned reservations and that the authors should mitigate their conclusions.
 Splendiani A, Corridore A, Torlone S et al (2019) Visible T1-hyperintensity of the dentate nucleus after multiple administrations of macrocyclic gadolinium-based contrast agents: yes or no? Insights Imaging 10(1):82
 Splendiani A, Perri M, Marsecano C et al (2018) Effects of serial macrocyclic-based contrast materials gadoterate meglumine and gadobutrol administrations on gadolinium-related dentate nuclei signal increases in unenhanced T1-weighted brain: a retrospective study in 158 multiple sclerosis (MS) patients. Radiol Med 123(2):125-134
 Raynaud JS, Darmon-Kern E, Lancelot E, Laurent AC, Desché P (2018) Re: Effects of serial macrocyclic-based contrast materials gadoterate meglumine and gadobutrol administrations on gadolinium-related dentate nuclei signal increases in unenhanced T1-weighted brain: a retrospective study in 158 multiple sclerosis (MS) patients. Radiol Med 123(6):432-433
 Roccatagliata L, Vuolo L, Bonzano L, Pichiecchio A, Mancardi GL (2009) Multiple sclerosis: hyperintense dentate nucleus on unenhanced T1-weighted MR images is associated with the secondary progressive subtype. Radiology 251(2):503-510
 Forslin Y, Shams S, Hashim F et al (2017) Retention of Gadolinium-Based Contrast Agents in Multiple Sclerosis: Retrospective Analysis of an 18-Year Longitudinal Study. AJNR Am J Neuroradiol 38(7):1311-1316
 Quattrocchi CC, Ramalho J, van der Molen AJ, Rovira À, Radbruch A (2019) Standardized assessment of the signal intensity increases on unenhanced T1-weighted images in the brain: the European Gadolinium Retention Evaluation Consortium (GREC) Task Force position statement. Eur Radiol 29(8):3959-3967