Symptoms Associated With Gadolinium Exposure

Abstract

Objective:

Some patients who received multiple administrations of gadolinium-based contrast agents (GBCAs) have been reported to develop “symptoms associated with gadolinium exposure” (SAGE). The aim of this study was to analyze pharmacovigilance data and to explore the various SAGE patterns of linear and macrocyclic GBCAs among patients exhibiting 3 or more SAGE symptoms.

Materials and Methods:

SAGE were identified from a review of the scientific literature, and the corresponding preferred terms (PTs) were searched in each system organ class recorded in the FDA Adverse Event Reporting System (FAERS). To ensure the comparability of data, 3 macrocyclic and 3 linear extracellular GBCAs currently approved for intravenous administration in the United States were considered. Only patients with 3 or more SAGE symptoms were included. SAGE weights, representing the percentage of SAGE symptoms among all adverse events collected over a 6-year period from 2014 to 2019, were calculated for each GBCA. The frequency of these symptoms to occur in sets of “3 PT combinations” was also analyzed. The 3 PT combinations were calculated by first selecting the PT with the highest occurrence for a GBCA and then combining it with all the PTs accounting for 5% or more of the total adverse events reported for the respective GBCA. This led to identify the most prevalent 3 PT combinations per GBCA. Moreover, in order to determine whether or not SAGE symptoms were specific to GBCAs, data for 4 water-soluble iodinated contrast media were also extracted from the FAERS database over the same period, using the SAGE list of symptoms as reference.

Results:

The analysis of FAERS data revealed a significantly higher SAGE weight for the linear GBCAs (20%–24%) than for the macrocyclic GBCAs (5%–9%). For the linear agents, the most prevalent 3 PT combinations of SAGE symptoms were reported in 152–164 occurrences, whereas for the macrocyclic agents, this range was significantly lower (1–13 occurrences). Moreover, all these agents could be categorized in 3 groups with different patterns of 3 PT combinations (ie, [gadodiamide and gadopentetate dimeglumine], [gadobenate dimeglumine and gadoteridol], and [gadoterate dimeglumine and gadobutrol]). The most prevalent PTs were found to be “pain,” “arthralgia,” and “headache” in each group, respectively.

Conclusions:

The global SAGE weights were significantly lower for the macrocyclic GBCAs as compared with the linear GBCAs. Moreover, the frequency of occurrence of 3 PT combinations was notably lower with the macrocyclic agents and comparable to the iodinated contrast media, indicating that SAGE may be negligible for this class of GBCAs. Different patterns of 3 PT combinations were also observed among the GBCAs involved in this study. A causal relationship could not be established between SAGE and the corresponding GBCAs, therefore, further research on this topic and routine pharmacovigilance are warranted.

Gadolinium is a rare earth metal belonging to lanthanide group of elements.¹ It has exceptional properties for enhancing magnetic resonance imaging signals at low concentrations.² This is achieved by reducing the longitudinal relaxation time (T1) of tissues resulting in increased signal intensity on T1-weighted images.³ Gadolinium is toxic in its free form; therefore, gadolinium chelates, referred to as gadolinium-based contrast agents (GBCAs), are synthesized in which gadolinium ion is bound to a chelator molecule to favor elimination and reduce toxicity.⁴ Based on the configuration of the chelate, GBCAs can be classified as either macrocyclic or linear molecules (Fig. 1).⁵ The macrocyclic GBCAs are known to exhibit very high kinetic stability, which enables their chelates to withhold gadolinium ion for significantly longer period as compared with the linear GBCAs (Fig. 1).⁶

FIGURE 1.

Structure and kinetic stabilities of routinely used extracellular GBCAs approved for intravenous use in the United States. â  Dissociation half-life (T _1/2 ) in acidic conditions (HCl, pH 1.0) at 25ºC.⁶

Currently, almost 50 tons of gadolinium is administered annually worldwide for diagnosis and follow-up of several diseases.^7–9 GBCAs are considered safe when used at doses approved for clinical use (0.1–0.3 mmol/kg).¹⁰ The rate of immediate adverse events is very low (<1 per 1000 injections), and these events are usually mild, with severe adverse events occurring in about 1 per 40,000 injections.¹¹ However, GBCAs are not without risks. Alongside hypersensitivity reactions and acute nephrotoxicity, nephrogenic systemic fibrosis (NSF), presenting with skin thickening and reddening, has been identified as a potentially debilitating condition associated with GBCA exposure, in patients with impaired renal function.^12–14

Moreover, gadolinium deposition in some brain areas, namely, globus pallidus and dentate nucleus, and other organs, primarily skin and bone, have raised some concerns about the long-term safety of GBCAs, even in patients with normal renal function, especially after multiple administrations of linear GBCAs.^15–25 Recently, a pattern of symptoms similar to NSF, but less severe, was associated with a history of GBCA exposure in patients with normal renal function. The term gadolinium deposition disease (GDD) has been suggested, encompassing a collection of nonspecific symptoms such as headache, fatigue, bone and joint pain, subcutaneous tissue thickening, and so on.²⁶ Although a few independent groups have published reports in the peer-reviewed literature,²⁷ GDD has not been fully accepted by the medical community.²⁸

In 2022, a new term, namely “symptoms associated with gadolinium exposure” (SAGE), was proposed by the American College of Radiology (ACR) as an alternative nomenclature, which does not presume a causal relationship between those symptoms and GBCA administration.²⁹ To date, the scientific and clinical data suggest a coincidental rather than causal relationship. SAGE is defined as a list of symptoms that may occur irrespective of kidney function and are unrelated to established early-onset adverse events (occurring <24 hours after GBCA exposure, such as acute allergic-like and physiologic reactions) and late-onset adverse events (occurring >24 hours after GBCA exposure, such as NSF).²⁹

In our recently published study, we explored SAGE by conducting an extensive analysis of the adverse events associated with 4 GBCAs and recorded in the European Medicines Agency and Food and Drug Administration (FDA) databases of suspected adverse drug reaction reports.³⁰ The aim of the current study was to approach the topic of SAGE with more optimized inclusion criteria and assess the frequency of adverse events reported to the FDA for 6 GBCAs currently approved for intravenous use in the United States.

MATERIALS AND METHODS

This study assessed the spontaneous reports of adverse events in the publicly available pharmacovigilance database in the United States, namely, the FDA Adverse Event Reporting System (FAERS), which contains reports of undesirable side effects potentially associated with the use of medicines since 1968. All the events received by the FDA have been coded using the Medical Dictionary for Regulatory Activities (MedDRA), assigned a preferred term (PT), and grouped by system organ classes (SOCs) in the FAERS database.

Selection Criteria

A search was undertaken in electronic databases of scientific and medical journals (PubMed and EMBASE) up to June 2023. The aim was to identify all the published articles (clinical studies or case reports) reporting persistent clinical signs or symptoms of potential gadolinium toxicity among patients exposed to GBCAs, independent of their renal function. The keywords used for the search were “symptoms associated with gadolinium exposure,” “SAGE,” “gadolinium deposition disease,” “gadolinium poisoning,” “gadolinium disease,” “gadolinium and chelation therapy,” and “small fiber neuropathy.”

The strategy for including or rejecting a study was the same as reported previously.³⁰ The adverse events selected from the retained articles were then searched for in the appropriate SOCs and PTs in the FAERS database.

For the current study, the data extraction was performed from the FAERS database for each GBCA, during a 6-year period ranging from January 2014 to December 2019. The rationale behind this selection was based on the following criteria: (i) the period before 2012 was excluded due to the higher prevalence of NSF cases with linear agents, which present common features with SAGE symptoms; (ii) the period before 2013 was excluded as it represented very limited or no use of gadobutrol and gadoterate meglumine, as they were new agents licensed in the United States; and (iii) the period from January 2020 onward was also excluded due to the significant drop of spontaneous notifications of adverse events to worldwide health authorities during the coronavirus disease 2019 (COVID-19) pandemic.

Moreover, in order to be consistent with the diagnostic criteria proposed for the so-called “gadolinium deposition disease,”³¹ only patients who exhibited 3 or more SAGE symptoms were included, thus reinforcing the specificity for SAGE in this study.³²

Contrast Agents

All extracellular GBCAs approved and marketed in the United States for the intravenous route of administration were included in this study. These included the macrocyclic GBCAs, namely, gadoteridol (Prohance; Bracco), gadobutrol (Gadovist/Gadavist; Bayer Healthcare), and gadoterate meglumine (Dotarem; Guerbet and Clariscan; GE Healthcare), as well as the linear agents, namely, gadobenate dimeglumine (Multihance; Bracco), gadopentetate dimeglumine (Magnevist; Bayer Healthcare), and gadodiamide (Omniscan; GE Healthcare). The remaining GBCAs either exhibited a notably different pharmacokinetic profile than the extracellular GBCAs or were approved for nonintravenous route of administration, or their manufacturing was discontinued. The selected linear and macrocyclic GBCAs were injected for single or multiple administrations. The period for the manifestation of SAGE after GBCA exposure ranged between 1 day and 9 years.

Moreover, in order to determine whether or not SAGE symptoms were specific to GBCAs, data for 4 water-soluble iodinated contrast media (ICM) were also extracted from the FAERS database, using SAGE list of symptoms as reference, covering the years 2014–2019. These ICM included ioversol (Optiray; Guerbet), iohexol (Omnipaque; GE Healthcare), iopamidol (Niopam/Isovue; Bracco), and iopromide (Ultravist; Bayer Healthcare).

Calculations

In order to determine the proclivity of the GBCAs to cause at least 3 SAGE symptoms, SAGE weights were determined per PT, by calculating the ratio between the number of adverse events in the PT and the total number of adverse events recorded in the database.³⁰ Then SAGE weights per SOC were obtained by calculating the ratio between the number of adverse events in the relevant PTs of this SOC and the total number of adverse events. All SAGE weights were presented as percentages. Descriptive statistics were applied to compare the SAGE weights per SOC between the different GBCAs.

Furthermore, additional analysis was undertaken to determine the most prevalent combinations of 3 SAGE symptoms per patient per GBCA. This was achieved by first selecting the PT with the highest occurrence for each GBCA and then combining it with all the PTs accounting for 5% or more of the total adverse events reported per GBCA. This led to identify the most prevalent 3 PT combinations per GBCA. Moreover, 3 PT combinations were also produced for the ICM and compared with both linear and macrocyclic GBCAs using descriptive statistics.

RESULTS

Overall, 127 publications of interest were identified in the scientific and medical literature. After a preliminary exclusion of papers not focusing on the SAGE topic, 28 potentially relevant articles were examined, from which 15 publications were selected. These included 9 clinical studies, 5 case reports, and 1 review article.^29,31–44

A total of 35,314 adverse events were found in the FAERS database over the study period with the following distribution: gadobutrol (n = 8565), gadobenate dimeglumine (n = 7167), gadopentetate dimeglumine (n = 6326), gadoterate meglumine (n = 5791), gadodiamide (n = 5548), and gadoteridol (n = 1917). As per the inclusion criteria, only cases exhibiting 3 or more SAGE symptoms were selected from these adverse events to calculate the SAGE weights (ie, number of SAGE events/total number of adverse events). Figure 2 provides the SAGE weights for each GBCA calculated globally and for each SOC. Globally, gadodiamide showed the highest SAGE weight (23.9%) of the 6 GBCAs included in this study. This was followed by gadopentetate dimeglumine (22.8%), gadobenate dimeglumine (20.0%), gadoteridol (9.5%), gadobutrol (5.5%), and gadoterate meglumine (4.7%).

FIGURE 2.

SAGE weights determined from FAERS database. SAGE indicates symptoms associated with gadolinium exposure.

When analyzed per SOC, the SAGE weights ranged between 0.0% and 9.3% for the linear agents and 0.0% and 4.2% for the macrocyclic agents. For the SOCs “investigations” and “skin and subcutaneous tissue disorders,” all GBCAs showed similar SAGE weights (<1.0%). For the SOC “psychiatric disorders,” greater SAGE weights were observed for the linear GBCAs (2.4% to 3.0%) as compared with the macrocyclic GBCAs (0.2% to 0.4%). For the remaining SOCs, gadobenate dimeglumine presented the lowest SAGE weight among the linear GBCAs; gadoteridol exhibited systematically a higher SAGE weight among the macrocyclic GBCAs.

For the SOC “musculoskeletal and connective tissue disorders,” the highest SAGE weights were reported for gadodiamide (7.4%) and gadopentetate dimeglumine (7.1%) followed by gadobenate dimeglumine (6.3%) and gadoteridol (4.2%). The SAGE weights of gadobutrol (2.2%) and gadoterate meglumine (1.2%) were notably lower (Fig. 2). The most prevalent PTs in the safety experiences recorded were “arthralgia” for the linear GBCAs, and to a lower extent “muscular weakness” and “pain in extremity” for gadoteridol (Fig. 3).

FIGURE 3.

SAGE weights by relevant PTs in the SOCs: “skin and subcutaneous tissue disorders” (A), and “musculoskeletal and connective tissue disorders” (B). The figures on the graphs represent SAGE weights. Others: pain in extremity, muscle tightness, muscle fatigue, muscle spasm, muscle twitching, musculoskeletal chest pain, fibromyalgia, ligament, and tendon pain.

For the SOC “general disorders and administration site conditions,” the linear GBCAs showed much higher SAGE weights (7.5%–9.3%) than the macrocyclic GBCAs (1.0%–2.1%) (Fig. 2). The most common PTs for this SOC were “fatigue,” “asthenia,” and “pain” showing the highest prevalence with the linear GBCAs, followed by gadoteridol (Fig. 4).

FIGURE 4.

SAGE weights by relevant PTs in the SOCs: “general disorders and administration site conditions” (A), “psychiatric disorders” (B), and “nervous system disorders” (C). The figures on the graphs represent SAGE weights.

For the SOC “nervous system disorders,” the SAGE weights of linear GBCAs were in the range of 3.1%–3.8%, which was about 2 times higher than the macrocyclic GBCAs (ranging between 1.3% and 2.0%) (Fig. 2). The most frequently reported PTs for this SOC were “paraesthesia,” “headache,” and “cognitive disorder” (Fig. 4). For the SOC “psychiatric disorders,” SAGE weights for linear GBCAs ranged between 2.4% and 3.0%, whereas for the macrocyclic GBCAs, this range was between 0.2% and 0.4% (Fig. 4). The linear GBCAs reported a notably higher SAGE weight for the PT “confusional state.”

Additional analysis was undertaken to assess the frequency of PTs, which occurred in combinations of 3 SAGE symptoms, referred to as 3 PT combinations. For the linear GBCAs, the most prevalent 3 PT combinations were reported in 152–164 occurrences, whereas for the macrocyclic GBCAs, this range was significantly lower (1–13 occurrences). Moreover, the 4 water-soluble ICM showed a very low frequency of 3 PT combinations for SAGE-like symptoms, with 1–4 occurrences (Fig. 5).

FIGURE 5.

Number of occurrence of the most frequent “3 PT combinations” among GBCAs and ICM.

Figure 6 displays the most frequent 3 PT combinations obtained for the linear GBCAs. The 3 PT combination “arthralgia/asthenia/pain” exhibited the highest occurrence as it occurred 164 times with gadobenate dimeglumine, 163 times with gadopentetate dimeglumine, and 160 times with gadodiamide. This was followed by the 3 PT combination “arthralgia/paraesthesia/pain,” which occurred 162, 161, and 153 times among patients exposed to gadopentetate dimeglumine, gadobenate dimeglumine, and gadodiamide, respectively. The 3 linear GBCAs displayed profoundly different patterns of 3 PT combinations: gadobenate dimeglumine was most frequently associated with combinations involving “arthralgia,” whereas gadopentetate dimeglumine and gadodiamide were more associated with combinations involving “pain.” Moreover, when considering the 3 PT combinations which were common to the 3 linear GBCAs (highlighted in gray in Fig. 6), it appears that “arthralgia” and “pain” were systematically reported.

FIGURE 6.

Number of SAGE representing 3 PT combinations for the linear GBCAs. C.S., confusional state; As, asthenia; Ar, arthralgia; Fa, fatigue; M.S., muscle spasms; Pa, paraesthesia; Pn, pain.

With regards to macrocyclic GBCAs, the most frequent 3 PT combinations were “headache/asthenia/fatigue,” which occurred 13 times in patients exposed to gadobutrol, and “arthralgia/fatigue/pain,” which occurred 12 times in patients exposed to gadoteridol (Fig. 7). Gadoteridol displayed a significantly different pattern of 3 PT combinations as it was most frequently associated with “arthralgia,” whereas gadobutrol and gadoterate meglumine were more associated with “headache.” Moreover, when considering the 3 PT combinations which were common to the 3 macrocyclic GBCAs (highlighted in gray in Fig. 7), it appears that “arthralgia” and “headache” were systematically reported.

FIGURE 7.

Number of SAGE representing “3 PT combinations” for the macrocyclic GBCAs. Pa, paraesthesia; Pn, pain; Ha, headache; M.W., muscle weakness; As, asthenia; Ar, arthralgia; Fa, fatigue; P.E., pain in extremity; B.P., bone pain; M.T., muscle twitching.

DISCUSSION

This study compared the proclivity of SAGE using real-world data from FAERS, which is one of the largest databases of adverse events designed to support the FDA postmarketing safety surveillance program for approved drugs. As FAERS is a spontaneous reporting system, therefore, the data obtained from this database cannot be used to calculate the real incidence rates of the adverse events in the exposed population. Moreover, since the adverse events recorded in FAERS are not medically validated, the demonstration of a causal relationship between GBCA administration and SAGE cannot be established. As an alternative, we considered the proclivity of the selected GBCAs to cause SAGE and presented them as SAGE weights as reported previously.³⁰

The evidence of prolonged gadolinium excretion and deposition in animal and human tissues after GBCA administration has long been recognized, both in preclinical and clinical studies.^45–54 Moreover, long-term gadolinium accumulation in the body and urinary elimination over a few weeks or months after the administration of GBCAs have also been reported.^55,56 Until now, gadolinium retention in the brain and other organs did not lead to any clinical consequences.²⁸

To date, we have insufficient data showing toxicity resulting from gadolinium deposition in tissues.^28,29 Nevertheless, hypotheses of a causal association between gadolinium retention after GBCA administration have been proposed. In 2016, Semelka and colleagues²⁶ suggested the term GDD to describe a clinical condition in which patients with normal kidney function developed long-lasting symptoms after GBCA exposure. They published several case series of patients who complained of intense pain, burning sensation, headache, and clouded mentation, as well as erythematous and thickened skin and subcutaneous tissue.^{31–35,38–40} However, in 2017, the FDA and the European Medicines Agency evaluated the existing data and found no scientific evidence to suggest a causal association between GBCA exposure and the development of symptoms in patients with normal kidney function.^57,58 In 2022, the ACR Committee on Drugs and Contrast Media stated that a disease-specific term, such as GDD, was deemed inappropriate. Therefore, they proposed a new term, namely SAGE, to replace GDD or any associated terms.²⁹

Our analysis, involving 3 or more adverse events, showed that the proportion of SAGE among all the recorded adverse events was significantly greater for the linear agents, ranging between 20.0% and 23.9%, as compared with the macrocyclic GBCAs, ranging between 4.7% and 9.5%. However, among the latter, gadoteridol showed SAGE weights that were almost 2 times greater as compared with gadoterate meglumine and gadobutrol.

With regards to the SOCs “psychiatric disorders,” “general disorders and administration site conditions,” and “musculoskeletal and connective tissue disorders,” the 3 linear GBCAs reported dramatically higher SAGE weights which were 10, 6, and 3 times higher than the 3 macrocyclic GBCAs, respectively. However, among the macrocyclic GBCAs, gadoteridol showed twice greater SAGE weights for all of these SOCs, as compared with other agents of the class.

Overall, “confusional state (clouded mentation/brain fog),” “pain,” “fatigue,” “asthenia,” “arthralgia,” “paraesthesia,” and “skin burning sensation” resulted in the highest SAGE weights in their respective SOCs, whereas “bone pain,” “muscular weakness,” “headache,” “cognitive disorder,” “skin discoloration,” and “skin induration” were of notable mention. Among the published data, similar symptoms have been reported in patients with normal renal function, after exposure to GBCAs. In a case series published by Semelka and colleagues,³⁴ 3 patients experienced “pain” affecting the central torso, hands, legs, and feet, 2 patients experienced “headache” and “confusional state” like symptoms such as “clouded mentation,” whereas others experienced “arthralgia,” “skin induration,” and “skin discoloration.” All these patients were exclusively administered linear GBCAs, namely, gadopentetate dimeglumine, gadodiamide, and gadobenate dimeglumine. Furthermore, Burke et al³³ reported survey results of 50 patients who experienced “bone pain,” “arthralgia,” and “headache” in over 75% of the cases and skin changes in approximately 60% of the cases. Linear agents were administered exclusively to 62% of the patients, whereas only 1 patient was administered a macrocyclic GBCA.

Moreover, in an observational study involving 42 patients, “bone pain,” “joint stiffness,” and “fatigue” were experienced by more than 75% of the patients, and about 70% of the patients experienced “clouded mentation” and “skin discoloration.”³¹ A majority of these patients (38%) received exclusively linear GBCAs, whereas 1 patient received a macrocyclic GBCA alone.

In the present study, in addition to selecting SAGE cases with 3 or more symptoms, we also analyzed the frequency of these symptoms in sets of 3 PT combinations. The frequency of these 3 PT combinations was significantly higher for the linear GBCAs as compared with the macrocyclic GBCAs. However, no notable difference could be observed between the macrocyclic GBCAs and the ICM, indicating that SAGE-like symptoms may not represent a risk for patients receiving macrocyclic GBCAs. Interestingly, for the linear GBCA gadobenate dimeglumine and the macrocyclic GBCA gadoteridol, the majority of 3 PT combinations were unique to these agents. On the contrary, among the linear GBCAs, gadodiamide and gadopentetate dimeglumine showed a similar pattern of 3 PT combinations. Among the macrocyclic GBCAs, gadobutrol and gadoterate meglumine also showed a similar pattern. These results suggest that all GBCAs do not present the same patterns of SAGE events.

It is noteworthy that, among the 3 PT combinations, “arthralgia” was consistently reported with gadobenate dimeglumine and gadoteridol, “pain” was more commonly reported with gadodiamide and gadopentetate dimeglumine, and “headache” was predominantly reported with gadoterate meglumine and gadobutrol. These results are consistent with the combination of symptoms described in the published literature.^31–40

The specific reason behind higher SAGE weights associated with linear GBCAs as compared with macrocyclic GBCAs is unclear. However, the results of the current study are in alignment with NSF and gadolinium deposition profiles of the 2 classes of GBCAs. Like SAGE weights, much more cases of NSF and gadolinium retention in the body have been reported with linear GBCAs as compared with macrocyclic GBCAs.^28,59 One plausible explanation behind this could be the kinetic stability of these GBCAs, which is a measure of dissociation of gadolinium from the chelate of the GBCA molecule.⁵ It is well established that the linear GBCAs have significantly lower kinetic stabilities (a few seconds) as compared with the macrocyclic GBCAs (a few hours to days).⁶ However, kinetic stability measurements are conducted in in vitro conditions; therefore, they are limited in predicting the behavior of the GBCAs in vivo. The pathological mechanism(s) remain incompletely understood and merits further investigation.

This study bears some limitations. First, because FAERS is a voluntary reporting system, it is likely that a large proportion of adverse events is missed. This precludes from accurately calculating the importance of SAGE. Second, as the overall exposure of patients to each GBCA is unknown, it is impossible to determine the real incidence of SAGE using FAERS data. However, we addressed this problem by calculating SAGE weights. Third, due to the lack of medical validation of the pharmacovigilance cases recorded in FAERS, a causal relationship between GBCAs and SAGE could not be established. However, we were able to appreciate the proclivity of SAGE using a more robust method to select the patients, over a period of 6 years. Fourth, duplication of data could not be avoided in FAERS database. However, as we selected cases who reported 3 or more symptoms, our analysis included a very selective patient population where duplication of adverse events could have been avoided. Fifth, since FAERS database is a 1-shot reporting system, there is no way of assessing the persistence or durability of symptoms. In addition, patients' history and injection protocol were not accessible. The database does not provide information if the patients were suffering from other medical conditions, which could be a key confounding factor. Finally, no information regarding the total number of GBCA injections and the type of GBCAs previously received by the patients was available. Therefore, the impact of total gadolinium load among the affected patients could not be assessed.

In conclusion, we provided a thorough analysis of the SAGE profile of GBCAs in FAERS database, over a period of 6 years, in the form of 2 datasets. First, we selected cases with 3 or more symptoms qualifying the definition of SAGE and presented their proclivity as SAGE weights. Second, the most prevalent of these symptoms were shortlisted to determine their frequency to occur in sets of 3 PT combinations. Globally, the linear GBCAs showed significantly higher SAGE weights as compared with the macrocyclic GBCAs; however, gadoteridol acquired more of an intermediate position among all GBCAs. With regards to the most frequent 3 PT combinations, all GBCAs could be categorized into 3 groups showing similar patterns of symptoms: (i) gadodiamide and gadopentetate dimeglumine, (ii) gadobenate dimeglumine and gadoteridol, and (iii) gadoterate meglumine and gadobutrol. The most prevalent PTs were found to be “pain,” “arthralgia,” and “headache” in each group, respectively. Moreover, the SAGE weights of macrocyclic GBCAs were very close to those of ICM, indicating that SAGE may be negligible for this class of GBCAs. Overall, a causal relationship could not be established between SAGE and the corresponding GBCAs, therefore, further research on this topic and routine pharmacovigilance are warranted.

ORCID ID

Imran Shahid https://orcid.org/0000-0002-5852-3037