Gadolinium Chelate Safety in Pregnancy: Barely Detectable Gadolinium Levels in the Juvenile Nonhuman Primate after in Utero Exposure

After gadoteridol undergoes transplacental transfer to the fetus, it is almost completely cleared, with nearly undetectable levels in the juvenile macaque.

Abstract

Purpose

To determine whether gadolinium remains in juvenile nonhuman primate tissue after maternal exposure to intravenous gadoteridol during pregnancy.

Materials and Methods

Gravid rhesus macaques and their offspring (n = 10) were maintained, as approved by the institutional animal care and utilization committee. They were prospectively studied as part of a pre-existing ongoing research protocol to evaluate the effects of maternal malnutrition on placental and fetal development. On gestational days 85 and 135, they underwent placental magnetic resonance imaging after intravenous gadoteridol administration. Amniocentesis was performed on day 135 prior to administration of the second dose of gadoteridol. After delivery, the offspring were followed for 7 months. Tissue samples from eight different organs and from blood were harvested from each juvenile macaque. Gadolinium levels were measured by using inductively coupled plasma mass spectrometry.

Results

Gadolinium concentration in the amniotic fluid was 0.028 × 10^-5 %ID/g (percentage injected dose per gram of tissue) 50 days after administration of one gadoteridol dose. Gadolinium was most consistently detected in the femur (mean, 2.5 × 10^-5 %ID/g; range, [0.81–4.1] × 10^-5 %ID/g) and liver (mean, 0.15 × 10^-5 %ID/g; range, [0–0.26] × 10^-5 %ID/g). Levels were undetectable in the remaining sampled tissues, with the exception of one juvenile skin sample (0.07 × 10^-5 %ID/g), one juvenile spleen sample (0.039 × 10^-5 %ID/g), and one juvenile brain (0.095 × 10^-5 %ID/g) and kidney (0.13 × 10^-5 %ID/g) sample.

Conclusion

The presence of gadoteridol in the amniotic fluid after maternal injection enables confirmation that it crosses the placenta. Extremely low levels of gadolinium are found in juvenile macaque tissues after in utero exposure to two doses of gadoteridol, indicating that a very small amount of gadolinium persists after delivery.

^© RSNA, 2017

Introduction

The use of gadolinium-based contrast agents (GBCAs) for magnetic resonance (MR) imaging in pregnant women remains controversial. The latest version of the American College of Radiology (ACR) Manual on Contrast Media recommends that in pregnant or potentially pregnant patients, “GBCAs should only be used if their usage is considered critical and the potential benefits justify the potential unknown risk to the fetus” (1). Similarly, the European Society of Urogenital Radiology (ESUR) Contrast Media Safety Committee guidelines also distinguish between GBCAs, stating that “when there is a very strong indication for enhanced MR, the smallest possible dose of one of the most stable gadolinium contrast agents may be given to the pregnant female” (2).

The Food and Drug Administration lists all GBCAs as class C drugs. This means “animal reproduction studies have shown an adverse effect on the fetus and there are no adequate and well-controlled studies in humans, but potential benefits may warrant use of the drug in pregnant women despite potential risks” (3). However, this listing is based on toxicity studies in which very high doses (between three and 20 times greater than a standard clinical dose) were studied in animal models (4,5). Adverse effects related to GBCA use have drawn substantial attention in recent years. After infusion, GBCAs are generally excreted rapidly and, as a result, they were once considered to pose no substantial toxicity risk (6). About 10 years ago, some, but not all, GBCAs were found to be associated with the development of nephrogenic systemic fibrosis in patients with impaired renal function (7). Although no causative mechanism has been established, it is widely accepted that slow excretion (arising from renal impairment) and GBCAs that are more labile to dechelation of the gadolinium(III) (Gd³⁺) ion increase the risk of nephrogenic systemic fibrosis. Linear chelates of Gd³⁺ (gadopentate, gadodiamide, and gadoversetamide) are the GBCAs that present the highest risk. The substituted linear chelates (gadobenate, gadofosveset, and gadoxetate) pose an intermediate risk, whereas the macrocyclic chelates (gadobutrol, gadoterate, and gadoteridol) pose the lowest risk (8).

When a GBCA is to be used in a pregnant or potentially pregnant patient, the ACR states that one of the agents believed to have low risk for development of nephrogenic systemic fibrosis (gadobenate dimeglumine, gadobutrol, gadoterate meglumine, or gadoteridol) should be used (1). The ESUR also considers gadoteridol, the chelate of gadolinium with the macrocyclic ligand 10-(2-hydroxy-propyl)-1,4,7,10-tetraazacyclododecane-1,4,7-triacetate (HP-DO3A), one of these low-risk contrast agents because of its kinetic inertness (2). This chelate was the GBCA of choice in this study, as discussed in our prior article (9).

A small number of studies with a limited sample size have shown no evidence of teratogenic or mutagenic effects in pregnant women who received a GBCA (10–12). However, the long-term effects of in utero exposure to GBCAs are unknown. The toxicity of GBCAs is widely accepted to occur only after dissociation of the Gd³⁺ ion from the chelate in which it was administered. Reports have shown evidence of deposition of Gd⁺³ in several tissues, including bones (13), brain (14–16), kidney (17), and skin (17), especially after repetitive exposure to a GBCA. Given the increasing data available in regard to tissue deposition in patients without renal impairment, the Food and Drug Administration published a safety announcement in July 2015 that it was investigating the risk of brain deposits associated with repeated use of GBCAs with MR imaging (18). The clinical importance of this deposition has yet to be established. For these reasons, GBCA use has been historically avoided during pregnancy. However, use of GBCAs during pregnancy has slowly increased for clinical indications, such as imaging the morbidly adherent placenta (19–22), and for nonfetal indications when a woman requiring contrast material–enhanced MR imaging is found to be incidentally pregnant (23–25).

Previously, we have shown that in gravid nonhuman primates, a limited portion of the administered dose of gadoteridol was able to cross the placenta and reach fetal circulation but at levels substantially lower than those in maternal tissue. Gadoteridol was found to be excreted through the fetal renal system and into the amniotic fluid, and only traces of gadolinium could be detected in the fetal tissue 19–21 hours after maternal injection (9).

The aim of this study was to assess gadolinium levels in juvenile nonhuman primate tissues after in utero exposure to a low-risk GBCA, gadoteridol, during the second and third trimesters of pregnancy.

Materials and Methods

Animals

All protocols were approved by the institutional animal care and utilization committee of the Oregon National Primate Research Center (ONPRC), and guidelines for humane animal care were followed. The ONPRC abides by the Animal Welfare Act and Regulations enforced by the U.S. Department of Agriculture. Rhesus macaques (Macaca mulatta) were maintained on either a control chow diet consisting of 26% protein or a protein-modified diet containing 13% protein. Animals were socially housed in indoor-outdoor pens, with ad libitum access to food and water. There were 10 female macaques and one male macaque per diet-determined cohort. Animals were allowed to breed naturally, and pregnancies were identified early in the first trimester with routine two-dimensional ultrasonography (US) (Voluson 730 Expert; GE Medical Systems, Kretztechnik, Austria), with fetal biometry measurements used for gestational dating. These animals were part of a study to determine the effects of gestational protein restriction on pregnancy outcomes and placental function (26).

Term in the Rhesus macaque is about 168 days. At gestational days 85 and 135 and after an overnight fast, animals were sedated via intramuscular injection of 10 mg/kg of ketamine. The macaques were then intubated and maintained under anesthesia with 1%–2% inhaled isoflurane gas for the duration of each imaging study. The first MR imaging examination was performed during the equivalent of the second trimester of pregnancy (day 85), and the second MR imaging examination was performed in the early third trimester of pregnancy (day 135). MR imaging of the gravid macaque was performed in accordance with a pre-existing study protocol designed to assess placental function (Figure) (27). In the two animal cohorts (ie, the control and protein-restricted groups), there were seven pregnant animals per group. These animals underwent imaging studies with administration of a gadolinium chelate. Later in pregnancy, prior to term delivery, one animal in each cohort experienced spontaneous preterm pregnancy loss. This resulted in 12 juvenile macaques for the follow-up study. Of these 12 animals, one juvenile from the protein-restricted group had failure to thrive and died at 1 month of age. Additionally, one juvenile control animal was sacrificed early because of an injury that caused an animal welfare concern. Thus, tissue samples from 10 animals were available.

Axial MR images transecting the placenta and uterus in a pregnant rhesus macaque on gestational day 135. The two visible placental lobes are outlined in green. (a) T2-weighted half-Fourier rapid acquisition with relaxation enhancement image (repetition time msec/echo time msec, 1200/103; flip angle, 150°). (b) T1-weighted gradient-recalled echo image (2.03/0.75, α = 20°) acquired 30 seconds after administration of 0.1 mmol/kg gadoteridol.

Amniocentesis

On gestational day 135, prior to contrast-enhanced MR imaging, animals underwent US-guided amniocentesis in the surgical unit of the ONPRC to obtain 5 mL of amniotic fluid. Samples were centrifuged at 1800×g for 15 minutes at 4°C, aliquoted, and stored at −80°C for later analysis.

Gadolinium Chelate Injection

Gadoteridol (ProHance; Bracco Diagnostics, Princeton, NJ) was injected intravenously into the cephalic vein of the sedated gravid macaque prior to dynamic contrast-enhanced MR imaging. Gadoteridol is the chelate formed between Gd³⁺ and HP-DO3A. The administered dose of gadoteridol was based on a standard clinical dose of 0.1 mmol per kilogram of maternal weight. Animals underwent whole-body 3-T dynamic contrast-enhanced MR imaging (Tim Trio; Siemens, Erlangen, Germany) with a T1-weighted three-dimensional gradient-recalled echo sequence, with a circular polarization transmit 15-channel receive radiofrequency coil (Quality Electrodynamics, Cleveland, Ohio), as described in our previously published placental MR imaging methods (18,27).

Tissue Collection

The animal cohorts used for gadoteridol analysis were part of a larger study in which offspring were followed for 7 months, which is when they are typically weaned. At 7 months of age, juvenile animals were sedated with 10 mg/kg ketamine and were taken to the ONPRC pathology unit for full necropsy. Animals were anesthetized with sodium pentobarbital administered intravenously at 25 mg/kg to achieve a deep plane of anesthesia (assessed by loss of palpebral, corneal, pain, and gag reflexes). The abdomen was incised after adequate anesthesia had been established, and the aorta was severed to cause exsanguination and euthanasia. A blood sample was obtained from the aorta and was processed in a vacutainer tube containing 10.8 mg K₂EDTA, inverted to mix, and a 1-mL whole-blood sample was snap frozen in liquid nitrogen. Tissue samples from the blood, brain, cerebellum, liver, kidney, spleen, lung, femur, and skin of 10 animals were harvested, weighed prior to dissection, and immediately snap frozen in liquid nitrogen for later analysis. The brain samples were taken from the caudal portion of the occipital lobe, as this region was available for analysis (other portions of the brain parenchyma were used for ongoing studies of this animal model and were not able to be obtained in this study). The liver samples were consistently collected from the center of the right lobe, the articular head of the femur was removed at the hip, and a portion of bone was consistently collected from the midportion of the femur.

Inductively Coupled Plasma Mass Spectrometry Assay

The Gd³⁺ content of the amniotic fluid and in the juvenile macaque brain, cerebellum, liver, kidney, spleen, lung, femur, and skin was measured with inductively coupled plasma mass spectrometry, as described previously (9).

Statistical Analyses

The quantitative levels of Gd³⁺ in the juvenile tissue were expressed as the percentage of the maternal injected dose (ID) detected per gram of juvenile tissue and were calculated with Microsoft Excel 2013 software (Microsoft, Redmond, Wash). P < .05 was considered indicative of a significant difference.

A two-tail test was performed to determine statistical significance. Post hoc power analysis was performed on the data for which no significance was found between dietary groups (femur and amniotic fluid) by using GPower software (Heinrich-Heine-Universität Düsseldorf).

Results

Animals

Mean maternal weight at gestational day 85 was 6.53 kg (range, 5.55–7.65 kg). At gestational day 135, the mean maternal weight was 6.68 kg (range, 5.55–7.8 kg). The gadolinium doses received were 0.1 mmol/kg per injection, corresponding to average doses of 102.6 mg and 105.1 mg at gestational days 85 and 135, respectively. The total average dose during pregnancy was 207.8 mg.

Gd³⁺ Analyses

The quantity of Gd³⁺ found in the tissues sampled 7 months after birth is given in the Table as the percentage of ID per gram of tissue. Gd³⁺ was consistently detectable in only the liver and femur. The limit of detection for the samples was determined from calibration standards and was 0.6 parts per trillion, which resulted in a limit of detection of 0.0003 × 10^-5 %ID/g. Three juveniles were found to have detectable levels of Gd³⁺ in areas other than the liver or femur. In one animal fed a control diet, Gd³⁺ levels of 0.095 × 10^-5 %ID/g and 0.13 × 10^-5 %ID/g were detected in the brain and kidney, respectively. In a second animal also fed a control diet, a Gd³⁺ level of 0.07 × 10^-5 %ID/g was detected in the skin. In a third animal fed a protein-restricted diet, a Gd³⁺ level of 0.039 × 10^-5 %ID/g was detected in the spleen. No other animals had detectable levels of Gd³⁺ in any of these organs. The mean gadolinium concentration in amniotic fluid at gestational day 135 (50 days after administration of the first dose of gadoteridol) was 0.028 × 10^-5 %ID/g.

Total Maternal Dose of Gadoteridol and Gd³⁺ Levels in Juvenile Rhesus Macaque Tissues at 7 Months and in Amniotic Fluid 50 Days after Initial Administration

Note.—For each column, there are two units of measure: ×10^-5 percentage ID per gram of tissue and, in parentheses, parts per billion.

*Comparison between different diet group means.

^†Below limits of detection. The Gd³⁺ levels are expressed in %ID/g × 10⁻⁵, which is the percentage injected maternal dose per gram × 10⁻⁵.

The control diet group was found to have, on average, lower levels of residual Gd³⁺ in the liver: 0.079 × 10^-5 %ID/g (2.9 × 10^-5 %ID per organ) versus 0.21 × 10^-5 %ID/g (7.0 × 10^-5 %ID per organ) in the protein-restricted group (P < .005). No significant difference was found between the residual Gd³⁺ levels of the two dietary groups for either the amniotic fluid (P = .47) or the femur (P = .32). However, a post hoc power analysis enabled us to confirm that the sample sizes were too small to show significance (1-β = 0.05 and 1-β = 0.149 for amniotic fluid and femur, respectively; 666 482 and 63 subjects would be required for amniotic fluid and femur, respectively).

Discussion

The results of this study support our previous conclusion that gadoteridol can be excreted intact from primate fetuses (9). In our prior study, biodistribution data in fetuses collected 1–2 days after maternal administration of gadoteridol were consistent with limited transplacental transfer of gadoteridol to the fetus, followed by renal excretion from the fetus into the amniotic fluid (9). Once in the amniotic fluid, gadoteridol can presumably return to the fetal circulation through various mechanisms, including absorption in the gastrointestinal tract with fetal swallowing (28–30), with a very small concentration of Gd³⁺ remaining in the amniotic fluid 50 days after maternal injection. Prolonged recirculation of the contrast agent in fetal tissue could raise health concerns; therefore, it is important that gadoteridol can be reabsorbed by the placenta and eventually excreted by the mother. This study shows that this is indeed what happens. At 50 days after administration of one dose of gadoteridol, the amount of Gd³⁺ in the amniotic fluid was four orders of magnitude lower than what was measured in our prior study 1 day after administration (9). This clearly indicates that gadoteridol is transferred from the fetal circulation back to the maternal bloodstream (31).

Excretion of gadoteridol from fetal tissue can also be seen when we compare the amount of residual Gd³⁺ in the juvenile macaque 7 months after delivery with that in the fetus 1 day after administration (9). In the majority of juveniles, Gd³⁺ was found to have cleared to below detectable limits in the blood, brain, cerebellum, kidney, lung, skin, and spleen. Only the liver and femur show any consistently detectable levels of residual Gd³⁺ in the juvenile. The levels of Gd³⁺ in the femur and liver are extremely low, corresponding to nanomolar (2.5 × 10^-5 %ID/g) and subnanomolar (0.15 × 10^-5 %ID/g) concentrations, respectively, and they are a tiny fraction of the total administered maternal dose. We acknowledge that maternal diet affects overall health and well-being. The observation that higher levels of Gd³⁺ were present in the liver after gestational protein restriction may raise concerns about the use of contrast-enhanced MR imaging in malnourished mothers. It is of particular importance that, despite the diet-related differences, the overall Gd³⁺ levels are extremely low (19 pmol/g) in the juvenile liver.

Inductively coupled plasma mass spectrometry is incapable of enabling distinction between intact gadoteridol and Gd³⁺ that is liberated from the chelate. However, previous preclinical studies in adult mammals indicate that dechelated Gd³⁺ tends to accumulate first in the liver and then eventually in bone (32). Many of these biodistribution studies in adult animals have limits of detection that are much higher than those in this study because the other studies use scintigraphy with a ¹⁵³Gd³⁺-labeled contrast agent for analysis; this makes it difficult to compare those results with our results, which were obtained with much lower detection limits (32–35). Deposition of dechelated Gd³⁺ in bone has been previously shown after administration of GBCAs (13) and can persist for many years (36). It is reasonable to infer that the residual Gd³⁺ in bone at a time point long after administration is mineralized dechelated metal. Minuscule residual levels of Gd³⁺ in bone are inevitable after administration of any GBCA (36); however, in more than 30 years of use, this residual level of Gd³⁺ has never been shown to manifest itself clinically. Gd³⁺, a known calcium (Ca²⁺) analog, could be incorporated into the carbonated calcium hydroxyapatite mineral phase of bone after being released from the gadolinium chelate (36). If we include all the data from femoral head specimens reported by Darrah et al, although the amount of Gd³⁺ found in the femurs of the juvenile macaques in this study (0.3 nmol/g) is a factor of three higher than they found in individuals with no GBCA exposure, it is a factor of 30–50 lower than the levels determined for cortical and trabecular bone in adult humans who had been exposed to GBCAs (36).

In an effort to gauge the clinical relevance of in utero exposure to a GBCA, a large retrospective study was recently published that suggested there is an increased risk of “a broad set of rheumatological, inflammatory, or infiltrative skin conditions and of stillbirth or neonatal death” in the offspring of women who undergo gadolinium-enhanced MR imaging during pregnancy when compared with risk in women who do not undergo this procedure (37). In their study, it is important to note that Ray et al had no data on the type of administered GBCA. As we have discussed, the ESUR and ACR recommend use of the more stable macrocyclic contrast agents, when required during pregnancy, which would likely mitigate the amount of dechelated Gd³⁺ in the fetal tissues. Further prospective studies are required to demonstrate any causality implicated by the aforementioned study.

Our study had several limitations. The sample size was small. Tissue collection was from a small portion of the organ of interest, and it is possible that differential deposition of gadolinium occurs within organs. For example, the basal ganglia in the brain and dentate nucleus in the cerebellum were not able to be specifically sampled, as only tissue from the occipital lobe was used in this study (16). Gadoteridol infusion was timed to the second and third trimesters of the macaque pregnancy; however, we acknowledge that our results could be affected if exposure had occurred in the first trimester, when the most active organogenesis occurs.

We show that after gadoteridol undergoes transplacental transfer to the fetus, it is almost completely cleared, with nearly undetectable levels in the juvenile macaque. By 7 months after delivery, Gd³⁺ is below detectable levels in nearly all the tissues we sampled in the juvenile macaque, with only trace amounts in the liver and bone. The levels of residual Gd³⁺ in bone are substantially lower than those found in adults after administration of one GBCA dose. While the toxicity of GBCAs has been widely associated with in vivo dechelation of Gd³⁺, the rapid and almost complete clearance from fetal tissues, combined with low levels of dechelation, is reassuring regarding risk during pregnancy. Given the similarities between human and nonhuman primate placental physiology, we suggest there could be relatively little deposition in human fetal tissues after maternal gadoteridol injection. However, the long-term risk of such low levels of gadolinium deposition is still unknown.

Advances in Knowledge

• ■ By 7 months after delivery, gadolinium 3+ (Gd³⁺) is below detectable levels in almost all tested tissues (blood, brain, cerebellum, kidney, lung, skin, and spleen) in the juvenile macaque, with only traces of Gd³⁺ remaining in the liver and bone.

• ■ In three of 10 animals, trace amounts of Gd³⁺ were detected in tissues other than liver and bone; these were the skin (n = 1), spleen (n = 1), and brain and kidney (n = 1).

Implication for Patient Care

• ■ Our results have implications for the safety of contrast-enhanced MR imaging in pregnancy, as nearly undetectable levels of gadolinium are found in the tissues of juvenile offspring after in utero exposure to gadoteridol from maternal dosing; however, long-term effects of these low doses are unknown.